Molybdenum(VI) complexes with a chiral L-alanine bisphenol [O,N,O,O’] ligand : Synthesis, structure, spectroscopic properties and catalytic activity

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Molybdenum(VI) complexes with a chiral L-alanine bisphenol [O,N,O,O’] ligand : Synthesis, structure, spectroscopic properties and catalytic activity © 2023 The Authors. Published by Elsevier B.V. Published version Peuronen, Anssi; Kivelä, Henri; Salonen, Pasi; Eskonen, Ville; Karman, Marta; Lahtinen, Manu; Romanowski, Grzegorz; Lehtonen, Ari Peuronen, A., Kivelä, H., Salonen, P., Eskonen, V., Karman, M., Lahtinen, M., Romanowski, G., & Lehtonen, A. (2023). Molybdenum(VI) complexes with a chiral L-alanine bisphenol [O,N,O,O’] ligand : Synthesis, structure, spectroscopic properties and catalytic activity. Inorganica Chimica Acta, 553, Article 121519. https://doi.org/10.1016/j.ica.2023.121519 2023 Inorganica Chimica Acta 553 (2023) 121519 Available online 7 April 2023 0020-1693/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Molybdenum(VI) complexes with a chiral L-alanine bisphenol [O,N,O,O’] ligand. Synthesis, structure, spectroscopic properties and catalytic activity Anssi Peuronen a , b , Henri Kivel¨ a a , Pasi Salonen a , Ville Eskonen a , Marta Karman c , Manu Lahtinen b , Grzegorz Romanowski c , Ari Lehtonen a , * a Department of Chemistry, University of Turku, FI-20014 Turku, Finland b Department of Chemistry, University of Jyv¨ askyl¨ a, P.O. Box 35, FI-40014 JY, Finland c Faculty of Chemistry, University of Gda´ nsk, Wita Stwosza 63, PL-80308 Gda´ nsk, Poland ARTICLE INFO Keywords: Molybdenum L-alanine Epoxidations Catalysis ABSTRACT Dioxidomolybdenum(VI) compound [MoO 2 Cl 2 (dmso) 2 ] reacts with a chiral tetradentate O 3 N-type L-alanine bisphenol ligand precursor (Et 3 NH)H 2 L ala to form an oxidochloridomolybdenum(VI) complex [MoOCl(L ala )] (1) as two separable geometric isomers with phenolate groups in cis or trans positions. The single crystal X-ray and NMR analyses of cisand trans-1 reveal that the complexes are formed of monomeric molecules, in which the ligand has a tetradentate coordination through three oxygen donors and one nitrogen donor. The reaction of Na 2 MoO 4 ⋅2H 2 O with the same ligand precursor in an acidic methanol solution leads to the formation of an anionic dioxido complex (Et 3 NH)[MoO 2 (L ala )] (2) with a trans coordination of the tetradentate ligand. Trans-1 and 2 were studied as active catalysts for olefin epoxidation: i.e. styrene, cyclohexene, S(–)-limonene and (–)- α -pinene using H 2 O 2 and tBuOOH as oxidants. 1. Introduction In transition metal chemistry, diverse ligands are utilized to modify the surroundings of the metal. Especially, the reactivity of the metal species is substantially dependent on the steric properties of the ligands, whereas a number of bulky anionic ligands with hard oxygen donor atoms have been used to get robust complexes which are stable under ambient conditions. To this end, several phenolato ligands, particularly, have shown great potential [1–3]. For example, amine bisphenols with various donors in the pendant arm form a family of readily available and adjustable tetradentate ligands, which can form chelates through hard nitrogen and oxygen donor atoms. They are combined with metals from across the periodic table to form molecular complexes, which can be used e.g. as active catalysts for a number of