Fluorescent probes and nanostructured materials for the detection of environmental toxins and catalysts development
Abstract
La tesis trata sobre la síntesis y utilización de materiales modificados, para el reconocimiento de distintas sustancias por fluorescencia o para el desarrollo de catalizadores de reacciones orgánicas de interés. En la primera parte de la tesis se desarrollaron sensores de sustancias tóxicas como mercurio (II) y plomo (II), midiendo en muestras de pescado y disolución respectivamente. Por otro lado, se perfeccionó un sensor para la detección de TATP en fase gas, explosivo sin uso militar utilizado en ataques terroristas. Además, y en complemento a lo anterior, se trabajó con sensores para detectar toxinas alimentarias como la cereulida, o moléculas de interés biológico/toxicológico como el monóxido de carbono. En una segunda parte, se modificaron una serie de materiales con partículas metálicas de oro o paladio para la catálisis de reacciones orgánicas de importancia en la industria; presentando ventajas como ser reciclables, alto rendimiento de reacción y gran selectividad en los productos.
Full text
UNIVERSIDAD DE BURGOS FACULTAD DE CIENCIAS Departamento de Química Área de Química Orgánica Fluorescent probes and nanostructured materials for the detection of environmental toxins and catalysts development AthesissubmittedforthedegreeofDoctorofPhilosophy by JoséGarcíaCalvo Burgos,2018
UNIVERSIDAD DE BURGOS FACULTAD DE CIENCIAS - DEPARTAMENTO DE QUÍMICA Dr. TOMÁS TORROBA PÉREZ, Professor of Organic Chemistry at the Organic Chemistry Department of the University of Burgos, HEREBY CERTIFIES: That the research work included in the dissertation: “Fluorescent probes and nanostructured materials for the detection of environmental toxins and catalysts development” performed by Mr. José García Calvo, Bsc. (Hons) Msc., in order to apply for the degree of Doctor of Philosophy (PhD) in Chemistry by the University of Burgos, constitutes an original the research contained in the PhD Thesis has been developed under my supervision and I give my approval for its submission to be defended as a PhD Thesis. Burgos, 15th June 2018 SGD: Tomás Torroba Pérez
Agradecimientos Han sido muchos años desde que comencé el doctorado y mucho trabajo el que he realizado en este tiempo hasta completar la tesis. Por todo ello, a continuación, quiero agradecer a todas las personas que han ayudado o participado en el desarrollo de la misma. Comenzando con mi director de tesis, Tomás Torroba, siempre disponible para discutir cualquier duda y cuya dirección, ayuda y paciencia me han permitido llegar hasta aquí. Para mi es muy importante agradecer también a todos los miembros y exmiembros del grupo con los que he trabajado directamente; a Patricia Calvo, con la cual compartí muchos años de tesis y cuya ayuda ha sido fundamental; a Borja Díaz de Greñu. por enseñarme a trabajar en el laboratorio cuando estaba empezando y a Alberto Díez por su ayuda en los análisis de masas y muchas otras dudas en los últimos años. Por supuesto, también quiero expresar mi agradecimiento a los muchos otros compañeros que he tenido durante estos años, a Daisy C. Velásquez. con la que he compartido mucho tiempo como doctorandos; a los estudiantes que realizaron sus trabajos de Fin de grado y/o Máster en el grupo, Clara Antón, Marcos Ibáñez, Virginia Renuncio, Cristina Viyuela y Andrea Revilla; así como a los técnicos, cuya ayuda es inestimable, Sergio Domingo, Nerea Jalón, Miryam Asensio y los últimos en unirse Juan Ingelmo e Irene Abajo. También al resto de compañeros que me han ayudado y/o con los que he colaborado y trabajado en el laboratorio en distintas partes de la tesis, Pablo Peña, Saúl Vallejos, Andrea Sancho o Natalia Busto entre otros muchos. Especialmente, también quiero mencionar y agradecer a James D. Wilton-Ely y los miembros de su grupo que me permitieron realizar la estancia en el Imperial College, que siempre se mostraron amables y con ganas de ayudar. Por último y más importante, todo esto no sería posible sin mi familia, mis padres, mi hermana y, sobre todo, mi hermano Víctor García, que trabajó en laboratorio conmigo y me ha ayudado siempre que lo necesito; y a mi pareja, Esther López, sin la cual no hubiera sido posible esta tesis, no sólo por leérsela y corregirla, sino por su apoyo diario, su ayuda y ánimo. Gracias a todos.
ABBREVIATIONS SOLVENTS: Ac: Acetone 1,2-DCE: 1,2-Dichloroethane. CH: Cyclohexane. DCM: Dichloromethane. DMF: N,N-Dimethylformamide. DME: Dimethoxyethane. DMSO: Dimethyl sulfoxide. EtOAc: Ethyl acetate. EtOH: Ethanol. Et2O: Diethylether. Hex: Hexane. MCH: Methylcyclohexane. MeCN: Acetonitrile. MeOD: Deuterated methanol. MeOH: Methanol. NMP: N-Methylpyrrolidone. PhCl: Chlorobenzene. THF: Tetrahydrofuran. Tol: Toluene. REAGENTS: DABCO: 1,4-Diazabicyclo[2.2.2]octane. DADP: Diacetone diperoxyde. DCTB: Trans-2-[3-(4-tert-Butylphenyl)-2-methyl-2-propenylidene]malononitrile. DIPEA: N,N-Diisopropylethylamine. DIT: Dithranol. DMAP: 4-Dimethylaminopyridine. DMPA: 2,2-Dimethoxy-2-phenylacetophenone. EDCI: N-(3-Dimethylaminopropyl)-N′-ethylcarbodiimide. Et3N: Triethylamine. HEPES: 2-[4-(2-Hydroxyethyl)piperazin-1-yl]ethanesulfonic acid. HOBt: Hydroxybenzotriazole. HMTD: Hexamethylene triperoxidediamine. NBS: N-Bromosuccinimide. MCPBA: 3-Chloroperoxybenzoic acid. NIS: N-Iodosuccinimide. Oxone: Potassium peroxymonosulfate. PBS: Phosphate buffer solution. PDA: Perylenedianhydride. PDI: Perylenediimide. PMI: Perylenemonoimide. pTsOH: p-Toluenesulfonic acid. PyBOP: Benzotriazol-1-yl-oxytripyrrolidinophosphonium hexafluorophosphate.
TATP: Triacetone triperoxide. TAMRA: Tetramethylrhodamine. TBTA: Tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine. TFA: Trifluoroacetic acid. TECHNIQUES. 1H NMR: Proton Nuclear Magnetic Resonance. 13C NMR: Carbon Nuclear Magnetic Resonance. 19F NMR: Fluorine Nuclear Magnetic Resonance AFM: Atomic Force Microscopy. EDX: Energy Dispersive X-Ray spectroscopy. ESI: Electrospray Ionization. ETAAS: Electrothermal Atomic Absorption Spectrometry. FAAS: Fame atomic absorption spectrometry. FT-IR: Fourier-Transform Infrared Spectroscopy. HRMS: High Resolution Mass Spectrometry. MALDI: Matrix-Assisted Laser Desorption/Ionization. TOF: Time-of-Flight. LSIMS: Liquid Secondary Ion Mass Spectrometry. HMBC: Heteronuclear Multiple Bond Correlation. HMQC: Heteronuclear Multiple-Quantum Correlation. ICP-MS: Inductively Coupled Plasma Mass Spectrometry. ICP-OES: Inductively Coupled Plasma Optical Emission Spectrometry. HPLC: High Performance Liquid Chromatography. COSY: Correlation Spectroscopy. (NMR analysis) NOESY: Nuclear Overhauser Spectroscopy. (NMR analysis) DEPT: Distortionless Enhancement by Polarization Transfer. (13C-NMR analysis) TEM Transmission Electron Microscopy. TGA: Thermogravimetric Analysis. UPLC: Ultra Performance Liquid Chromatography. TLC: Thin Layer Chromathography. XPS: X-Ray Photoelectron Microscopy. OTHER ABBREVIATIONS: CORM: CO Releasing Molecules. Cq: Quaternary Carbon. DFT: Density Functional Theory. EDG: Electron Donating Group. EWG: Electron Withdrawing Group. FBS: Fetal Bovine Serum (for cell culture). FRET: Förster Resonance Energy Transfer. NPs: Nanoparticles. PET: Photoelectronic Transference. PCT: Photoinduced Charge Transfer. Ps: Particles. τ = FLD: Fluorescence Lifetime Decay. ΦF = FQY: Fluorescence Quantum Yield.
INDEX INTRODUCTION. CONTEXT AND OBJECTIVES OF THE THESIS ……………………………………………………………1-6 CHAPTER 0. INTRODUCTION TO FLUORESCENT PROBES …………..7-50 1. SENSORS, DEFINITION AND TYPES ..................................................................................... 9 1.1. Types of chemical sensors .......................................................................................................... 9 1.2. Sensors, characteristics and advantages. The role of Absorbance + Photoluminescence sensors. ..................................................................................................................................... 11 2. CHARACTERISTICS OF FLUORESCENT PROBES .......................................................... 13 2.1. The mechanism of fluorescence ............................................................................................... 13 2.2. Factors that have influence over fluorescence .......................................................................... 14 2.3. Fluorescence quenching: .......................................................................................................... 15 2.4. ON-OFF vs OFF-ON fluorescent sensors ................................................................................ 17 2.5. Types of molecular fluorescent probes ..................................................................................... 17 2.6. Characteristics of an ideal fluorescent probe ............................................................................ 21 2.7. How to work with molecular sensors: ...................................................................................... 23 3. PURPOSE AND PROCEDURE WHEN PERFORMING FLUORESCENT STUDIES ..... 24 3.1. Solvatochromism ...................................................................................................................... 24 3.2. Tests with different species; cations, anions, oxidative, reductive species, amines… ............. 26 3.3. Work concentration and molar extinction coefficients (ε) ....................................................... 28 3.4. Kinetic effects ........................................................................................................................... 29 3.5. Titration methods ...................................................................................................................... 30 3.6. Stoichiometry determination .................................................................................................... 31 3.7. Thermodynamic equilibrium constant calculation (K) ............................................................. 33 3.8. Limits of detection (LODs) ...................................................................................................... 38 3.9. Fluorescence quantum yields (ΦF) ............................................................................................ 43 3.10. Fluorescence decay lifetime (τ) ................................................................................................ 45 4. GENERAL SCHEME OF FLUORESCENT PROBES; PHOTOPHYSICS AND SUPRAMOLECULAR CHEMISTRY .............................................................................................. 47 5. RESUMEN DEL CAPÍTULO .................................................................................................... 48
6. EXPERIMENTAL SYNTHESIS OF GOLD NPs .................................................................. 297 6.1. Antecedents ............................................................................................................................ 297 6.2. Synthesis and characterization of gold NPs in solution .......................................................... 297 6.3. Synthesis and characterization of supported gold particles .................................................... 299 7. APPLICATION IN CATALYSIS OF GOLD-NPs ................................................................ 307 7.1. Synthesis and yields ............................................................................................................... 307 7.2. Recyclability of the catalyst ................................................................................................... 308 7.3. Turn-over number (TON) and turn-over frequency (TOF) .................................................... 310 7.4. Conclusions of the material as catalyst ................................................................................... 310 8. EXPERIMENTAL SYNTHESIS OF PALLADIUM-NPs ..................................................... 311 8.1. Characterization of the films .................................................................................................. 313 8.2. Deep study of PB20_80A1 and PB80_20 modifications with palladium .............................. 315 9. PALLADIUM SUPPORTED POLYMERS AS CATALYSTS ............................................. 322 9.1. Reduction of DMAD with palladium modified films ............................................................. 324 9.2. Leaching of the polymers ....................................................................................................... 325 9.3. Other important characteristics studied .................................................................................. 325 10. CATALYTIC REDUCTION OF COMPOUNDS WITH BIOLOGICAL INTEREST . 326 10.1. Conditions of the reaction and yields obtained ...................................................................... 328 10.2. Results for Pd films reduction and comparison with Pd/C catalyst ........................................ 328 11. PALLADIUM-SUPPORTED REDUCTION. MECHANISM AND CALCULATIONS 334 12. SUPPORTED GOLD AND PALLADIUM PARTICLES. SUMMARY .......................... 336 13. RESUMEN DEL CAPÍTULO .............................................................................................. 338 FINAL CONCLUSIONS……………………………………339-342 ANNEX……………………………………………………….343-354 REAGENTS AND SOLVENTS……………………………………………………………………345 APPARATUS......................................................................................................................................345 MATERIAL AND SAMPLES FOR SENSORING STUDIES.......................................................346 PUBLICATIONS................................................................................................................................348
INTRODUCTION. CONTEXT AND OBJECTIVES OF THE THESIS
INTRODUCTION.CONTEXTANDOBJECTIVESOFTHETHESIS|3 JoséGarcíaCalvo|PhDThesis CONTEXT Throughout the years, the research topics of the group have changed so as to achieve the objectives of different projects and to improve previous results. In this way, the work has been characterized for being open to explore new areas of chemistry instead of being focused in one topic. Therefore, this thesis explores the applications of new dyes in some research areas such as chemical sensors, catalysis and new materials. Initially, the main scope of the research group was the development of fluorescent sensors for detection of environmental contaminants and dangerous/toxic substances. In this regard, the research was devoted to the elaboration of fluorescent chemical probes for a quick, easy and cheap detection. This was the topic of the first chapter of the thesis (Chapter 0) which includes a background search to understand how fluorescence works, what are their advantages and which would be the proper way to use fluorescent probes for sensing. Thus, the laboratory work of this thesis started by the development and improvement of fluorescent probes for detection of Hg(II) derivatives. Previous members of the group had synthetized and tested blue/yellow fluorescent probes soluble in organic-aqueous media. However, the research needed for completion, which was achieved by synthetizing a derivative soluble in water. Furthermore, the possibility of creating a material with water affinity and enhanced sensitivity to Hg(II) was explored. The development of the study is thoroughly explained in Chapter 1. Alongside the development of the probes for Hg(II) detection, the research group became involved in a European Research Project. The project was entitled “Sensory devices network for food supply chain security”, SNIFFER, whose objectives were focused on the development of tools for detection of CBR agents (chemical, biological and radiological). As a consequence, and taking into account previous results, two objectives were raised for the rest of the research: the improvement of the characteristics of the fluorogenic probes (emission, solubility and selectivity) and their adaptation for detection of some chemical and biological threats. First synthetized probes (Hg(II) selective probes) were based on the fluorescence of indanone derivatives. As a consequence, the working range was very limited (generally blue fluorescence) and the emission intensity was quite poor (inferior to 10 % in fluorescence quantum yield). To solve these issues, the first step was the search for a fluorogenic backbone with more suitable characteristics. The selected dye was the perylenemonoimide (PMI) family of dyes. These structures are usually fluorescent in the region of green-red and show fluorescence quantum yields close to 100 %, in most cases; therefore, they are perfect candidates for fluorescent probes. Chapter 2 describes the properties, advantages and some of the applications of this kind of derivatives. Using the new fluorescent backbones in the synthesis of sensors was the next task to address. It was divided into three parallel research topics. First, a PMI was adapted for the detection of explosives, specifically TATP (Chapter 3A). Second, PMI derivatives were used as a metallic ligand so as to develop sensitive probes to carbon monoxide (in collaboration with the Imperial College London, Chapter 3B). Finally, they were modified to be sensitive to potassium and lead cations, and indirectly, for detecting foodborne toxins, such as cereulide (Chapter 3C), being an objective directly associated to the SNIFFER project. This part represents the core of the thesis covering all the studied applications for the synthetized PMIs and presenting the results in Chapters 2 and 3. Additionally, during the search for specific materials for new Hg(II) sensors, some outstanding properties of several of the modified materials were discovered. They possessed the ability to perform
4|INTRODUCTION.CONTEXTANDOBJECTIVESOFTHETHESIS JoséGarcíaCalvo|PhDThesis one-step synthesis and modification of the surface with metallic nanoand microparticles. Hence, the work proceeded throughout the optimization of such materials for their modification, due to the lack of previous results in literature about such properties in a material. Furthermore, the study was complemented with a research about the applications of these materials, mainly for heterogeneous organic catalysis (Chapter 4). Finally, and complementary to all the results explained in the different chapters of the thesis, two extra parts are included. On one side, the Annex contains the data about reagents, materials and technical characteristics of the used equipment, as well as a summary of publications and symposiums in which this work was presented. On the other side, and in order to clarify some parts, the Experimental Appendix (digital format) includes all the reaction schemes, a complete characterization of the different compounds and materials, some extra experiments (from Chapters 1, 3B and 3C) and the DFT calculation parameters. OBJECTIVES Different aims were proposed for the different chapters of the thesis: Chapter 0: Introduction to fluorescent probes. Explanation about the different kinds of chemical sensors. Introduction to fluorescent sensors. Elaboration of a protocol to work with fluorescent probes. Chapters 1, 2 and 3 of the thesis are oriented to the development of new fluorescent probes for several applications. Chapter 1: Fluorescent probes for the detection of Hg(II) derivatives. Introduction of Hg(II) and its derivatives, the importance of MeHg(II). Description of new fluorescent probes for detection of Mercury(II) derivatives. Synthesis of water soluble probes for Hg(II) detection. Detection of Hg(II) and MeHg(II) in cellular environments. Creation of materials that are capable to detect Hg(II) presence in water. Detection and quantification of Hg(II) from fish samples. Chapter 2: Perylenemonoimides. Introduction and general properties. Introduction to the properties of perylene imide derivatives Optimization of the synthesis of PMIs. Comparative study of PMIs substituted in different ways. Introduction to some general applications for PMIs: cellular imaging, solar cells, sensors and biological markers.
INTRODUCTION.CONTEXTANDOBJECTIVESOFTHETHESIS|5 JoséGarcíaCalvo|PhDThesis Chapter 3A. PMI derivatives for detection of explosives. Introduction to the interests of detecting TATP. Synthesis of a fluorescent probe with a selective response for TATP detection. Optimization of a material to perform the detection of TATP in the gas phase. Evaluation of the properties of the material for TATP detection. Chapter 3B. PMI-Ru(II) complexes for detection of CO. Introduction to the advantages of Ru(II) complexes for CO sensing. Synthesis of complexes Ru(II)-PMI. Evaluation of the properties / sensitivity of the complexes. Application of the complexes in CO detection. Chapter 3C. PMI derivatives for K+ and Pb2+ sensing. Introduction to the importance of K+, Pb2+, cereulide and derivatives; their role and the interest of their detection. Synthesis of fluorescent PMI probes for K+ and Pb2+. Synthesis of cereulide, a natural ionophore of K+. Evaluation of the sensors properties. Detection of cereulide from B. cereus cultures. Preparation of cereulide derivatives and comparison of their properties. Design and evaluation of a material for sensing of cations, intended for Pb2+. Chapter 4. Supported gold and palladium nanoparticles for catalysis. Introduction to the applications of metal-nanoparticles modified materials. Modification of a material with metal nanoparticles in one-step. Optimization of conditions for getting homogeneous surfaces covered by gold and palladium nanoparticles and characterization of the materials. Using the materials for heterogeneous catalytic processes.
6|INTRODUCTION.CONTEXTANDOBJECTIVESOFTHETHESIS JoséGarcíaCalvo|PhDThesis General scheme of the Chapters of the thesis and their content
CHAPTER 0 INTRODUCTION TO FLUORESCENT PROBES ABSTRACT Due to their many applications, one of the aims of this thesis was the development of highly efficient fluorogenic probes, which may be useful for detection of different species or as biological markers, amongst other possibilities. With this aim, this chapter is intended to define what the fluorescent sensors are, what properties they have, how to design them properly and what data can be obtained from them, being an introduction for Chapters 1, 2 and 3.
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|9 JoséGarcíaCalvo|PhDThesis 1. SENSORS, DEFINITION AND TYPES A sensor is defined as an interface capable of receiving and translating information across physical, chemical and biological samples. Chemical sensors are characterized for being based on a chemical interaction that leads to one or several analytical responses; therefore, a molecular sensor1 is a system that is capable of detecting the presence of a specie in a molecular level by a signal that can be measured. Throughout history there have been many methods to perform sensing procedures and some of them, such as liquid or gas chromatography, have become standards. Although such methods are very sensitive, their cost is usually prohibitive to low scale industries, for instance in some food and pharmaceutical quality control. As a consequence, since 1970s there has been an increasing interest in developing methods with lower costs, faster response times, possibility of miniaturization and greater accessibility. All these characteristics may be achieved by designing new and more efficient molecular sensors, which has become a very important issue not only in the field of chemistry but in industry too, fundamentally because of the high level of different substances they handle. The main purpose for molecular sensors is detecting the presence of some toxic or dangerous species in waste but it can be extended to other purposes. Some examples of what molecular sensors could perform are the detection of changes in the DNA structure or the distribution of metabolites within cells. In conclusion, research in this field is of high interest for biomedical, environmental, safety or food related purposes. 1.1. Types of chemical sensors 2 There are many ways of developing and classifying chemical sensors. Nevertheless, the process followed for detection is common for all of them and is divided into three steps: recognition of the analyte, transduction of the signal and measurement/analysis (Figure 1). Figure 1. Recognition process of a sensor. 1 J. X. J. Zhang, K. Hoshino, Molecular Sensors and Nanodevices: Principles, Designs and Applications in Biomedical Engineering, Elsevier, Chapter 1, pp. 1-42, 2014. 2 P. Gründler, Chemical Sensors: An Introduction for Scientists and Engineers, Springer-Verlag Berlin Heidelberg, Chapter 1, pp. 1-13, 2007.
16|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis Figure 6. Scheme of the two different kinds of fluorescent quenching and how they work. Dynamic quenching is something intrinsic to every fluorescent material or solution. It is defined as the decrease in fluorescence because of collisions in a molecular level. Thus, if there are many species, the final fluorescence decreases, and if the species have accessible energy levels (high molecular weight, transition metals, ions, very conjugated…) the fluorescence decreases even faster with concentration. Moreover, because of how it works, it is highly affected by temperature and viscosity of the media. The variation on the emission intensity is related to the concentration of the quencher by the SternVolmer equation:7 1𝑘𝜏𝑄 Equation [A] 𝑘8𝑅𝑇/3𝜂 Equation [B] Where Φ is the quantum yield, I the emission intensity, τ the fluorescence decay lifetime, [Q] the concentration of the quencher and kq the quenching constant. While kq is directly proportional to temperature and inversely to the viscosity of the reaction (η). In consequence, this previous equation gives a linear dependence between the variation in emission and the concentration of the quencher, provided that the solvent and temperature are constant. Static quenching consists of the specific interaction between the fluorophore and the analyte. When it occurs, it is far more noticeable, compared to dynamic quenching, and less dependent on temperature. An equation that relates the molecular interaction in the equilibrium with the variation of the emission and concentration may be defined to calculate the thermodynamic equilibrium constant (Section 3.6). In practice, there are several experiments that may be performed so as to distinguish static from dynamic quenching. For example, when it is dynamic, the fluorescence decreases linearly with the quencher concentration (except if the variation in quencher concentration is too high) but, when it is static, it has downward curvature when having enough amount of the quencher. Another way to distinguish between them is the fact that dynamic quenching is highly affected by temperature. What is more, some techniques, such as the study of the lifetime of the fluorophore in the presence of the analyte, are useful to distinguish between dynamic and static quenching (Section 3.10). 7 J.R. Lakowicz, G. Weber, Biochem. 1973, 12, 4161-4170.
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|17 JoséGarcíaCalvo|PhDThesis 2.4. ON-OFF vs OFF-ON fluorescent sensors Detecting the presence of an analyte by fluorescence may be done by two processes: By a decrease in fluorescence (ON-OFF). This method is based on the specific quench of fluorescence by interaction with the analyte, static quenching. The main drawback is that quenching is not a selective process. In fact, the presence of other substances (such as cations and/or anions) or the probe in high concentration leads to OFF processes (dynamic quenching). As a consequence, all of them must be taken into account in the results, eventually leading to possible false positive results if they are not considered. By an increase in fluorescence (OFF-ON). It is associated to an interaction with specific analytes, provided that temperature and solvent are set. It may occur through different mechanisms; an increase from a non-fluorescent species, a decrease in fluorescence by increasing in another wavelength, or just an increase in different wavelengths to the initial one. This is the most trustworthy method because of the easy distinction from quenching effects. In addition, the selectivity may give different responses depending on the analyte/analytes. Some authors may consider a third variation when measuring fluorescence. A change in the wavelength of emission. However, this might be considered as an OFF-ON probe, being an increase of fluorescence at a different wavelength. Besides that, it is important to remark that there are ways in which the fluorescent properties of the probe may be tuned differently. For example, there are some OFF-ON processes that work through a mechanism with lower selectivity than some ON-OFF processes. To illustrate the idea, it might be the case of an isolated non-fluorescent complex in which a ligand is displaced by the analyte under study, increasing the final fluorescence, and this process could be performed with very low selectivity. 2.5. Types of molecular fluorescent probes There are many ways to classify molecular fluorescent probes, depending on how they work. Most authors classify fluorescent probes by the interaction Fluorophore-Receptor, being the most straightforward distinction. There are two main possible interactions, if the probe works by chemical reaction or by complexation. In addition, at the same time, they may be classified depending on the type of complexation or reaction, as it is explained in this section. First, and in order to simplify the classification, this section is described from the point of view of OFF-ON probe. The equivalent process may be done the other way around, for the ON-OFF samples, with the differences previously explained.
18|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis 2.5.1. The recognition process is based on a complexation8 a) Photoelectron transfer (PET): 9 Fluorescent PET molecular sensors consist of a fluorophore linked to a recognition unit via a non-conjugated sigma bond (spacer). When the receptor interacts with its analyte, the electron transfer is hindered and an enhancement of fluorescence is observed (Figure 7A). From the point of view of the orbitals (Figure 7B) of the molecule, the electrons from the HOMO level of the receptor are between the HOMO level and the LUMO level of the fluorophore, inhibiting the fluorescence from the excited state by non-radiative relaxation processes; until the analyte is recognized. Figure 7. PET process mechanism and fluorescent response (A) and orbitals diagram (B). b) Photoinduced charge transfer (PCT): PCT molecular sensors consist of a sensor in which the recognition unit is part of the fluorophore. Usually, the fluorophore contains an electron - donating group (for example an amino group) conjugated to an electron - withdrawing group. As a consequence, after excitation with light, it undergoes intramolecular charge transfer from the donor to the acceptor. (Figure 8) The change in dipole moment (change in polarity/conjugation), when recognizing a species by the receptor unit, results in a Stokes shift that depends on the fluorophore, giving the opportunity to design probes on this basis. 8 B. Valeur, Molecular Fluorescence: Principles and Applications, Wiley-VCH, Weinheim, Chapter 14, 2002. 9 A. Prasanna de Silva, H. Q. N. Gunaratne, T. Gunnlaugsson, A. J. M. Huxley, C. P. McCoy, J. T. Rademacher, T. E. Rice; Chem. Rev. 1997, 97, 1515-1566. A B
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|19 JoséGarcíaCalvo|PhDThesis Figure 8. PCT process mechanism and fluorescent response (up) and energy levels of frontier orbitals diagram (down). c) Excimer formation: This case involves two or more side chains bounded to the recognition unit, usually two fluorophores, that interact by π-π stacking or other polarity involved forces. The fluorophores are joined by the recognition unit and the presence of the analyte modifies the interaction between them, altering the fluorescence. When this interaction between fluorophores gives a different wavelength of emission to the monomers it is called excimer, which allows ratiometric measurements. (Figure 9) When developing sensors with this kind of mechanism the fluorophores must be carefully chosen. Groups with high likeness to interact with themselves are usually the best to choose, groups such as perylene derivatives.10 10 K. V. Balakin, V. A. Korshun, I. I. Mikhalev, G. V. Maleev, A. D. Malakhov, I. A. Prokhorenko, A. Y. Berlin, Biosensors Bioelectron. 1998, 13, 771-778.
20|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis Figure 9. Excimers separation process mechanism when recognising an analyte and its fluorescent response. d) Förster Resonance Energy Transfer (FRET) In a similar way to excimers, there are two joined species involved, two different fluorophores in this case. One of the fluorophores acts as an acceptor, absorbing the light and transferring the energy to the other, which emits. The presence of a recognition unit between both fluorophores may alter the conformation and, as a consequence, the response. (Figure 10) Figure 10. FRET process mechanism and fluorescent response. e) Combination of several processes: It is not usually taken into account when explaining how to design a molecular probe, but most of them work by several of the previously explained mechanism at once. Many probes are a mixture between PET a PCT processes, in which the fluorescence is inhibited by electron transfer (PET) but when the analyte is recognized the conjugation of the system also changes (PCT) what varies the emission wavelength. This is the case of the probe JG76, explained in Chapter 3C.
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|21 JoséGarcíaCalvo|PhDThesis 2.5.2. The recognition process is based on a reaction: It is also possible to detect the presence of analytes by having species that modify their fluorescence after a chemical reaction with the analyte. This kind of method is not usual for quantifying and detecting analytes because of its characteristics. A reaction destroys the samples, and the media must be strictly controlled to assure reproducibility. It usually has greater time dependence and it is very uncommon not having by-products. In spite of the drawbacks, they present some advantages, such as greater changes in the signal, usually much more than in complexation processes. This characteristic leads to the possibility of lower limits of detection under ideal conditions. Undoubtedly, a particular case is based on the existence of catalytic reactions; it may lead to limits of detection far lower than any complexation reaction could reach. However, the quantification would not be so reliable, being capable of detecting traces of an analyte, but not being possible to determine exact quantities with certainty. 2.6. Characteristics of an ideal fluorescent probe11 In order to develop fluorescent probes that fulfil their purpose appropriately, they must reach certain standards, and possess suitable photophysical and analytical properties to the role they have: 2.6.1. Photophysical properties: High thermal and photochemical stability, to avoid easy and/or fast degradation. High molar absorptivity (ε): this value shows the absorption of the solution, if it is high, it means that it is easier to promote the electrons from the ground state, which could allow higher sensitivity and higher fluorescence. High final fluorescence quantum yield (ΦF), when the analyte is detected. It decreases the influence of dynamic quenching effects, and the measurements are more accurate and with less background noise. Excitation wavelength superior to 400 nm, far for UV - blue excitation wavelengths. That is because high energy irradiation is most likely to degrade some samples, especially when they are biological. Emission wavelength, for experimental measurements it is better when the emission is above 450 nm. The emission detectors are more sensitive in the region until 750 nm and, biological samples, have less interferents between 500 – 900 nm (many proteins emit in the UV-blue region). Large stokes shift between excitation – emission spectra. It would avoid reabsorption processes and simplify data treatment (overlapped signals). It becomes of upmost importance in cases such as FRET systems, to avoid HOMO-FRET between fluorophores. High fluorescent increase: so as to measure the fluorescence properly, the fluorescence must increase typically more than 100 % in intensity; at some wavelength, once the analyte is detected. In literature, it is very common to find publications in which the most remarked fact is how much increases the fluorescence, frequently underestimating the rest of the parameters. In reality, having 2 11 M. D. Heagy, Chemosensors 2011, 13, 253–273.
22|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis or 40-fold increase in fluorescence does not make it better or worse without taking into account the proper additional data. Figure 11 illustrates a representative example. Figure 11. Two examples of fluorescence increase. Example 1: 40-fold increase (from 0.05 %). Example 2: 5-fold increase (from 5 %). In the example 1, a blue fluorescent probe has an initial Fluorescence Quantum Yield (ΦF) very low (0.05 %) which increases to 2 %, meaning 40 times increase, although with a very low fluorescent emission. In consequence the experimental error in measurements is likely to be very high and the possibility of having interferents would be high too. On the other side, the example 2 shows an orange fluorescent probe with a 5-fold increase in fluorescence (ΦF from 5 to 25 %). Apparently, the increase is much lower, mainly because the initial fluorescence was already high. However, if the fluorescence quantum yield is taken into account, the fluorescence is much higher and the possible experimental errors in the measurements far lower, being likely to be a much better probe than the one from Example 1. Additionally, as it was previously explained, a yellow-orange fluorescent probe is more appealing than a blue one, in regard to possible applications in biological samples and sensitivity of the detection systems. 2.6.2. Analytical properties: Direct measurements: The measurements are more reliable if they occur after the interaction molecule-analyte without needing a third specie. Having more than two species in equilibrium hinders the interpretation and decreases the reliability of the results. High selectivity: The probe has to be as selective as possible. Although 100 % selectivity is not possible for any probe, especially when based on a complexation process, the best conditions must be studied and always taken into account. High sensitivity, reaching a low limit of detection. It is usually related with the selectivity to a certain extent, being less trustworthy when working close to the limit. For this reason, it is important that the detection limit was low, in order to work outside the limit conditions. ON fluorescence: The fluorescence is preferred to be ON, as it is more selective and trustworthy than OFF, because the fluorescence can be always inhibited by many species, especially when working with concentrated solutions, as it was previously explained (Section 2.4). Water solubility (for most probes): water soluble probes are necessary for bio-applications. For instance, measuring cations in water (from rivers or living organisms) or to introduce them into cells. Despite the difficulties of doing it, because of the high hydration spheres of many ions and molecules in water, they are very interesting.
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|23 JoséGarcíaCalvo|PhDThesis Recyclability: the probes that work by a complexation process are reusable in many occasions, by displacing the equilibrium. However, it is not usually possible, or much more difficult, when working with probes that react with the analyte. Taking into account all these characteristics, the probes have to fulfil as many of them as possible. The importance of each one would be different depending on the final application. 2.7. How to work with molecular sensors In the development of this thesis the processes followed to develop fluorogenic sensors may be summarized into a series of steps that were followed for every probe developed. In addition, although the study is about designing fluorescent molecular probes, many steps are the same for other type of sensors, especially colorimetric probes. 1) Synthesis of a fluorescent backbone. 2) Synthesis of the recognition part. 3) Joining the fluorescent moiety with the recognition part. 4) Testing and choosing an appropriate material as support (if required). 5) Characterization of the compounds, using techniques such as NMR, IR, melting point, Mass spectrometry analysis, elemental analysis, IR, SEM or EDX. 6) Studies of fluorescence: This part may be different depending on what is the purpose for the synthetized molecule or material. In general, for sensors that are useful in solution, the process starts with a solvatochromism; after that, some qualitative tests with different species, followed by the calculation of the work concentration, and next the existence of possible kinetic effects. Finally, the qualitative tests are performed, for instance calculation of the stoichiometry, and some other parameters, such as the limits of detection for different analytes, equilibrium constants, fluorescence quantum yields or fluorescence lifetimes decays are measured.