reactions [4]. The research on the coordination chemistry of molybdenum is often inspired by its role in several biological oxidation and oxotransfer (OAT) reactions [5]. As a result, a number of molybdenum-containing model compounds for OAT reactions have been studied [6–9]. Similarly, a number of artificial dioxidomolybdenum(VI) compounds have been used in important industrial processes, such as olefin epoxidation [10–12]. Generally, the molybdenum catalysts involve metal ions in high oxidation states, and thus the reactivity of many active species can be modelled with isolated oxidomolybdenum species, whereas such model compounds are commonly prepared using multidentate phenolato ligands. For example, the reactions of MoO 2 Cl 2 and its soluble derivatives with amine trisphenols, aminoalcohol bisphenols or aminoacid bisphenols are known to yield neutral, mononuclear oxidochloridomolybdenum(VI) [MoOCl (L)] type complexes which can catalyze various oxidation reactions, for instance sulfoxidation, epoxidation and haloperoxidation [13–15]. A tetrapodal ligand can coordinate to the metal center in a fashion resembling a three-bladed propeller, which generates a chiral environment around the metal. This, however, typically leads to the formation of two enantiomers instead of enantiopure products. When introducing a ligand-based chirality to the metal complexes, natural α -amino acids are useful building blocks for the preparation of enantiopure ligands. As epoxides are valuable precursors for a wide variety of organic functionalities in the synthesis of fine chemicals and pharmaceutical intermediates, the catalytic oxidation of alkenes is one of the most extensively studied reactions in organic chemistry [16,17]. Furthermore, transformation of epoxides is also a key step for numerous processes important for both synthetic organic chemistry and biochemistry, due to the ring-opening reaction of epoxides generating new useful * Corresponding author. E-mail address: [email protected] (A. Lehtonen). Contents lists available at ScienceDirect Inorganica Chimica Acta journal homepage: www.elsevier.com/locate/ica https://doi.org/10.1016/j.ica.2023.121519 Received 16 February 2023; Received in revised form 5 April 2023; Accepted 6 April 2023 Inorganica Chimica Acta 553 (2023) 121519 2 carbon–carbon bonds [18,19]. From the industrial point of view, there is an interest on some plant-derived monoterpenes which can be accumulated in bulk amounts from biogenic waste streams. Particular attention has recently been paid on monocyclic and bicyclic monoterpenes, such as abundant natural products limonene and pinene, but also on inexpensive by-products from technical forestry resins and wood pulp by-produced in the cellulose manufacture [20] and the citrus fruit juice industry [21], respectively. Although R(+)-limonene is the primary enantiomer in most plants that produce limonene, the resin from a few plants such as Pinus strobus L. contains substantial amounts of the S(–)-enantiomer, as well. α -Pinene is the major monoterpene of pine essential oils widely used as a food flavoring ingredient and also accepted as a safe food additive [22]. In recent years, there has been a growing interest in the synthesis of their oxidation products especially optically pure epoxides which have been found as important building blocks in asymmetric synthesis and are more particularly used as chiral precursors [23]. In this study, we used an amine bisphenol derivative of L-alanine ((Et 3 NH)H 2 L ala ) to prepare new oxidomolybdenum(VI) complexes with a chiral chelating ligand. The epoxidation activities for styrene, cyclohexene, S(–)-limonene and (–)- α -pinene were tested using H 2 O 2 and tBuOOH as oxidants. 