24|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis 3. PURPOSE AND PROCEDURE WHEN PERFORMING FLUORESCENT STUDIES When working with fluorescent sensors is recommendable to follow a series of rules and to use probes adapted to the purpose they have. Hence, during the development of this thesis, a protocol has been developed for fluorescent measurements.12 It was elaborated and followed in order to ensure that all measurements were performed to obtain all required information through a useful, trustworthy and efficient method. 3.1. Solvatochromism Whenever the probe is going to be used in solution, the first step is to check the solubility and behaviour in different solvents, which is fundamental to optimize the probe for the applications that are pursued. As a consequence, there is a double purpose: Finding out the best solvent: the method is useful for studying the solubility in different solvents, which gives an idea about the possibilities of the probe. That is of upmost importance when the probes are made to work in specific solvents, (such as water). Furthermore, it is interesting for studying possible mixtures of solvents, whenever the desired is not possible. For example, mixtures Ethanol:Water. Studying the behaviour in different solvents: once the probes are dissolved, measuring the spectra of absorption and fluorescence and taking pictures under visible and UV-light, the change in colour and fluorescence, with polarity or using protic or aprotic solvents, is shown. Along the development of the Thesis, a general procedure was followed: For each compound, a mother solution was prepared, the solvent had to be quite volatile and one in which the probe was highly soluble and stable; chloroform was a usual candidate when working with organic molecules. From this solution, an aliquot was evaporated in different vials, one for each solvent that was measured. Once evaporated, the probe was redissolved in each one of the chosen solvents, under sonication if the probe was not soluble by stirring, and there was no risk of degradation. The final concentration was selected between 5 – 100 µM, depending on the characteristics of the probe. Finally, the absorbance and the fluorescence spectra were measured. As an example, the study of the highly solvatochromic compound JG125 (deeply studied in Chapters 2 and 3A) is showed in Figures 12, 13 and 14. It was performed in different solvents (ordered by the Snyder polarity index): 12 Several pictures are taken from the results shown in the next chapters, as an example to illustrate what kind of information is obtained. This is only a previous example for a better understanding and hence, they are deeply explained in their corresponding chapters.
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|25 JoséGarcíaCalvo|PhDThesis 1. Water 2. MeOH 3. DMSO 4. DMF 5. MeCN 6. Acetone 7. AcOEt 8. THF 9. CHCl3 10. CH2Cl2 11. Toluene 12. Et2O 13. Hexane 14. Cyclohexane Figure 12. Solvatochromism of JG125 (10 μM solutions), pictures under visible (up) and UV light (down) in different solvents. 400 450 500 550 600 650 0.0 0.1 0.2 0.3 0.4 Absorbance Wavelength (nm) H2O MeOH DMSO DMF MeCN Acetone AcOEt THF CHCl3 DCM Toluene Et2O Hexane CH 400 450 500 550 600 650 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) MeOH DMSO DMF MeCN Acetone AcOEt THF CHCl3 DCM Toluene Et2O Hexane CH Figure 13. Solvatochromism of JG125 (10 μM solutions), absorbance in different solvents. The fluorescence was measured by excitation at 512 nm: 550 600 650 700 0 500 1000 1500 2000 2500 3000 3500 4000 Emission intensity (a.u.) Wavelength (nm) MeOH DMSO DMF MeCN Acetone AcOEt THF CHCl3 DCM Tol Et2O Hexane CH 550 600 650 700 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) DMSO Acetone AcOEt THF DCM Toluene Et2O Hexane CH Figure 14. Solvatochromism of JG125 (10 μM solutions), fluorescent response when λexc = 512 nm, in different solvents. 1234567891011121314 1234567891011121314
32|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 300 400 500 600 700 800 900 X Pb(II) ( F-Fo ) X Pb(II) Figure 19. Job´s plot of a 1:1 process, probe JG76 (5 µM) for Pb2+ detection in EtOH (Chapter 3C). Nonetheless, in the last years the validity of the Job’s plot method for calculating stoichiometry has been often questioned. In this regard, several papers have been published showing experimental results and simulations that demonstrate the limitations of this method, and how it is usually applied incorrectly.15 In the paper by Hibbert and Thordarson15 it is thoroughly explained why it is not possible to apply this method in every case. In particular, by referring to the work of Jurczak and coworkers,16 they simulated the results of Job’s Plot 1:2 and 2:1 with different relations between (molar fraction):(equilibrium constants). Surprisingly, the Job’s Plot analysis gave information about the relation between constants, but the condition in which it is related to stoichiometry is limited, and impossible to predict without knowing the equilibrium constant previously to the titration. Additionally, some authors point out that the results may still have some reliability when the relation between the equilibrium constant and the concentration is: 𝐾𝑒𝑞1/𝐻𝑜𝑠𝑡 In practice, where either both K1 and K2 (hypothetical 1:2 equilibrium) are large or one of them is relatively large compared to the other, Job-plots appear to be valid. But this would imply knowing the stoichiometry before calculating it, which makes the method quite unfit for such calculations. Considering the results, the authors conclude that the best way to proceed is to calculate the stoichiometry by fitting the titrations to different models 1:1, 1:2, 2:1… The one with the “best-fitting regression model” is likely to be the real stoichiometry. In order to clarify concepts, the results are considered as the “best-fitting regression model” when the error is low, there is no tendency in the residues (the scatter of the residual plot) and statistic tests, that validate the relation data-fitting (such as the F-test for the sum-of squares regression), are performed with positive results. In conclusion, the best method to determine the stoichiometry proceed by adjusting the equilibrium constant to several possible models. The most likely correct results are related to the best fitting, being the Job’s Plot a reassuring calculation but never determining. 15 D. B. Hibbert, P. Thordarson, Chem. Commun. 2016, 52, 12792-12805. 16 F. Ulatowski, K. Dabrowa, T. Balakier and J. Jurczak, J. Org. Chem. 2016, 81, 1746–1756.
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|33 JoséGarcíaCalvo|PhDThesis 3.7. Thermodynamic equilibrium constant calculation (K) The equilibrium constants between analyte-probe in solution may be measured under different conditions. Equilibrium constants not only depend on the solvent, but also on the ionic strength and temperature. In order to ensure reliability in their determination, the ionic strength and temperature must be constant during the titration and the concentrations must be as low as possible (avoiding dimers or other interactions). As it was previously explained, the applied methods for titrations work at high dilution conditions by increasing the concentration of the analyte. In addition, the results may be compared by using different variations such as changing the concentration of the probe, the analyte additions or even doing the titration of the analyte with probe. In any case, if the probe is not likely to have side processes, such as aggregation, the results are usually comparable and with high repeatability and robustness. Moreover, the calculation of equilibrium constants when having slow association and/or dissociation constants must be done by waiting the necessary amount of time until the species reach equilibrium, taking into account that the necessary amount of time is usually dependent on the concentration of the species (first order or more). There are several methods for performing experimental calculations, what is more, there is plenty of software available17 to make this kind of calculation fast and simple. The software is continuously updated and with plenty of explanations about how to adapt it to the most common applications, such as fluorescence, absorbance, circular dichroism or NMR titrations. These methods work by iterative fitting, however, there are some alternatives. Nowadays, the use of outdated calculation methods that rely in approximations is still very spread. The most common is based on Benesi-Hildebrand equation.18 Nevertheless, their use makes no sense nowadays, because of the easy access to more accurate methods without needing for relying on approximations; surprisingly, BenesiHildebrand and similar equations are frequently found in recent publications in the topic of sensors.19 The next part will consist on the theory behind the constant calculation, briefly explained from the point of view of a fluorescent probe, starting with 1:1 complexes. 3.7.1. Equilibrium constant of the complex ML (1:1) The complexation reaction can be stated by the following equilibrium: Where P represents the probe and A the analyte By taking into account the mass balance and the fluorescence: 𝐶𝑃𝑃𝐴 (1) 17 An example of webpage adapted to calculate equilibrium constants by iterative methods: http://supramolecular.org, accessed 23th June 2018. It might be also calculated by software, such as the used for calculations: Origin v2016. The equations were adapted for the software with the help of professor Saturnino Ibeas Cortés from University of Burgos. 18 B. H. Hildebrand. J. Am. Chem. Soc. 1949, 71, 2703–07. 19 Two examples, A) A. Ghosh, S. Das, S. Kundu, P. K. Maiti, P. Sahoo, Sensors and Actuators B: Chem. 2018, 266, 80-85. B) S. Fernández-Alonso, T. Corrales, J. L. Pablos, F. Catalina, Sensors and Actuators B: Chem. 2018, 270, 256-262.
34|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis 𝐶𝐴𝑃𝐴 (2) 𝐼𝑓𝑃𝑓𝑃𝐴 (3) Where CA, CP, IF, fP and fPA are the total concentrations of the probe (CP) and analyte (CA), the intensity of fluorescence and the proportional fluorescence factors of the probe and the complex PA, respectively. The rest of the parameters are the concentrations of the species in the equilibrium. By calculating the concentration of the probe in the equilibrium on the equation (1) and substituing on the equation (3), the following equation is obtained: 𝐼𝑓𝐶𝑓𝑓𝑃𝐴 (4) The definition of the equilibrium constant states that: 𝐾 (5) By calculating [PA] on the previous equation (5), it may be replaced on equation (4), obtaining equation (6): 𝐼𝑓𝐶 𝐶𝐶 𝐶𝐶 4𝐶𝐶 (6) This equation is used by iterative fitting in function of CA-IF (experimental data), solving a nonlinear least square regression, starting by giving initial values of K1, fP and fPA. The results obtained are similar to the example from Figure 20 (JG76, Chapter 3C). 0246810 0 500 1000 1500 2000 Emission intensity 571 nm (a.u.) C s, ( M) Figure 20. Fitted fluorescent emission of a titration with K(CF3SO3) of a 2 µM solution of JG76 in EtOH. In practice, the fitting calculation of the complexation constants is repeated several times and the value of K is estimated from it.
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|35 JoséGarcíaCalvo|PhDThesis 3.7.2. Complexes M2L (2:1) Similarly, the equations for the equilibria 1:2 (analyte:probe) and 2:1 might be calculated, an extensive explanation might be found in literature.20 This leads to obtain the expressions: 𝐴 (Equation [E]) Where β represents the cumulative or overall constant. This β is the constant for the formation of a complex from reagents and can be expressed as the product of each constant, which considers the formation of the complex step by step. For instance, the cumulative constant for the formation of M2L is given by β = K1K2, b is the molar extinction coefficient of the intermediate complex ML. 3.7.3. Complexes ML2 (1:2) 𝐴 𝑏𝐾𝑀 (Equation [F]) where 𝑁1𝐾𝑀1𝐾𝑀8𝐶𝐾𝐾𝑀 3.7.4. Indirect calculation for thermodynamic equilibrium constant: During the development of the thesis it was reached a point in which it was interesting to calculate the equilibrium constant of a complex which was not fluorescent. To do so, the same equation may be applied without major changes in, for example, NMR titrations or circular dichroism. However, the main drawback of these methods is the high work concentration, which usually makes the results not valid when diluted, and that they are not always applicable. In this regard it has been developed a new way to calculate thermodynamic equilibrium constants, due to an indirect calculation with a fluorogenic probe throughout a displacement in the complexation equilibrium. 20 K. A. Connors, Binding Constants: The Measurement of Molecular Complex Stability, J. Wiley & Sons, New York, 1987. b) Doctoral Thesis performed by Dr. Daniel Moreno. Supervised by Prof. Tomás Torroba. Nuevas sondas cromo-fluorogénicas a partir de derivados de organopaladio y compuestos indénicos, Universidad de Burgos (Spain), February 2011. M + L M L 1 [] [][] M L K M L M L + M M 2 L 2 2 [] [][ ] M L K M ML M + L M L 1 [] [][] M L K M L M L + L M L 2 2 2 [] [][ ] M L KLML
36|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis In the work performed for this Thesis, this method has been successfully applied along Chapter 3C. The calculation was performed for compounds that acted as ligands of potassium cation, a synthetic fluorescent crown ether (JG76 = P) and some non-fluorescent natural depsipeptides, such as valinomycin and cereulide (X). Then, the thermodynamic equilibrium constant JG76-K+ (P-A) was studied and calculated. With the data it was proposed the possibility of having an equilibrium to calculate the constant X-K+ (X-A). Now it was necessary to include a new equilibrium: Mass and fluorescence balance: 𝐶𝑃𝑃𝐴 (7) 𝐶𝐴𝑋𝐴𝑃𝐴 (8) 𝐶𝑋𝑋𝐴 (9) 𝐼𝑓𝑃𝑓𝑃𝐴 (10) Where CP, CA, CX, IF, fP y fPA are, concentrations of probe, analyte and not fluorescent ligand, the fluorescence intensity (only depends on the probe and the complex probe-potassium) and the factors of proportion between probe and PA complex. The rest are the concentration of the species on the equilibrium. By taking into account K1 and K2 and the equations (7) and (9): 𝐾 (11) 𝐾 (12) a) First option, solving the equation: By solving (11), the concentration [VK], in (12): 𝑃𝐴𝐾𝐾𝐾𝑃𝐴1𝐾𝐶2𝐶𝐾1𝐾𝐶𝐶𝐶𝐾 𝑃𝐴𝐾𝐶1𝐾2𝐶𝐶𝐾𝐶𝐶𝐾𝐶𝐶0 (13) This cubic function could be solved by the next procedure: Being the equation (14): 𝑎𝑥𝑏𝑥𝑐𝑥𝑑0 (14) Dividing between “a” and replacing x = z-b/3a, z3+pz+q = 0 (Tschirnhaus transformation); where p and q are calculated as:
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|37 JoséGarcíaCalvo|PhDThesis 𝑝 (15) 𝑞 (16) To know the number of real roots, the discriminant Δ is calculated as: ∆4𝑝27𝑞 (17) With these data, further calculations for the constant were done, but at the end: The equation turned out to be too much complicated, with many parameters to adjust. The results of adjusting data to the equation were more dependent on the initial values than on the variation of them. It was decided that, in order to make a more accurate calculation of the constant, it was necessary to make some approximations and simplify the equation. b) Simplification of the equation: The experiment started with a solution of not fluorescent compound (X) and the analyte (A), subsequently at the start of the titration there was complex (XA) in the equilibrium. When the probe (P) was added it formed a complex with the free analyte (A), creating the new complex (PA) and replacing the previous complex (XA). In conclusion, the concentration of XA decreased, whereas the concentration of X increased. Afterwards, a possible approximation could be done: CX - [XA] CX And this simplification was more realistic when the initial proportion X/A was as high as possible. The new equation obtained, from (11) and (12) was: 𝑃𝐴 (18) As it is explained in Chapter 3, this method was developed specifically for a potassium probe that increases its fluorescence in presence of potassium cations, although it can be applied to other equilibria. The validity of the method was evaluated: First, the test was repeated several times with different initial proportions of X:A, a proportion 1:1, 1:0.25 and 1:0.1. It was checked that the results were slightly different, but it had the best fitting when the proportions were 1:0.1, in which the approximation was more valid, because of the simplification. Second, the method was also compared with values calculated by other methods from literature. In our case, there were plenty of studies with circular dichroism. The results were comparable and even more reliable, due to some issues that were found in the literature calculations.
38|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis The equilibrium constant with the fluorescent molecule (K1) and with the nonfluorescent one (K2) should of the same order. Methods for calculating equilibrium constants have usually errors and deviations of at least 10 %. If the previously calculated K1 was much bigger than K2 the error would be higher than the value of the constant. In addition, if it was the other way around (K2>>K1) the approximation would not be valid. As an example, the graphs in Figure 21 were obtained by changing between cereulide/valinomycin and EtOH (see Chapter 3C for further information) getting very similar results to the already studied circular dichroism:21 04812 0 250 500 750 I F ( a.u. ) [Valinomycin], ( M) 048 0 400 800 1200 1600 I F ( a.u. ) [Cereulide] ( M) Figure 21. Fitted fluorescent titration emissions with JG76 of a 2 µM solution of K(CF3SO3) and 20 µM of valinomycin and cereulide solution in EtOH (A and B). 3.8. Limits of detection (LODs) The IUPAC defines the detection limit as the smallest quantity of an analyte that can be detected with reasonable certainty for a given analytical procedure, being distinguishable from the blank. The method used during this thesis is not widely spread among scientist working in the field of sensors. Therefore, the reasons for using it are explained by giving an example of one limit of detection calculated by this method, and compared with the results of some of the most common methods found in literature. The explanation about how the expressions used to calculate the LODs work are deeply elaborated and discussed in literature, being beyond the aim of the thesis. Figure 22 represents an example for the titration of probe JG25 with Hg2+(from Chapter 1); the measurements at low concentrations of Hg2+ were fitted to a linear regression in order to calculate the LODs. 21 M. C. Rose and R. W. Henkens, Biochim. Biophys. Acta 1974, 372, 426−435. [A] [B]
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|39 JoséGarcíaCalvo|PhDThesis 0.00.20.40.60.81.01.21.4 10000 20000 30000 40000 50000 60000 Integral 379-650 (a.u.) Hg2+ Concentration (mM) 0 2 4 6 8 10 12 14 16 18 20 16800 16900 17000 17100 17200 17300 17400 17500 17600 Integral 379-650 (a.u.) Hg 2+ Concentration (M) Figure 22. Titration of JG25 with adding Hg2+. Global titration (left) and linear regression for calculation of the LOD (right). In addition, the blank (sample only containing probe) was measured three times obtaining an intensity value of 16876 a.u. and a standard deviation of 15.7 a.u. A) Method 1: Calculation of the LOD based on measurements of the blank. The method is based in what IUPAC defines as the equation so as to measure LODs: 𝐿𝑂𝐷𝑥𝑘𝑠 Where xb is the media of the measurements of the blank (only containing probe), k is a constant dependent on the reliability given to the method (usually 3) and sb is the standard deviation of a region with low concentration of the analyte. 𝐿𝑂𝐷16876315.716923 This equation gives the LOD in signal units. This signal is transformed into concentration by adjusting it to the fitted linear regression with the analyte. y = 38.8x + 16895 x = LOD = 0.7 µM B) Method 2: By using a linear regression at low concentrations of the analyte. In most papers from literature,22 the limit of detection (LOD) was estimated by the following equation: LOD 3.3 SD/s Where SD is the standard deviation of a blank figure sample and s is the slope of the calibration curve in a region of low analyte content. Depending on the specific paper, they also use SD as the standard deviation of the linear regression. A) Using SD as the standard deviation of the blank: 22 S. Ahuja, M. Dong, Handbook of Pharmaceutical Analysis by HPLC, Elsevier, New York, 2005. y=38.8x+16895 R²=0.995
40|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis 𝐋𝐎𝐃 3.3 . .𝟏.𝟑𝟑 μM B) Using SD as the standard deviation of the linear regression: 𝐋𝐎𝐃 3.3 . . 𝟏.𝟒𝟐 μM Furthermore, some authors also distinguish between limit of detection and limit of quantification, substituting the factor 3.3 by 10, to give more reliability to the calculation. C) Method 3: By linear regression + False positive and negative. Based on the results of several authors23 there are many factors that must be taken into account when calculating the limit of detection: • Adjusting to a mean square linear regression. • Removing the “outliers”. Understanding them as points that are significantly different from the rest, within a 95 % of probability. If it is not possible to fulfil this part, the measurements should be repeated. • Adjusting to a linear regression and checking that the slope is significantly different from 0 (95%). P-Value >0.05. • Calculation of the LOD when the probability of false positive (α) and false negative (β) is equal or inferior to 5 %; or the value that the author is looking for. To do so, some software such as, “R” could provide the fitting and give the results. This method allowed to obtain a LOD = 6.6 µM. From the comparison of the methods a discrepancy may be seen between the most commonly used in literature and the method used in this thesis. 1) The obtained values of the LODs from most of the literature methods are, at least, between 5 to 10 times lower than the ones obtained by Method 3. The main issue is the dependence on the blank. Between the many issues with literature methods, in the first place, the deviation of the blank may not be the same when adding the analyte. In addition, it should be possible to get far lower values when measuring it repeatedly. Sometimes it should be theoretically possible to get values 100 or even 1000 lower than the values calculated by other methods. 2) Many of the methods from literature have not measured the quantity they put as a limit of detection. If they are not capable of measuring it in laboratory conditions, then it shouldn’t be given as the LOD. As a general rule, the LOD is higher or around the value of the secondthird point given when presenting a reliable linear regression used for the calculation, and always higher than the first value that is considered different from 0. 3) Papers usually do not give details of many of the parameters they calculate. For example, about how many times they have measured the blank, or they do not specify the way in which they have calculated the LOD. 23 M. C. Ortiz, L. A. Sarabia, M. S. Sánchez, Anal. Chim. Acta. 2010, 674, 123-142.
CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES|41 JoséGarcíaCalvo|PhDThesis In addition, there are more methods apart from the ones exposed, and some of them may be trustworthy if correctly explained. In order to compare results between probes there are many factors that should be taken into account, not only the numbers, but also the method used to get the values and the data that is provided. Deeper information about this topic has been studied by many analytical chemists. In fact, it has been recently published a book in which many conclusions fit to the findings explained in this thesis.24 The publication is focused in the analytical procedures to perform limits of detection correctly. Of course, it is a much deeper analysis that delves into mathematical explanations to face this issue. However, although it might be very complex for many chemists with no experience in the area, some of the conclusions about the mistakes when determining LODs are easily understandable and may be summarized as:25 Using the statistical parameters wrongly. In this part are included practices such as: o Defining the LODs as signal noise ratio = 3. o Ignoring possible errors in estimating the value of the blank. o Using the IUPAC as guidance, being barely updated since 1970s. o In the formula from the IUPAC, defining “k” as 3 without particular justification. o Accepting previous results from literature without checking. Calculating LODs out of the Content Domain. Inferred estimations of the LOD would not be valid; only numerical values recorded and reported might be taken as true. Meaning that the LODs must be values within the region measured, always superior to the lowest measurement that is different from the blank (with no analyte). In general, the methods used in literature to perform LODs are very controversial, especially for chromo-fluorogenic probes. Introducing more reliable methods is the first step for the development of standardized procedures for the future, but it must start by rejecting outdated old techniques, that give unreliable information, and giving more data about the procedures followed for the calculations. A particular case of LOD calculation, when the thermodynamic equilibrium is not reached: Another topic that could be addressed when calculating LODs is what to do when the titration and calculations are performed for species that had not reached the equilibrium (usually because it would take minutes or hours). In most papers it is an issue that is ignored, although it is a problem easily solved if the data are properly explained for the conditions in which the limit is valid. This kind of titrations (equilibrium reached after too long periods of time) may be addressed in several ways. First, by waiting the necessary amount of time; the different points of the titrations are prepared at the same time and after waiting the necessary amount of time, it may give reliable results. However, this could be not possible when the quantity of probe is limited or the kinetic is unknown to a certain extent (which could be a consequence of working with supported probes, Chapter 1). Furthermore, LODs that are very low but need several hours/days to give signal could not be useful. In spite of the difficulties, there are some alternatives. For example, performing the titrations similarly to when there is no time dependence but with constant time lapse between additions. 24 E. Voigtman, Limits of detection in chemical analysis, John Wiley & Sons, Inc: Hoboken, New Jersey, 1st Edition, Chapter 1, 2017. 25 E. Voigtman, Limits of detection in chemical analysis, John Wiley & Sons, Inc; Inc: Hoboken, New Jersey, 1st Edition, Chapter 24, 2017.
48|CHAPTER0.INTRODUCTIONTOFLUORESCENTPROBES JoséGarcíaCalvo|PhDThesis 5. RESUMEN DEL CAPÍTULO El propósito de este capítulo es introducir el uso de sensores fluorescentes como método para la detección de analitos de interés. Para alcanzar este objetivo, se muestra que este tipo de sensores presentan ventajas en factores importantes como el tiempo o el coste, respecto a otros métodos clásicos analíticos. También se hace hincapié en el mecanismo de fluorescencia, las propiedades, los tipos, las distintas formas de crear sensores moleculares fluorescentes y sus limitaciones. Complementando todo lo anterior, se describen los procesos para el monitoreo de las distintas propiedades y parámetros que definen a las sondas fluorescentes. Al mismo tiempo, se explica el modo en que se han evaluado estas propiedades y qué clase de experimentos en el laboratorio y en el tratamiento de datos son necesarios para ello. Algunos de estos experimentos son el estudio del solvatocromismo para evaluar el comportamiento en distintos disolventes, métodos para el cálculo de la estequiometría de un complejo y cálculo de distintas constantes y parámetros propios (constantes de equilibrio, límites de detección, rendimientos cuánticos de fluorescencia o tiempos de vida). En resumen, este capítulo sirve de introducción para los Capítulos 1, 2 y 3 en los cuales se diseñaron sondas moleculares y materiales fluorescentes modificados para la detección de especies de interés, así como su utilización como marcadores biológicos, para terapia fotodinámica o en otras posibles aplicaciones.
CHAPTER 1 FLUORESCENT PROBES FOR THE DETECTION OF Hg(II) DERIVATIVES OBJECTIVES This chapter aims to explain the work performed in the search of new fluorogenic sensors for the detection of mercury cationic species, with high environmental impact because of their toxicity. Specifically, Hg(II) and MeHg(II) fluorescent sensors were synthetized; based on indanone derivatives, working in water solution (soluble or supported sensor), in cellular media1 and in fish extracts.2 1 B. Díaz de Greñu, J. García-Calvo, J. Cuevas, G. García-Herbosa, B. García, N. Busto, S. Ibeas, T. Torroba, B. Torroba, A. Herrera, S. Pons, Chem. Sci., 2015, 6, 3757-3764. 2 J. García-Calvo, S. Vallejos, F. C. García, J. Rojo, J. M. García, T. Torroba, Chem. Commun., 2016, 52, 1191511918.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|51 JoséGarcíaCalvo|PhDThesis 1. INTRODUCTION. THE IMPORTANCE OF MERCURY 1.1. Mercury, history and applications Mercury, as many other heavy elements of the periodic table, is often associated to danger and toxicity. Although it is usually true, mercury also has many applications and has proven to be very useful when correctly handled. Mercury derivatives, in the history of humankind, have been used in paintings, jars or even with medicinal purposes, generally leading to little results in terms of health. Old applications still remain in use, such as the chlor-alkali industry, the use in batteries or the mercury lamps although they are now banned or disappearing.3 Because of the many applications of the metal derivatives and the toxicity for living organisms of mercury compounds, it is of great importance to know where the different mercury species come from and detecting and quantifying their presence, which is critical for toxicological and environmental issues. The origins of mercury species in the environment are mainly two, volcanic and human action.4 Whereas volcanic processes occur without human interference and have not had significant increase, the concentration of mercury species in the sea has tripled compared to pre-anthropogenic interference,5 which gives an idea of the impact of human action. Mercury salts and sulfurs are the main source when mining this element for different purposes, however, the disposals have not always been correctly treated, specially until the first news about their dangerousness came to light in the 20th century. Furthermore, over the last years, the realization of the consequences of dealing with high levels of mercury, and the development of new ways to detect and stop irresponsible disposals, have been very important for avoiding massive introduction of this kind of waste in the environment. Although it is also true that sometimes the information is confusing and it is not clear how dangerous it could be, generating unnecessary social alarm.6 1.2. The cycle of mercury There are two very important organometallic derivatives of mercury, methylmercury (MeHg(II)) and dimethylmercury (Me2Hg). Dimethylmercury is a liquid and very volatile species that possesses the highest toxicity for animals of all mercury derivatives, a single drop of this organometallic compound can go through most of the clothes and laboratory protection and it is immediatelly absorbed through the skin, causing neuronal damage and being mortal within a few days after skin absorption.7 For this reason, and being the less common of all the species, it is not studied for detection purposes. 3 a) M. S. Bank, Mercury in the Environment: Pattern and Process. 1st ed., University of California Press, 2012. b) G. Liu, Y. Cai, N. O’Driscoll, X. Feng, G. Jiang. Environmental Chemistry and Toxicology of Mercury, G. Liu, Y. Cai, N. O'Driscoll, (Editors), John Wiley & Sons, Hoboken, New Jersey, Chapter 1, pp. 1-12, 2012. 4 P. A. Ariya, M. Amyot, A. Dastoor, D. Deeds, A. Feinberg, G. Kos, A. Poulain, A. Ryjkov, K. Semeniuk, M. Subir, K. Toyota, Chem. Rev. 2015, 115, 3760. 5 C. H. Lamborg, C. R. Hammerschmidt, K. L. Bowman, G. J. Swarr, K. M. Munson, D. C. Ohnemus, P. J. Lam, L.-E. Heimbürger, M. J. A. Rijkenberg, M. A. Saito, Nature 2014, 512, 65. 6 J. Burger, M. Gochfeld, Rev. Environ. Health. 2013, 28, 129-143. 7 D. W. Nierenberg, R. E. Nordgren, M. B. Chang, R. W. Siegler, M. B. Blayney, F. Hochberg, T. Y. Toribara, E. Cernichiari, T. Clarkson, N. Eng. J. Med. 1998, 338, 1672–1676.
52|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis From the other different mercury species, apart from metallic mercury, which also emits toxic vapours, the cationic Hg(II) and MeHg(II) are of upmost importance. The cation Hg(II) has been widely studied and controlled. Nevertheless, Hg(II) is an inorganic salt, being barely absorbed by living beings. As a consequence, its toxicity is far lower than the organometallic cations, from what the most representative is MeHg(II). The existence of MeHg(II) comes from the mercury cycle, which explains how this lypophilic cation is introduced in the trophic chain. Mercury cycle consists of an abiotic sulfuration-desulfuration equilibrium in bacteria (Figure 1) that transforms Hg(II) in water to MeHg(II); after that, some plants absorb this compound. The lipophilicity of MeHg(II) makes it easy to be absorbed by living beings. In addition, since this complex of MeHg(II) and cysteine resembles the structure of large and neutral amino acid methionine, it can enter the cell and exit as a complex with reduced glutathione, thus forming water soluble complexes in tissues.8 The special characteristics of this cation makes it bioaccumulative, which means that going up in the trophic chain, the concentration is bigger each time, reaching dangerous concentrations in some big aquatic animals, such as swordfish or shark. Figure 1. Cycle of Hg(II) - MeHg(II).9 1.3. Mercury presence and its risks in our society There are records of several disasters related with massive contamination with mercury cationic species, the two examples with more impact were, the leaking of MeHg(II) in Minamata’s river in the 1950s and the contamination of seed grain in Iraq in 1971, caused by its use as pesticide. As a result of these disasters, the consequences of introducing Hg(II) and MeHg(II) to the environment are widely 8 Reviews: (a) M. Farina, J. B. T. Rocha, M. Aschner, Life Sci. 2011, 89, 555–563; b) M. Yamashita, Y. Yamashita, T. Suzuki, Y. Kani, N. Mizusawa, S. Imamura, K. Takemoto, T. Hara, M. A. Hossain, T. Yabu, Touhata, Mar. Biotechnol. 2013, 15, 559–570; c) G. J. Lu, Y. Tian, N. Vora, F. M. Marassi, S. J. Opella; J. Am. Chem. Soc. 2013, 135, 9299−9302. 9 H. Hintelmann. Metal. Ions Life Sci. 2010, 7, 365-401.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|53 JoséGarcíaCalvo|PhDThesis documented and studied.10 In spite of these facts, the most important and constant source of mercury cationic species in food, in our society, comes from fish.11 The explanation for it comes from the already explained cycle of mercury, so that some fish are capable to accumulate concentrations close, or even superior to 5 ppm of this element. The concentration of the different mercury species in living beings consists of the highly toxic derivative, MeHg(II), (70-90 %) and only a 10-30 % of Hg(II). Food with mercury concentrations in the range of ppm is toxic if regularly introduced on human diet, even if it is eaten once a week.12 As a consequence, many American and European international institutions do not recommend to eat fish with more than 1 ppm of this metal. The results of a long period of ingestion go from long-lasting neurological damage to physical and psychological deficiencies;13 being especially toxic for children and the foetus14 in pregnant women. For all these reasons, nowadays detection of mercury cationic derivatives has drawn a lot of attention for health, environmental and industrial issues. In this regard, many examples of detection may be found in literature, especially for Hg(II), in spite of the greater importance of the other species. That is the reason because the efforts in this area of research aim to develop new selective, highly accurate and fast methods to detect and discriminate different mercury species. 1.4. Standardized techniques for detecting mercury presence in a sample There are many ways to quantify Hg(II) derivatives present in solutions, or extracted from a sample to a solution, the most common are analytical techniques based on the use of classical analytical devices: Inductive coupled plasma mass spectrometry (ICP-MS): it is the most common and reliable. This method is applied after the degradation of the sample in concentrated acid 10 See for example: a) T. A. Douglas, L. L. Loseto, R. W. Macdonald, P. Outridge, A. Dommergue, A. Poulain, M. Amyot, T. Barkay, T. Berg, J. Chetelat, P. Constant, M. Evans, C. Ferrari, N. Gantner, M. S. Johnson, J. Kirk, N. Kroer, C. Larose, D. Lean, T. G. Nielsen, L. Poissant, S. Rognerud, H. Skov, S. Sørensen, F. Wang, S. Wilson, C. M. Zdanowicz, Environ. Chem. 2012, 9, 321–355; b) C. R. Hammerschmidt, M. B. Finiguerra, R. L. Weller, W. F. Fitzgerald, Environ. Sci. Technol. 2013, 47, 3671−3677; c) A. L. Soerensen, R. P. Mason, P. H. Balcom, E. M. Sunderland, Environ. Sci. Technol. 2013, 47, 7757−7765; d) D. Liu, S. Wang, M. Swierczewska, X. Huang, A. A. Bhirde, J. Sun, Z. Wang, M. Yang, X. Jiang, X. Chen, ACS Nano 2012, 6, 10999–11008; e) J. Chen, S. Zhou, J. Wen, J. Anal. Chem. 2014, 86, 3108−3114. 11 a) R. Wang, X. B. Feng, W. X. Wang, Environ. Sci. Technol. 2013, 47, 7949−7957; b) W. F. Fitzgerald, C. H. Lamborg, C. R. Hammerschmidt, Chem. Rev. 2007, 107, 641−662, c) H. Hintelmann, Organomercurials. Their Formation and Pathways in the Environment, in A. Sigel, H. Sigel, R. K. O. Sigel: Organometallics in Environment and Toxicology: Metal Ions in Life Sciences, Chapter 11, vol. 7, pp. 365-401; 2010; d) I. Lehnherr, V. L. St. Louis, H. Hintelmann, J. L. Kirk, Nature Geosci. 2011, 4, 298-302; e) J. M. Parks, A. Johs, M. Podar, R. Bridou, R. A. Hurt, S. D. Smith, S. J. Tomanicek, Y. Qian, S. D. Brown, C. C. Brandt, A. V. Palumbo, J. C. Smith, J. D.Wall, D. A. Elias, L. Liang, Science 2013, 339, 1332-1335. 12 https://www.fda.gov/food/foodborneillnesscontaminants/metals/ucm2006760.htm; accessed on June 2018. 13 a) J. Z. Byczkowski, Methyl Mercury Toxicity: Pharmacokinetics and Toxicodynamic Aspects, in R. R. Watson, V. R. Preedy, Reviews in Food and Nutrition Toxicity, CRC Press, Taylor & Francis Group, Boca Raton, Florida, Chapter 2, Vol. 3, 2005; b) M. Aschner, N. Onishchenko, S. Ceccatelli, Toxicology of Alkylmercury Compounds, in: A. Sigel, H. Sigel, R. K. O. Sigel, Organometallics in Environment and Toxicology: Metal Ions in Life Sciences, Chapter 12, vol. 7, pp. 403-434, 2010. 14 Reviews: a) P. Aggarwal, S. Gaur, P. Gauba, Environ. Dev. Sustain. 2014, 16, 71–78; b) S. Díez, Rev. Environ. Contam. Toxicol. 2009, 198, 111-132; c) J. E. Sonke, L.-E. Heimbürger, A. Dommergue, C. R. Geoscience 2013, 345, 213–224.