2. Experimental 2.1. General remarks The ligand precursor (Et 3 NH)H 2 L ala was made of L-alanine, 2–4-ditert-butylphenol, paraformaldehyde and triethylamine using the known procedures. [24,25]. MoO 2 Cl 2 (dmso) 2 was prepared following the published method [26]. Other chemicals and solvents were obtained from commercial sources and were used as purchased. All syntheses were carried out under ambient atmosphere. The IR spectra were measured with Bruker Optics, Vertex 70 device with a diamond ATR setup. The UV spectra were recorded in CHCl 3 solutions. A Shimadzu GC-2025 gas chromatograph with a Zebron ZB-5 capillary column (30 m ×0.25 mm ×0.25 mm) and FID detector were used to analyse the reaction products of the oxidation of olefins. The identity of the products was confirmed using a GC–MS model Shimadzu GCMS-QP2010 SE. Single crystal X-ray data were collected by Rigaku Oxford Diffraction SuperNova diffractometer equipped with micro-focus dual-source (Mo/ Cu) and Atlas detector using Cu K α (λ =1.54184) radiation. The crystals were kept under a nitrogen stream at 123 K during data collection. Data collection and reduction were carried out in CrysAlis Pro software whereas Olex 2 GUI was used for crystal structure solving (SHELXS) and refinement (SHELXL). For cis-1 and trans-1 all non-hydrogen atoms were located from the difference density map and refined using anisotropic displacement parameters whereas a standard SHELXL riding atom model was used for hydrogen atoms. Disorder is observed for crystal structures of both cis-1 and trans-1. For cis-1, which crystallised as an acetonitrile solvate, one of the two acetonitrile molecules in the asymmetric unit was refined with two positions with ca. 0.7:0.3 occupancies. In case of trans-1, one of the two distinct complex molecules in the asymmetric unit shows disorder of one of the four tert-butyl groups (central carbon C25B). Consequently, the tert-butyl group was refined with two distinct positions with ca. 0.83:0.17 occupancies while the C – C distances and anisotropic parameters were restrained using SADI and SIMU. CheckCIF/PLATON procedure finds a potential higher symmetry for trans-1 (space group P2 1 /c). The data could be solved and refined in this centrosymmetric space group, but resulted in a disagreeably high R-values as well as disorder of the L-alanine moiety suggesting that the initial space group P2 1 is correctly assigned and no racemisation of the chiral centre has occurred during the synthesis. The crystal structure of 2 was solved in space group P1 and only partially refined to determine the molecular geometry and conformation of the complex (see further details in ESI). 2.2. NMR spectroscopy The 1 H and 13 C NMR spectra were measured with a 500 MHz Bruker Avance ( 1 H: 500.13 MHz, 13 C: 125.76 MHz) or Avance III ( 1 H: 500.08 MHz, 13 C: 125.75 MHz) spectrometer. Deuterated chloroform (CDCl 3 ) at 25 ◦C was used as solvent for cisand trans-1, and DMSO‑d 6 at 30 ◦C for 2. Proton and carbon chemical shifts were referenced to internal tetramethylsilane (δ TMS =0.00 ppm). The 1D 1 H NMR spectra were measured with a single-pulse-acquire sequence (Bruker pulse program “zg30”) with a 30◦flip angle and a 4.3–5.1 s recycle delay. For the 1D 13 C NMR spectra (“zgpg30”), a 30◦flip angle, 3.1 s recycle delay, and broadband 1 H decoupling (“waltz16”) were used. The gradient-selected 1 H– 1 H COSY spectra were recorded in a double-quantum filtered mode (“cosygpmfqf”). 