54|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis solution (usually nitric acid). A plasma source ionizes the sample and allows the quantification of elemental ions, the results show a relation molecular weight – signal intensity. Flame atomic absorption spectroscopy (FAAS): it is performed by the sample ionization, measuring the absorption of the sample when it is transformed to gaseous state by a flame. It is capable of measuring elemental composition. Both analytical techniques are the most reliable way to detect the presence of an element in a sample and quantify it. Nevertheless, they also have important drawbacks such as the needing for a specialist to perform the measurements, the high price, long time, necessity of a pre-treatment of the samples (generally with strong acids) or the impossibility to distinguish between the different derivatives, which might be of extreme importance in the case of mercury. In contrast, the use of fluorescent and colorimetric methods for the quantification have increased its popularity because of their simplicity, lower prices, faster results and the potential for discriminating between derivatives. 1.5. Fluorogenic and chromogenic probes for Hg(II) detection in literature15 Mercury cations are considered as big, soft and metallic because of the atomic weight of mercury and the position in the periodic table. This makes them a good target for detection, since there are many possibilities to design probes with high affinity. There is a huge number of publications about detecting mercury cations by using colorimetric and/or fluorometric probes. So then, the next part serves as an example of some common characteristics from the most cited probes in literature. Probes containing sulfur atoms.16 Most probes for mercury cations contain several sulfur atoms because of the high affinity between mercury-sulfur. Probes based on macrocycles. There are many examples in literature (see Figure 2) and they work by many different mechanisms. There are examples of ON-OFF17 or OFF-ON systems,18 working usually by PET or PCT mechanisms. In addition, using macrocycles with O/S atoms increases the solubility in water and they are easy to combine with fluorescent molecules. 15 Chromogenic or fluorogenic probes for Hg(II) and other cations: Reviews: a) X. Li, X. Gao, W. Shi, H. Ma, Chem. Rev. 2014, 114, 590−659; b) D. Sareen, P. Kaur, K. Singh, Coord. Chem. Rev. 2014, 265, 125–154; c) M. Formica, V. Fusi, L. Giorgi, M. Micheloni, Coord. Chem. Rev. 2012, 256, 170–192; Specific fluorogenic probes for Hg(II): Reviews: d) M. J. Culzoni, A. Muñoz de la Peña, A. Machuca, H. C. Goicoechea, R. Babiano, Anal. Methods. 2013, 5, 30–49. e) M. Tian, L. Liu, Y. Li, R. Hu, T. Liu, H. Liu, S. Wang, Y. Li, Chem. Commun. 2014, 50, 2055-2057; Colorimetric nanoprobes for Hg(II): f) I. Ratera, A. Tárraga, P. Molina, J. Veciana, in T. Torres, G. Bottari, Editors: Organic Nanomaterials: Synthesis, Characterization, and Device Applications, John Wiley & Sons, Hoboken, New Jersey, Chapter 24, pp. 529-548, 2013. Recent examples: g) Q. Wei, R. Nagi, K. Sadeghi, S. Feng, E. Yan, S. J. Ki, R. Caire, D. Tseng, A. Ozcan, ACS Nano 2014, 8, 1121–1129; h) Q. Yue, T. Shen, J. Wang, L. Wang, S. Xu, H. Li, J. Liu, Chem. Commun. 2013, 49, 1750-1752 (and references therein); Lifetime fluorescence: i) D. Huang, C. Niu, M. Ruan, X. Wang, G. Zeng, C. Deng, Environ. Sci. Technol. 2013, 47, 4392−4398. 16 D. Riccardi, H. B. Guo, J. M. Parks, B. Gu, A. O. Summers, S. M. Miller, L. Liang, J. C. Smith, J. Phys. Chem. Lett. 2013, 4, 2317−2322. 17 A. B. Descalzo, R. Martínez-Máñez, R. Radeglia, K. Rurack, J. Soto, J. Am. Chem. Soc. 2003, 125, 34183419. 18 S. Yoon, A. E. Albers, A. P. Wong, C. J. Chang, J. Am. Chem. Soc. 2005, 127, 16030-16031.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|55 JoséGarcíaCalvo|PhDThesis O OS S ON N O Figure 2. Examples of molecular probes with macrocycles containing sulfur atoms for the detection of mercury cations described in publications from Martínez-Máñez17 and Chang,18 respectively. Probes based on recognition by a specific PET – excimer system.19 Some groups have designed molecules with special affinity for mercury cations because of supramolecular recognition and specific changes in the structure, affecting fluorescence. For instance, Guo and coworkers developed a probe (Figure 3) as an V shape coordinating molecule capable of changing its fluorescence with high selectivity in water buffered solution. Figure 3. Structure of the molecule synthesized by Guo and co-workers for the identification of Hg(II) in solution. Probes that work by a chemical reaction,20 in which, for instance, the mercury cations may work by producing a redox reaction and a complexation. The molecules of Figure 4, which contain an azadiene group, are capable of complexing selectively Hg(II) over other cations. The fluorescence of the product increases by adding Hg(II). 19 X. Guo, X. Qian, L. Jia, J. Am. Chem. Soc. 2004, 126, 2272-2273. 20 Fluorimetric chemodosimeters: Example: A. Caballero, R. Martínez, V. Lloveras, I. Ratera, J. Vidal-Gancedo, K. Wurst, A. Tárraga, P. Molina, J. Veciana, J. Am. Chem. Soc. 2005, 127, 15666-15667. Reviews: a) Yang, Y.; Zhao, Q.; Feng, W.; Li, F. Chem. Rev. 2013, 113, 192−270; b) X. Chen, T. Pradhan, F. Wang, J. S. Kim, Q. Yoon, J. Chem. Rev. 2012, 112, 1910–1956; c) D. T. Quang, J. S. Kim, Chem. Rev. 2010, 110, 6280–6301; d) K. Kaur, R. Saini, A. Kumar, V. Luxami, N. Kaur, P. Singha, S. Kumar, Coord. Chem. Rev. 2012, 256, 1992–2028; e) J. Du, M. Hu, J. Fan, X. Peng, Chem. Soc. Rev. 2012, 41, 4511–4535; desulfurization in cells: K. Bera, A. K. Das, M. Nag, S. Basak, Anal. Chem. 2014, 86, 2740−2746.
56|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis N N Fe Fe NN O Figure 4. Fluorescent molecules described in the paper from Caballero et al, emission in the presence of Hg(II) in MeCN:H2O 7:3 solutions (blue) or just MeCN (green). Probes using nanoparticles or nanosystems.21 They are characterized for having molecular recognition subunits, such as aptamers joined to highly fluorescent molecules / nanomaterials. In particular, there are aptamers with specific interactions with mercury cations, being capable to develop nanosystems with high sensitivity. In addition to fluorescent detection of the substance in a quantitative way, some of the probes have also application in cellular imaging; which might be useful to visualize the probes in cells, as it is shown in Figure 5. ON N N N SO ON N NN S O Hg 2+ Hg 2+ Na 2 S Figure 5. Rhodamine sensor ThioRh-1 in A549 cells 10 µM probe, 30 µM Hg(II), synthesized and tested by the group of Jiang and coworkers.22 In spite of the great amount of papers in the topic, there was a series of common problems for most of the probes from literature: 21 Graphene aptamers/current change: a) J. H. An, S. J. Park, O. S. Kwon, J. Bae, J. Jang, ACS Nano 2013, 7, 10563–10571; b) J. Li, W. Tu, H. Li, M. Han, Y. Lan, Z. Dai, J. Bao, Anal. Chem. 2014, 86, 1306−1312; Fluorogenic arrays: c) S. S. Tan, S. J. Kim, E. T. Kool, J. Am. Chem. Soc. 2011, 133, 2664–2671; up-conversion nanophosphors: d) X. Li, Y. Wu, Y. Liu, X. Zou, L. Yao, F. Lia, W. Feng, Nanoscale 2014, 6, 1020-1028; e) Q. Liu, J. Peng, L. Sun, F. Li, ACS Nano 2011, 5, 8040–8048. 22 H. H. Wang, L. Xue, C. L. Yu, Y. Y. Qian, H. Jiang, Dye. Pigment. 2011, 91, 350–355.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|57 JoséGarcíaCalvo|PhDThesis There were no probes 100 % soluble in water media, some of them were soluble in waterorganic solvent mixtures with high percentages of water but there were no examples of 100% water soluble probes. The LODs calculated in most of the papers are not reliable, usually the values are calculated by outdated methods that give unreliable results. There are only a few examples of using probes for detecting both Hg(II) and MeHg(II),23 and none of them were able to speciate between both until the publication of our research was done,24 which is in strong contrast to the enormous interest that MeHg(II)-induced neurotoxicity promotes25 and its imaging in living systems.26 There were no examples of materials that change its colour/fluorescence in presence of MeHg(II) cations.2 In this regard, the synthesized systems had the objective to overcome these issues. 23 Fluorogenic probes for MeHg(II): a) M. Santra, D. Ryu, A. Chatterjee, S. K. Ko, I. Shin, K. H. Ahn, Chem. Commun. 2009, 2115–2117; b) I. Costas-Mora, V. Romero, I. Lavilla, C. Bendicho, Anal. Chem. 2014, 86, 4536−4543 (and referenced methods therein); d) E. Climent, M. D. Marcos, R. Martínez-Máñez, F. Sancenón, J. Soto, K. Rurack, P. Amorós, Angew. Chem. Int. Ed. 2009, 48, 8519–8522; C. Coll, A. Bernardos, R. MartínezMáñez, F. Sancenón, Acc. Chem. Res. 2013, 46, 339–349. 24 Speciation Hg[II]/MeHg[II]: Y. Li, Y. Yin, G. Liu, Y. Cai, in: Environmental Chemistry and Toxicology of Mercury, G. Liu, Y. Cai, N. O'Driscoll, (Editors), John Wiley & Sons, Hoboken, New Jersey, Chapter 2, pp. 1558, 2012. 25 S. Ceccatelli, M. Aschner, Editors: Methylmercury and Neurotoxicity, Current Topics in Neurotoxicity, Vol. 2, Springer, New York, 2012. 26 Methylmercury imaging: reaction-based examples in reviews: a) M. J. Pushie, I. J. Pickering, M. Korbas, M. J. Hackett, G. N. George, Chem. Rev. 2014, 114, 8499−8541; b) Z. Guo, S. Park, J. Yoon, I. Shin, Chem. Soc. Rev. 2014, 43, 16-29; an example of upconversion bioimaging of MeHg(II): c) Y. Liu, M. Chen, T. Cao, Y. Sun, C. Li, Q. Liu, T. Yang, L. Yao, W. Feng, F. Li. J. Am. Chem. Soc. 2013, 135, 9869−9876.
64|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis 3.2. Supported probes for detection of Hg(II) and MeHg(II) Next step after designing water soluble molecular probes was to develop materials with similar properties, which would improve the applications significantly. In this regard, the work was oriented in using two kinds of materials; modified silica nanoparticles and polymeric films. In the same way used for molecular probes, they were fully tested following the next steps: Synthesis and characterization. Evaluation of the qualitative properties in water, response to cations and anions. Quantitative analysis with Hg(II) and MeHg(II), kinetic effect, titrations, LODs calculation and Fluorescence Quantum Yield. Measurements with fish samples. Synthesized probes: First, silica NPs were modified with JG10 (Figure 17). Figure 17. Scheme of silicon nanoparticles JG23, which were modified with PEG chains and JG10. The results were not satisfactory for mercury detection in preliminary tests. Hence, the next step was to improve the sensitivity and water affinity by using other materials, hydrophilic polymers. Two kinds of polymers were synthetized that reached optimal results (Figure 18). Figure 18. Structure of the polymers JG32 and JG25. JG32 is a polymer slime soluble in water, whereas JG25 is a film with water affinity.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|65 JoséGarcíaCalvo|PhDThesis 4. SYNTHETIC PROCEDURE FOR MOLECULAR PROBES 4.1. Synthesis of the thiophosphinate derivatives The first step was the development of derivatives from the probes BD116 and BD119, following the same synthetic route than their precursors, a SN2 reaction of the thiophosphinate chloride with the amine group (Figure 19). Figure 19. Synthesis scheme of JG7 and JG30. The ketone derivative JG7 was, as BD116, a white creamy solid that gave colourless solutions whereas the dicyanomethylene derivative JG30, as BD119, was a yellow powder which gave very solvatochromic pale yellow solutions in most organic solvents. 4.2. Synthesis of modifiable mercury sensors: In the same way used for thiophosphinate derivatives, the reaction was performed at room temperature by dissolving the reagents in dichloromethane or chloroform (Figure 20). Figure 20. Synthesis of JG9 and JG10. The reaction takes place through a nucleophilic addition of the amine to the isothiocyanate. The purification was performed by column chromatography in DCM:MeOH (2 %) obtaining the products as white powders in 56-58 % yields.
66|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis 4.3. Synthesis of PEG chains Two synthetic PEGs and one commercial (Figure 21) were used in order to obtain water soluble compounds and, at the same time, to increase the permeability in cells as it had been reported previously by other authors:31 Figure 21. Triple bond substituted PEGs used for water-soluble derivatives. Tri-PEG JG44 was synthesized starting from tetraethylene glycol monomethyl ether and methyl gallate ester, following the scheme explained in Figure 22. Figure 22. Synthesis of the 3,4,5-tri-PEG benzyl alcohol for the synthesis of tri-PEG JG44. The reaction started with a nucleophilic addition of the alcohol to p-toluenesulfonyl chloride, under basic conditions in DCM solution. The product was purified by column chromatography DCM:MeOH (4%) as eluent. Afterwards, a nucleophilic substitution of the toluenesulfonic acid group by the alcohol group and potassium carbonate as a base, under heating, led to the tri-PEG with an ester group derivative. The ester was also purified by column chromatography, DCM:MeOH (4%) as eluent. Finally, the ester group was reduced to alcohol by lithium aluminium hydride in THF solution, obtaining the tri-PEG with a hydroxyl group in almost quantitative yield as a pale yellow oil, by filtration of the product and evaporation of solvent. Finally, the triple bond-containing PEGs were synthesized from propargyl bromide, Figure 23. 31 a) A. X. Zhang, R. P. Murelli, C. Barinka, J. Michel, A. Cocleaza, W. L. Jorgensen, J. Lubkowski, D. A. Spiegel, J. Am. Chem. Soc. 2010, 132, 12711–12716. b) D. K. Tosh, K. Phan, F. Deflorian, Q. Wei, L. S. Yoo, Z.-G. Gao, K. A. Jacobson, Bioconjugate Chem. 2012, 23, 232−247. C) C. Deraedt, N. Pinaud, D. Astruc, J. Am. Chem. Soc. 2014, 136, 12092−12098.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|67 JoséGarcíaCalvo|PhDThesis Figure 23. Synthesis of JG14 and JG44, PEG chains with a triple bond. Propargyl bromide (excess) was dissolved in THF under nitrogen. Next, sodium hydride was added at 0 ºC and after five minutes the PEG was also added. Then, it was stirred for 45 minutes and heated under reflux overnight. The product was purified by neutralization with HCl, extraction with DCM-Water (3×DCM) and column chromatography DCM:MeOH (6 %) of the residue, obtaining the product as a yellow-orange liquid. 4.4. Click chemistry for developing water soluble compounds based on PEG chains The click reactions (Figure 24) were performed between the azide group JG10 and the triple bond from the PEG chain. Figure 24. Synthesis scheme of JG15, JG47 and JG45, PEG substituted indanone molecular derivatives. The reaction was performed in DMF, under nitrogen atmosphere and in presence of a Cu(I) catalyst (5%), stirred at 30ºC overnight. The reaction yields were improved by adding TBTA ligand to the reaction for avoiding copper retention in the final product32 in the same molar concentration than the catalyst. The purification was performed by column chromatography DCM: MeOH and the products obtained in 49-65 % were waxy solids. 32 P. S. Donnelly, S. D. Zanatta, S. C. Zammit, J. M. White, S. J. Williams. Chem. Commun., 2008, 21, 2459– 2461.
68|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis 5. TESTS WITH MOLECULAR PROBES PEG molecular probes were developed to increase the final solubility in water media and the cell permeability. With this aim, the studies started by registering the response to the presence of Hg(II) and MeHg(II) (selectivity and sensitivity) and checking the intracellular response to the presence of these cations. Different experiments were performed. 5.1. Solvatochromisms In order to characterize the properties, the first step was to perform a solvatochromism study of the mercury sensing products JG7, JG30, JG15, JG47 and JG45. The conditions were as follows: Solvents in order of polarity: 1. Water 2. MeOH 3. DMSO 4. DMF 5. MeCN 6. Acetone 7. AcOEt 8. THF 9. CHCl3 10. CH2Cl2 11. Toluene 12. Et2O 13. Hexane 14. Cyclohexane The solvatochromic studies (Figures 25 and 26) were performed for the probes, they showed a clear influence from polarity and hydrogen bonding interactions, contributing to the Stokes shifts. Figure 25. Normalized absorbance and emission spectra of JG7and JG30, 50 µM in different solvents. Normalized absorbance Normalized fluorescence JG750μM λexc=325nm JG3050μM λexc=383nm 250 300 350 400 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) Water MeOH DMSO DMF MeCN EtOAc THF CHCl3 DCM Toluene Et2O Hexane CH 350 400 450 500 550 600 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) Water MeOH DMSO DMF MeCN THF CHCl3 DCM Toluene Et2O Hexane EtOAc 300 400 500 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) Water MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O 450 500 550 600 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane CH
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|69 JoséGarcíaCalvo|PhDThesis Figure 26. Normalized absorbance and emission spectra of JG15, JG47 and JG45, 50 µM in different solvents. Pictures of the solvatochromism were also taken under visible and UV light (366 nm); Figure 27. In this case, they were colourless under visible light, therefore, it was only showed the response under UV light. Normalized absorbance Normalized fluorescence JG1550μM λexc=325nm JG4750μM λexc=325nm JG4550μM λexc=325nm 250 300 350 400 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) Water MeOH DMSO DMF MeCN EtOAc THF CHCl3 DCM CH 400 500 600 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) Water MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM 300 350 400 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) Water MeOH DMSO DMF MeCN EtOAc THF CHCl3 DCM Toluene Et2O 350 400 450 500 550 600 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) Water MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene 300 350 400 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) Water MeOH DMSO DMF MeCN Acetone THF CHCl3 DCM Et2O Toluene 400 500 600 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) Water MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene
70|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis Figure 27. Pictures of probes of JG7, JG30, JG15, JG47 and JG45, 50 µM in different solvents under UV light at 366 nm. From these tests a series of conclusions were obtained: Less polar solvents such as hexane or cyclohexane were not capable of solubilising the products. All the products had a hypsochromic effect with the decrease of polarity, insignificant for absorbance and as much as 80 nm for fluorescence, in the range MeOH-Toluene. JG30 was also characterized for having a stokes shift that changes up to 150 nm depending on the solvent. In diethyl ether and toluene, compounds showed fluorescence within the long UV for some of the derivatives, for that reason their emission was not observed by the naked eye.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|71 JoséGarcíaCalvo|PhDThesis Only JG45 and JG47 were soluble in water, JG15 was not soluble in spite of having a PEG chain, it was not enough to solubilise the molecule, which was the main reason for synthesizing JG47 and JG45. The probes with a PEG chain had a hypsochromic shift compared with JG7. The fluorescence was maximized in DMSO for all the probes, where there was no increase independently of the analyte added. The best solvents, high solubility and not very high fluorescence, were methanol and acetonitrile. 5.2. Molar extinction coefficients To get a more accurate value, each of the molar extinction coefficients for the different probes were calculated by doing a regression between 10 to 50 μM concentrations in MeOH. JG7 MeOH λ (ε) = 318 nm (33000 M-1cm-1) log(ε) = 4.52 JG30 MeOH λ (ε) = 383 nm (38000 M-1cm-1) log(ε) = 4.58 JG15 MeOH λ (ε) = 320 nm (27000 M-1cm-1) log(ε) = 4.43 JG47 MeOH λ (ε) = 315 nm (25000 M-1cm-1) log(ε) = 4.40 JG45 MeOH λ (ε) = 310 nm (37000 M-1cm-1) log(ε) = 4.57 5.3. Ions tests As it was explained in Chapter 0, the probes were tested following a general procedure, taking pictures of solutions of the probes in the presence of different cations and anions under visible and ultraviolet light (366 nm): The concentration of the probes was 0.1 mM. The solvents were methanol and water or mixtures between them; while having as high percentage of water as possible, without experimenting precipitation. It was also tested the mixture acetonitrile:water, but with worse results in terms of solubilitydetection capabilities (See Experimental Appendix 5) Non-coordinant counterions were used; perchlorate, triflate and tetrabutylammonium, except for the salt AuCl3 (See Annex). The probes did not have any visible colour nor change under ambient light, as in solvatochromisms, with the exception of Au3+. After adding ever cation, the pictures are showed only under UV light at 366 nm, Figures 28, 29 and 30. The case of Au3+ is special and it was treated separately in section 5.3.2.
72|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis 5.3.1. Response of the probes to cations in water or mixtures MeOH:water: Figure 28. Response to the presence of different cations by probes JG7 and JG30 in solutions 0.1mM, in mixtures water/methanol under a 366 nm light. Figure 29. Response to the presence of different cations by probes JG9 and JG10 in solutions 0.1mM, in mixtures water/methanol under 366 nm light. Figure 30. Response to the presence of different cations by probes JG15, JG47 and JG45 in solutions 0.1mM, in mixtures water/methanol under 366 nm light.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|73 JoséGarcíaCalvo|PhDThesis From these images many conclusions were obtained. All of the probes, with the exception of JG9, had selective response to Hg(II). JG7 and JG30 had similar response to the previously synthesized compounds bearing a dimethylthiophosphinate (BD116 and BD119). However, the solubility in water was lower, precipitating when there was more than 40 % water in the mixture of solvents. As a consequence, they were not so deeply studied as the previous probes. The fluorescence also increased in presence of silver and gold cations, except for the PEG derivatives in which decreased (JG45 and JG47) or remained constant (JG15). The reason is explained in Section 5.3.2. PEG derivatives presented a side effect, greater pH dependence, being very sensitive to the presence of Lewis acid cations such as Fe(III) or Sn(II). PEG derivatives presented kinetic effects in presence of Hg(II), in which the fluorescence increased upon the addition of Hg(II). Adding high excess of Hg(II) to the PEG solution caused the precipitation of the compound (providing an apparent decrease in fluorescence). 5.3.2. Au3+ and Ag+ cations effect in water solution containing PEG probes: Concurrently to the studies of Hg(II) effect, it was observed an interaction effect of some probes with gold and silver cations. In the case of the thiophosphinate probes and the one with an azide group, only an increase in fluorescence was observed. This fact was easily explained by different interaction processes, what is normally due to pH effect and the thiophilicity of gold and silver cations. In contrast, additional changes occurred for the solutions that possessed a PEG chain, having a decrease in fluorescence instead of an increase and, what is more remarkable, a change of colour (Figures 31 and 32). Figure 31. Silver (left) and gold (right) cations, 0.2 mM, in JG47 solution, 0.5 mM. After 48 hours. Figure 32. Silver (0.1 and 0.2 mM, left) and gold (0.2 and 0.4 mM, right) cations in JG45 solution, 0.1 mM. After 24 hours. Moreover, these probes were useful not only for the formation of gold nanoparticles (deeply explained in Chapter 4), but in order to stabilise both gold and silver nanoparticles. It can be justified due to the presence of PEG chains and a reductive agent, presumably the carbothioamide moiety. Ag+ Ag+ Au3+ Au3+ Ag+ Au3+ 0.1 mM 0.2 mM 0.2 mM 0.4 mM
80|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis Figure 44. JG19 IR spectra and peak picking. IR (KBr, cm-1): 3397 (OH), 2924 (C-H), 2882 (C-H), 1693, 1636, 1455, 1352, 1092, 952, 799, 466. Figure 45. JG23 IR spectra and peak picking. IR (KBr, cm-1): 3399 (OH), 2943 (C-H), 2889 (C-H), 1693, 1636, 1546 (CAr-CAr), 1452, 1092, 954, 799, 468. The infrared characterization (Figures 43, 44 and 45) showed the presence of signals between 2950-2850 cm-1 indicated the existence of C-H groups and, around 1550 cm-1, the presence of aromatic compounds. It should be a characteristic signal from triple bonds/triazole groups at 2200-2000 cm-1, the low proportion of these groups may cause that it was not observed.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|81 JoséGarcíaCalvo|PhDThesis 6.2. Polymer supported probes Starting from probe JG10, it was proposed the synthesis of a polymer capable to react by click chemistry with the probe. To achieve that purpose, there were a series of conditions to be fulfilled: The polymeric film had to have high water affinity. Depending on the its specific aim, the polymer had to be or soluble in water or remaining as a film maintaining, with good mechanical properties. There had to be a significant and selective increase of fluorescence in presence of mercury(II) derivatives. In this regard, a series of monomers, specified in Figure 46, were selected. Figure 46. Monomers, crosslinker and photo-initiator used for the synthesis of polymers modified with mercury(II) sensitive probes. 2-Hydroxyethyl acrylate gave water affinity to the polymer. Propargyl methacrylate allowed the binding by click chemistry with the azide group of JG10. Finally, the proportions of ethylene glycol methacrylate, the crosslinker, gave the polymer its mechanical properties, not being soluble in water when there was enough of this component. The process of polymerization consists of a mixture of the monomers, polymerized by photochemical radical polymerization with DMPA35. For the development of films, the polymerization was performed in a 100 μm thick silanized glass hermetic mould upon irradiation with a UV mercury lamp, which was designed for that purpose by the Group of Polymers from Burgos University. 35 a) B. Redondo-Foj, M. Carsi, P. Ortiz-Serna, M. J. Sanchis, S. Vallejos, F. García, J. M. García, Macromolecules 2014, 47, 5334; b) B. Redondo-Foj, M. Carsi, P. Ortiz-Serna, M. J. Sanchis, F. García, J. M. García, J. Phys. D: Appl. Phys. 2013, 56, 295.
82|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis 6.2.1. Synthesis, modification and characterization of the polymers: This kind of polymer was synthetized with a low percentage of probe, in order to not saturate the samples, which would have led to black-brown samples. If the polymers are very colourful it usually leads to no sensitivity to any cation. Therefore, and after testing with several percentages, the quantity that gave good results was having 5 % of propargyl groups for the films and 1% for the soluble polymers. A) Water soluble polymer synthesis (JG32): As it is indicated in Figure 47, in order to introduce the fluorescent core JG10, it was proceeded as a normal catalytic reaction of triazole formation by click chemistry. Figure 47. Synthesis of a mercury sensitive polymer derivative from 2-hydroxyethyl acrylate soluble in water (JG32) The initial polymer was provided by the group of polymers of Burgos University. This polymer was a mixture of 99% molar of 2-hydroxyethyl acrylate and 1% molar of propargyl methacrylate, which resulted in a colourless slime. The polymeric slime was dissolved in dry DMF under nitrogen atmosphere, and JG10 was added to the mixture, in a proportion 1:1 with the amount of propargyl groups. Finally, once everything is dissolved, the Cu(I) catalyser was added and it was left under stirring at 30ºC for 24 hours. For purification, the slime was precipitated by adding diethylether to the solution, and the slime was washed several times until obtaining a yellow slime with blue fluorescence under UV light.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|83 JoséGarcíaCalvo|PhDThesis B) Synthesis of a fluorescent crosslinked polymer of 2-hydroxyethyl acrylate (JG25): Figure 48. Synthesis of a fluorescent crosslinked polymer of 2-hydroxyethyl acrylate (JG25). The initial polymer (JG25_SA2) was also provided by the group of polymers of Burgos University. This polymer was a mixture of 95% molar of 2-hydroxyethyl acrylate and 5% molar of propargyl methacrylate. In addition, it presented a 5% of ethylenglycolmetacrylate as crosslinker; resulting in a colourless film. The water-swelling percentage (WSP) of the membrane was 60% and the DMF swelling 300 %. Following the scheme from Figure 48, a flask, adapted to contain the polymer, was put under nitrogen and JG10 was dissolved in dry DMF. After that, the film was put into the flask; once the polymer is swelled with DMF, the catalyst containing Cu(I) was added. The polymer in solution was stirred with an orbital shaker for 72 hours, while it was observed that the yellow colour of the solution turned pale and the polymer yellowish (Figure 49). Finally, the film was cleaned by washing it with DMF, DMFwater and water, increasing the percentage of water little by little. This process must be performed carefully because of the different swelling depending on the solvent DMF>Water>dry; any drastic change would lead to a break of the polymer. Figure 49. Picture of membranes JG25 once the reaction has finished. C) Characterization of the polymers: The synthesized polymers were characterized by IR (Figures 50 and 51), SEM (Figure 52A) + EDX (Figure 52B).
84|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis Figure 50. IR JG25 (ATR, cm-1): 3529-3356 (O-H), 2968-2859 (C-H), 1708 (C=O), 1514 (CAr-CAr), 1441, 1393, 1264, 1167, 1077, 1043, 893, 839 (fingerprint zone). Figure 51. IR JG32 (ATR, cm-1): 3506-3330 (O-H), 2953-2878 (C-H), 1720 (C=O), 1444, 1392, 1167, 1076, 888, 839 (fingerprint zone). In the case of JG25 (Figure 50) there were some signals at 1514 and 1598 cm-1 associated to the presence of the probe (aromatic groups). In contrast, there is no presence of these signals on the initial IR spectrum, their intensity was low due to the low percentage (5%) (Inset in Figure 50). In comparison, JG32 (Figure 51) signals at 1600-1500 cm-1 were barely distinguished from the noise. It was because, for JG32, the percentage of the fluorogenic probe was too low (1%), which agrees with previous results. 100 200 300 400 500 600 700 800 0 20 40 60 80 100 Temperature (ºC) Weight (%) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Deriv. Weight/Temp Air 100 200 300 400 500 600 700 800 0 20 40 60 80 100 Temperature (ºC) Weight (%) 0.0 0.5 1.0 1.5 2.0 Deriv. Weight/Temp N2 Figure 52. TGA of JG25, decomposition at 440ºC. JG25_SA2 JG25
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|85 JoséGarcíaCalvo|PhDThesis Coating %Sreal/%Steor Gold 79.4 Gold 100.7 Gold 83.5 Carbon 84.9 Carbon 68.5 Carbon 93.7 Figure 53. [A] SEM image of JG25, [B] X-ray fluorescence analysis from JG25, [C] Comparison between theoretical proportions of sulfur atoms and amounts detected by EDX. The SEM-EDX analysis were performed with gold and carbon recap. The atomic proportion was indicated by the EDX on different areas of the polymer. The proportion between oxygen or carbon/sulfur atoms was very similar to the theoretical results associated to a 100 % stoichiometric reaction (60-100 % depending on the area). [A] [B] [C]
86|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis 7. QUALITATIVE TESTS OF MATERIAL SUPPORTED PROBES 7.1. Tests with substituted nanoparticles The nanoparticles were exposed to solutions of different cations. To do so, 15 mg of silica nanoparticles JG23 were put in 0.5 mL water solutions of cations (0.1 mM). Afterwards, the response was checked in visible and under UV-light (see Figure 54). Figure 54. Pictures of vials adding cations in water solution over 15 mg of silica nanoparticles modified (JG23). 366 nm light. The pictures showed no remarkable changes in fluorescence. There was a disappearance of fluorescence in presence of Au(III) and a little increase for Ag(I). In addition, there was a change in colour for these two cations. For Ag(I) the silica nanoparticles became pink with low concentrations (Figure 55), and grey when no water or when the concentration was high. For Au(III) the silica nanoparticles become almost black (Au-NPs formation, see Chapter 4). Figure 55. Pictures of Ag+ addition zoom for modified JG23, 0.2 mM under visible light, right. JG23 reference with no cations, left.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|87 JoséGarcíaCalvo|PhDThesis In conclusion, the synthetized nanoparticles are not useful for detection of Hg(II) as they were synthetized, but they have other properties that are worthy to be studied, such as the formation of silver or gold nanoparticles in their surface (See Chapter 4). 7.2. Tests with soluble polymer JG32 In order to check the fluorescence of the soluble polymer JG32, it was compared to the azide probe JG10. The solution of the polymer was prepared and the concentration in a concentration of 0.012 g/L of the polymer (1% of the probe, 10-6 mol probe/L). Then, the cations were added, see Figure 56. Figure 56. Fluorescence under UV light of JG10 and JG32 in solution with cations. 366 nm light. In case of JG10 the fluorescence increased very little in presence of Ag(I) and increased highly in the presence of Hg(II) and Au(III). The polymer was in water and there were not any changes under visible light. The fluorescence under UV light (366 nm) increased for Hg(II) and decreased for Au(III). The fluorescence of JG32 was also checked in the presence of other cations. Figure 57, sequence: Water, Hg(II), MeHg(II), Au(III), Pd(0), Rh(III), Ir(III), Pt(II), Co(II), Pd(II). (As chlorides, except Pd(0) that was Pd(dba)2). Figure 57. JG32 in water, fluorescence under UV light in presence of different cations. 366 nm light. Two processes were observed, as it is shown in Figures 56 and 57, an increase of the fluorescence under UV light for Hg(II) and MeHg(II) and a decrease for Au(III) and Pd(II). The increase in fluorescence in presence of different cations was also quantified in a fluorometer, as it is shown in Figures 58, 59 and 60:
88|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis Nothing Water Ag+ Ni2+ Sn2+ Cd2+ Zn2+ Pb2+ Cu2+ Fe3+ Sc3+ Al3+ Hg2+ MeHg+ 0 20406080 Emission increase (%) Cation Figure 58. Variation in fluorescence of JG32 in water solution, 12 mg/L. After adding cations, 0.2 mM, and waiting 60 minutes. λexc = 320 nm, λem = 455 nm Nothing Water Ag+ Ni2+ Sn2+ Cd2+ Zn2+ Pb2+ Cu2+ Fe3+ Sc3+ Al3+ Hg2+ MeHg+ 0 20406080100 Emission increase (%) Cation Figure 59. Variation in fluorescence of JG32 in water solution, 12 mg/L. After adding cations, 0.2 mM, waiting 60 minutes and increasing 0.2 mM the concentration of Hg(II) in all of them. λexc = 320 nm, λem = 455 nm
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|89 JoséGarcíaCalvo|PhDThesis Nothing Water Ag+ Ni2+ Sn2+ Cd2+ Zn2+ Pb2+ Cu2+ Fe3+ Sc3+ Al3+ Hg2+ MeHg+ 0 20406080 Final emission/Initial emission Cation Figure 60. Variation of fluorescence of JG32 in water solution, 12 mg/L. After adding cations, 0.2 mM, waiting 60 minutes and increasing 0.2 mM the concentration of MeHg(II) in all of them λexc = 364 nm, λem = 445 nm. A major increase in fluorescence occurred in presence of Hg(II) and MeHg(II), moreover there were minor increases in the presence of Sn(II) and Fe(III), associated to the Lewis acidity of these cations. In addition, a total inhibition of the signal for Au(III) or partial inhibition of the signal for Cu(II) were also observed. 7.3. Tests with polymeric film JG25 Several pieces of polymer 0.3×0.3 cm (approximately) were added to different vials and 60 µL of the cations solutions (5 mM) were added, enough to cover all the surface of the polymer. Then, the polymers were left to dry and a picture was taken under UV light, Figure 61, and the changes in fluorescence registered in the fluorometer, Figure 62. Figure 61. Fluorescent response of JG25 in presence of 60 µL of cations (UV light). Ag+ Ni2+ Sn2+ Cd2+ Zn2+ Pb2+ Cu2+ Fe3+ Sc3+ Al3+ Hg2+ MeHg+ Reference
96|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis After adding Hg(II) the fluorescent emission increased very quickly for the first five minutes, (Figure 72) then, increased linearly for 70 minutes (Figure 73A). The process was very slow and supposed a 54 % increase in the global fluorescence; reaching 6 % in the first 5 minutes and 33 % in the first 90 minutes. Moreover, this process was dependent on the concentration of Hg(II) added, but the results were very similar when the time was very short, less than twenty minutes. (Figure 73B) 20 40 60 80 200 220 240 Emission intensity (a.u.) Time ( min ) 0 20406080100 190 200 210 220 230 240 250 Emission intensity (a.u.) Time (min) 0.1 mM 0.08 mM Figure 73. Kinetic response of JG25 in an aqueous solution of Hg(ClO4)2 0.1 mM [A] and comparison with 0.08 mM [B]. λexc = 369 nm. λem = 445 nm, 100 minutes. 8.2.2. Titration experiments by adding Hg(II): After the kinetic results, a deep study on the equilibrium was not performed because of the difficulty and low interest, the need of a long waiting time and the relation between concentration to it. In spite of this fact, there were some alternatives; it is noteworthy that the changes at the beginning of the additions were linear; as a consequence, adding quantities of Hg(II) every 5 minutes resulted in a graphic plot that was very similar to the typical plot for species studied in equilibrium, Figure 74. 400 500 600 0 100 200 300 400 500 600 700 Emission intensity (a.u.) Wavelength (nm) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 10000 20000 30000 40000 50000 60000 Integral 379-650 (a.u.) [Hg(II)] (mM) Figure 74. Left: Fluorescence curves by addition of increasing concentrations of Hg(ClO4)2, λexc = 369 nm. Right: Titration plot by using integral surfaces of the fluorescence curves between 379-650 nm in response to increasing concentrations of Hg(ClO4)2 λexc = 369 nm. In this way, the limit of detection obtained was a little higher than taking measurements when the interaction Hg(II)-probe reached the equilibrium. Nevertheless, the value was perfectly valid for an assigned concentration and time. The saturation of the signal was reached when the concentration of Hg(II) was near 1 mM (two hours and a half after the first addition) and, in a concentration below 10 µM, the variation of intensity between additions was linear. y = 0.66x + 201.05 R² = 0.990 1mMHg(II) [A] [B]
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|97 JoséGarcíaCalvo|PhDThesis 8.2.3. Results of the LOD calculation: 0 2 4 6 8 10 12 14 16 18 20 16800 16900 17000 17100 17200 17300 17400 17500 17600 Integral 379-650 nm (a.u.) [Hg(II)] (M) Figure 75. Linear regression by using integral surfaces of the fluorescence curves between 379-650 nm in response to increasing concentrations of Hg(ClO4)2, λexc = 369 nm, LOD calculation. Therefore, with the data showed in Figure 75, the LOD was calculated to be 6.6 µM or 1.3 ppm of Hg(II) in water. Value reached 15 minutes after the first measure, so this was the average time that was necessary to detect a noticeable increase of the fluorescence of the solution. 8.2.4. Kinetic response to MeHg(II): The behaviour of the probe in presence of MeHg(II) turned out to be different to the previous behaviour in presence of Hg(II). Figure 76 shows the changes in fluorescent emission on time by normalizing the spectra: 0 200 400 600 800 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 Normalized Emission Intensity Time (min) 250 M 100 M 5 M Figure 76. Normalized kinetic response of JG25 to MeHgCl aqueous solution. λexc = 364 nm. λem = 445 nm Several conclusions were obtained for the kinetic study. The increase on the emission was more or less the same independently of the MeHg(II) concentration. However, when the concentration was high, the maximum of emission was reached in 90 minutes, after that the concentration decreased. In contrast, when the concentration was low, the emission increase was slower, but the decrease started after 8 hours, which was observed for MeHg(II) 5 µM. y = 38.8x + 1.69ꞏ104 R² = 0.995
98|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis The presence of two processes was a possible explanation for these facts. The first one was the complexation of the cation, which was very fast and needs a very low amount of MeHg(II) to reach the saturation. Then, there was another process in which the MeHg(II) modified the structure of the complex between the polymer and the MeHg(II) cation (possibly a reaction) which was slower and depended on the concentration. 8.2.5. Titration experiments by adding MeHg(II): By the same way than in the case of Hg(II), the titration experiments with the polymer where done by adding the solution of the cation every 5 minutes (Figure 77). 400 450 500 550 600 0 50 100 150 200 250 300 350 Emission intensity (a.u.) Wavelength (nm) 0 50 100 150 200 18000 20000 22000 24000 26000 28000 30000 32000 34000 Integral 380-650 nm (a.u.) [MeHg(II)] ( M ) Figure 77. JG25 titration. Left: Fluorescence curves by addition of increasing concentrations of MeHgCl, λexc = 364 nm. Right: Titration plot by using integral surfaces of the fluorescence curves between 380-650 nm in response to increasing concentrations of MeHgCl, λexc = 364 nm. During the measurements the emission intensity increased faster than in the case of Hg(II). 8.2.6. Results of the LOD calculation: 0.5 1.0 1.5 2.0 2.5 3.0 18400 18600 18800 19000 19200 19400 19600 19800 Integral 380-650 (a.u.) Concentration (M) Figure 78. JG25 titration. Regression plot by using integral surfaces of the fluorescence curves between 380-650 nm in response to increasing concentrations of MeHgCl, λexc = 364 nm. LOD calculation. From Figure 78, the LOD calculated was 1.5 µM or 0.3 ppm of MeHg(II). This value was reached in less than 20 minutes. 0.2mM y = 456x + 1.83ꞏ104 R² = 0.988
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|99 JoséGarcíaCalvo|PhDThesis 8.2.7. Effect of pH on JG25: Although the quantitative measurements were performed in deionized water (pH = 8 approximately), it was important to study the pH effect, in order to do further studies, such as the measurements from fish samples, which were the final objective of the work. In this regard, a solution buffered at pH = 7.8 (HEPES buffer, 5 mM) was acidified little by little with HCl 1M and changes in pH – fluorescence were evaluated (Table 3 and Figure 79). The results indicated that an increase of fluorescence occurred while lowering pH. Then, by using the same process, the increase in fluorescence was measured in a sample that contained 10 µM of Hg(II). pH Emission intensity (a.u.) Emission intensity + Hg(II) (a.u.) 7.8 39 89.2 7.5 43.5 100.6 7.1 49.4 116.7 6.6 60.9 148.4 5.4 69.7 173 4.6 96.4 243 4 110.1 282.7 3.5 121.9 318.7 Table 3. Emission intensity, with and without Hg(ClO4)2, at different pH, of JG25. pH 7.8 pH 7.5 pH 7.1 pH 6.6 pH 5.4 pH 4.6 pH 4 pH 3.5 0 50 100 150 200 250 300 350 Emission intensity (a.u.) Without Hg 2+ With Hg 2+ Figure 79. Emission intensity of JG25, with and without Hg(ClO4)2, at different pH; λexc = 365 nm, λem = 455 nm The proportional increase in fluorescence with Hg(II) was higher in acidic pH, being lower if the buffer was basic; see Figure 79. Therefore, it was determined that the best pH for evaluating Hg(II) presence was a controlled slightly acidic or neutral pH. (λexc =365 nm, λem = 450-460 nm)
100|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis 9. ANALYSIS OF FISH SAMPLES In order to find the best conditions to have reproducible and reliable measurements of the presence of mercury in fish samples, some preliminary tests were necessary: Lyophilization of the fish; to avoid interference from the quantity of water in the sample. Extracting the samples; having homogeneous samples and with less interferents give reliability to the measurements. Qualitative measurements; so as to check if the fluorescence changes before any other analysis. ICP analysis, from the fish and the extracted samples. Fluorescent response of JG25 when in contact to fish and mercury extracts. Comparison ICP-JG25 results. The scheme from Figure 80 represents a guide for the process. Figure 80. Scheme of treatment for analysis of fish samples. 9.1. Extraction of mercury species from fish Two methods were tested: acidic extraction and silica extraction. It was also possible to perform a basic extraction, but it was discarded because of the high temperatures, the extracts presented an intense yellow colour and the probe had no good results at pH superior than 8. In addition, all the measurements were done with lyophilized fish because the quantity of water is vital in the concentration calculation. Acid extraction: 5 ml of HCl 5 M in water and 5 ml of NaCl 0.25 M in water were added to 0.5 g of lyophilized fish in a sealed vial. The mixture was sonicated for 10 minutes and heated for other 10 minutes at 60 ºC. Then, the mixture was centrifuged at 4000 rpm for 10 minutes and at 7000 rpm for other additional 10 minutes. Afterwards, the liquid phase was filtered in a glass fibre filter of 0.22 μm pore. Finally, the samples to be measured by fluorescence were concentrated to 1 ml of water.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|101 JoséGarcíaCalvo|PhDThesis Silica extraction: 0.5 g of fish and 2 ml of water were mixed in a mortar and the mixture was grinded, 1g of silica was then added and mixed. Next, the mixture was treated with HCl 5 M in water following the same procedure used for the acid extraction. Finally, the samples to be measured by fluorescence were concentrated to 1 ml of water. 9.2. Determination of mercury by ICP-Mass analysis For determination of metals by ICP it was standardized a method in which the organic material was digested under highly concentrated HNO3 solutions. Afterwards, the samples are diluted and introduced in the system. The ICP was calibrated with standards36 and the concentration determined by adjusting the signal into a regression done with this calibration. The samples were lyophilized, causing a loss in weigh of 70-80 %, which is shown so as to compare the real concentration in fish. Conc. Hg (ppm) lyophilized Conc. Hg (ppm) fresh Sample Acid extraction Silica extraction % (Water) Acid extraction Silica extraction Swordfish 5.1 ± 0.1 6.0 ± 0.6 73 1.4 ± 0.1 1.6 ± 0.6 Tuna 3.1 ± 0.1 5.1 ± 1.4 71 0.9 ± 0.1 1.5 ± 1.4 Panga 1.1 ± 0.4 1.3 ± 0.6 78 0.2 ± 0.1 0.3 ± 0.1 Salmon 0.015 ± 0.016 0.07 ± 0.01 72 0 0 Conger eel 2.4 ± 0.1 2.2 ± 0.2 80 0.5 ± 0.1 0.4 ± 0.1 Dogfish 6.8 ± 0.3 3.9 ± 0.1 75 1.7 ± 0.1 1.0 ± 0.1 Table 4. Amount of mercury detected by ICP-Mass analysis on fish samples, lyophilized fish (left), corresponding amount to fresh fish (right). Swordfish Tuna Panga Salmon Conger Dogfish 0 1 2 3 4 5 6 7 [Hg] (ppm) Acid extraction Silica extraction Swordfish Tuna Panga Salmon Conger Dogfish 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 [Hg] (ppm) Acid extraction Silica extraction Figure 81. Representation of the amount of mercury detected by ICP-Mass analysis on fish samples, (left) lyophilized, (right) the corresponding amount to fresh fish. The data from Table 4 and Figure 81 shows that the concentration was higher when the fish was bigger. Another important result was that the concentration of mercury in salmon could be 36 The process was performed entirely by technicians at the Research building from University Burgos, given a pre-treated sample and the approximated concentration to expect.
102|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis considered as 0. This was normal because salmons were obtained from a fish farm, not from wild sea fish, so there was no possible bioaccumulation. The LOD of the polymer JG25 was 1.3 ppm for Hg(II) and 0.3 ppm for MeHg(II), therefore, a direct measurement would be theoretically possible, depending on the interferents and the proportion g(fish)/water volume. The order of magnitude in mercury concentration means that, in swordfish, tuna, conger eel and dogfish it could be over the LOD, especially in case of swordfish, which is known to have around the maximum amount of mercury ingestion per week recommended by the FDA, (1.3 ppm). 9.3. Measurements of mercury extracts with JG25 1 mL of each extract was added to a cuvette with JG25. Then the fluorescence was checked. The variation of intensity was measured with water (blank) and then by adding the extract, the difference is the value given in Table 5. ∆ Emission intensity, λexc = 365 nm (a.u.), λem = 455 nm (a.u.) Extraction Method Swordfish Tuna Panga Salmon Conger eel Dogfish Acid 301.6 229.6 0.91 5.5 152.1 377.3 Silica 354.2 307.2 27.7 6.4 162.5 171.7 Table 5: Emission intensity variation of JG25 in contact with fish samples. Swordfish Tuna Panga Salmon Conger eel Dogfish 0 100 200 300 400 Emission intensity (a.u.) Acid extraction Silica extraction Figure 82. Emission intensity variation of JG25 in contact with fish samples, λexc = 365 nm, λem = 455 nm. From the representation in Figure 82 it is clear that the fluorescence was similar to the results of the ICP analysis, evidencing the relation between emission and ppm of Hg(II) + MeHg(II).
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|103 JoséGarcíaCalvo|PhDThesis 9.4. Comparison ICP – JG25 To compare the results, the process followed to elaborate Figure 83 was: Normalizing the graphs, taking as reference the ICP results of the acid extraction of dogfish. Considering the results of the ICP-Mass in ppm and converting the fluorescence values to ppm, from the ICP-Mass values. Swordfish Tuna Panga Salmon Conger Dogfish 0 2 4 6 [Hg] (ppm) ICP Acid JG25 Acid ICP Silica JG25 Silica Figure 83. A comparison between the values of mercury in fish samples obtained by ICP-Mass analysis and fluorescent measurements. The different results were a consequence of multiple factors such as the treatment of the samples or the fish matrix, although they were very close when comparing methods. The increase of fluorescence in solution was enough to see the signal of fluorescence for the samples with higher concentrations of mercury, namely swordfish, tuna and dogfish. None of the salmon samples showed an increase in fluorescence. For panga, the values were very close to the LOD, therefore, it was expected that there was no signal in fluorescence measurements. As a result, the method might be useful in order to measure quantitatively the mercury concentration when measuring above the LOD. 9.5. Direct analysis of mercury on fish samples with JG25 First, some preliminary tests were performed with fish samples that contained high quantities of mercury such as tuna and swordfish. After that, several samples of fish, tuna, swordfish, conger and panga were measured by homogenization with fish samples. As qualitative test, it gave good results but with low repeatability because of the different percentage of water, among other reasons. Finally, the tests were performed by using directly the same samples originally used for extractions, 0.5 g of lyophilized fish were mixed with 2 ml of water. Then, a piece of the polymeric sensor JG25 was added. To check the difference in fluorescence every polymer fragment in contact with fish samples was measured at different waiting times in the fluorometer, obtaining the results in Table 6 and Figure 84. The results may be compared with the corresponding results from the extraction by normalizing to one of them (dogfish in this case) (Figure 85). Normalization
104|CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES JoséGarcíaCalvo|PhDThesis ∆ Emission intensity 365 nm (a.u.) Sample/ time (h) Swordfish Tuna Panga Conger eel Dogfish 0.5 220.5 212.2 21.3 32 230.93 1 260.8 237.7 21.97 65.79 272.2 24 301.6 280 30.68 136.3 318.4 Table 6. The relation between the concentration of mercury and the obtained values of fluorescence for fish samples and the polymeric sensor. Swordfish Tuna Panga Conger Dogfish 0 50 100 150 200 250 300 350 Emisssion intensity (a.u.) 30 minutes 60 minutes 24 hours Figure 84. Emission intensity variation with JG25 in fresh fish samples (λexc = 365 nm, λem = 455 nm) at different waiting times. Swordfish Tuna Panga Conger Dogfish 0.0 0.5 1.0 1.5 2.0 [Hg](ppm) ICP Acid JG25 Acid ICP Silica JG25 Silica JG25 Fresh Figure 85. Emission intensity variation in experiments with fresh fish samples JG25 (λexc = 365 nm, λem = 455 nm) compared with the results from the extracts. Therefore, a relation between the concentration of mercury and the obtained values of fluorescence was confirmed.
CHAPTER1.FLUORESCENTPROBESFORDETECTIONOFHg(II)DERIVATIVES|105 JoséGarcíaCalvo|PhDThesis 10. DATA SUMMARY JG7 and JG30 They were soluble in Methanol:Water 80:20. JG7 was sensitive to Hg(II) and MeHg(II) whereas JG30 was sensitive only to Hg(II). JG7 MeOH λ (ε) = 318 nm (33000 M-1cm-1) JG30 MeOH λ (ε) = 383 nm (38000 M-1cm-1) JG15 It was soluble in Methanol:Water 9:1. JG15 MeOH λ (ε) = 320 nm (27000 M-1cm-1). Φ MeOH (JG15) = 0.18 ± 0.02 Φ MeOH (JG15 + Hg(II)) = 0.44 ± 0.02 (1 equivalent no waiting time) Hg(II) addition studies in MeOH were performed, the 1H-NMR evidenced complexation+reaction process. JG47 JG47 was soluble in 100 % water. JG47 MeOH λ (ε) = 315 nm (25000 M-1cm-1) It was sensitive to Hg(II), but the response was highly dependent on the solvent. Φ MeOH (JG47) = 0.08 ± 0.02 // Φ H2O (JG47) = 0.1 ± 0.02 Φ MeOH (JG47 + Hg(II)) = 0.64 ± 0.02 // Φ H2O (JG47 + Hg(II)) = 0.16 ± 0.02
CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES|113 JoséGarcíaCalvo|PhDThesis 1. INTRODUCTION. PERYLENE DERIVATIVES, STRUCTURE AND PROPERTIES The word perylene comes from rylene. Rylenes are a family of dyes based on naftalene groups connected via peri-positions (Figure 4). As showed in Figure 1, the general structure of these compounds depends on the number of naftalene groups, when n = 0 the dye is called perylene, n = 1 terrylene and n = 2 quaterrylene. Being these 3 derivatives considered as the best options for organic synthesis. Figure 1. Rylene structure scheme. Perylene derivatives are the most studied rylene dyes because of their characteristics. The highly conjugated structure gives great photoelectronic properties and diminishes the drawback of low solubility, a common issue when n > 1. Perylene derivatives have been used in a wide range of applications within supramolecular chemistry,1 organic electronics2 or the development of chemical sensors.3 Firstly, in regard to chemical sensors and the use of perylenes as colorants, it is important to distinguish between pigments and dyes. They are called pigments when their physical and chemical properties depend highly on aggregation; a fact directly associated to the way of synthesis, the presence of different solvents and temperature. In contrast, dyes are more independent from their environment; usually due to the inability to interact between them. These characteristics become of upmost importance in the case of perylene derivatives, that may behave as pigments or dyes depending on the structure of each specific derivative. The starting materials when working with perylene derivatives are two, being both commercially available. Perylene, a yellow powder with strong blue fluorescence. Perylene dianhydride (PDA), the synthesis of which is standardized nowadays (Figure 2). In short, since the development by Kardos in 1912 from acenaphtene;4 the synthesis consisted of 4 steps to obtain the bisanhydride, which is deeply studied in literature. Furthermore, it is usually finished with an imidization process. 1F. Würthner, Chem. Commun. 2004, 14, 1564-1579. 2C. Huang, S. Barlow and S. R. Marder, J. Org. Chem. 2011, 76, 2386-2407. 3X. Zhang, S. Rehm, M. M. Safont-Sempere, F. Würthner, Nat. Chem. 2009, 1, 623–629. 4M. Kardos, D.R.P. 276357, 1913.
114|CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES JoséGarcíaCalvo|PhDThesis Figure 2. PDA synthetic scheme. Between perylene and PDA, the second one has been the most interesting starting material for a variety of applications, due to its properties. The modification of the anhydride groups with amines and the substitution in peri, bay and ortho positions gives them a wide range of possibilities to modulate aggregation and/or electronical properties. Therefore, different substitution leads to changes in characteristics related with each other; such as colour, fluorescence or solubility. As a consequence, there is a huge quantity of possible variations and, due to the interesting applications, plenty of literature about the topic. Starting from PDA, the most common way to proceed has been the introduction of imide groups,5 within the peri positions of the PDA. This reaction allows to alter the solubility/aggregation, minimizing changes in electronical properties. Concurrently, there are different possible perylene imide derivatives (PIs), depending on the substituents in the core. There are two positions susceptible to have the imide group, which is used as a method to classify PIs, Figure 3. Figure 3. The two most common perylene imide derivatives (PIs): perylenediimide (PDI) and perylenemonoimide (PMI). 1.1. Perylene imides (PIs) synthesis Physical and chemical properties of PIs rely on substitution. Then, the first step when using PIs is always to adapt the system to the objective to be fulfilled, which leads to introducing groups that are most likely to achieve particular goals. From the many variations, the usual positions to modify are divided into several groups (Figure 4): 5a) T. Maki, H. Hashimoto, J. Chem. Soc. Jap., 1951, 54, 544. b) Y. Nagao, Prog. Org. Coat., 1997, 31, 43-49.
CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES|115 JoséGarcíaCalvo|PhDThesis N N O O OO N O O RRR bay ortho ortho bay bay peri imide imide imide bay orthoortho ortho ortho Figure 4. Possible substituted sensitive positions in PDIs and PMIs. In this regard, there are several routes to functionalized PDIs and PMIs, the most common and remarkable are summarized in the Figures 5 and 6, respectively: Figure 5. Some common synthetic routes for functionalization of PDIs. Briefly summarized, PDIs are functionalized through several routes: Imide modification: The groups in the imide positions are not directly conjugated to the perylene core. Then, its modification alters solubility but does not affect the electronical properties to a large extent. However, it is possible to develop fluorogenic materials by introducing recognition units in those positions, by changing the fluorescence quantum yield and giving them a potential use as PET sensors.6 Bay functionalization: Being the most common, it is directly related with their electronic properties. There are two main products, the bis-functionalization (1,6 and 1,7) and the tetrafunctionalization. 6I. Georgiev, A. R. Sakr, V. B. Bojinov, Dyes Pigm. 2011, 91, 332-339.
116|CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES JoséGarcíaCalvo|PhDThesis Ortho functionalization:7 The most “novel” procedure to modify PDIs, it may alter the electronical properties too, but distinctly to bay functionalization. It is performed by using iridium or ruthenium catalysts. The scheme showed in Figure 5 gives an idea of the possibilities, although not the specific order of the reactions. Bay functionalization may start by bromination of the PDA and the imide groups may be changed after introducing different substituents in bay position. In addition, modification of groups alters reactivity and yields, making possible the modulation and optimization of a variety of PI derivatives. PMIs functionalization8 is, in many occasions, similar to PDIs as showed in Figure 6. Figure 6. Some common synthetic routes for PMIs functionalization. Being an extensive part of the work developed during the thesis, it is widely explained in section 3.3. in which the procedures from literature were optimized for the objectives of the research group. In general, the process starts with the imidization-decarbonylation of the PDA, in one or several steps. Then, depending on the purpose, it is substituted in peri, bay and/or ortho positions. In addition, 7a) X. Li, H. Wang, J. A. Schneider, Z. Wei, W.-Y. Lai, W. Huang, F. Wudl, Y. Zheng, J. Mater. Chem. C. 2017, 5, 2781-2785. b) J. E. Bullock, M. T. Vagnini, C. Ramanan, D. T. Co, T. M. Wilson, J. W. Dicke, T. J. Marks, M. R. Wasielewski, J. Phys. Chem. B 2010, 114, 1794-1802. 8 Y. Hu, S. Chen, L. Zhang, Y. Zhang, Z. Yuan, X. Zhao, Y. Chen, J. Org. Chem. 2017, 82, 5926-5931.
CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES|117 JoséGarcíaCalvo|PhDThesis it is possible to change the imide group, provided that the other substituents of the PMI are stable to the process. Imidization: starting from PDA there are two methods to obtain PMIs, by one-pot decarbonylation-imidization,9 or divided into three steps.10 First, the synthesis of the PDI; then, formation of the perylene monoanhydride-monoimide and, finally, decarbonylation of the monoanhydride. Peri substitution: once the PMI substrate is obtained, the most common step is monobromination in peri. The conditions are “soft” so as to obtain the mono-substituted product selectively. Bay + peri substitution: The process is usually performed starting by a bromination under “strong” conditions (high excess of bromine and high temperatures), which gives a mixture of two tri-brominated products with a low quantity of the tetra-brominated. Reimidization: When working with PMIs, different imide substituents provide several properties, so there could be many reasons to change them in later steps of the synthesis. Among the multiple possibilities, some examples are the groups that could not resist the conditions for the direct formation of the monoimide, from PDA (such as boc-protected amines) or some amines that are not bulky enough to make PMIs soluble in common solvents. For these or other reasons, it might be of interest changing the imide group. The process undergoes by the monoanhydride formation11 followed by the introduction of a primary amine. Ortho functionalization: It is the same procedure than when working with PDIs, which involves a Ruthenium catalyst in the process. 1.2. Modulating PIs properties, the stacking of PIs Some interesting features of perylene imide derivatives (PIs) are the electronical properties, which led to exceptionally high fluorescence quantum yield, close to 100 % in many occasions. Furthermore, the emission and absorbance are in the range of visible-NIR, and it is easily tuneable with the variation of substituents,12 increasing the applicability. So as to fully understand the properties and applications of PIs, it is necessary to explain to some extent one of their most important sources, their ability to perform homo-stacking between molecules. The aggregates of PIs have been studied for decades, and yet, nowadays it is difficult to predict it before experimental testing. It is important not only because of the relation aggregationsolubility, but for the optoelectronical properties too, affecting directly the wavelength of absorbanceemission (energy levels) and the fluorescence quantum yields (more aggregation implies more possible non-radiative routes for relaxation). The search on literature about the stacking of perylene derivatives usually leads to a massive amount of information because of the dependence on multiple factors. In this regard, Figure 7, which is based 9 a) L. Feiler, H. Langhals, K. Polborn. Eur. J. Chem. 1995, 7, 1229–1244. b) L. Pleux, A. L. Smeigh, E. Gibson, Y. Pellegrin, E. Blart, G. Boschloo, A. Hagfeldt, L. Hammarström, F. Odobel, Energy Environ. Sci., 2011, 4, 2075-2084. 10 Y. Geerts, H. Quante, H. Platz, R. Mahrt, M. Hopmeier, A. Böhm, K. Müllen; J. Mater. Chem. 1998, 8, 2357– 2369. 11 T. Dentani, K. Funabiki, J.-Y Jin, T. Yoshida, H. Minoura, M. Matsui, Dyes and Pigments. 2007, 72, 303-307. 12 C. Li, H. Wonneberger, Adv. Mater. 2012, 24, 613–636.
118|CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES JoséGarcíaCalvo|PhDThesis on a review by Würthner et al,13 shows, schematically, the consequences of different ways of aggregation, and halfway situations. In addition, it is not only valid for perylene derivatives but for every molecule with capability to perform π-π stacking. Figure 7. Energy diagram for 2D formation of dimer and the effect over the energy levels. To simplify the situation, this section is explained from the point of view of perylene derivatives. Figure 7 considers the effect of having a monomer, and the different interaction between 2 molecules (dimers) in a 2D model. The relation aggregation-luminescent properties might be explained from the energy diagram in Figure 7: Monomers: The fluorescence is the one associated to the perylene derivatives. J-aggregates (θ = 0º): when the perylene molecules are aggregated in the same plane. As a consequence, it creates a more accessible excited energy level for the electrons. It implies less energy for the fluorescent emission, giving a batochromic shift. H-aggregates (θ = 90º): when perylene molecules are aggregated on top of each other. The more excited energy level for the electrons becomes less accessible, so the wavelength of emission increases, giving a hypsochromic shift. Apart from what is shown, there are also many possible and more complex variations. Some of these properties are briefly summarized in this section, so as to understand the complexity of the topic. Nevertheless, most properties and interactions are widely explained in specific sources, focusing the topic on the situation and the molecule-system under study. This interpretation is based on Kasha’s model, which is a qualitative explanation from the point of view of the interaction of transition dipole moments of chromophores, with respect to the spatial arrangement when the photoexcitation takes place. Halfway situations also exist depending on the angle θ, in which it is aggregated. Situations between J-H aggregates are not only possible but the most common situation. In particular, it is possible to have very similar wavelength of emission, comparing with the monomer, if the right angle between molecules is achieved (54.7º). 13 F. Würthner, C. R. Saha-Möller, B. Fimmel, S. Ogi, P. Leowanawat, D. Schmidt; Chem. Rev. 2016, 116, 962−1052.
CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES|119 JoséGarcíaCalvo|PhDThesis To explain why the change in behaviour of perylene derivatives occurs when stacking, two opposite situations may be found (the examples were taken from the own research of our group): Substituents with low polarity (for example, aliphatic chains): the polarity of the core is higher than the polarity of the substituents. In consequence, when the solvent is highly polar (for instance, DMSO), the core would tend to be facing the solvent (J-aggregate), giving a batochromic shift. However, if the solvent is less polar (methylcyclohexane for example), the cores will tend to aggregate and the substituents will be facing the solvent, giving a hypsochromic shift. See example in Figure 8. N O O JG62 Aliphatic substituent Figure 8. PMI JG62 solvatochromism in different solvents. Polar substituents (for example, hydroxyl or amino groups): when the polarity of the substituents is high enough, the behaviour may be the opposite, having H-aggregates when the solvent is more polar and J-aggregates in the opposite case. See example in Figure 9. N N O O O O N N N O O PC63 Polar substituent A liphatic substituent Aliphatic substituent Figure 9. PDI PC63 solvatochromism in different solvents.14 Apart from this, it is important to remark that it is not only a matter of polarity of the coresubstituents. The position, conjugation and bulkiness of the substituents may have even a more important role; considering that, in reality, aggregation occurs in a 3D distribution. So as to fulfil what 14 P. Calvo-Gredilla, J. García-Calvo, J. V. Cuevas, T. Torroba, J.-L. Pablos, F. C. García, J.-M. García, N. ZinkLorre, E. Font-Sanchis, Á. Sastre-Santos, F. Fernández-Lázaro, Chem. Eur. J. 2017, 23, 13973-13979.
120|CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES JoséGarcíaCalvo|PhDThesis was explained above, it would be necessary to explore the possibility of stacking in a different position; having very polar substituents does not mean that it will tend to give H-aggregates in polar solvents (Figure 10). N O O N NN O O JG125 Polar substiuent Aliphatic substituent Figure 10. PMI JG125 (10 μM) solvatochromism in different solvents Additionally, talking about the aggregation, the size of the substituents introduced must be taken into account. Having bulky substituents reduces stacking and gives the products properties of a dye instead of a pigment. What is more, the presence of bulky substituents in the imide position increases dramatically the solubility of the PIs which, among other factors, leads to enhanced reactivity. Likewise, if it is not soluble enough, it might be impossible to perform new synthetic processes. The introduction of substituents in other positions, besides the imide group, also alters the aggregation of the PIs but, in contrast, it changes deeply the optical properties. For PDIs, the introduction of different substituents in bay positions is the most common way to proceed and for PMIs, is the peri substitution. This technique allows to modulate, apart from the electronical properties, the possible ways of aggregation between them, which is directly related with solubility, absorbance and fluorescence. Moreover, there is a noteworthy distinction between PMI and PDI; although PMIs have more and easier synthetical variations, the lower symmetry may hinder getting a straightforward prediction-explanation for their stacking properties. Besides, the great relation aggregation-concentration may be analysed not only by changes in luminescence. For example, applied to PMIs, the effect of having different concentrations was observed for some synthetized derivatives (JG2L from Chapter 3B), getting variations in the 1H NMR signals depending on how concentrated is the sample, Figure 11.
CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES|121 JoséGarcíaCalvo|PhDThesis 1.52.02.53.03.54.04.55.05.56.06.57.07.58.08.59.0 f1 (ppm) 12.15 3.07 3.11 1.04 4.01 1.00 4.08 2.03 1.85 1.64 1.66 1.67 1.72 1.75 1.83 1.86 2.00 7.52 7.54 7.54 7.56 7.58 7.84 7.86 8.20 8.21 8.21 8.22 8.23 8.24 8.24 8.28 8.29 8.30 8.31 8.32 8.32 8.34 8.34 8.42 8.42 8.44 8.44 8.45 8.46 8.47 8.48 8.86 Figure 11. PMI JG2L, 1H NMR in CDCl3, effect of dilution between 1 – 0.1 mM. In Figure 11, the dependence on concentration of JG2L illustrates the aggregation properties. This reliance is not always so noticeable, but it evidences the importance of the substitution. In addition, increasing concentration leads to a decrease in fluorescence emission. The studies from Chapter 2 and 3 of this thesis are focused in using PMIs as chemical sensors, ligands and/or biological markers. Hence, the explanation of the stacking is usually complex, although there is plenty of literature applications related to the stacking.15 15a) L. Huang, V. J. Catalano, S.-W. Tam-Chang, Chem. Commun. 2007, 2016–2018. b) L. Huang, S.- W. TamChang, W. Seo, K. Rove, Adv. Mater. 2007, 19, 4149–4152. +Conc ‐Conc
128|CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES JoséGarcíaCalvo|PhDThesis 3.2. Bromination of PMIs Bromination reactions are extensively used and an easy way to modify PMIs, being usually the first step to tune their fluorescence and other properties.10 If the bromination reaction is performed under mild conditions, the monobrominated product is the only one obtained, in quantitative yields. However, using high temperatures and high excess of bromine lead to polybrominated products. Figure 16. Synthetic route for bromination of PMI derivatives. Poly-bromination (up) and monobromination (down). For some years, using the tri-brominated monoimide was the most common procedure. That was the case because bromines in bay position react easily with alcohol groups in basic media29 (sometimes by a copper(I) catalysed reaction) and the peri position bromine does not react. In contrast, C-C coupling through Pd catalysed reactions or a reaction with an amine in basic media,30 occur in bay and peri position. Despite the apparent advantages of the tri-brominated product, during the last years the brominated PMI derivatives that people work with have changed, from the tri-brominated to the mono-brominated product. For a while, tri-brominated PMIs were synthetized as they were a pure one-step major product from strong bromination conditions, until some authors reported the unspecificity of the reaction.31 When using strong conditions, like the ones specified in Figure 16 up, the PMI is not only tri-brominated but tetra-brominated too; what is more, the tri-bromination is not specific, but a mixture of isomers. In addition to this fact, the process of purification is tedious and very complex, needing several days, columns and crystallizations to separate the tri-brominated isomers. In fact, it is worth to remark that after the report about the existence of the isomers, the number of papers published working with tri-brominated PMIs decreased dramatically. Nowadays, most researchers work with mono-brominated PMIs. However, it is true that there would be many possibilities for obtaining derivatives with groups in bay position. A possibility, would be leaving the purification until the bromine is substituted (Figure 17). Performing purification in later steps is likely to be a more efficient method, faster, with better yields and less waste (solvents and other purification material). In fact, it is likely to be the method that it has been followed by most authors, although not specified. 29 P. Shao, N. Jia, S. Zhanga, M. Bai, Chem. Commun. 2014, 50, 5648-5651. 30 T. Dentani, K. Funabiki, J.-Y. Jin, T. Yoshida, H. Minoura, M. Matsui, Dyes Pigm. 2007, 72, 303-307. 31 A. Keerthi, Y. Liu, Q. Wang, S. Valiyaveettil, Chem. Eur. J. 2012, 18, 11669–11676.
CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES|129 JoséGarcíaCalvo|PhDThesis Figure 17. Schematic synthesis of a tri-substituted pure PMI. Example of the steps that may simplify the purification process. During the development of the thesis only mono-brominated products were used (Figure 16 down), so as to simplify the synthetic routes. The synthesis only required 3 hours in DCM in a ten-fold excess of bromine; an extraction with saturated bisulfite solution and no further purification, obtaining the product as a red powder in a quantitative yield. This reaction was also scalable, tested between 50-500 mg.
130|CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES JoséGarcíaCalvo|PhDThesis 3.3. Reimidization of PMIs Figure 18. Synthesis scheme of reimidization process in a PMI. Reimidization procedure of PMIs32 and PDIs is widely studied in literature; it consists of a saponification treatment followed by an imidization process. For different reasons, the imide group may be substituted by a more appealing group. In a representative reaction, our starting material consisted of the N-1-adamantylethyl-PMI that was changed by a group containing a boc-protected amine (Figure 18). The free amine group was useful for the afterwards anchoring reaction to other species, such as a group with water affinity or a polymeric material (Chapter 3C). On the other side, the presence of a group (R) in peri (Figure 18) is not casual. Having bulky groups in different positions (it could be also in ortho or bay) may increase the solubility of the monoanhydride intermediate and exclude the tendency of perylene derivatives to stack and precipitate. For instance, a bromine group in peri was enough to obtain the product in high yield. There are some limitations, for instance, the R group must resist the saponification conditions. Everything considered, the synthesis is straightforward. The PMI is dissolved in t-BuOH, next NaOH (or KOH) was added in high excess to the solution (50 equivalents) and left under stirring at 80ºC for 15 hours. After that, the mixture was quenched with high excess of glacial acetic acid and after one hour under stirring it was filtered and washed several times with methanol. It allowed to obtain the monoanhydride in quantitative manner. In the second step, the monoanhydride was transformed into monoimide by following a classical procedure for imidization. The monoanhydride was mixed with imidazole with a 5% mol of zinc acetate and heated at 100ºC. Once the imidazole became liquid and the monoanhydride was dissolved, the amine was added and the mixture was left under stirring for 2 hours at 100ºC. The residue was dissolved in DCM and washed with water (several times). After column chromatography, the product was obtained in high yield >80%, which could vary depending on the particular PMIs (starting reagent and product). 32A.Bolag,N.Sakai,S.Matile,Chem.Eur.J.2016,22,1‐10
CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES|131 JoséGarcíaCalvo|PhDThesis 3.4. Borylation of PMIs Suzuki-Miyaura couplings are one of the most common procedures in order to create aryl-aryl bonds in peri position. To perform the reaction it is necessary to have an aryl halide and an arylboronate (or boronic acid). Taking this into account, some reactions may require having the boronic acid/ester of the perylene either because of the impossibility to have it in the other reagent or to increase the general yield of the coupling reaction. The procedure is similar to the general preparation of aryl boronates, and there is plenty of literature about it.33 (Figure 19) Figure 19. Synthesis scheme for borylation of PMIs. As it is schematized in Figure 19, the brominated reagent was dissolved in 1,4-dioxane (around 10 mM) under nitrogen atmosphere. After that, the palladium(0) catalyst was added to the solution (5% mol) and the mixture was refluxed for 12-24 hours. The product obtained was purified (usually DCM:MeOH mixtures, 1–5 %) by column chromatography, obtaining yields around 50%, being a process that competes with the homo-coupling. 3.5. Introduction of groups in peri position by Suzuki reaction The Suzuki reactions have been performed following the classic conditions previously reported in literature. The most important issue when working with perylene derivatives is that the compound might not be soluble in some solvent mixtures because of the low solubility of PIs (especially when they have an imide group which is not bulky enough), which decreases the final yield. Several synthetic conditions were tested for different synthetized PMIs. As a general procedure, see Figure 20. Figure 20. General scheme for a Suzuki reaction in PMIs. The brominated derivative, boronic acid derivative and the base (Na2CO3, 5 equivalents) were added to a flask under nitrogen atmosphere, then they were dissolved in a mixture of solvents (see below) until a concentration 1-5 mM and left under reflux for 16-24 hours. The product was purified by column chromatography (by using mixtures DCM:MeOH from 0 to 5 %) obtaining products from red to purple, depending on the substituent. 33Seeforexample;P.Shao,M.Bai;Chem.Commun.,2012,48,9498‐9500.
132|CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES JoséGarcíaCalvo|PhDThesis From the multiple possibilities when choosing the solvents, three different mixtures were tested for PMIs. Using THF:water (8:1), is one of the most common mixtures used for performing Suzuki coupling reactions, such as the ones described Chapter 1. However, it does not work properly with most PMIs, probably because of the low solubility. Using a mixture Toluene:BuOH:H2O (4:1:0.4). This mixture has been used by our research group when working with many different products, including PMIs. It works for a wide variety of reagents (all that we have tested), however, the yield is usually lower than when optimized with other solvents. Mixture DME:water (2:1). It has been tested with some PMIs giving good yields, similar or better to the previous one. There are plenty of conditions apart from the ones already explained. However, the performed reactions lead to yields equal or superior to 40 % and they were not further optimized.27
CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES|133 JoséGarcíaCalvo|PhDThesis 4. RELATION STRUCTURE-LUMINESCENCE OF THE SYNTHETIZED PMI DERIVATIVES Modification of peri and bay positions in PMIs are used for tuning their optical properties.34 Previous to the introduction of specific details and applications for particular compounds in Chapter 3, this section explains and compares some details related to the absorbance-fluorescence properties of the synthesizing PMI derivatives. The purpose of this study is to show how the solvent and the structure affects each other when changing them, which serves as an explanation for some of the possible and potential uses they may have, such as the fluorescent sensors that have been developed. The study was divided into three sections. First, the influence of the solvent was studied for the different PMI derivatives. Starting materials: which includes PMIs JG62, JG73, JG75, JG2L and JG7L. PMI derivatives with phenyl, pyridine and pyrimidine piperazine substituents: JG125, JG116, JGphen, JG125d, JG117 and JGphend. Probes for K(I), Pb(II) detection and cellular imaging: JG76, JG121, JG119c1 and JG119c2. PMI-Ru(II) derivatives: JG10L and JG11L. Secondly, there was an especial case that is worth mentioning, the synthesis and testing of a combination of Bodipy-PMI probe. Their properties are explained as introduction for future research in FRET systems giving the molecule special properties, mentioned in Section 4.2. Finally, a table with a comparison between molar absorptivity, fluorescence quantum yields and fluorescence lifetime decay is presented and several conclusions are explained in this regard. 34C. Li, J. Schöneboom, Z. Liu, N. G. Pschirer, P. Erk, A. Herrmann, K. Müllen, Chem. Eur. J. 2009, 15, 878 – 884.
134|CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES JoséGarcíaCalvo|PhDThesis 4.1. Influence of solvent on the synthesized PMIs 4.1.1. Solvatochromism of starting materials (Figures 21-25): Figure 21. Molecular structure of PMIs JG62, JG73, JG75, JG2L and JG7L. Figure 22. Normalized spectra of JG62, JG73 and JG75, 10 μM in different solvents. Normalized absorbance Normalized fluorescence JG6210μM λexc=475nm JG7310μM λexc=475nm JG7510μM λexc=475nm 500 550 600 650 700 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane MCH 500 550 600 650 700 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane MCH 400 450 500 550 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane MCH 400 450 500 550 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane MCH 400 450 500 550 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane MCH 500 550 600 650 700 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane MCH
CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES|135 JoséGarcíaCalvo|PhDThesis Figure 23. Normalized spectra of JG2L and JG7L, 10 μM in different solvents. Figure 24. Picture under UV light of JG62, JG73 and JG75, 10 μM in different solvents. Normalized absorbance Normalized fluorescence JG2L10μM λexc=475nm JG7L10μM λexc=475nm 400 450 500 550 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane MCH 500 550 600 650 700 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane MCH 400 450 500 550 0.0 0.2 0.4 0.6 0.8 1.0 Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane MCH 500 550 600 650 700 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane MCH
136|CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES JoséGarcíaCalvo|PhDThesis Figure 25. Pictures under UV light of JG2L and JG7L, 10 μM in different solvents. The properties and conclusions obtained may be summarized as follows: The effect of changing solvents on the absorbance was basically the same for all the probes, and not very remarkable for any of them. The maximum of absorbance showed a batochromic shift with polarity of around 20 nm from most polar solvents (MeOH) to aliphatic solvents (MCH). Fluorescent emission was more affected by changes in polarity than absorbance, it also has a batochromic shift with polarity, but higher, around 50-60 nm for the same range of solvents. None of them were soluble in water. JG75 had red fluorescence in solid state. 4.1.2. PMI derivatives with piperazine substituted phenyl, pyridine and pyrimidine substituents (Figures 26-30): N O O N N NN O ON O O N N N HN N O O N NN O O N O O NN O ON O O NHN N O O N NHN JG116 JG117 JGphen JGphend JG125 JG125d Figure 26. Molecular structure of PMIs JG125, JG116, JGphen, JG125d, JG117 and JGphend.
CHAPTER2.PERYLENEMONOIMIDES.INTRODUCTIONANDGENERALPROPERTIES|137 JoséGarcíaCalvo|PhDThesis Figure 27. Normalized spectra of JG125, JG116 and JGphen, 10 μM in different solvents. Normalized absorbance Normalized fluorescence JG12510μM λexc=511nm JG11610μM λexc=508nm JGphen10μM λexc=515nm 400 450 500 550 600 650 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane CH 550 600 650 700 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) DMSO Acetone EtOAc THF DCM Toluene Et2O Hexane CH 400 450 500 550 600 65 0 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane CH 400 450 500 550 600 65 0 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane CH 550 600 650 700 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane CH 550 600 650 700 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission intensity Wavelength (nm) MeOH DMSO DMF MeCN Acetone EtOAc THF CHCl3 DCM Toluene Et2O Hexane CH
240|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis Figure 37. JG76 in ethanol, in a mixture with a buffer solution of different pH (20 µM), buffer 20 mM HEPES. Under visible and 366 nm UV light. (λexc=500 nm). In conclusion, the probe was very sensitive to the pH of the solution, especially when far from buffer regulation. A buffer solution, with pH higher than 7, was enough to avoid the pH effect in the fluorescence. 8.2.6. JG76 with ions at controlled pH Due to the previous results, it was necessary to check the behaviour of the probe in buffer solution and in the presence of cations. The compound JG76 was first dissolved in EtOH, from which the final solution was [JG76] = 20 µM in 70 % EtOH - 30 % H2O (v/v) buffer solution, 20 mM of HEPES. The buffer pH was 7. A picture was taken by adding 5 equivalents of different species (100 μM), Figure 38. Additionally, the increase in emission was also registered (Figure 39). Figure 38. JG76 in ethanol-water mixture (30% water, 20 µM), buffer solution of pH 7, 20 mM of HEPES. Addition of different cations (0.1 mM), under visible and 366 nm UV light.
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|241 JoséGarcíaCalvo|PhDThesis 550 600 650 700 0 100 200 300 400 500 600 700 800 900 Emisssion intensity (a.u.) Wavelength (nm) Water Fe(ClO4)3 Be((NO3)2 Zn(ClO4)2 Sn(ClO4)2 Al(ClO4)3 Pb(ClO4)2 Ba(NO3)2 K(CF3SO2) Zn(ClO4)2 Al(ClO4)3 Fe(ClO4)3 Cu(ClO4)2 Be((NO3)2 Sn(ClO4)2 KCF3SO2 Ba(NO3)2 Pb(ClO4)2 01234567 Figure 39. JG76 in ethanol-water mixture (30% water, 20 µM), buffer solution of pH 7, 20 mM of HEPES. Addition of different cations, 0.1 mM. Fluorescence spectra (left) and increase in total emission (right). λexc = 500nm. λem = 571 nm. Fluorescence of JG76 in 70 % EtOH - 30 % H2O (v/v) buffer solution increased selectively for K+, Ba2+ and Pb2+: Around 30 % in the presence of Lewis acid cations. Around 130 % in the presence of K+. Around 350 % in the presence of Ba2+. Around 500 % in the presence of Pb2+. 8.3. Analyte detection comparison, JG76 against JG103 JG76 working conditions were optimized in ethanol, qualitatively and quantitatively. Moreover, to improve selectivity, 30 % of HEPES buffer in water at pH 7 lead to no response when in presence of acidic cations. The solution is selective to K+, Pb2+ and Ba2+. JG103 was not good for quantitative measurements. It possessed low solubility in water miscible solvents and the selectivity was not better than for JG76, (See Experimental Appendix 5) In consequence, quantitative measurements, such as thermodynamic equilibrium constants, stoichiometry calculation or limits of detections were only performed for probe JG76.
242|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis 1234567 0 25 50 75 100 125 150 Emission intensity 573 nm (a.u.) [JG76] (M) 123456 0.025 0.050 0.075 0.100 0.125 0.150 0.175 Absorbance 500 nm [JG76] (M) 9. DETECTION TESTS FOR K+, Pb2+ AND CEREULIDE IN SOLUTION First, some solutions of the probe were prepared in ethanol and the absorbance (Figure 40) and fluorescence (Figure 41) were checked at high dilution. Absorbance at 500 nm: [JG76] (µM) Absorbance [JG76] (µM) Absorbance 6.5 0.182 2 0.046 4.8 0.126 1.6 0.044 3.6 0.09 1.2 0.026 2.8 0.06 0.8 0.006 Figure 40. Absorbance at different concentrations of JG76. Fluorescence (λexc = 500 nm, λem= 573 nm): Figure 41. Fluorescent emission at different concentrations of JG76. In conclusion, the absorbance and fluorescence change linearly between 1.6 to 6.5 µM. The chosen concentration for the tests was between 2 µM to 6 µM. 9.1. JG76 Job’s Plot, stoichiometric determination of the complex A group of solutions was measured with a molar fraction between 0-1 of cations/JG76. The fluorescence was measured with λexc = 500 nm and λem = 571 nm. The molar fraction of cations (Xc) was represented versus the peak of emission (F) minus the emission when XAnalyte = 0, (F0) multiplied per the molar fraction (Xc). Xc vs Xc (Fo-F). The Job’s Plot analysis was performed to estimate the stoichiometry of the complex; taking into account the limited range of application of the technique explained in Chapter 0. The most likely stoichiometry of the complex was easily deduced for JG76:K+ (Figure 42) and JG76:Pb2+ (Figure 43). [JG76] (µM) Em. Int. 573 nm (a.u.) 6.5 140.7 4.8 98.5 3.6 66.3 2.8 53.0 2 34.8 1.6 28.2 1.2 34.5 0.8 33.1 y=0.0289x–0.0108 R²=0.988 y=23.06x‐11.62 R²=0.996
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|243 JoséGarcíaCalvo|PhDThesis 0.2 0.4 0.6 0.8 200 300 400 500 600 700 X K (F-F o ) X K 0.2 0.4 0.6 0.8 300 400 500 600 700 800 EtOH BnOH X K (F-F o ) X K Figure 42. Job’s Plot of JG76:K+ complex, fluorescence analysis in EtOH (left) and EtOH and BnOH comparison (right) The plot was represented several times as it is shown in Figure 42, obtaining always the maximum centred in 0.5, which meant that the complex JG76:K+ was 1:1. 0.20.30.40.50.60.70.80.9 300 400 500 600 700 800 900 X Pb ( F-Fo ) X Pb Figure 43. Job’s Plot of JG76/Pb2+ complex, fluorescence analysis in EtOH. In the case of the complex JG76:Pb2+, which was represented in Figure43, the stoichiometry was also 1:1. 9.2. Fluorescence Quantum Yields (ΦF) Fluorescent quantum yields were determined by using an integration sphere (see example in Figure 44), this is an absolute method, as it was explained in Chapter 0.
244|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis Figure 44. Example of the measurement of JG76 in EtOH solution with the software from FLS980 Edinburgh instrument to measure a quantum yield. Keeping in mind that the error associated to the method is at least 1 % the samples were repeated several times obtaining variations in the results around this number as it is shown in Figure 45. EtOH ΦF JG76 K+ Pb2+ error ΦF,K+/ ΦF,0 ΦF,Pb2+/ ΦF,0 0.12 0.48 0.84 0.02 4 7 BnOH ΦF JG103 K+ Oxone* error ΦF,K+/ ΦF,0 0.01 0.05 0.13 0.01 5.2 MeCN ΦF JG103 K+ error --- 0.14 0.01 *Oxone(Potassium peroxymonosulfate) behaves as acid and contains K+ at the same time. Figure 45. Table of quantum yields of JG76 and JG103 in different solvents and in presence of different species. 9.3. Fluorescence decay lifetimes (τ) Fluorescence decay lifetimes were measured using a time-correlated single photon counting instrument (FLS980 Series, Edinburgh instruments) with a 510 nm pulsed LED (Edinburgh instruments, EPL-510) light source having a 177.4 ps. Decays were recorded at 510 nm for each probe. The probes JG76 and JG103 were tested in the studied solvents and in presence of lead and potassium cations. Calculating the lifetime decays from the graphs in Figure 46 and represented in Figure 47.
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|245 JoséGarcíaCalvo|PhDThesis 0 1020304050 0.00 0.25 0.50 0.75 1.00 IRF JG76 EtOH Normalized photon counts Deca y time , ( ns ) 10 20 30 40 50 0.0 0.2 0.4 0.6 Normalized photon counts (ns) 10 20 30 40 50 1.0 1.5 2.0 2.5 3.0 3.5 Photon counts (ns) 10 20 30 40 50 1.0 1.5 2.0 2.5 3.0 Photon counts (ns) 10 20 30 40 50 1.0 1.5 2.0 2.5 3.0 3.5 Photon counts (ns) Figure 46. Lifetime decay fitting of JG76 dissolved in EtOH, IRF and fitting (up); JG76 in Acetone (middle left), JG103 BnOH (middle right) and JG103 acetone (down). Probe Solvent τ (ns) χ2 JG76 EtOH 3.58 1.138 JG103 BnOH 3.22 1.192 JG76 Acetone 3.95 1.075 JG103 Acetone 4.22 1.099 Figure 47. Lifetime decay, JG76 and JG103 in different solvents. Both JG76 and JG103 had lifetime decays between 3-4.5 ns. In the same solvents, the fluorescence decay lifetime was very similar comparing JG103 and JG76, due to the perylene structure, responsible of the fluorescence. Measuring with and without potassium/lead cations no change in τ was observed. JG76EtOH τ=3.58ns Χ²=1.138 JG76Acetone τ=3.95ns Χ²=1.075 JG103BnOH τ=3.22ns Χ²=1.192 JG103Acetone τ=4.21ns Χ²=1.099
246|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis 9.4. Thermodynamic equilibrium constants 9.4.1. Measuring parameters and method: Equilibrium constants were measured for the complexes with JG76. By preparing a solution 2-5 µM of JG76 and increasing the concentration of K(CF3SO3) or Pb(ClO4), without changing the concentration of JG76 in solution. 9.4.2. Equilibrium constant of K+ and Pb2+ with JG76: JG76-K+ complex was 1:1, in agreement with literature and the Job’s Plot results. Additionally, the experimental results for lead cations were similar. In consequence, the equation to calculate thermodynamic equilibrium constants was the one explained in Chapter 0. 𝐼𝑓𝐶 𝐶𝐶 𝐶𝐶 4𝐶𝐶 ꞏ Equation [1] Being Cp the concentration of JG76, CA the concentration of the analyte and fP/fPA the proportional fluorescent factors of JG76 and the complex JG76-analyte, respectively. The fluorescence titration was performed under a constant concentration of probe and the fluorescence values fitted by nonlinear least squares regression,59 starting in an initial value of K1, fP and fPA and calculating the value of K1 and the error associated by iteration, Figure 48 for JG76+K+ and Figure 49 for JG76+Pb2+. 02468 0 500 1000 1500 2000 Emission intensity 571 nm (a.u.) [KCF 3 SO 3 ] ( M) 0246810 0 500 1000 1500 2000 2500 S (EtOH) S (BzOH) Max. Emission intensity (a.u.) KCF 3 SO 3 ( M) Figure 48. Fitted fluorescent emission of a titration with K(CF3SO3) of a 2 µM solution of JG76 in EtOH (left) and EtOH and BnOH compared (right). The calculation of the complexation constants was repeated 3 times for each solvent. Besides, the results were compared by doing a titration of K+ with JG76 and JG76 with K+. The same results were obtained. Ethanol solution: K (JG76+K+) = (2.2 0.1) ×106 M-1 Log K(JG76+K+) = 6.34 0.03 Benzylic alcohol solution: K (JG76+K+) = (1.3 0.1) ×106 M-1 59 The software for the calculation was Origin v2016.
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|247 JoséGarcíaCalvo|PhDThesis Log K(JG76+K+) = 6.11 0.03 0 5 10 15 20 25 30 1000 2000 3000 Emission intensity 585 nm (a.u.) [Pb(ClO 4 ) 2 ] (M) Figure 49. Fitted fluorescent emission of a 2 µM JG76 solution in EtOH titrated with Pb(ClO4)2. The fitting calculation of the complexation constants was repeated 3 times: K (JG76+Pb2+) = (1.55 0.1) × 106 M-1 Log K(JG76+Pb2+) = 6.19 0.02 9.4.3. Equilibrium constant Valinomycin-K+ calculated by circular dichroism: There are some procedures described in literature to measure the thermodynamic equilibrium constant of natural potassium ionophores, such as valinomycin.60 Due to the chiral properties of these cyclic depsipeptides, circular dichroism is the one taken as standard for the thermodynamic constant calculation of these cyclic depsipeptides. Once a complex between the peptide and a cation is formed, the deviation of polar light may be measured. In a reference experiment, the measurements were performed starting from a concentration of 0.3 mM of valinomycin; then, the equivalents of potassium were gradually increased until 3.75 equivalents (1.13 mM), see Figure 50. 210 220 230 240 250 260 -2000 0 2000 4000 6000 8000 10000 12000 14000 [ ], deg cm2 dmol-1 , nm Figure 50. Circular dichroism of a valinomycin 0.3 mM solution in ethanol, increasing the concentration of potassium cations from 0 to 1.13 mM. 60M. C. Rose, R. W. Henkens, Biochim. Biophys. Acta 1974, 372, 426−435 0eq.K+ 3eq.K+
248|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis The concentration of valinomycin was chosen because of the optimal concentration to work with the dichroism signal. Although the results were, apparently, the same explained in literature, it turned out that, checking the fitting at different wavelengths, the constant seemed to change depending on the wavelength. Adjusting between 230 nm-250 nm the graphs in Figure 51 were obtained: 02468 0.000 0.002 0.004 0.006 0.008 0.010 230 , deg [K + ], 10 4 M 02468 0.000 0.002 0.004 0.006 0.008 232 , deg [K + ], 10 4 M 02468 -0.002 0.000 0.002 0.004 0.006 0.008 234 , deg [K +], 104 M 02468 -0.004 -0.002 0.000 0.002 0.004 0.006 238 , deg [K +], 104 M 02468 -0.004 -0.002 0.000 0.002 0.004 , deg [K +], 104 M 242 02468 -0.002 -0.001 0.000 0.001 0.002 246 , deg [K + ], 10 4 M 02468 -0.0010 -0.0005 0.0000 0.0005 0.0010 250 , deg [K + ], 10 4 M Figure 51. Molar ellipticity of a 0.3 mM solution in ethanol of valinomycin, increasing the concentration of potassium cations from 0 to 1.13 mM at different wavelengths (230-250 nm).
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|249 JoséGarcíaCalvo|PhDThesis The results could be represented by the same equation used for fluorescence which allowed to calculate the thermodynamic constant exposed in Figure 52. λ. (nm) K × 10-5 (M-1) Log K 230 10.3 6.01 232 8.62 5.94 234 8.07 5.91 238 6.55 5.82 242 5.88 5.77 246 6.33 5.8 250 8.51 5.93 Figure 52. Table of equilibrium constant valinomycin-K+ calculated at different wavelengths. Although in literature the given equilibrium constant was calculated at 238 nm, and taken as independent from the wavelength, it turned out to be slightly dependent on it. This characteristic was observed in literature when the complexation occurs in DNA structures (different binding sites),61 but it was not the case. Therefore, for valinomycin, it was understood as a consequence of the limitations of the method. In conclusion, the equilibrium constant in pristine ethanol was between (0.6 to 1.0) ×106 M -1 or (LogK) between 5.8 to 6.0; calculated with a 0.3 mM of valinomycin. 9.4.4. Valinomycin and Cereulide derivatives, equilibrium constant calculation by fluorescence: By using the indirect method explained in Chapter 0, the equilibrium constants were calculated by fluorescence. The procedure is schematized in Figure 53. Figure 53. Scheme for the titration of the complex Valinomycin-K+ with JG76 to calculate the complexation constant Valinomycin-K+. The experiment started with a solution in ethanol of Valinomycin or Cereulide (V) and Potassium cations (K), giving a complex between them (VK) in equilibrium. When the probe (S) was added, it formed a complex with free K+ creating the new complex (SK) and replacing the previous complex (VK). In conclusion, the concentration of VK decreased, whereas the concentration of V increased. Then, the approximation used was: CV - [VK] CV 61P. Kumar, R. Barthwal, Biochimie, 2018, 147, 153-169.
256|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis 9.5.4. Limit of detection of Cereulide: In a solution 2 µM of JG76, the concentration of K(CF3SO3) followed the proportions [JG76]/[K+] = 1.25. The concentration of cereulide was increased in several additions, and the fluorescence spectra were registered (Figure 64). 0 1020304050607080 200 400 600 800 1000 1200 Emission intensity 571nm (a.u.) Cereulide concentration ( M) Figure 64. Regression of a titration with cereulide of JG76 2 µM and K+ 1.5 µM in EtOH studying the decreasing fluorescent emission. λexc = 500 nm The detection limit calculated for cereulide was 0.21 µM or 240 ppb. 9.5.5. Summary, LODs using JG76 in EtOH: EtOH LOD (μM) LOD (ppb) [JG76]/[K+] K+ 0.06 2.3 - Pb2+ 0.03 6 - Valinomycin* 0.54 600 1.25 Cereulide* 0.21 240 1.25 *JG76 concentration between 25 μM Figure 65. LODs of K+, Pb2+, valinomycin and cereulide calculated in EtOH solution with probe JG76, quantities experimentally measured with 5 % of false positive/negative. y=‐257x+1100 R²=0.979 0.1 0.2 0.3 0.4 0.5 950 975 1000 1025 1050 1075 Emission intensity 571 nm (a.u.) Cereulide concentration (M)
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|257 JoséGarcíaCalvo|PhDThesis 10. MEASUREMENTS OF EXTRACTED CEREULIDE SAMPLES BY FLUORESCENCE JG76 was very sensitive and selective to potassium cations, with a complexation constant even higher than cereulide (in EtOH solution), 2.2 × 106 M-1 against 0.9 × 106 M-1. Therefore, it was chosen as fluorescent probe for detection of cereulide extracted from rice extracts. 10.1. Culture and extraction of natural cereulide62 The extraction of the natural cereulide was performed from cultures of B. cereus F4810/72 strains (see Figure 66), following the methodology developed for cooked rice.63 Briefly summarized, it consisted of the inoculation of rice with 300 CFU (Colony forming units), average value found in rice dishes. The CFU was determined at several time points, along with cereulide production, by UPLC-TOF-MS. The extraction proceeded by using acetonitrile and several purifications, involving heating, extraction and centrifugation to obtain cereulide in a concentration around 0.2 to 3.5 µM in acetonitrile. Figure 66. Cereulide extracts were placed in PCA culture (Plate count agar) before (left) and after (right) centrifugation to evaluate the presence of bacterial spores. No spores were detected after centrifugation. 10.2. Fluorescence measurements of natural cereulide extracts a) Conditions: Cereulide extracts were provided from rice cultures and extracted in acetonitrile with a concentration calculated by HPLC analysis. The idea was to measure the change in fluorescence in presence of: A constant concentration of JG76. A constant concentration of potassium cation. 62The procedure was performed by Wilson Antunes at Laboratório de Bromatologia e de Defesa Biológica (LBDB) do Exército, Lisboa, Portugal. 63a) A. Z. Muratovic, R. Tröger, K. Granelli, K.-E. Hellenäs, Toxins 2014, 6, 3326-3335; b) M. Yamaguchi, T. Kawai, M. Kitagawa, Y. Kumeda, Food Microbiol. 2013, 34, 29-37; c) M. Decleer, A. Rajkovic, B. Sas, A. Madder, S. De Saeger, J. Chromatogr. A 2016, 1472, 35–43.
258|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis The sample of cereulide, concentration unknown, was titrated by adding synthesized cereulide or valinomycin. These samples were provided in acetonitrile, which had concentrations between 0.2 - 3.5 µM. b) Measuring directly from extracted samples: First of all, the easiest way to perform the measurements would have been by making a solution of the samples and studying the effect of increasing cereulide concentration. This straightforward method led to high fluorescent results without response to increasing potassium cations or cereulide. To explain this fact, several reasons were suggested: The cereulide samples contained an unknown concentration of potassium or species that acted as Lewis acid. Because of that, high initial concentrations of potassium or some Lewis acids led to inaccuracy in the determination of cereulide. The cereulide samples had matrix contribution, that interfered in the measurements; therefore, there was a background fluorescence from the matrix, which may be easily eliminated by subtraction of the fluorescence when there was no probe JG76. It can be other ways to interfere in the measurements. If that was the case, the matrix would need to be completely removed before measuring. All these issues were tested and solved when possible. c) Eliminating the excess of potassium in solution: Due to the higher solubility of cereulide in organic solvents, the solution was extracted by liquid-liquid extraction (DCM-Water); due to it, the effect of water-soluble interferents was prevented. Finally, the sample was evaporated and redissolved in EtOH, avoiding the interference of any watersoluble ions before the titration. To check if the cereulide remained dissolved after the extraction, the final solution in EtOH solution was measured by UPLC to compare results, verifying its presence, Figure 67. The UPLC method of calculation was: Studying the elution time of synthetic cereulide. Measuring the intensity of a sample of extracted cereulide. Making a calibration with synthetic cereulide at different concentrations. Figure 67. UPLC chromatogram of cereulide samples; Synthetic cereulide (left) and extracted sample (right). The extracted cereulide samples were purified and concentrated in EtOH to a final solution containing around 1.2 µM of cereulide. Cereulidepeak Cereulidepeak
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|259 JoséGarcíaCalvo|PhDThesis d) Measuring the fluorescence of extracted samples: It must be taken into account that, obtaining cereulide from cultures was a difficult task that not always ends up with a growing culture of the bacteria. In consequence, it usually finished by having solutions of a few milliliters, and many times with concentrations around 1 µM or less. Due to this fact, the stock of solution was very low. To perform the measurements, after the extraction the variation in the signal of emission was measured, similarly to a regular standard addition analysis. Cereulide was added to a solution of [JG76] = 2 µM and [K+] = 0.75 µM and the fluorescent response was compared between pure ethanol solution and extract solution. 0 102030405060708090100110 100 200 300 400 500 600 700 Reference Rice sample Emission intensity 571 nm (a.u.) Cereulide concentration (M) Figure 68. Emission intensity of rice sample (red) and reference (black) with increasing quantities of cereulide. The concentrations were [JG76] = 2 µM and [K+] = 0.75 µM. λexc = 500 nm and a λem = 571 nm. Figure 68 shows a comparison and results with cereulide samples and EtOH in which synthetic cereulide was added. Sample and reference were done several times the same day in order to minimize the experimental error. e) Interpretation of the results for a representative experiment: Before obtaining any conclusions from fluorescence, one of the most important factors that had influence over the results was the matrix of the rice sample rice. This matrix was supposed to be a mixture of proteins that may affect the measurements. The only possibility to make reliable fluorescence quantification, or at least an approximation of the presence of cereulide in the samples, was to check how the emission of the probe was affected by the matrix, when the concentration of probe (K+ constant) and the concentration of potassium (probe constant) changed. To do so, the fluorescence of the matrix was measured without adding JG76 and after adding different concentrations of JG76, to discard interferences matrix-probe, Figure 69.