1 H– 1 H NOE spectra were acquired with 1D (“selnogp”) and 2D (“NOEsygpph”) transient NOESY experiments using a 0.30 s mixing time. The 1 H– 13 C HSQC spectra (“hsqcedetgpsisp2”) were recorded with multiplicity editing (resulting in positive CH, CH 3 and negative CH 2 correlation signals) and were optimized for a 145 Hz onebond 1 H– 13 C coupling constant. The 1 H– 13 C HMBC long-range correlation experiments (“hmbcgplpndqf”) were optimized for a 10 Hz (longrange) 1 H– 13 C coupling constant while utilizing a 145 Hz low-pass Jfilter. The dynamic lineshape analyses were done with the DNMR Lineshape Analysis module ver. 1.1 included in Bruker’s TopSpin 4.0.2 software. 2.3. Computational chemistry The molecular model building and basic molecular mechanics (MM) geometry optimizations were done with the ChemDraw 20.0 and Chem3D 20.0 software, using the MM2 force field. For MM-based conformational search for complex 2, Materials Studio Conformers simulation tool included in the BIOVIA Materials Studio environment was used. The DFT geometry optimizations and GIAO NMR chemical shift calculations were done with Gaussian 09 or Gaussian 16 software [27] using the B3LYP functional and an atom-specific basis set with def2-TZVP and effective core potential (ECP) for Mo and 6–31 +G(d,p) for the other atoms. The chemical shifts were calculated for the DFToptimized structures both in a gas phase and in solution, with the PCM method used to model the chloroform (for cis-1 and trans-1) and DMSO (for 2) solvents. The Cartesian coordinates of the DFT-optimized geometries of cis-1, trans-1, trans-2-A and trans-2-B are given in ESI Table S1. 2.4. Molybdenum(VI) complexes 2.4.1. Synthesis of 1 1.0 mmol batches of [Et 3 NH]H 2 L ala (0.63 g) and MoO 2 Cl 2 (dmso) 2 (0.36 g) were mixed with 40 ml of toluene and the reaction mixture was refluxed for 6 h. Blue trans-1 and purple cis-1 were isolated from the reaction mixture by silica column chromatography using CH 2 Cl 2 as an eluent. The synthesis was repeated several times, while the yields for trans-1 were typically 50–70% and for cis-1 the yields were 5–10 %. Small amounts of 2 were isolated from the reaction mixture by adding 2% of Et 3 N in the eluent. Crystals of cis-1 and trans-1 for X-ray studies were obtained from hot acetonitrile upon cooling to room temperature. Cis-1: IR: 2961 s, 2930 m, 2861 m, 1718 vs, 1593 s, 1458 s, 1393 w, 1381 m, 1360 m, 1310 m, 1283 w, 1257 w, 1240 s, 1200 s, 1169 vs, 1114 m, 1066 m, 1025 w, 1008 w, 961 vs, 953 vs, 940 m, 929 w, 914 s, 879 vs, 869 vs, 847 s, 826 m, 808 s, 761 m, 752 m, 704 m, 654 w cm −1 . UV: ε 4450 (320 nm), 6080 (503 nm), 9010 (685 nm) dm 3 mol -1 cm −1 . 1 H NMR (CDCl 3 ): δ 7.42 (2H, m, 3-H & 13-H), 7.22 (1H, d, J =2.2 Hz, 11H), 7.16 (1H, d, J =2.1 Hz, 5-H), 4.79 (1H, d, J =13.8 Hz, 7-H syn ), 4.01 (1H, quartet, J =7.0 Hz, 16-H), 3.73 (1H, d, J =13.8 Hz, 7-H anti ), 3.57 (1H, d, J =13.6 Hz, 9-H syn ), 3.42 (1H, d, J =13.6 Hz, 9-H anti ), 1.53 (9H, s, 2-tBu), 1.51 (3H, d, J =7.0 Hz, 16-CH 3 ), 1.47 (9H, s, 14-tBu), 1.33 (9H, s, 4-tBu), 1.29 (9H, s, 12-tBu). 