260|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis 0.00.51.01.52.0 0 250 500 750 1000 Rice sample Ethanol Emission intensity (a.u.) JG76 concentration (M) Figure 69. Comparison between ethanol and matrix solutions increasing JG76 concentration in ethanol rice samples and pure ethanol solution. λexc = 500 nm, λm = 571 nm. In conclusion, the changes in fluorescence were because there was a background fluorescence, which can be calculated when [JG76] = 0, and it’s barely affected by [JG76] at the work concentration. f) Influence of the matrix correction for the determination of cereulide concentration The background was subtracted from the titration. Then, due to the quantity of cereulide, the initial value of fluorescence, showed an approximation of the cereulide quantity in the sample. (See examples in Figures 70, 71 and 72) 0 20406080100 100 200 300 400 500 600 700 Reference Rice sample - Background Emission intensity (a.u.) [Cereulide] (M) Figure 70. Fluorescent emission of a rice sample versus the reference by titration with increasing quantities of cereulide. The concentration of JG76 was 2 µM, and concentration of K+ was 0.75 µM. Cereulide in the rice sample was 1 µM. λexc = 500 nm, λm = 571 nm. y=519.81x+16.507 R²=0.9978 y=528.54x+78.63 R²=0.9989 012345 300 400 500 600 700 Reference Rice sample - Background Emission intensity (a.u.) [Cereulide] (M)
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|261 JoséGarcíaCalvo|PhDThesis 0 1020304050607080 100 200 300 400 500 600 700 Reference Synthetic Cereulide Emission intensity (a.u.) [Cereulide] () Figure 71. Fluorescent emission of a rice sample spiked with cereulide 1.75 µM, versus the reference by titration with increasing quantities of cereulide. The concentration of JG76 was 2 µM, and concentration of K+ was 0.75 µM. λexc = 500 nm, λm = 571 nm. 0 1020304050607080 100 200 300 400 500 600 700 Reference Sample Rice Sample - Background Emission intensity (a.u.) [Cereulide] (M) Figure 72. Fluorescent emission of a rice sample with no cereulide versus the reference by titration with increasing quantities of cereulide. The concentration of JG76 was 2 µM, and concentration of K+ was 0.75 µM. λexc = 500 nm, λm = 571 nm. Three cases summarize the possibilities: Spiked samples (Figure 71) showed results in agreement with the cereulide added. The extracts gave an average of 1.0 ± 0.2 µM compared to a sample that was checked to be 1.2 µM by mass spectrometry (Figure 70). A rice sample, without containing cereulide, gave no difference in the titration (Figure 72). 012345 200 300 400 500 600 700 Reference Synthetic Cereulide Emission intensity (a.u.) [Cereulide] () 012345 300 400 500 600 700 Reference Sample Rice Sample - Background Emission intensity (a.u.) [Cereulide] (M)
262|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis For the calculation of cereulide concentration it would not be necessary to perform the whole titration curve to make the calibration, however, it was advisable to ensure that the procedure was reliable. Additionally, the process had to be repeated several times to be reliable. g) Summary about how to process data: Once the method was validated it was simplified to 5 steps: The cereulide is extracted from the culture by the optimized procedure.62 Acetonitrile solutions are evaporated, extracted in DCM:water and redissolved in ethanol solution. The fluorescence of the background is registered. The probe JG76 and K(CF3SO3) are added (prefixed concentration) to the sample, and the fluorescence intensity is measured. The intensity of the background is subtracted to the intensity of JG76 sample. If the value obtained was not significantly different from a blank, the quantity of cereulide is considered as 0. If it is different, further analysis is necessary. o To obtain a concentration value from fluorescence, it is calculated a linear regression around the result by adding synthetic cereulide or valinomycin. o Mass spectrometry analysis is required for a higher precision in the determination. In this case, cereulide was measured indirectly and by fluorescence. The method was more sensitive than mass spectrometry analysis to interferents, so the conditions were controlled and the results carefully evaluated. In any case, measuring by fluorescence is likely to give different results than UPLC-MS. The elution by UPLC gives peaks not only associated to cereulide but for its many possible derivatives. In contrast, the results from fluorescence were associated to their different potassium affinities; therefore, it introduces uncertainty in the measurements and it would not give the same results than pure cereulide.
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|263 JoséGarcíaCalvo|PhDThesis 11. BIOLOGICAL DETECTION OF POTASSIUM, COMPARISON BETWEEN IONOPHORES 11.1. Potassium transport of cyclic depsipeptides and their comparison In Section 9.4 it was calculated that the different potassium ionophores had different affinity for potassium cations. It is also well known from literature that natural ionophores, such as valinomycin or cereulide, have a direct response involving intra-extracellular equilibrium of potassium (See Section 2). For instance, it was studied that whereas valinomycin acts regulating the equilibria, independently of potassium global concentration, in case of cereulide it destabilizes the cells when the concentration of potassium is low.21 One of the causes of the different behaviours may be associated to small changes in the structure, which affects the potassium transport. In consequence, the possibility of influencing the potassium transport by changing the structure is worthy of being studied. The fluorescent assays were performed by the group of Barboiu at Montpellier (France),64 they used what is called the Fast Filter method, in order to read quasi instantaneously the emission at 510 nm under alternate excitation at 403 and 460nm.65 The experiments were performed in unilamellar vesicles (LUV): Inside the LUV: aqueous solution of 10 mM sodium phosphate, pH 6.4, 100 mM NaCl. Outside the LUV: aqueous solution (1.85 mL) of 10 mM sodium phosphate, pH 6.4, 100 mM KCl. Procedure: Upon initiating the experiment, 0.02 mL of 1 mM compound in DMSO was added to the measurement cell containing HPTS-loaded LUV (8-hydroxypyrene-1,3,6-trisulfonic acid) after 50 s. Then, 29 μL of 0.5 M aqueous NaOH were added after 100 s bringing external pH to 7.4. Finally, maximal changes in dye emission were obtained at the end of each experiment by lysis of the liposomes with detergent (0.04 mL of 5% aqueous Triton x100). 0 100 200 300 400 500 600 0.0 0.2 0.4 0.6 0.8 1.0 Normalized emission Time (s) Blank Valinomycin JG115 JG115B JG100 JG121 Figure 73. Intensity ratio (I460 / I403) over time of solution containing HPTS-loaded LUV followed by a) compound addition (@ 50 sec.) of valinomycin (for reference), DMSO (blank), JG115, JG115B, JG100 or JG121; b) base addition (@ 100 sec.) and finally c) Triton for lysis (@ 100 sec.). All spectra are normalized (max/initial emission) 64 Institut Européen des Membranes, Montpellier, France. The measurements were performed by Yves-Marie Legrand and Li Yuhao. 65 Using a Perkin Elmer fluorometer and measuring at 20ºC. sterichindrance
264|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis JG115 and JG115B, JG100 (Cereulide) and JG121were tested and compared to the reference compound, Valinomycin, Figure 73. The results in terms of potassium transport had the same order than the stability constants: K(Valinomycin) ≈ K(Cereulide) ≈ K(JG115B) > K(JG115) > K(JG121) This fact was really interesting due to the implications in biomedicine, valinomycin is a drug used for controlling processes related with potassium equilibria. Then, a deep study of this kind of ionophores may lead to understand how to modulate transport with different derivatives, which may open a new field for the development of new drugs with straightforward pharmaceutical applications. 11.2. Intracellular measurements Another interesting aspect of the cereulide detection is the visualization of the action of cereulide in living cells. For this purpose, cellular location studies were performed by the group of López-Fanarraga66 in HeLa cells (human cervical carcinoma cells), cultured under standard conditions.67 With the fluorogenic probe JG76. In a combination of JG76 and Cereulide (JG100). With the fluorescent cereulide derivative JG121. Procedure and data: Hela cells were incubated with the probe in the culture medium. Cells were fixed with 4% paraformaldehyde before taking images. The nuclei of fixed cells were stained with Hoechst dye (bisbenzimide), fluorescent in blue, before performing high-resolution confocal microscopy imagining. All confocal cell images were pseudo-coloured. excitation at 488 nm, and emission in green/red/near red. For the images of JG76: (18 µM in 1% DMSO/culture medium (v/v)) – Figure 74 After 12 hours exposure, the probe stained intracellular vesicular structures that resembled endo-lysosomes (Figure 74 upper). After 24 hours exposure, the probe JG76 displayed a pattern clearly localized within cytoplasmic and endosomal membranes (arrows) (Figure 74 middle). After 120 h of staining the probe, JG76 was also localized in the cytoplasmic membrane (green arrow) (Figure 74 lower). HeLa cells did not display detectable toxicity signs when grown in the presence of the potassium ions fluorescent probe JG76 for up to 120 h. 66 M. López-Fanarraga and Eloisa López Lavado from Universidad de Cantabria, Santander (Spain). 67 a) L. Rodriguez-Fernandez, R. Valiente, J. Gonzalez, J. C. Villegas, M. L. Fanarraga, ACS Nano 2012, 6, 6614-6625; b) L. García-Hevia, R. Valiente, R. Martín-Rodríguez, C. Renero-Lecuna, J. González, L. Rodríguez-Fernández, F. Aguado, J. C. Villegasa, M. L. Fanarraga, Nanoscale 2016, 8, 10963–10973; c) B. Sanz, M. P. Calatayud, T. E. Torres, M. L. Fanarraga, M. R. Ibarra, G. F. Goya, Biomaterials 2017, 114, 62-70.
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|265 JoséGarcíaCalvo|PhDThesis 50 m Z projection 10 m Z projection 2 20 m inset 2 5 m 2 20 m inset 2 5 m Figure 74. Confocal microscopy projection images of probe JG76 in HeLa cells 12 h after staining. (Upper). 24 h after staining (Middle), 120 h after staining (Lower). Nuclei were stained with Hoechst dye (blue channel). The different emissions were obtained by exciting the probe sequentially with the 488, 562 and 638 nm lasers. Different fluorophore emissions are pseudo-coloured in their respective wavelengths (green = 500-550; red = 570-620; purple = 662-737 nm). nucleus/JG76 nucleus/JG76 nucleus / JG76 24 h exposure 120 h exposure 12 h exposure
272|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis Figure 85. Composition obtained from the EDX representation of the SEM images of the studied polymers. Elemental proportions (weight) O/C Pb/O Pb/C Theoretical JGIF 0.359 0 0 Theoretical JG151dp 0.343 0.077 0.026 JGIF 0.5 0 0 JGIF + Pb(II) 0.463 0 0 JG151dp 0.499 0 0 JG151dp + Pb(II) 0.442 0.094 0.042 Figure 86. Table of proportions O/C/Pb. Pb was only detected for JG151dp + Pb(II), and the proportions of Pb/O were very close to the theoretical proportions, what is more, no Pb(II) was detected in any of the other samples (Figure 86). However, the method presented important limitations. The samples had to be covered with gold, and the EDX analysis gave relative composition within the penetration depth of the laser (usually around 2 μm). 12.3. JG151dp ions test Pieces of polymer were introduced in different cations solutions in water or buffer solutions (0.5 mL, 50 µM) and pictures were taken (Figures 87, 88, 89 and 90). In the same way observed for molecular probes solutions (JG76), not using buffered solutions led to response to acidic pH and Lewis acids.
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|273 JoséGarcíaCalvo|PhDThesis Figure 87. JG15dp in buffer water HEPES buffer solution (10 mM) at different pH values, under visible (up) and 366 nm UV light (down). In HEPES 10 mM solutions, the response to different analytes was studied: Figure 88. JG151dp in water, in presence of different cations (50 µM) under visible light (up) and under UV light (down). 0.2 mL of pH buffer 7 HEPES 10 mM. The counterions were noncoordinative species like CF3SO3-, ClO4and Clin case of Pd2+. The response to anions was negative by testing F- - Cl- - Br- - I- - BzO- - NO3- - H2PO4- - HSO4- - AcO- -CN- - SCN- , being the cation Bu4N+. Figure 89. JG151dp in water, in the presence of different species (50 µM) under visible light (up) and under UV light (down). Ions sequence: Reference - WaterLi+- Na+ - K+ - Rb+ - Cs+ - Mg2+ - Ca2+ - Sr2+ - Ba2+ - NH4+. Counterions: ClO4- , CO32- (Cs+), NO3- (Rb+) Figure 90. JG151dp in water at buffer pH 7 HEPES solutions (10 mM), changing the pH and in presence of cations at controlled pH 7. Under visible (up) and 366 nm UV light (down). Sequence: Reference - pH 12 - pH 7.2 - pH 1 - K(CF3SO3) - Cs(CO3)2 - Ba(NO3)2 - Pb(ClO4)2 - Sn(ClO4)2 RAg+Ni2+Sn2+Cd2+Zn2+Pb2+Cu2+Fe3+Sc3+Al3+Hg2+Au3+Co2+Pd2+Ir3+ RpH12pH7.2pH1K + Cs 2+ Ba 2+ Pb 2+ Sn 2+ 2 5.5 7 7.4 9 12 RNa+K+Be2+Cs+Ba2+Li+Mg2+OxoneCa2+Sr2+NH4+
274|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis Characteristics of the polymer JG151dp: JG151dp is very sensitive to extreme pH. Cations with Lewis acid behavior result in increased fluorescence if the pH was not controlled. Potassium cations did not increase fluorescence of the polymer, in contrast with the soluble probe JG76. The fluorescence increased under extreme acidic pH or in presence of Ba2+ or Pb2+, therefore, the pH had to be controlled. JG151dp emission was also registered quantitatively in the presence of the most representative species in solution and compared before and after adding lead cation, Figure 91. Ag+ Ni2+ Sn2+ Cd2+ Zn2+ Cu2+ Fe3+ Sc3+ Al3+ Hg2+ Co2+ Na+ K+ Be2+ Cs+ Ba2+ 0.0 0.2 0.4 0.6 0.8 1.0 1.2 (I F /I 0 )-1 Cation Cation + Pb 2+ Figure 91. Fluorescent emission increase in the presence of different cations (50 µM), dissolved in pH 7 HEPES water solutions (10 mM). (λexc = 500 nm, λem = 593 nm). 12.4. Fluorescence quantum yields (ΦF) and decay lifetimes (τ) of JG151dp The polymer fluorescence was measured dry, before and after being in a concentrated Pb(ClO4)2 solution in water, obtaining the next fluorescence quantum yields (calculated with an integration sphere): ΦF JG151dp = 0.22 ± 0.02 ΦF JG151dp + Pb2+ = 0.39 ± 0.02 *The fluorescent response is not the same dry than wet. LODs and the experiments for Pb(II) detection were measured with the sample in water solution. Ag+Ni2+Sn2+Cd2+Zn2+Cu2+Fe3+Sc3+Al3+Hg2+Co2+Na+K+Be2+Cs2+Ba2+
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|275 JoséGarcíaCalvo|PhDThesis The fluorescence lifetime decay of JG151dp was calculated and compared with JG76 dissolved in EtOH (Figure 92): 10 20 30 40 50 0.0 0.2 0.4 0.6 Normalized photon counts (ns) JG76 EtOH Fluorescence Lifetime decay 10 20 30 40 50 1.0 1.5 2.0 2.5 3.0 3.5 JG15dp Fluorescence Lifetime decay Normalized Photon counts (ns) Probe Solvent Lifetime (ns) χ2 JG76 EtOH 3.58 1.138 JG151dp - 4.35 1.151 Figure 92. Lifetime decay, JG76 dissolved in EtOH (left), JG151dp (right). Table of Lifetime decays, JG76 and JG151. (down) The conclusions of the analysis were: Both JG76 and JG151dp had lifetime decays between 3-4.5 ns. It wasn’t observed any change between measuring with and without lead cations. 12.5. Limit of detection of Pb 2+ 12.5.1. With JG151dp in deionized water: 520 540 560 580 600 620 640 660 680 700 0 2 4 6 8 10 12 14 16 18 20 Emission intensity x 10 -7 (a.u.) Wavelength (nm) 0.4 0.6 0.8 1.0 1.2 1.4 15 16 17 18 19 Emission intensity x 10 -7 (a.u.) Concentration (M) Figure 93. Fluorescence spectra of JG151dp in water solution while increasing Pb2+ concentration (left) and linear regression of a titration of JG151dp with Pb2+ in water, studying the increase in the fluorescent emission. (λexc = 500 nm, λem = 580 nm). The limit of detection for Pb2+ with JG151dp in deionized water was 610 nM or 130 ppb. (From Linear Regression in Figure 93) y = 4.02x + 13.32 R² = 0.963 2µMPb 2+ 0.2µMPb 2+
276|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis There were two main issues to take into account from this titration: Pb(II) in water solution precipitated when pH was higher than 7.8. Pb(II) presence affected the media (pH), decreasing the fluorescence when the concentrations of the metal were low. (Figure 94) 0123456 400 450 500 550 600 650 700 Emission intensity at 580 nm (a.u.) Pb 2+ Concentration (M) Figure 94. Fluorescence spectra of JG151dp in water solution while increasing Pb2+ concentration (left) and linear regression of a titration of JG151dp with Pb2+ in water, studying the increase in the fluorescent emission. (λexc = 500 nm, λem = 580 nm). As a consequence, measurements of the presence of Pb2+ were repeated in buffered media. 12.5.2. Limit of detection of Pb2+ with JG151dp in buffer pH 7 The LOD was calculated in HEPES buffered water solution (0.5 mM). 0 5 10 15 20 25 300 400 500 600 700 800 900 1000 1100 Pb(II) Fitted curve Emission intensity at 593 nm(a.u.) Pb 2+ Concentration (M) 0.2 0.4 0.6 0.8 1.0 1.2 420 440 460 480 Emission intensity at 539 nm (a.u.) Pb 2+ Concentration (M) Figure 95. Linear regression of a titration of JG151dp with Pb2+ in HEPES pH 7 solution, studying the increase in the fluorescent emission (λexc = 500 nm, λem = 593 nm). The limit of detection for Pb2+ with JG151dp in deionized HEPES buffer was 290 nM or 66 ppb. (From Linear Regression in Figure 95) y = 0.608x + 405.38 R² = 0.985 0µMPb 2+ 2µMPb 2+ 20µMPb 2+
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|277 JoséGarcíaCalvo|PhDThesis 12.6. Recyclability of the polymer JG151dp The recyclability of the polymer was checked by reusing different pieces of polymer. First, the polymer was placed in a Pb(ClO4)2 solution in water 0.1 mM. The fluorescence before and after being in solution was registered. After that, JG151dp was placed in an EDTA solution, 0.1 M. Then, it was washed up with distilled water (Figure 96). This procedure allowed to use the polymer again having equivalent results. The recyclability was checked seven times by calculating the quantum yield before and after being in presence of Pb(II) cations (Figure 97). No significant change was observed. Figure 96. Scheme of the recycling process for the polymer JG151dp. I1 Pb1 I2 Pb2 I3 Pb3 I4 Pb4 I5 Pb5 I6 Pb6 I7 Pb7 -- 4.0 4.5 5.0 5.5 6.0 6.5 Emission intensity x 10 5 (a.u.) Figure 97. Fluorescence increase of a recycled piece of JG151dp. (λexc = 500 nm, λem = 593 nm)
278|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis 13. DETECTION OF Pb(II) IN COLOURED SOLUTIONS The interests of detecting Pb2+ in solution were numerous: in tap water, in poisoned beverages, mud or balsamic vinegar. In case of colourless tap water, the probe JG76 would be a very sensitive option, because it was able to detect traces of less than 6 ppb, although the solvent should be changed to other but water. However, for measuring other solutions, such as coffee or muddy water, it is not possible to introduce a dissolved fluorescent probe to quantify the cation in solution, because of the intrinsic colour. A solution to this issue was found by the use of polymer supported probes, following the procedure schematized in Figure 98. The polymer was hold between two magnetic surfaces. This polymer was introduced in a distilled water solution and the fluorescence was registered. The polymer was taken from this solution and introduced in the solution contaminated with Pb2+. After 5 minutes in this solution, the polymer was retired from the solution and washed with distilled water. The presence of Pb2+ was detected by measuring the fluorescence in the same conditions than initially. Distilled water Distilled water Pb 2+ containing solution solution with no Pb 2+ Figure 98. Reaction scheme of Pb(II) detection in colored solutions with JG151dp solutions.
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|279 JoséGarcíaCalvo|PhDThesis Solutions tested The samples were tested in solutions of coffee and tea (Figure 99). 0 5 10 15 20 14 16 18 20 22 24 26 28 30 Pb(II) in coffee samples exc = 500 nm em = 570 nm Emission intensity x 10 -4 (a.u.) [Pb(II)] (M) Figure 99. Coffee solutions containing one of them a concentration of Pb(II) 0.02 mM and the other as a blank (left) and fluorescence titration results increasing Pb2+ concentration (right). This procedure allowed us to measure the presence of Pb2+ when the concentration was higher than 1 µM. In any case, it was an example and it should be adapted to each different matrix, which could lead to different limits when working in low concentrations. Limits of the technique The possibility of measuring by this process is limited by the solution to measure. If the solution contains complexed Pb(II), such as high concentration of acetate, or pH too acid or basic, the measurements are not so reliable. Additionally, to ensure the possibility of the quantification of the process, the concentration must be compared with a blank reference with no Pb(II) in solution and by a calibration with a linear regression. For example, it was checked to work by comparison between coffee and tea samples with and without lead (Figure 99).
280|CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING JoséGarcíaCalvo|PhDThesis 14. SUMMARY OF THE CHAPTER In summary, the fluorescent probe JG76 may be considered as a very sensitive probe for detection of lead and potassium cations in alcohol-water mixtures. Moreover, it was a useful tool for the visualization of cereulide in live cells as well as the localization of potassium rich structures, from live cells. Cereulide and the rest of cyclic depsipeptides with potassium affinity have also demonstrated to be a research field to be exploited. During this chapter it has been shown how little modifications may change their behaviour, opening the possibility to tune the potassium affinity and, therefore, the applications. Finally, a new material, JG151dp was synthetized with the possibility to detect lead cations in water colourful samples. In addition, it was optimized to have an outstanding recyclability, going up to 7 times without showing any degradation. Between the many possible applications, it may be useful for monitoring industrial processes, the possible pollution of a river or the water from lead containing pipes. Data summary Soluble probe JG76. JG76 Solvent Value Units Error Fluor. Lifetime Decay τ (JG76) EtOH 3.58 ns 1.138 (χ2) Therm. Eq. constant K (JG76 + K+) EtOH 22 M-1 × 10-5 1 M -1 × 105 K (JG76 +Pb2+) EtOH 15.5 M-1 × 10-5 1 M -1 × 105 Fluor. Quantum Yield Φ (JG76) EtOH 0.12 - 0.02 - ΦF (JG76 + K+) EtOH 0.48 - 0.02 - ΦF (JG76 +Pb2+) EtOH 0.84 - 0.02 - Limits of detection LOD (K+) EtOH 60 nM - - LOD (Pb2+) EtOH 30 nM - - LOD (Val) EtOH 540** nM - - LOD (Cer) EtOH 210** nM - - **JG76 concentration between 25 μM and [JG76]/[K+] =1.25 Figure 100. Table that summarizes the calculated parameters for JG76, including fluorescent lifetime decays, thermodynamic equilibrium constants, fluorescence quantum yields and limits of detection.
CHAPTER3C.PMIDERIVATIVESFORK+ANDPb2+SENSING|281 JoséGarcíaCalvo|PhDThesis Cyclic depsipeptide potassium ionophores: Cyclic depsipeptide potassium ionophores Solvent Value Units Error Fluor. Lifetime Decay τ (JG121) EtOH 3.75 ns 1.07 (χ2) Therm. Eq. constant K (Val. + K+) EtOH 9.4 M-1 × 10-5 0.2 M -1 × 105 K (Cer. + K+) EtOH 9.7 M-1 × 10-5 0.2 M -1 × 105 K (JG115 + K+) EtOH 9.3 M-1 × 10-5 0.2 M -1 × 105 K (JG115B + K+) EtOH 0.4 M-1 × 10-5 0.01* M-1 × 105 K (JG121 + K+) EtOH <JG115B M-1 × 10-5 - - Fluor. Quantum Yield ΦF (JG121) EtOH 0.99 - 0.01 - *The error was given by the fitting curve Figure 101. Table that summarizes the calculated parameters for different cyclic depsipetides, including fluorescent lifetime decays, thermodynamic equilibrium constants, fluorescence quantum yields and limits of detection. Fluorescent material JG151dp: JG151dp Solvent Value Units Error Fluor. Lifetime Decay τ (JG151dp) Water* 3.75 ns 1.15 (χ2) Fluor. Quantum Yield ΦF (JG151dp) Air 0.22 - 0.01 - ΦF (JG151dp+Pb2+) Air 0.39 - 0.01 - Limits of detection LOD (Pb2+) Water 610 nM - - LOD (Pb2+) Water* 290 nM - - *HEPES buffer water solution 10 mM. Figure 102. Table that summarizes the calculated parameters for the film JG151dp, including fluorescent lifetime decays, fluorescence quantum yields and limits of detection.
288|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis to control polydispersity, shape and size (Figure 3); by smart control of several parameters simultaneously; such as proportions, concentration, solvent, reductive agent…13 Figure 3. Different morphologies for gold nanoparticles changing reduction reagent, stabilizer and solvent. Picture by Liz-Marzán et al.13 2.3. Physical and chemical properties-applications of gold nanoparticles Chemical properties: gold NPs might have different stability depending on their particular characteristics. Being usually synthetized by reduction of Au(III), gold NPs are not highly reactive to oxidative reagents. Nevertheless, they are particularly sensitive to the presence of some compounds; such as cyanide, thiol groups, highly acidic solution (“aqua regia”) or to the presence of some other cations (for example, Ag+ or Hg2+). As it was previously mentioned the sensitivity-shape/size is directly related to the stabilizers of the nanoparticles. For example, when surrounded by a sulfur-PEG stabilizer (very common in medical research) the sensitivity is much lower against other gold-appealing groups than when, for example, having only citrate as stabilizer. Additionally, the “stabilizers” may have other functions due to the capacity to surround gold NPs. For instance, it is possible to put fluorescent molecules or drugs with pharmacological activity. This feature is useful for using the NPs as drug carriers or as markers / sensors.14 Other related and interesting applications are associated to the electronical properties, their high surface (compared to bulk atoms ratio) and their overall chemical inertness. These characteristics confer to gold nanoparticles catalytical properties for redox processes. For instance, gold nanoparticles have been used successfully for oxidation of CO and H2, reduction of NO and other catalytic reactions,15 such as a variety of C-C coupling reactions. Physical properties: they have significantly different properties compared to bulk gold; for instance, lower melting temperature (related with diameter of the particle)16 or higher thermal conductivity (high surface/volume). On top of them, the electronical properties are directly related with what is called surface plasmon absorptions (SPA), which is simultaneously related to their size. 13 a) M. Grzelczak, J. Pérez-Juste, P. Mulvaney, L. M. Liz-Marzán, Chem. Soc. Rev. 2008, 37, 1783–1791. b) T. H. Ha, H.-J. Koo, B. H. Chung, J. Phys. Chem. C, 2007, 111, 1123. 14 L. M. Liz-Marzan. Materials Today, 2004, 2, 26-31. 15 K, Kvitek, R. Prucek. J. Mat. Sci. 2005, 22, 2461–2473. 16C. Burda, X. Chen, R. Narayanan, M. A. El-Sayed. Chem. Rev. 2005, 105, 1025-1102.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|289 JoséGarcíaCalvo|PhDThesis Moreover, the variation in the SPA has not only influence over the final absorption-colour, but other properties such as the conductivity are also affected.13 In consequence, when working with solutions of nanoparticles, studying their absorption might be directly related with the size and shape of the particle, which might be very useful for getting a raw estimation of their characteristics. As an example, for gold (Figure 4), the absorption between 20-100 nm particles produces an increase in its wavelength of emission, simultaneously to size. Figure 4. Normalized UV-vis absorption of nanoparticles of different size in water solution (left), and variation of the plasmon bandwith as a function of particle diameter (right). From Link et al.17 2.4. Supported gold NPs Gold nanoparticles (Au-NPs) display a range of physical and chemical properties that are promising for the development of optical, electronic, and chemical devices. In doing so, many of these devices would require from immobilization of gold nanoparticles in a single layer, or in multilayers, over surfaces.18 Up to now, the process has been performed in two steps. First, the gold salt is reduced by adding a reductive reagent to the solution (for example, citrate). Secondly, the immobilization is performed by modification of the surface with functional groups that provide attractive interaction to gold nanoparticles.19 However, these procedures have no lead to optimal routes for the formation of monolayer ensembles on various substrates. In contrast, a homogeneous nanostructured system (Figure 5) may get unique optical and electronic properties, which make them a good prospect for future application in microelectronics, solid state chemical, biological sensors or catalysis.20 In spite of the potential applications, the formation of continuous films of metal nanoparticles on a solid substrate is not a simple task, because the size of metal clusters and their concentration on the surface are rather difficult to control. 17S. Link. M. A. El-Sayed, J. Phys. Chem. B 1999, 103, 4212-4217. 18A. X. Wang, X. Kong, Materials 2015, 8, 3024-3052. 19a) M. S. Onses, C. J: Thode, C. -C. Liu, S. Ji, P. L. Cook, F. J. Himpsel, P. F. Nealey, Adv. Funct. Mater. 2011, 21, 3074–3082; b) F. L. Yap, P. Thoniyot, S. Krishnan, S. Krishnamoorthy, ACS Nano 2012, 6, 2056– 2070. 20a) M. A. Mahmoud, D. O’Neil, M. A. El-Sayed, Chem. Mater. 2014, 26, 44−58. b) L. Prati, A. Villa, Acc. Chem. Res. 2014, 47, 855–863; c) E. C. Dreaden, A. M. Alkilany, X. Huang, C. J. Murphy, M. A. El-Sayed, Chem. Soc. Rev. 2012, 41, 2740–2779. d) F. Mitschang, H. Schmalz, S. Agarwal, A. Greiner, Angew. Chem. Int. Ed. 2014, 53, 4972–4975.
290|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis Figure 5. Au nanocubes layer over a silicon wafer, synthetized by Mahmoud and coworkers.20 Until now, the adsorption of the metallic nanoparticles from their colloidal solutions has been employed for modifying surfaces with nanoparticles,21 but this method presents major drawbacks. Highly dispersed gold nanoparticles are difficult to obtain in a solution phase without using organic stabilizers and reducing reagents such as citric acid, alkylamine, alkylthiol, and cationic surfactants, complicating the adsorption step. Therefore, the approaches with prebound reductants and stabilizers on the same material are quite desirable. 21 Z. Zhang, C. Liu, J. Bai, C. Wu, Y. Xiao, Y. Li, J. Zheng, R. Yang, W. Tan, ACS Appl. Mater. Interf. 2015, 7, 6211-6219.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|291 JoséGarcíaCalvo|PhDThesis 3. PALLADIUM NANOPARTICLES 3.1. Historical background Pd(0) derivatives have been mostly used in the fields of catalysis and hydrogen detection, purification and storage.22 As a consequence, Pd nanoparticles with a high relation surface/volume have the potential to provide a cost-effective solution to the requirements of evolving catalytical and electrochemical applications. In contrast with gold or silver, Pd(0) nanoparticles are not so well known by the different colour variations with size, usually black or brown (Figure 6), but they also have different morphologies that may tune the Pd(0) applications. Figure 6. Absorbance of Pd nanoparticles formed from [Pd3(OAc)6] and n-dodecyl sulphide by heating from room temperature to 80ºC and waiting from 0 (a) to 40 minutes (e), by Obare et al.23 The band at 400 nm indicates the presence of Pd-NPs. Due to the difficulty of distinguishing the nano-structure from the palladium bulk or molecular derivatives it was not until the end of 90s when the development of highly accurate characterization systems, such as transmission electron microscopy (TEM), allowed a deeper study of this kind of metallic particles. Therefore, in the last 30 years, the process has been perfectioned to relate internal structure - synthesis methods (Figure 7) and electronical properties, managing to achieve the optimization of its use for many applications. 22 A. Chen, C. Ostrom, Chem. Rev. 2015, 115, 11999−12044 23 M. Ganesan, R.G. Freemantle, S. O. Obare, Chem. Mater. 2007, 19, 3464–3471.