13 C NMR (CDCl 3 ): δ 172.5 (C17), A. Peuronen et al. Inorganica Chimica Acta 553 (2023) 121519 3 162.9 (C15), 157.1 (C1), 151.3 (C4), 149.4 (C12), 140.7 (C2), 140.4 (C14), 129.6 (C6), 129.1 (C10), 125.5 (C11), 125.4 (C5), 124.5 (C13), 124.1 (C3), 63.0 (C16), 60.7 (C7), 52.1 (C9), 35.8 (2-C(CH 3 ) 3 ), 35.2, 35.1 (4-C(CH 3 ) 3 & 14-C(CH 3 ) 3 ), 34.9 (12-C(CH 3 ) 3 ), 31.5 (12-C(CH 3 ) 3 ), 31.4 (4-C(CH 3 ) 3 ), 30.4 (2-C(CH 3 ) 3 ), 29.9 (14-C(CH 3 ) 3 ), 9.2 (16-CH 3 ). Trans-1: IR: 2956 s, 2925 vs, 2856 s, 1690 vs, 1593 s, 1461 s, 1392 m, 1375 m, 1363 m, 1330 w, 1306 w, 1290 w, 1259 s, 1239 vs, 1201 s, 1169 vs, 1122 m, 1094 m, 1067 m, 1040 m, 1023 s, 953 vs, 917 s, 880 s, 862 vs, 807 s, 775 m, 757 s, 695 m, 657 w cm −1 . UV: ε 1580 (365 nm), 4940 (515 nm), 13,800 (683 nm) dm 3 mol -1 cm −1 . 1 H NMR (CDCl 3 ): δ 7.44 (2H, m, 3-H & 13-H), 7.15 (1H, d, J =2.2 Hz, 11-H), 7.08 (1H, d, J =2.1 Hz, 5-H), 4.82 (1H, d, J =14.4 Hz, 7-H anti ), 4.22 (1H, d, J =13.6 Hz, 9H anti ), 3.74 (2H, m, 7-H syn & 9-H syn ), 3.57 (1H, quartet, J =7.0 Hz, 16H), 1.57 (9H, s, 14-tBu), 1.54 (9H, s, 2-tBu), 1.34 (9H, s, 4-tBu), 1.33 (9H, s, 12-tBu), 1.26 (3H, d, J =7.0 Hz, 16-CH 3 ). 13 C NMR (CDCl 3 ): δ 173.7 (C17), 162.0 (C15), 159.9 (C1), 152.0 (C4), 151.0 (C12), 141.5 (C2), 139.7 (C14), 128.8 (C6), 127.5 (C10), 126.1 (C11), 125.8 (C5), 124.4 (C13), 124.2 (C3), 60.7 (C7), 58.7 (C16), 56.5 (C9), 35.7 (2-C (CH 3 ) 3 ), 35.5 (14-C(CH 3 ) 3 ), 35.2, 35.1 (4-C(CH 3 ) 3 & 12-C(CH 3 ) 3 ), 31.4, 31.4 (4-C(CH 3 ) 3 & 12-C(CH 3 ) 3 ), 30.3, 30.2 (2-C(CH 3 ) 3 & 14-C(CH 3 ) 3 ), 8.9 (16-CH 3 ). 2.4.2. Synthesis of 2 0.50 mmol of Na 2 MoO 4 ⋅2H 2 O (0.19 g) was mixed with a 0.43 mmol of (Et 3 NH)(H 2 L ala ) (0.27 g) in 10 ml of MeOH. Acetic acid (0.1 ml) was added and the reaction mixture was heated to the reflux temperature for two hours, cooled to the room temperature and stored at +4 ◦C overnight to obtain 200 mg (62 %) of an orange-yellow solid product. IR: 3090 m(br), 3000 m, 2955 m, 2900 m, 2864 m, 1670 s, 1580 m, 1470 s, 1440 m, 1412 m, 1380 m, 1361 m, 1330 m, 1305 w, 1275 m, 1255 s, 1237 vs, 1203 m, 1168 s, 1120 m, 1079 m, 1041 w, 1009 m, 979 w, 920 s, 885 vs, 840 s, 821 s, 806 m, 776 m, 759 m, 740 m, 698 w, 674 w, 649 w, 622 m, 601 m, 550 s cm −1 . 1 H NMR (DMSO‑d 6 ): δ 8.88 (1H, br s, NH (CH 2 CH 3 ) 3 ), 7.13 (1H, d, J =2.4 Hz, 13-H), 7.05 (1H, d, J =2.3 Hz, 11H), 7.03 (1H, d, J =2.3 Hz, 3-H), 6.90 (1H, d, J =2.3 Hz, 5-H), 4.77 (1H, d, J =15.1 Hz, 7-H anti ), 3.81 (1H, d, J =15.1 Hz, 7-H syn ), 3.63 (1H, br d, 9-H anti ), 3.48 (1H, d, J =12.3 Hz, 9-H syn ), 3.29 (1H, under the H 2 O peak, 16-H), 3.08 (6H, quartet, J =7.3 Hz, NH(CH 2 CH 3 ) 3 ), 1.35 (9H, s, 14tBu), 1.28 (9H, s, 2-tBu), 1.25 (9H, s, 12-tBu), 1.23 (9H, s, 4-tBu), 1.17 (9H, t, J =7.3 Hz, NH(CH 2 CH 3 ) 3 ), 1.13 (3H, d, J =7.1 Hz, 16-CH 3 ). 13 C NMR (DMSO‑d 6 ): δ 174.9 (br, C17), 162.2 (br, C15), 160.3 (C1), 139.7 (C4), 139.4 (br, C12), 136.5 (C2), 135.4 (br, C14), 124.8 (C11), 124.0 (br, C10), 123.2 (C5), 122.6 (C13), 122.1 (C6), 120.6 (C3), 58.3 (2C, br, C7 & C16), 54.9 (C9), 45.7 (NH(CH 2 CH 3 ) 3 ), 34.6 (2-C(CH 3 ) 3 ), 34.2 (14C(CH 3 ) 3 ), 33.7 (2C, 4-C(CH 3 ) 3 & 12-C(CH 3 ) 3 ), 31.5 (12-C(CH 3 ) 3 ), 31.4 (4-C(CH 3 ) 3 ), 29.9 (14-C(CH 3 ) 3 ), 29.6 (2-C(CH 3 ) 3 ), 8.6 (NH(CH 2 CH 3 ) 3 ), 8.4 (16-CH 3 ). 