292|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis Figure 7. Scheme shape transformation of Pd nanoparticles when growing. From Chen and Ostrom review.22 3.2. Synthesis of Pd nanoparticles 3.2.1. Synthesis procedures: Pd-NPs synthesis, just as for other metals, are synthetized by physical or chemical techniques. Within the physical techniques there are sputtering methods, ionand electron-beam-induced deposition or laser ablation. All of them, combined with stabilizing surfactants may achieve controlled compositions, morphologies and other attributes. Within the chemical methods, the most common is the electrochemical deposition which, controlling the potential or current density, produces deposition of Pd(0) from the oxidized form in solution. Other method is what is called hydrothermal deposition, in which a chemical reaction occurs above the solvent boiling point and at pressures higher than 1 bar. Finally, there are synthetic procedures involving photoreductants and/or chemical reducing agents. These electroless deposition methods are based on displacement deposition or autocatalytic deposition. Displacement deposition is reliant on the reduction potential of metallic precursors (such as Fe2+).24 While autocatalytic deposition proceeds by using chemical reducing agents, including ascorbic acid, ethylene glycol, citric acid or sodium borohydride;25 along with stabilizers, such as PVP. Chemically obtained Pd-NPs may have a wide variety of shapes, usually as octahedral or truncated octahedral nanocrystals (Figure 7); that are attached to one another depending on the media and stabilizers. 3.2.2. Shape and size regulation: Shape controlled Pd-NPs need careful nucleation and growth conditions,26 which is achieved throughout the manipulation of kinetic parameters;27 especially in aqueous solutions.28 For controlling growth conditions, the presence of an organic stabilizer/additive is associated to the control on the reaction kinetics and hence, the shape/size of the nanostructures:29 24 R. Ojani, Z. Abkar, E. Hasheminejad, J. -B. Raoof, Int. J. Hydrogen Energy 2014, 39, 7788−7797. 25V. Raghuveer, P. Ferreira, A. Manthiram, Electrochem. Commun. 2006, 8, 807−814. 26Y. Xia, Y. Xiong, B. Lim, S. E. Skrabalak, Angew. Chem. Int. Ed. 2009, 48, 60–103. 27Y. Wang, H. -C. Peng, J. Liu, C. Z. Huang, Y. Xia, Nano Lett. 2015, 15, 1445−1450. 28B. Lim, M. Jiang, J. Tao, P. H. C. Camargo, Y. Zhu, Y. Xia, Adv. Funct. Mater. 2009, 19, 189–200. 29J. Watt, S. Cheong, M. F. Toney, B. Ingham, J. Cookson, P. T. Bishop, R. D. Tilley, ACS Nano, 2010, 4, 396– 402.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|293 JoséGarcíaCalvo|PhDThesis Using additives (such as polyol and sulfate): they guide the reaction kinetics and give control over shape.30 Furthermore, they often serve as reducing and capping agents, for instance hexacarbonylmetals,31 acids, amines and CO32 or EDTA.33 Stabilizing agents, like PVP, the function of which is directly related to the control of NPs size.34 Some agents act as reducing, stabilizing and additives simultaneously; for instance, some peptides. It usually helps for the generation of monodisperse, water-soluble palladium nanoparticles of controlled size.35 Controlling nucleation and using seed-mediated growth of palladium nanocrystals is a well stablished way to prepare size-controlled NPs.36 Nevertheless, significant achievements have been obtained by the seedless growth of palladium nanocrystals getting tuneable structures37 and ultrathin palladium nanosheets.38 3.3. Supported palladium nanoparticles, synthesis and applications To take advantage of Pd-NPs the next step is the immobilization in solid supports, as it would permit recovery when involved in catalysis, and easier work-up and purification processes.39 Until now, most Pd(0) supported materials have been synthesized by electrochemical methods and, thanks to their characteristics, used with catalytical purposes. In contrast, when they are not directly supported by electrochemical methods, the usual way to do it is similar to other nanoparticles (as gold NPs). Preformed Pd-NPs are put in presence of a material that also acts as stabilizer. Some of the most common applications of Pd-NPs, apart from hydrogen storage, has been as catalyst in a variety of chemical reactions, some examples are: Carbon-carbon coupling reaction40 (especially in water).41 Hydrogenation of oxoderivatives,42 with recyclable materials.43 Electrocatalysis,17 often as bimetallic nanocrystals.44 For ethanol oxidation,45 nitroaryl reduction46 or CO2 storage.47 30H. Huang, Y. Wang, A. Ruditskiy, H.-C. Peng, X. Zhao, L. Zhang, J. Liu, Z. Ye, Y. Xia, ACS Nano 2014, 8, 7041–7050. 31Y. Li, Y. Yan, Y. Li, H. Zhang, D. Li, D. Yang, CrystEngComm, 2015, 17, 1833–1838. 32X. Yin, J. Wu, P. Li, M. Shi, H. Yang, ChemNanoMat 2016, 2, 37 – 41. 33C. Shang, W. Hong, Y. Guo, J. Wang, E. Wang, Chem. Eur. J. 2017, 23, 5799 – 5803. 34C. Evangelisti, N. Panziera, A. D’Alessio, L. Bertinetti, M. Botavina, G. Vitulli, J. Catal. 2010, 272, 246–252. 35S. Corra, U. Lewandowska, E. M. Benetti, H. Wennemers, Angew. Chem. Int. Ed. 2016, 55, 8542 –8545. 36Y. Xia, K. D. Gilroy, H.-C. Peng, X. Xia, Angew. Chem. Int. Ed. 2017, 56, 60–95. 37Y. Zhang, M. Wang, E. Zhu, Y. Zheng, Y. Huang, X. Huang, Nano Lett. 2015, 15, 7519−7525. 38X. Yin, X. Liu, Y.-T. Pan, K. A. Walsh, H. Yang, Nano Lett. 2014, 14, 7188−7194. 39D. Astruc, F. Lu, J. Ruiz Aranzaes, Angew. Chem. Int. Ed. 2005, 44, 7852–7872. 40A. Fihri, M. Bouhrara, B. Nekoueishahraki, J.-M. Basset, V. Polshettiwar, Chem. Soc. Rev. 2011, 40, 5181– 5203. 41G. Yun, Z. Hassan, J. Lee, J. Kim, N.-S. Lee, N. H. Kim, K. Baek, I. Hwang, C. G. Park, K. Kim, Angew. Chem. Int. Ed. 2014, 53, 6414–6418. 42A. Balouch, A. A. Umar, A. A. Shah, M. M. Salleh, M. Oyama, ACS Appl. Mater. Interfac. 2013, 5, 9843−9849. 43E. Hariprasad, T. P. Radhakrishnan, ACS Catal. 2012, 2, 1179−1186. 44K. D. Gilroy, A. Ruditskiy, H.-C. Peng, D. Qin, Y. Xia, Chem. Rev. 2016, 116, 10414−10472.
294|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis When using supported Pd-NPs as catalysts, the molecular mechanisms are linked to the characteristics of the nanoparticles.40-47 Hence, the performance of palladium nanocatalysts is influenced by all the parameters that characterize the NPs; shape, size, stabilizers and support. A particular case of catalytic procedure by supported palladium is the reduction by hydrogenation of alkenes48 and alkynes.49 A very efficient method but with low selectivity.50 In contrast, the importance of a selective, clean and cheap hydrogenation of internal alkynes to (Z)-alkenes is of great importance in pharmaceutical and industrial compounds.51 Regarding that selectivity, some methods with Pd nanoparticles have achieved high selectivity in DMF dispersion52 or supported with other nanomaterials.53 Existing extensive research in the mechanism of the (Z)-selectivity54 or the Z/E interconversion of the initially obtained olefins.55 In summary, having well-defined structures, in shape and size, is the goal to achieve for new palladium nanomaterials, giving them clear-cut properties for innovative catalytic processes.22 45L. Ren, L. Yang, P. Yu, Y. Wang, L. Mao, ACS Appl. Mater. Interfac. 2013, 5, 11471−11478. 46 E. D. Sultanova, V. V. Salnikov, R. K. Mukhitova, Y. F. Zuev, Y. N. Osin, L. Y. Zakharova, A. Y. Ziganshina, A. I. Konovalov, Chem. Commun. 2015, 51, 13317—13320. 47 A. Modak, M. Pramanik, S. Inagakib, A. Bhaumik, J. Mater. Chem. A, 2014, 2, 11642–11650 48 a) S. K. Mahato, R. U. Islam, C. Acharya, M. J. Witcomb, K. Mallick, ChemCatChem 2014, 6, 1419–1426. b) M. Iwanow, J. Finkelmeyer, A. Söldner, M. Kaiser, T. Gärtner, V. Sieber, B. König, Chem. Eur. J. 2017, 23, 12467–12470. c) S. K. Surmiak, C. Doerenkamp, P. Selter, M. Peterlechner, A. H. Schäfer, H. Eckert, A. Studer, Chem. Eur. J. 2017, 23, 6019–6028. 49 Y. Gao, C.-A. Chen, H.-M. Gau, J. A. Bailey, E. Akhadov, D. Williams, H.-L. Wang, Chem. Mater. 2008, 20, 2839–2844. 50 A. S. Reddy, K. C. K. Swamy, Angew. Chem. Int. Ed. 2017, 56, 6984–6988. 51 K. Chernichenko, A. Madarász, I. Pápai, M. Nieger, M. Leskelä, T. Repo, Nat. Chem. 2013, 5, 718-723. 52 J. Hori, K. Murata, T. Sugai, H. Shinohara, R. Noyori, N. Arai, N. Kurono, T. Ohkuma, Adv. Synth. Catal. 2009, 351, 3143–3149. 53 T. Mitsudome, Y. Takahashi, S. Ichikawa, T. Mizugaki, K. Jitsukawa, K. Kaneda, Angew. Chem. Int. Ed. 2013, 52, 1481–1485. 54 F. Zaera, ACS Catal. 2017, 7, 4947−4967. 55 a) S. Furukawa, T. Komatsu, ACS Catal. 2016, 6, 2121−2125. b) K. Tokmic, A. R. Fout, J. Am. Chem. Soc. 2016, 138, 13700−13705.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|295 JoséGarcíaCalvo|PhDThesis 4. CHARACTERIZATION OF SUPPORTED NPs The reference technique for studying NPs is transmission electron microscopy (TEM) and its high resolution version (HRTEM). This kind of analysis gives accurate data of shape, size and dispersity of the particles, or even the crystalline structure in high resolution. There are even more precise methods, such as atomic force microscopy (AFM), in which the structure of the nanomaterials could be observed with atomic resolution. Figure 8. TEM apparatus from Valladolid University. When the particles are deposited over a surface, the shape and size is observed and measured by scanning electron microscopy (SEM), usually combined with X-Ray photoelectron microscopy (XPS) for the composition of the external layer or energy-dispersive X-ray spectroscopy (EDX) for getting composition with more penetration into de material. SEM analysis has the disadvantage of lower resolution than TEM, but the sample preparation is simpler. Moreover, XPS and EDX analysis give the atomic composition, which is related to the amount and number of stabilizers presented in the sample (such as sulfur containing molecules). Cheaper and more straightforward techniques are UV-Vis and NIR absorption. Metal nanoparticles have characteristic plasmons (SPA) that give different absorption depending on shape and size, giving raw information when measured. Additionally, techniques such as Raman dispersion or fluorescence give complementary information to the absorption and dynamic light scattering may give information about aggregation. Besides, other techniques could give additional information related to specific properties of different nanoparticles, solubility, additives, stabilizers and stability; techniques such as NMR, mass spectrometry, melting point or calorimetric studies.
296|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis 5. OBJECTIVES OF THE CHAPTER For some time, the group had been working in the detection of Hg(II), which is a known thiophilic cation. As it was explained in Chapter 1, detection of Hg(II) was performed by fluorogenic reagents having sulfur atoms. If the results of qualitative tests are carefully observed (Section 5.3.3 of Chapter 1) the characteristic colour of Au-NPs was obtained in solution. In addition, the change in colour also happened in solid materials, like silica NPs or polymeric films. The colour and its intensity depended on many factors such as the composition of the material, Au(III) concentration, solvent, the relation between volume (solution)/ size (material) and the time of contact. Once checked in literature, the interest of getting Au-NPs supported in a material became clear. In consequence, the purpose was to reach different objectives: Optimization of materials to act as reductants, stabilizers and support for the nanoparticles. Evaluation and characterization of the synthetized gold nanoparticles. Optimization of the material-conditions to obtain gold nano-microparticles, controlling time, concentration and solvents while following the effect over size and shape of the particles; in solution and supported. Looking for a practical application of the supported gold nanoparticles. This experimental was aimed to get applications as a catalyst for organic synthesis; in particular, C-C coupling by substitution of the Pd catalyst used in Suzuki-Miyaura reactions. After obtaining satisfactory results, the idea of getting similar synthetical procedures for a variety of supported metallic nanoparticles led us to try with other metals and copolymers. From all of them, there was a series of polymeric derivatives from what Pd(0) nanoparticles were easily obtained, giving to the material outstanding applications as it was subsequently noticed. The objectives for supported Pd-NPs were: Simplify the preparation, by just adding the polymers to a Pd(II) solution in water. The aim was to improve pre-existent procedures in which the nanoparticles were preformed and stabilized in a surface. Adaptation and control of the synthesis to obtain monodisperse polycrystalline palladium nanoparticles uniformly distributed over the surface of polymers. Evaluation and characterization of the results. Search for a straightforward application of the supported NPs. In this case, it was potentially viable to use them as an efficient, portable and reusable catalyst for the stereoselective semihydrogenation reaction of internal alkynes to (Z)-alkenes.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|297 JoséGarcíaCalvo|PhDThesis 6. EXPERIMENTAL SYNTHESIS OF GOLD NPs 6.1. Antecedents In Chapter 1, it was introduced how an excess of PEG containing probe in solution was capable to reduce Au(III) in solution to gold NPs. What is more, the silica modified solids with terminal triple bonds led to the formation of gold nanoparticles in solution (Figure 9), although they precipitated with time. Figure 9. JG19 in presence of HAuCl4 (0.5 mM) after 5 hours. Later, the reduction synthesis was optimized for solutions and polymeric films. 6.2. Synthesis and characterization of gold NPs in solution In order to test the action of solutions containing triazole, PEGs and carbothioamide probe, several experiments were performed in solution. The water-soluble derivatives JG45 and JG47, (Hg(II) probes from Chapter 1) were tested in presence of Au(III) solutions. Figure 10. Au-NPs formed in presence of JG47 (0.5 mM JG47 and 0.2 mM Au(III)). After 48 hours. Figure 11. Au-NPs formed in presence of JG45 (0.1 mM JG45 and 0.2 mM and 0.4 mM Au(III)). After 24 hours. As it may be observed in Figures 10 and 11, this method required high concentration of the analyte to stabilize the NPs and, if the proportion of Au(III) added was too high, it precipitated (dark purple). In consequence, to get a proof of the compound acting not only as reductant but as stabilizer, the solution was added to HEPES buffer - Au-NPs. HEPES gold nanoparticles are characterised as not very stable and with tendency to agglomeration and precipitation.56 However, in the presence of JG45 solution their colour changed from purple to pink, evidencing stabilization of the AuNP by JG45. 56 R. Cheng, J. Wu, H. Li, G. Cheng, Z. Lu, C.-M. Che, Rare Metals, 2010, 29, 180–186. Au3+Au3+ Au3+ 0.2mM0.4mM
304|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis Cryofracture and SEM imaging of the transversal section were also performed to one of the samples. It was verified that the Au-NPs were, in this case, evenly distributed also inside the bulk of the polymer (Figure 20D). Nevertheless, they were much smaller (50-90 nm average) than the particles on the surface (200-250 nm). This fact explained the optical characteristics found for polymer membranes on which gold nanoparticles were grown by using DMF (Figure 21). 6.3.4. Additional experiments: The importance of having “triple bonds” in the material or adding very low volumes of gold solutions were facts that could give additional information. In a first experiment (Figure 22), it was checked whether a simple 2HEA (2hydroxyethylacrilate) polymer, a polymer with the same characteristics but without the 5% of propargyl methacrylate; could deposit Au-NPs. Negative results were obtained, even if the concentration and time were changed. Therefore, the presence of bonded JG10 or the triple bonds on the surface of the polymers were necessary for the formation of gold nanoparticles. Figure 22. 2HEA polymer (left) and 2HEA polymer after 2 days in contact to a water solution 0.5 mM HAuCl4. Additionally, an experiment was performed in which a drop of Au(III) from a concentrated solution was added to the corner of a polymer (Figure 23). The aim was to show in one single piece of polymer the importance and relation between [Au(III)]/surface – swelling – time. From the experiment, long nanorods up to 1.5 μm were obtained (blue) with very few flat gold nanoplates, triangles and hexagons.57 However, the distribution was not homogeneous in this case; the gold nanorods, with lengths that ranged from 100 nm to 1.5 microns, where the major product for some areas of the polymer while Au-NPs similar to the other tests, were obtained within the place where the drop was added (pink-purple). 57a) N. Li, P. Zhao, D. Astruc, Angew. Chem. Int. Ed. 2014, 53, 1756–1789. b) I. Pastoriza-Santos, R. A. Alvarez-Puebla, L. M. Liz-Marzán, Eur. J. Inorg. Chem. 2010, 4288–4297. c) Y. Wang, K. Sentosun, A. Li, M. Coronado-Puchau, A. Sánchez-Iglesias, S. Li, X. Su, S. Bals, L. M. Liz-Marzán, Chem. Mater. 2015, 27, 8032−8040.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|305 JoséGarcíaCalvo|PhDThesis Figure 23. A) SEM image of the polymer membrane surface coated by gold nanorods, nanoplates and nanoparticles (40 µL Au3+, HAuCl4 5 mM in water), scale bar 1 μm. B) Image showing the colour of the polymer by transmitted and reflected light. 6.3.5. Summary: The results of the different tests for the formation of film supported Au-NPs were: In the presence of the fluorescent probe there was higher dispersion in particle size. Using high concentrations of the salt increased faster the number of particles than the size. Low relation volume/surface led to more variation in the shape of the nanoparticles. The size of the particles was between 20 nm to 400 nm, depending on the conditions. The different shapes encountered were tridimensional, amorphous, spherical, tetrahedral or tetrahedral truncated; bidimensional, triangles; and monodimensional (or with a preferential grow direction), bars with different thickness. In water, the particles were located over the polymer while using DMF led to NPs inside the film. A B
306|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis Figure 24. Gold nanoparticles formation at different time and reaction conditions. SEMSatthesamescale. Comparisonbetweenthe nanoparticlesobtainedat differentconditions. (a)(b)(c) (d)(e) (f)(g) (h)(i) (a) JG25, 1 mL water + 40µL Au3+- 2 h. (b) JG25, 100µL Au3+ - 25 min. (c) JG25, 100µL Au3+ - 24 h. (d) JG25, 100µL Au3+- 2 h. (e) JG25_SA2, 1 mL water + 40µL Au3+ - 24 h. (f) JG25_SA2, 1 mL water + 100µL Au3+ - 24 h. (g) JG25_SA2, 40µL Au3+ - 24 h. (h) JG25_SA2, 100µL Au3+ - 24 h. (i) JG25_SA2, 100µL Au3+ - 25 min. Au3+ = 5 mM HAuCl4 in water
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|307 JoséGarcíaCalvo|PhDThesis 7. APPLICATION IN CATALYSIS OF GOLD-NPs Once the coating of the surface of the polymer by gold nanoparticles was sufficiently studied, practical applications were the next step to be addressed. Between the many reactions catalysed by solid supported gold nanoparticles,58 poly(2-aminothiophenol) supported gold nanoparticles had shown excellent catalytic activity for Suzuki-Miyaura cross-coupling reactions.59 The polymer supported Au-NPs material was found to be useful as a solid reusable catalyst. The conditions to perform these reactions were very similar to the classical reactions performed with Pd(PPh3)3 as the catalyst. 7.1. Synthesis and yields As a representative experiment, a 25 mL round bottom flask was filled with 30 mg (0.14 mmol) of 5-bromoindan-1-one, one equivalent of the boronic ester or acid, 75 mg (0.7 mmol) of Na2CO3 and a 0.3×0.3 cm polymer piece coated with gold nanoparticles obtained from previous synthesis. After that, a solvent mixture composed of 4 mL of THF and 0.5 mL of water was added. The mixture was refluxed for 20 hours, extracted by partition in DCM:Water and purified by column chromatography. (a) (b) (c) (d) Figure 25. An example of polymer used in the catalytic synthesis. (a) Preparation. (b) Colour under transmitted white light. (c) Round bottom flask with the reagents before adding the solvent. (d) SEM image of the polymer before the reaction. The corresponding spectral and physical characterization of every obtained product was checked in each case to be identical to previously reported compounds. Yields are compared in Figure 26 with the reported yields for every compound, 1-3,60 4,61 562 and 663. 58a) Y. Zhang, X. Cui, F. Shi, Y. Deng, Chem. Rev. 2012, 112, 2467–2505. b) M. Stratakis, H. Garcia, Chem. Rev. 2012, 112, 4469−4506. 59 J. Han, Y. Liu, R. Guo, J. Am. Chem. Soc. 2009, 131, 2060–2061. 60 B. Díaz de Greñu, J. García-Calvo, J. V. Cuevas, G. García-Herbosa, B. García, N. Busto, N.; S. Ibeas, T. Torroba, B. Torroba, A. Herrera, S. Pons, Chem. Sci. 2015, 6, 3757–3764. 61 B. Díaz de Greñu, D. Moreno, T. Torroba, A. Berg, J. Gunnars, T. Nilsson, R. Nyman, M. Persson, J. Pettersson, I. Eklind, P. Wästerby, J. Am. Chem. Soc. 2014, 136, 4125–4128. 62 T. Gómez, D. Moreno, B. Díaz de Greñu, A. C. Fernández, T. Rodríguez, J. Rojo, J. V. Cuevas, T. Torroba, Chem. Asian J. 2013, 8, 1271–1278. 63 M. D. Chordia, M. Zigler, L. J. Murphree, H. Figler, T. L. Macdonald, R. A. Olsson, J. Linden, J. Med. Chem. 2005, 48, 5131-5139.
308|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis Figure 26. Representative experiments of Suzuki reactions by using the Au-NPs modified polymer as catalyst. 7.2. Recyclability of the catalyst The modified polymers used as heterogeneous catalysts can be recovered from the solution and reused again in other synthetic processes. To know exactly how many times it could be used, and how it affected its catalytical properties, the polymer was recycled several times in different conditions. It was found that the polymeric catalyst worked properly after subsequent use of the polymer, by using 40 % of water and 60 % of THF as solvent. Albeit, having different water proportions of the solvent in the Suzuki reactions led different percentages of swelling: 8:1 THF:H2O, 330 % 6:4 THF:H2O, 310 % 4:6 THF:H2O, 280 % The reaction was tested at different percentages of THF in order to get the best results. However, although higher percentage of THF slightly increased the yield of the reaction, it decreased the recyclability of the catalyst, because the stiffness of the polymer was lower. In the end, if the percentage of THF was too high, the polymer broke easily into pieces and was impossible to recover it. No significant change in yield was found when using the same piece of polymer up to 4 times (4:6 THF:H2O). In order to use the polymer for more cycles, it would have been necessary to optimize the mechanical properties by using thicker layers, and/or adding more crosslinker to the polymer.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|309 JoséGarcíaCalvo|PhDThesis Additionally, it was noticed that polymers with high concentration of gold in the surface were more efficient in getting faster reactions with higher yields of the product than polymers with a lower surface concentration of gold nanoparticles. Complementary to the catalytic behaviour it was also evaluated the possibility that the catalysis occurred by the Au-NPs, that may dissolve into the solution. In fact, a small fraction of Au-NPs dissolved in the solution (Figure 27) when it was heated to temperatures over 70 ºC, and/or the percentage of organic solvent was too high with respect to water (more than 70 % THF). But, contrary to what was expected, for those tests, the yield of the reaction was lower than when it was performed under the standard conditions. Figure 27. Catalytic solution, 80% THF after 3 days at 70 ºC. In contrast, when conditions were optimal, the polymer was recovered almost unchanged (Figures 28 and 29) from the solution after the reaction had finished. Figure 28. Piece of polymer recovered after the catalysis process (16 hours, 65 ºC, THF:H2O 60:40) Figure 29. SEM image and EDS analysis of the piece of polymer recovered after the catalysis process (16 hours, 65 ºC, THF:H2O 60:40). Sodium was absorbed from the sodium carbonate used for the reaction. Au Na O C
310|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis 7.3. Turn-over number (TON) and turn-over frequency (TOF) The turn over number (TON) of a catalyst is defined as: 𝑇𝑂𝑁 Equation [1] The results with the area of the polymer. TON = 0.36 mmoles/0.18 cm2 = 2 mmoles/cm2, probably superior because the main problem was the fragility of the polymer. The turn over frequency (TOF) is defined as the turnover number per time unit. The calculation is performed through the equation: 𝑇𝑂𝐹 Equation [2] In order to calculate this number, it would be necessary to perform the reaction several times with different periods of reflux and different pieces of polymer. In case of the polymer, it was not possible more than an estimation, due to the limited amount and the difficulties on performing it with exactly the same conditions. With the available data it was concluded that the time is lower than 16 hours. So the TOF would be superior to 0.0021 mmoles/(cm2 × min). 7.4. Conclusions of the material as catalyst The polymer with supported Au-Ps was used as an efficient portable and reusable catalyst for Suzuki reactions in mixed organic-aqueous solvents. The yields obtained were similar to classic methods, that use Pd(PPh3)4 as catalyst. The working conditions were optimized to mixtures THF:Water (between 40 or 60 % THF), a base and temperature around 65ºC overnight. It was tested to work for at least 4 reactions with the same piece of polymer In conclusion, the simplicity in the preparation of the catalyst, with no need of additional reagents for reduction stabilization,64 makes the system competitive for Suzuki reactions against gold NPs suspended in solution,65 supported on paper,66 or to other solid supported gold NPs catalysts.67 Therefore, these polymer-supported gold NPs may be considered as a useful material for green catalysis in the synthesis of fine chemicals by heterogeneous catalysis.68 64G. Li, R. Jin, Nanotechnol. Rev 2013, 2, 529–545. 65 T. Chen, G. Li, H. Qian, R. Jin, Catalysis by Atomically Precise Gold Nanoclusters, in: Z. Wu, S. H. Overbury, Catalysis by Materials with Well-Defined Structures, Elsevier, New York, Chapter 8, pp. 239–262. 2015. 66 a) G. Zheng, L. Polavarapu, L. M. Liz-Marzán, I. Pastoriza-Santos, J. Pérez-Juste, Chem. Commun. 2015, 51, 4572–4575. b) G. Zheng, K. Kaefer, S. Mourdikoudis, L. Polavarapu, B. Vaz, S. E. Cartmell, A. Bouleghlimat, N. J. Buurma, L. Yate, A. R. de Lera, L. M. Liz-Marzán, I. Pastoriza-Santos, J. Pérez-Juste, J. Phys. Chem. Lett. 2015, 6, 230−238. 67 a) Y. Li, X. Fan, J. Qi, J. Ji, S. Wang, G. Zhang, F. Zhang, Mater. Res. Bull. 2010, 45, 1413–1418. b) M. G. Speziali, A. G. M. da Silva, D. M. V. de Miranda, A. L. Monteiro, P. A. Robles-Dutenhefner, Appl. Catal. AGen. 2013, 462– 463, 39– 45. 68 X. Liu, L. He, Y.-M. Liu, Y. Cao, Acc. Chem. Res. 2014, 47, 793–804.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|311 JoséGarcíaCalvo|PhDThesis 8. EXPERIMENTAL SYNTHESIS OF PALLADIUM-NPs After the satisfactory results with Au-NPs, the possibility of getting different supported metallic NPs was tested with other metals – crosslinked polymers. From these tests it is worth to remark the case of Pd(II). From the different tests, PdCl2ꞏ2NaCl salt had good response when reacted in the presence of some crosslinked polymeric films. The components of the polymers tested are shown in Figure 30. Figure 30. Components of the film monomers and photoinitiator (DMPA), “x”, “y” and “z” represent the relations between components; x+y= 100%, z is the percentage respect to x+y.
312|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis From the monomeric structures, several mixtures were tested, as it is detailed in the table below (Figure 31). Films (%) VP A12 A4 A1 M2 MEGMA MP A2HE C PBM2 50 x x x 50 x x x 0 PBMEGMA 50 x x x x 50 x x 0 PB0 60 x 40 x x x x x 0 PB20_80A12 20 80 x x x x x x 10 PB20_80A4 20 x 80 x x x x x 10 PB20_80A1 20 x x 80 x x x x 10 PB0_100A4 x x 100 x x x x x 10 PB0_100A1 x x x 100 x x x x 10 PB80_20A1 80 x x 20 x x x x 10 PB80_20A4 80 x 20 x x x x x 10 A2HE5 x x x x x x x 100 10 JG25_SA2 x x x x x x 5 95 10 Figure 31. Name and composition of some of the tested polymeric structures. These colourless films (1×1 cm) were put in PdCl2ꞏ2NaCl solution (5 mM, 3 mL) for 20 hours, and washed with water. Afterwards, changes in the colour of the polymer were observed for some of them, Figure 32: Figure 32. Pictures of the polymeric films after in presence of PdCl2ꞏ2NaCl solution.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|313 JoséGarcíaCalvo|PhDThesis Four different behaviours were observed. The polymer remained unaffected, it took Pd(0) on the surface or it absorbed Pd(II), or a mixture of Pd(0)+Pd(II). Figure 33. Visual appearance of the films after in presence of PdCl2ꞏ2NaCl 5 mM for 20 hours. From left to right; no effect, Pd(0) and Pd(II) or Pd(0)+Pd(II). The polymers that contained only acrylate derivatives (PB0_100) or a high percentage of the acrylate with a long aliphatic chain (PB20_80A4 and PB20_80A12) showed no change after being in the presence of palladium solutions. The film remained colourless, first picture on Figure 33. PB20_80A1 presented a black layer over its surface; once analysed, it was concluded that they were Pd(0) nanoparticles. Second picture in Figure 33. PB0, PB80_20A4, PB80_20A1 and A2HE5 acquired orange-brown colour. Being a mixture of Pd(II) and Pd(0). Third picture in Figure 33. PBM2, PBMEGMA and JGSA2 acquired a reddish colour. Afterwards, it was checked hat they also presented a mixture between Pd(II) and Pd(0). Fourth picture in Figure 33. 8.1. Characterization of the films The polymeric films were characterized by infrared and TGA, presenting a degradation point at 430 ºC in TGA, and not remarkable changes in the infrared although there was Pd(0) or Pd(II) absorbed. In contrast, UV-Vis absorption and SEM-TEM analysis gave information about the supported Pd: UV-Vis absorption spectra: 400 500 600 700 800 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance (a.u.) Wavelength (nm) PB20_80A1 PB80_20A4 JG25_SA PB0 PBM2 PBMEGMA PB80_20A1 400 500 600 700 800 0.0 0.2 0.4 0.6 0.8 1.0 Normalized absorbance Wavelength (nm) PdCl2 Figure 34. Normalized absorption spectra of different synthesized polymers (left) and PdCl2 ꞏ2NaCl 5mM solution in water (right).
320|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis Figure 46. TEM images of the Pd-NPs. First row: medium resolution TEM, scale bar 1 μm, 1 μm, 300 nm, 100 nmSecond row: high resolution TEM of individual nanoparticles, scale bar 80 nm. Figure 47. Left, TEM of an individual nanoparticle, the lattice spacing intervals of 19.5 and 22.3 Å correspond to blocks of 10 planes each and can be indexed as {200} and {111} of fcc Pd, respectively. Middle, high resolution TEM of individual nanoparticles showing the boundaries of the nanoparticle, scale bar 5 nm. Right up: FFT of the HRTEM image showing the interplanar distances corresponding to {111} and {222} family planes, a pattern corresponding to a f.c.c. crystal in the zone axis [011]. Right down: Distribution of the particle size.69 The NPs were deeply studied by HRTEM. These nanoparticles appeared as polycrystalline dendrites showing internal structure from radial growth and an average diameter of 75 nm. The Figure 47, left figure, shows the image of a Pd particle obtained by means of high resolution transmission electron microscopy (HRTEM). The fringes with lattice of 1.95 and 2.23 Å are calculated measuring the separation of 10 planes and dividing these values by 10. These values can be indexed as {111} and {200} of fcc Pd, respectively. The size distribution was the same than the obtained by SEM for Pd@PB80_20A4 and Pd@PB20_80A1, the small difference were likely to be because of the aggregation and the three-dimensional position of the particles in SEM, complicating an accurate measurement of isolated particles. 50 60 70 80 90 100 0 5 10 15 20 25 Percentage (%) Particle size (nm)
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|321 JoséGarcíaCalvo|PhDThesis 8.2.4. Summary of Pd-NPs structures: The results for the optimized Pd modified films from Section 8 might be summarized as follows: The films containing mixtures VP/A lead to Pd-NPs providing that the ester was a short aliphatic chain. It was only necessary to use a PdCl2ꞏ2NaCl solution in water. The oxidation state of Pd was detected by UV-Vis absorbance and EDX analysis. For PB20_80A1 the Pd-NPs were formed spontaneously and deposited in the surface of the polymer. It was possible to transfer to other surfaces the Pd-NPs from Pd@PB20_80A1. PB80_20 polymers absorbed Pd(II). They needed for H2 atmosphere to reduce Pd(II) to PdNPs. Pd@PB80_20 presented Pd-NPs distributed in the polymer with less aggregation than Pd@PB20_80A1, and not only in the surface. The Pd-NPs for these two polymers were polycrystalline nanodendrites with an average diameter of 75 nm.
322|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis 9. PALLADIUM SUPPORTED POLYMERS AS CATALYSTS Dendritic-like palladium nanostructures have found use in many fields, such as the preparation of electrodes for electrocatalytic ethanol oxidation,71 formic acid oxidation,72 or lithium–oxygen batteries.73 Albeit, there are more possible applications; regarding that, this work is oriented to the use of supported Pd catalyst on the highly interesting74 semihydrogenation of disubstituted alkynes with molecular hydrogen. The importance of this new catalyst comes from the easiness to prepare these materials and their special characteristics. An ideal Pd heterogenous catalyst should possess the next characteristic: Using green solvents (avoiding DCM or DMSO). Having high recyclability of the catalyst. Getting no leaching of Pd to the solutions. The relation [Pd (mol)]/polymer surface (cm2) has to be low. The turnover number (TON), (number of moles converted to product)/(surface of the heterogeneous catalyst), should be as high as possible. In literature, there are several examples of supported catalysts for the reduction of triple bonds. The most common one, and the industrial standard, was Pd/C, which has been widely use, although it is very expensive. In addition, most of these catalysts have some issues such as requiring a filtration, big amounts of the catalyst and low (if some) recyclability. As a consequence, it was of upmost importance the development of a new catalyst that overcome those problems. With this aim in mind, Pd(0) modified polymers were found as a solution for getting very efficient and recyclable heterogenic catalysts for selective reduction of triple bonds. After some preliminary testing, some variations of the different polymers were selected as representative examples. In particular, from all the polymers that contained palladium in its surface the ones with better properties were selected: Pd@PB20_80A1 which had Pd nanoparticles transferable to other surfaces. Pd@PB80_20A1 and A4, which had Pd nanoparticles anchored to the polymer. Pd@JG25SA, which was, apparently, very similar to PB80_20. Furthermore, a reagent was selected as a model to compare the results in the reduction, optimizing the reaction conditions as follows: The solvent had to be a good solvent for most organic compounds. With that purpose, the reaction was tested in MeOH (a green and common solvent), DCM and THF. The yields were calculated by NMR and/or column purification, depending on the byproducts. The film was 0.5×0.5 cm / 5 mL solution. 71A) K. Qi, Q. Wang, W. Zheng, W. Zhang, X. Cui, Nanoscale 2014, 6, 15090–15097. B) S. J. Ye, D. Y. Kim, S. W. Kang, K. W. Choi, S. W. Han, O. O. Park, Nanoscale 2014, 6, 4182–4187. 72 A) A. Klinkova, P. D. Luna, E. H. Sargent, E. Kumacheva, P. V. Cherepanov, J. Mater. Chem. A 2017, 5, 11582–11585. B) T. Huang, S. K. Moon, J.-M. Lee, Sustainable Energy Fuels 2017, 1, 450–457. 73 S. J. Ye, D. Y. Kim, D. W. Kim, O. O. Park, Y. Kang, J. Mater. Chem. A 2016, 4, 578–586. 74 D. B. Burueva, K. V. Kovtunov, A. V. Bukhtiyarov, D. A. Barskiy, I. P. Prosvirin, I. S. Mashkovsky, G. N. Baeva, V. I. Bukhtiyarov, A. Y. Stakheev, I. V. Koptyug, Chem. Eur. J. 2018, 24, 2547–2553.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|323 JoséGarcíaCalvo|PhDThesis For the initial experiments, the selected sample for reduction was dimethyl acetylenedicarboxylate (DMAD), that was added as much as 500 mg each 5 ml of solvent. DMAD was interesting for several reasons so as to check the capabilities of the polymer as a catalyst: o It was soluble in common solvents, such as methanol or dichloromethane. o It was simple to analyse, it is liquid and there is only one signal in the 1HNMR spectra. o It was not an isolated internal triple bond, DMAD is a Michael acceptor with an electrondeficient triple bond, conjugated with the ester groups. However, it is not a large conjugation which means that the reaction would not be totally favoured nor disfavoured. Specific conditions for the tests: Thus, in 10 mL vials, 500 mg of DMAD were dissolved in 5 mL of solvent, a piece of polymer, 0.5×0.5 cm, was added to the solution and the vial was placed in a reactor, then H2 was introduced to the chamber until reaching 5 atm and the mixture remained under H2 for 15 hours. After that, excess hydrogen was released, the solid catalyst was removed and the solvent evaporated (Figure 48). Figure 48. Structure of the possible products of the reduction of the triple bond from DMAD (up) and 1HNMR spectrum of the crude obtained from using the polymer as catalyst (down). The film used for the example was Pd(II)PB20_80A4, without previous cleaning nor reduction. In 1HNMR from the Figure 48 it can be seen the presence of a mixture between the starting material and the products cis (maleate) and reduced to the simple bond (succinate). The trans product (fumarate) was rarely obtained (THF solutions) and in very low yields. As a result, all the expected products were easily detected by 1HNMR.
324|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis 9.1. Reduction of DMAD with palladium modified films The following table (Figure 49) summarizes the results: Polymer Solvent Initial (%) Cis (%) Trans (%) Simple (%) Pd@PB80_20A4* - 93 7 0 0 MeOH 10 90 0 0 DCM 26 74 0 0 THF 37 46 9 8 Pd@PB20_80A1 - 90 10 0 0 MeOH 6 94 0 0 DCM 71 29 0 0 THF 83 8 9 0 Pd@JG25SA2 - 93 7 0 0 MeOH 56 40 0 4 DCM 97 3 0 0 THF 9 8 6 0 *Pd@PB80_20A4 palladium was previously reduced under H2 atmosphere 5 atm, 1 hour and washed 5 times with 10 mL of MeOH. Figure 49. Reduction of the triple bond from DMAD in presence of 3 different heterogenous catalysts, 3 different solvents and with no solvent. Previously to give an analysis of the results, there were several characteristics to take into account: Best yields (% reacted) were reached for PB20_80A1 and PB80_20A4 and using MeOH as solvent. In addition, the yields of several repetitions are shown in the table in Figure 50. Polymer Initial (%) Cis(%) Simple(%) 1st Pd(II)@PB80_20A4* 5 72 23 Pd@PB20_80A1 6 94 0 2nd Pd@PB80_20A4* 9 85 6 Pd@PB20_80A1 21 79 0 3rd Pd@PB80_20A4 5 95 0 Pd@PB20_80A1 32 68 0 4th Pd@PB80_20A4 8 92 0 Pd@PB20_80A1 65 35 0 *The films were not previously reduced and washed before the reactions. If done so, the yields are similar to the reactions done afterwards. Figure 50. Yields after repeating the reduction of the triple bond from DMAD, 4 times in MeOH as solvent. Pd@PB20_80A1 had the drawback (for catalytic purposes) of the transferable particles, if the polymer was scratched, or put in contact with some surfaces, the quantity of particles decreased. Also, there could be transference to the solutions (although they were not soluble in MeOH) loosing effectiveness and reducing the recyclability.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|325 JoséGarcíaCalvo|PhDThesis Pd(II)@PB80_20A4 had the simple bond (succinate) as a by-product in the first two reactions. This is the consequence of using the not reduced-not washed polymer. After the second reaction, it is constant at least for 4 reactions more with yields of the cis product (maleate) superior 90 %. Pd@PB80_20A4 has the best yield/recyclability but, in order to not having by-products, such as the succinate, the polymer should be carefully washed with MeOH, and previously subjected to hydrogen atmosphere. As a result, it can be used as a catalyst with high yields. Other molecules reached also quantitative yields in MeOH (see Section 10, “catalytic reduction of compounds with biological interest”). This yield changes depending on the solvents and the molecule to reduce (the surroundings of the triple bond), apart from being limited by its TON. Pd(II)JG25SA2 gave similar results to PB80_20A4, but worse yield. Additionally, the mechanical properties were worse; therefore, it was not further studied. 9.2. Leaching of the polymers One of the biggest problems of using Pd supported catalysts is the quantity of the palladium that may leach to the solutions. Several situations were studied, to do so an aliquot of the solution from the reaction explained in section 9.1. (1 mL) was analysed by ICP mass (three times each), calibrating with palladium solutions. The results were: Pd@PB20_80A1: the palladium in solution was negligible, due to the insolubility and despite being transferable. The quantity detected was inferior to 2 µM. Pd(II)@PB80_20A4, not reduced/washed film: the quantity detected was between 0.4 - 0.1 mM. Pd@PB80_20A4 after reducing and washing the polymer: the quantity detected was inferior to 6 µM. These data were also important in order to show the necessity of using washed polymers and the low concentration of palladium in solution when doing so. 9.3. Other important characteristics studied Reaction time: the reaction time was tested in MeOH solution for DMAD reductions. The yield was constant when the reactions lasted 5 hours of more. However, less time led to lower yields, although this parameter should be adjusted depending on the specific molecule to reduce, and also depended on other parameters like the concentration or temperature. Polymers without crosslinker: PBM2, PBMEGMA and PBMA did not contain crosslinker, that caused that they were soluble in most of the organic solvents. Despite this fact, they were tested in solution but the reaction did not work. Turnover number (TON): the TON was defined as (product moles)/surface. It was calculated based on the studied reduction reaction. In this regard, the catalyst turnover number was calculated considering a yield of 90 % and 6 reactions; obtaining a TON of 75 mmoles/cm2.
326|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis 10. CATALYTIC REDUCTION OF COMPOUNDS WITH BIOLOGICAL INTEREST With the polymer with better results (Pd@PB80_20A4) and the optimized conditions, a series of compounds were chosen as potential substrates for semihydrogenation reactions because of their properties and applications in different fields (Figure 51). For showing the importance of this kind of reactions, it was necessary to check if the procedure permitted the quantitative and selective transformation of several important pharmacological drugs or intermediates into related compounds. In this regard, the catalyst had to have no by-products and restricted reactivity to triple bonds, and not halogens or double bonds. Figure 51. Molecules with triple bonds studied Mifepristone: a steroidal progesterone75 and glucocorticoid receptor antagonist,76 used as contraceptive agent and in the treatment of breast cancer.77 By cis-semihydrogenation reduction, was transformed into aglepristone,78 a related progesterone antagonist used for the 75A. Yamada, Y. Kazui, H. Yoshioka, A. Tanatani, S. Mori, H. Kagechika, S. Fujii, ACS Med. Chem. Lett. 2016, 7, 1028−1033. 76 C. A. Sanhueza, M. M. Baksh, B. Thuma, M. D. Roy, S. Dutta, C. Préville, B. A. Chrunyk, K. Beaumont, R. Dullea, M. Ammirati, S. Liu, D. Gebhard, J. E. Finley, C. T. Salatto, A. King-Ahmad, I. Stock, K. Atkinson, B. Reidich, W. Lin, R. Kumar, M. Tu, E. Menhaji-Klotz, D. A. Price, S. Liras, M. G. Finn, V. Mascitti, J. Am. Chem. Soc. 2017, 139, 3528−3536. 77 Y. Lin, R. Liu, P. Zhao, J. Ye, Z. Zheng, J. Huang, Y. Zhang, Y. Gao, H. Chen, S. Liu, J. Zhou, C. Chen, H. Chen, Eur. J. Med. Chem. 2018, 146, 354-367. 78 A. Gogny, F. Fiéni, Theriogenology 2016, 85, 555–566.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|327 JoséGarcíaCalvo|PhDThesis treatment of various progesterone-dependent physiological or pathologic conditions in veterinary medicine. Efavirenz:79 a non-nucleoside reverse transcriptase inhibitor used as a first-line anti-HIV drug.80 It was transformed into (Z)-dihydroefavirenz,81 a predicted but yet unavailable efavirenz analogue. Thus, expanding the pharmaceutical possibilities of fluorine-containing pharmaceuticals.82 Tazarotene:83 a receptor selective retinoid that specifically binds to retinoid receptors in the skin after ester hydrolysis.84 It is currently used for topical treatment of psoriasis, it was then converted into (Z)-dihydrotazarotene [ethyl (Z)-6-(2-(4,4-dimethylthiochroman-6yl)vinyl)nicotinate] a new potential retinoid.85 8-Bromo-7-(2-butyn-1-yl)-3-methylxanthine (Br-R): a key intermediate for the synthesis of linagliptin,86 a xanthine dipeptidyl peptidase-4 (DPP-4) inhibitor for the treatment of type 2 diabetes.87 It was transformed into (Z)-8-bromo-7-(2-buten-1-yl)-3-methylxanthine, a new intermediate on the way to new DPP-4 inhibitors.88 The presence of the sensitive bromo substituent in the reduced product constitutes a remarkable proof of the selectivity of the catalytic semihydrogenation selectivity of the described process, although in this case a purification step was required. Pentacosa-10,12-diynoic acid (2t-Lipo): a carboxylic acid with a lepidic chain with triple bonds, it serves as an example of the capabilities for reducing triple bonds in molecules with potential applicability in creating lipidic membranes for vesicles. In view of the importance of all-(Z) polyunsaturated lipids for biological purposes, the available pentacosa-10,12-diynoic acid was selected and subjected the compound to semihydrogenation conditions, from which a scarce yield was obtained for the corresponding (Z),(Z)-pentacosa-10,12-dienoic acid due to the natural tendency of the starting material to form Langmuir–Blodgett structures, that polymerize under the light giving blue and red solutions.89 To compare the selectivity of the reaction with reported conditions of a common palladium/carbon catalyst, the same reactions were performed in the presence of commercial palladium on carbon (10%), obtaining in all cases mixtures of products coming from different hydrogenation patterns, usually triple bonds to saturated hydrocarbons as well as producing the reduction of internal double bonds, the dehalogenation products and mixtures of several of the possible products of reduction. In contrast, the selectivity of the heterogenic catalysis with modified films was verified. 79S. Li, J.-A. Ma, Chem. Soc. Rev. 2015, 44, 7439–7448. 80 M. M. Bastos, C. C. P. Costa, T. C. Bezerra, F. de C. da Silva, N. Boechat, Eur. J. Med. Chem. 2016, 108, 455-465. 81 D. D. Christ, A. J. Cocuzza, S. S. Ko, J. A. Markwalder, A. E. Mutlib, R. L. Jr. Parsons, M. Patel, S. P. Seitz, U.S. patent 1999, US 5874430 A 19990223. 82 Y. Zhou, J. Wang, Z. Gu, S. Wang, W. Zhu, J. L. Aceña, V. A. Soloshonok, K. Izawa, H. Liu, Chem. Rev. 2016, 116, 422−518. 83 I. S. Makarov, C. E. Brocklehurst, K. Karaghiosoff, G. Koch, P. Knochel, Angew. Chem. Int. Ed. 2017, 56, 12774 –12777. 84 A. M. Mansour, Polyhedron 2016, 109, 99–106. 85 R. Álvarez, B. Vaz, H. Gronemeyer, Á. R. de Lera, Chem. Rev. 2014, 114, 1−125. 86 M. Eckhardt, T. Klein, H. Nar, S. Thiemann, Discovery of Linagliptin for the Treatment of Type 2 Diabetes Mellitus, in: J. Fischer, D. P. Rotella, Eds., Successful Drug Discovery, Wiley-VCH Verlag, 2015, Ch. 7. 87 W.-L. Wu, J. Hao, M. Domalski, D. A. Burnett, D. Pissarnitski, Z. Zhao, A. Stamford, G. Scapin, Y.-D. Gao, A. Soriano, T. M. Kelly, Z. Yao, M. A. Powles, S. Chen, H. Mei, J. Hwa, ACS Med. Chem. Lett. 2016, 7, 498−501. 88 L. Juillerat-Jeanneret, J. Med. Chem. 2014, 57, 2197−2212. 89 S. Balakrishnan, S. Lee. J.-M. Kim, J. Mater. Chem. 2010, 20, 2302–2304.
328|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis 10.1. Conditions of the reaction and yields obtained The reaction conditions were similar to the DMAD reduction. 50-100 mg of the different analytes were dissolved in 2-5 mL of MeOH and put under 5 atm of H2 overnight. Pd@PB80_20A4 cis (%) Efavirenz >99 Mifepristone >99 Br-R 90 Tazarotene 85 2t-Lipo* ≥50 Figure 52. Yields (%) for the cis product for the selected molecules. These yields were calculated after the purification by column chromatography of the product except for 2t-Lipo. Characteristics of each reaction: Efavirenz and Mifepristone had a favoured reduction reaction, with yields superior to 99 % and no by-products. Br-R had several minor by-products, needing column chromatography for the separation of the cis product. Tazarotene was less reactive under the same reaction conditions. Therefore, an easy separation from the remaining starting material, that was reused, was also needed. 2t-Lipo was only slightly soluble in MeOH and chloroform. Nevertheless, the reaction occurred with high yields as it can be deduced from 1H NMR, which demonstrated its applicability for the purpose of reducing triple bonds in lipidic chains. In this case, the yield was estimated from the 1H NMR because a column chromatography was not possible due to its low solubility. 10.2. Results for Pd films reduction and comparison with Pd/C catalyst In this section, the results of using the polymeric heterogenic catalyst and Pd/C are compared. The main differences are showed by the purified NMR spectra and the results of high resolution mass spectra analysis. In addition, the compounds were fully characterized, although the specifics are more detailed in the Experimental Appendix 4. In this way, the reaction was performed with the catalysts under similar conditions to previous synthesis; 100 mg of analyte, 5 mL of MeOH and 100 mg of Pd/C or a 0.5×0.5 piece of Pd@PB80_20A4. All components were put at 5 atm pressure and stirred for 16 hours.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|329 JoséGarcíaCalvo|PhDThesis 10.2.1. Mifepristone: Figure 53. 1H NMR spectra (CDCl3), comparison between Mifepristone and Aglepristone, its reduced specie. Analysis performed for the product of Pd@PB80_20A4 reduction. In this case, the product obtained under Pd/C, was a mixture of different reductions, as it was showed by the mass spectra. Mifepristone was likely to be reduced in different positions due to the presence of double bonds in its structure. The major product was the reduction with six more hydrogen atoms (probably the triple bond to simple and one of the double bonds) and in less quantity the reductions adding four and eight hydrogen atoms. Pd@PB80_20A4: HRMS (ESI-TOF) m/z calcd for C29H38NO2 (M+): 432.2897; found: 432.2911. Pd/C: HRMS (ESI-TOF): (a) m/z calcd for C29H40NO2 (M+): 434.3054; found: 434.3057. (b) m/z calcd for C29H42NO2 (M+): 436.3210; found: 436.3217. (c) m/z calcd for C29H44NO2 (M+): 438.3367; found: 438.3367. HA+HB
336|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis 12. SUPPORTED GOLD AND PALLADIUM PARTICLES. SUMMARY Supported Au(0): Properties of the Au substituted polymer: -Composition: oJG25_SA2 is formed by hydroxyethylacrylate (2HEA, 95%) and propargyl methacrylate (PGM, 5%). ethylene glycol dimethacrylate (EGDMA, 5%) was used as cross-linking agent. oJG25 was as JG25_SA2, but JG10 substituted which formed a triazole. -In the presence of the fluorescent probe, JG25, the shape, size and dispersion of Au(0) particles were higher. -High concentrations of Au(III) solutions increased faster the number of particles than their size. -Low relation volume/surface led to more variation in the shape of the nanoparticles, not so much in their number. -The size of the nanoparticles was between 20 nm to 400 nm of diameter, depending on the conditions. -The different shapes encountered were tridimensional, amorphous, spherical, tetrahedral, tetrahedral truncated…; bidimensional, triangles; and monodimensional (or with a preferential grow direction), bars with different thickness. -In water Au-NPs were located over the polymer while, when using DMF, NPs inside the film were also obtained. Properties as catalyst of the supported Au(0) films: •They were checked to catalyse Suzuki reactions, with similar yields compared to general conditions used in published results (using Pd(PPh3)4 as catalyst). •They should be used under mild working conditions, THF:Water between 40 or 60 % THF and temperature around 65ºC overnight, to maximize the relation yield/recyclability. •They were tested to work for at least 4 reactions with the same piece of polymer. •The TON for the Suzuki coupling was 2 mmol/cm2. Properties of the Pd modified films: (best results) Composition of the films: oPB20_80A1 contained vinylpyrrolidone (VP, 20%) and methyl acrylate (A1, 80%). The crosslinker was ethylene glycol dimethacrylate (EGDMA, 5%). oPB80_20A4, contained vinylpyrrolidone (VP, 80%) and n-butyl acrylate (A4, 20%). The crosslinker was ethylene glycol dimethacrylate (EGDMA, 5%). PB20_80A1, when in presence of PdCl2ꞏ 2NaCl water solution formed a black Pd(0) layer over its surface; once analysed, it was concluded that they were Pd(0)-NPs.
CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS|337 JoséGarcíaCalvo|PhDThesis PB80_20A4, when in presence of PdCl2ꞏ 2NaCl solution was capable to absorb Pd(II). These Pd may be easily reduced to Pd(0)-NPs, by being in the presence of hydrogen gas for several minutes. The aggregation of the Pd-NPs is much higher for Pd@PB20_80A1, although they may be easily transferred to other surfaces. The nanoparticles formed were polycristalline nanodendrites with an average diameter of 75 nm. Properties as catalyst of supported Pd-NPs: They were checked to be an efficient heterogenic catalyst for semihydrogenation reduction of alkynes in green solvents,95 such as methanol. Dimethylacetylenedicarboxylate (DMAD) was studied as standard for regioand stereoselective reduction to its (Z) hydrogenated product, dimethyl maleate. It occurred under hydrogen atmosphere (5 atm) in less than 15 hours, with yields around 95 %. The recyclability of the product was checked to be of at least 6 times when using Pd@PB80_20A4. The leach of palladium to the solution was detected to be inferior to 6 μM. The calculated TON was 75 mmoles/cm2. The cis selective reduction was performed successfully to products with pharmaceutical interest. As example 5 products were tested, mifepristone, efavirenz, tazarotene, 8-Bromo7-(2-butyn-1-yl)-3-methylxanthine and pentacosa-10,12-diynoic acid. Obtaining 100 % selectivity in the transformation from alkynes to (Z)-alkenes, against alkanes. The drugs mifepristone and efavirenz were quantitively reduced. Tazarotene and the xanthine derivative needed for purification; despite this fact, they were obtained in high yields. In conclusion, the remarkable selectivity and the simplicity in the preparation of the catalysts, with no need for additional ligands, makes the systems competitive to known solid supported metallic nanoparticles catalysts for Suzuki coupling and the semihydrogenation reaction. The polymer supported metallic nanoparticles material can be very useful for heterogeneous catalysis in the synthesis of important drugs, or intermediates for drugs, in a highly sustainable chemistry. 95C. Capello, U. Fischer, K. Hungerbühler, Green Chem. 2007, 9, 927–934.
338|CHAPTER4.SUPPORTEDGOLDANDPALLADIUMNANOPARTICLESFORCATALYSIS JoséGarcíaCalvo|PhDThesis 13. RESUMEN DEL CAPÍTULO El objetivo de este capítulo consistía en mostrar un nuevo procedimiento para la modificación de superficies con partículas metálicas de oro y paladio; así como ilustrar alguna de las posibles aplicaciones de estos materiales modificados. Durante el desarrollo de materiales poliméricos modificados para la detección de Hg(II), (Capítulo 1) se observó que estos polímeros, en contacto con disoluciones acuosas de Au(III), eran capaces de producir espontáneamente nano y micro-partículas de oro (0) unidas a su superficie. Posteriormente, un estudio en profundidad permitió determinar que tanto la forma como el tamaño de estas partículas eran regulables cambiando las condiciones de trabajo. Así, entre otros parámetros, se estudió la dependencia entre el disolvente, la concentración de Au(III) o la proporción volumen/superficie lo cual permitió la modificación controlada de estos polímeros; posteriormente estudiados por técnicas como absorción o microscopia electrónica. El siguiente paso fue la búsqueda de una aplicación a estos materiales. De las múltiples posibilidades, la investigación se centró en la utilización de los mismos como catalizadores heterogéneos. Para ello, se eligió la reacción de Suzuki de acoplamiento C-C. Los resultados fueron satisfactorios, obteniendo rendimientos similares a los obtenidos por métodos clásicos, pero con claras ventajas como la posibilidad de reciclar el catalizador. De igual modo que para la modificación de superficies con oro se planteó la posibilidad de usar otros metales. Entre ellos paladio. Este metal se caracteriza por poseer propiedades que lo hacen de gran utilidad en aplicaciones como catálisis o almacenamiento de hidrógeno. Tras el testeo y la optimización de condiciones con diferentes polímeros se escogieron dos variantes que contenían mezclas de polivinilpirrolidona y un alquil-acrilato. Por un lado, estos polímeros fueron capaces de ser modificados de manera sencilla con nanopartículas de Pd(0), e incluso con la posibilidad de transferir estas partículas una vez formadas. En cuanto a la utilidad de los materiales modificados con nanopartículas de Pd(0), el estudio se centró en sus aplicaciones como catalizadores, en este caso para la semi-hidrogenación selectiva de triples enlaces internos a dobles (Z-alquenos). Esta propiedad, dio resultado para un conjunto de moléculas con distinto interés biológico (precursores de medicamentos y hormonas) presentando a su vez altos rendimientos y siendo un catalizador reutilizable y que funciona en disolventes verdes, principalmente alcoholes.
FINAL CONCLUSIONS
FINALCONCLUSIONS|339 JoséGarcíaCalvo|PhDThesis As it was explained in the thesis, the objectives and achievements were extensive. In the next part, the conclusions are summarized by chapters: Chapter 0: Introduction to fluorescent probes. The properties and ideal characteristics of molecular sensors were described. It was highlighted the importance of the specificity, selectivity, reusability, low-cost, fast operation mechanism, cheap maintenance and portability. The importance of fluorescent sensors and how they work was explained. From their many characteristics it is worth to remark the importance of having OFF-ON sensors for higher selectivity, as well as how the different types of sensors work (complexation or reaction) and how they may be properly designed (system fluorophore-receptor). A protocol was elaborated so as to work with fluorescent probes, for getting repeatable and reliable results. In doing so, different studies give characteristics and parameters that are of great importance: o Solvatochromism (response to different solvents). o Tests with different species, cations, anions, oxidative, reductive species, amines… o Work concentration and molar extinction coefficients (ε). o Kinetic effects. o Titration methods. o Thermodynamic equilibrium constant calculation (K). o Limits of detection (LODs). o Fluorescence quantum yields (Φ). o Fluorescence decay lifetime (τ). Chapter 1: Fluorescent probes for detection of Hg(II) derivatives. The importance of having probes for detection of Hg(II) derivatives was explained. It was remarked the importance of the organic species, such as MeHg(II), being a bioaccumulative cation presented in some fish samples and very toxic for living organisms. A series of water soluble fluorescent probes, highly selective to Hg(II) derivatives, were synthetized and characterized. Selective detection of Hg(II) and MeHg(II) in cellular environments was achieved. Polymeric modified materials were created, capable of detecting Hg(II) presence in water. o The LOD of the polymer soluble in water, JG32, for Hg(II) was 1.5 ppm and for MeHg(II) it was 6.5 ppm. o The film with water affinity, JG25, had a LOD for Hg(II) of 1.3 ppm and for MeHg(II) it was 0.3 ppm. The supported probe JG25 was adapted to work for detection and quantification of Hg(II) from fish. Concentrations around 1 ppm of mercury cations were detected on dogfish, swordfish and tuna. .
340|FINALCONCLUSIONS JoséGarcíaCalvo|PhDThesis Chapter 2: Perylenemonoimides. Introduction and general properties. The advantages and properties of perylene imide derivatives as fluorescent backbones were explained. It was of particular importance to explain their stacking properties, highly dependent on the different substituents and their position. The synthesis for PMI derivatives as starting materials was optimized. A comparison between PMIs substituted in different positions was performed in the same solvent (chloroform). o The wavelength with the maximum of absorption was around 500-520 nm. Except for the ruthenium complex JG10L (576 nm). o The molar extinction coefficient was maximized by creating a combination PMIBodipy (JGBod). It makes them a potential candidate for applications in sensing and photoelectronics. o The fluorescence lifetime decays were between 3.5-5.1 ns. o The fluorescence stokes shift, when comparing between more and less polar solvents, was maximized in probes such as JG116 and JG125 that were the examples containing a pyrimidine or pyridine group. JG119c2 resulted to have applications as cellular biomarker with interaction with Gquadruplex DNA. It was selected for future studies, currently under development. Chapter 3A. PMI derivatives for detection of explosives. TATP is a white powder that could act as explosive in terrorist attacks; because of that and some particular properties that it possesses, such as the difficulty to give a controlled explosion, its detection is of high interest. A PMI derivative, JG125, was selected as the fluorescent probe with response to TATP oxidation, by increasing fluorescence. o The characteristics for detection in solution were measured, and it was determined that it was possible to detect 0.27 mg of TATP with JG125 (2.5 μM) in CHCl3:MeOH 9:1, in 2.5 mL solution. o The detection was selective to other oxidants such as water peroxide, and with the possibility to be distinguished from acid vapours. A series of materials, silica derivatives, were modified for the detection of TATP in the gas phase, 0.12 mg was the limit of detection of TATP gas. Scheme showing the interpretation of the results and the application of the different probes.
FINALCONCLUSIONS|341 JoséGarcíaCalvo|PhDThesis Chapter 3B. PMI-Ru(II) complexes for detection of CO. The interest and antecedents for using Ru(II) complexes for CO sensing were: its toxicity as gas (being a semicombustion residue) and its crucial role in some biological processes. Ru(II) complexes for detection of CO have high selectivity in comparison to electrochemical methods, being more sensitive and cheaper. In addition, they are less toxic and with better results compared with other metallic complexes, such as Pd. It was performed the synthesis of three different complexes Ru(II)-PMI; JG10L, JG11L and JG12L; containing a PEG to increase the final solubility in water. The Ru(II) complexes were useful for CO detection, but not soluble in 100 % water media nor in cellular environments. It was explained the potential applicability of the complexes not only as CO detectors but as light sensitive derivatives. This characteristic may be useful for future research in photoelectronics and biological applications. JG10L presented selective response to other analytes of interest, such as isonitriles or cyanide. The next step would be the creation of derivatives changing the triphenylphosphine groups so as to improve the properties and applications. Chapter 3C. PMI derivatives for K+ and Pb2+ sensing. The importance of detecting K(I) was briefly summarized, being part of most biological processes; as a consequence, it is interesting not only in solution but in cellular environments too. Cereulide and similar derivatives from natural potassium ionophores are of great importance. They have direct influence in potassium transport between intra-extracellular media; acting by equilibrating concentrations or by disrupting the equilibria, which leads to the death of the cells. Pb(II) is a cation that is toxic for living organisms. Its presence may come from different sources, the most remarkable are the incorrectly treated tap water samples (“Flint water crisis”). Because of that, it was determined that measuring concentrations of Pb(II) in coloured water samples (containing solid from corrosion) would be of high interest. Two fluorescent PMI probes were synthetized, with selective sensitivity to K(I) and Pb(II); a triazacryptand (JG103) and a crown ether (JG76). Additionally, a natural ionophore of K+, cereulide, was synthetized in laboratory; along with some derivatives. They were employed to research about some of their possible biological applications and to determine how the properties of these ionophores could be modulated. The crown ether probe JG76 resulted to be an efficient probe for detection of K(I) and Pb(II) in ethanol solution (LOD = 2.3 ppb of K(I) and 6 ppb of Pb(II)) and in cell cultures. Additionally, it was possible to detect the presence of cereulide from contaminated cultures (LOD = 240 ppb). The studies of the effect of using several cereulide derivatives had promising results, being capable of tuning the potassium affinity and perform cellular location by fluorescent modification. The crown ether probe was modified to be anchored to a polymeric material (JG151dp). The modified polymer (JG151dp) was used successfully in detection of Pb(II) in coloured water samples with no interference from potassium cations. (LOD = 66 ppb)
342|FINALCONCLUSIONS JoséGarcíaCalvo|PhDThesis Chapter 4. Supported gold and palladium nanoparticles for catalysis. The different properties of gold and palladium nanostructured materials were summarized, being a state of the art topic with potential applications in catalysis, biomedicine (drug delivery of cancer therapy), biological and chemical sensors (cellular imaging and biosensors), electronics, environmental remediation and hydrogen storage (for Pd). The synthesis of supported gold and palladium particles was optimized to be a one-step synthesis in water over a modified surface. The polymer containing hydroxyethylacrylate (2HEA, 95%) and propargyl methacrylate (PGM, 5%), ethylene glycol dimethacrylate (EGDMA, 5%), and its derivative with a triazole (+ JG10), was capable of the spontaneous formation of gold particles in their surface, in the presence of Au(III) in water. The conditions were optimized. For the synthesis of Pd modified polymers two derivatives were optimized and fully characterized; they contained vinylpyrrolidone (VP), an alkyl acrylate (A) and different Pd substitutions depending on their percentages. The most useful for catalysis was called Pd@PB80_20A4, which was covered of homogeneous 75-80 nm Pd-NPs. Gold covered polymers were used as an alternative catalyst for Suzuki reactions, getting the same yields in similar conditions (THF:water mixtures). The main advantage was the recyclability of the polymer; up to 4 times. The TON for the Suzuki coupling was 2 mmol/cm2. Palladium covered polymers were used for selective semihydrogenation of triple bonds to obtain the (Z)-alkene derivative. The reduction process was tested to be selective against double bonds and halogenated molecules and it was successfully used for reduction of specific drugs with pharmacological interest. They presented high recyclability, up to 6 times, with no significant leaching of palladium to the reaction solvent and a TON of 75 mmol/cm2.
ANNEX
350 | ANNEX José García Calvo | PhD Thesis José García Calvo; Patricia Calvo Gredilla; Tomás Torroba Pérez; Nathalie Zink Lorre; Ángela Sastre Santos; Fernando Fernández Lázaro; Enrique Font Sanchís PCT/ES2016/0709; España. Register: 23/12/2015 4. Oral communications (12; 6 different symposium): Versatility of polysubstituted PDI: Synthesis, optical properties and biological applications. Alberto Diez-Varga, Tomás Torroba, Patricia Calvo-Gredilla, Daisy Romero-Velásquez, José GarcíaCalvo, Mónica L. Fanarraga, Eloísa G. Lavado. XXVII Biennial Meeting in Organic Chemistry. Santiago de Compostela, Galicia, España; 20/06/2018 – 22/06/2018. A smart material for detection of mercury in fish. José García Calvo; Patricia Calvo Gredilla; Víctor García Calvo; Jose Miguel García Pérez; Saúl Vallejos Calzada; Félix Clemente García; Maria José Rojo Cámara; Maria Teresa Rodríguez; Alberto Díez de la Varga; Daisy Carolina Romero Velásquez; Eva Clara Antón García; Marcos Ibáñez Llorente; Tomás Torroba Pérez. XXXVI Reunión bienal de la sociedad española de química. Sitges, Cataluña, España; 25/06/2017 - 29/06/2017. Fluorescence Quenching of Perylenediimides by Absorption with Graphene Oxide. Patricia Calvo Gredilla; José García Calvo; Daisy Carolina Romero Velásquez; Alberto Díez de la Varga; Eva Clara Antón García; Marcos Ibáñez Llorente; Víctor García Calvo; Tomás Torroba Pérez. XXXVI Reunión bienal de la sociedad española de química. Sitges, Cataluña, España; 25/06/2017 - 29/06/2017. Surface coating by gold nanoparticles on functional polymers. José García Calvo; Víctor García Calvo; Jose Miguel García Pérez; Saúl Vallejos Calzada; Félix Clemente García; Tomás Torroba Pérez. XXXVI Reunión bienal de la sociedad española de química. Sitges, Cataluña, España; 25/06/2017 - 29/06/2017. Development of new sensors for process of fluorogenic recognition of cations and oxidizers.
[Document text truncated for crawler view.]