2.5. Oxidation of olefins In a typical procedure, styrene or cyclohexene (1.00 mmol), an oxidant (3.00 mmol), i.e. aqueous 30% H 2 O 2 or 5.5 M of tBuOOH in decane, and catalyst (0.020 mmol) were heated at 80 ◦C for 1 h of reaction time in 10 ml of 1,2-dichloroethane (DCE). The reactions were monitored by GC and the yields were recorded as GC yield based on the starting styrene or cyclohexene. The oxidation products were characterized by GC–MS. The influence of amounts of catalyst and oxidant was also studied to check their effect on the conversion and selectivity of the reaction products. 3. Results and discussions 3.1. Syntheses. The ligand (Et 3 NH)H 2 L ala was made of L-alanine, 2–4-di-tert-butylphenol, formaldehyde and triethylamine modifying slightly the known procedures [24,25]. The reaction of stoichiometric amounts of MoO 2 Cl 2 (dmso) 2 and (Et 3 NH)H 2 L ala in refluxing toluene leads to a dark solution, which contained blue trans-1 as a major product and small amount of purple cis-1 as well as a trace amount of a yellow product 2 (Scheme 1). The components were isolated by column chromatography and identified by IR and NMR spectroscopy. The molecular structures were also verified by single crystal X-ray diffraction (see below). The IR spectra of cis-1 and trans-1 show strong IR absorptions at 914 and 917 cm −1 , respectively, which can be assigned as Mo=O stretch, but they lack the typical doublets for the MoO 2 moiety. On the contrary, complex 2 show strong IR-absorptions at 920 and 885 cm −1 – characteristic for the octahedral complexes carrying a MoO 2 function [28]. The reaction was repeated several times by applying different concentrations and reaction times but only trans-1 was obtained in practical yields whereas cis-1 and 2 could be isolated only in minor quantities. In general, the shorter reaction times yielded larger relative ratio of cis-1 even if the yields were lower, but cis-1 tends to isomerize to trans-1 upon purification steps. The yellow product 2 was prepared for the further analyses and catalyst tests in a higher yield by the reaction of Na 2 MoO 4 ⋅2H 2 O with (Et 3 NH)H 2 L ala in an acidic methanol solution (Scheme 1b). 3.2. Solution structures. The solution structures of the oxidochlorido complexes cis-1 and trans-1, and the dioxido complex 2 were determined with 1D and 2D 1 H and 13 C NMR spectroscopy. The 1 H and 13 C NMR spectra of cisand trans-1 dissolved in CDCl 3 feature sharp resonance peaks whose total number and intensities are consistent with the electrically neutral, diamagnetic, monomeric structures shown in Scheme 1. The NMR peaks were assigned, including stereochemistry of the diastereotopic geminal hydrogens, with help of cis trans Scheme 1. Formation of the complexes cis-1, trans-1 and 2 from L-alanine derived ligand precursor (Et 3 NH)H 2 L ala . Fig. 1. Solution structure of cis-1 in CDCl 3 as deduced by NMR spectroscopy. For clarity, the tBu and Me groups have been contracted. A. Peuronen et al. Inorganica Chimica Acta 553 (2023) 121519 4 standard 2D NMR correlation spectroscopy (COSY, NOESY, HSQC, HMBC) and the full list of assigned 1 H and 13 C chemical shifts is given in the Experimental section. The geminal hydrogens at C7 and C9 are denoted by “syn” if they are on the same side of their ring as the L-alanine bridge, and by “anti” otherwise (cf. Figs. 1 and 2). The presence of a chiral carbon (C16) in the ligand makes its two phenolate arms chemically inequivalent in both cis and trans isomer, and thus these arms display distinct 1 H and 13 C chemical shifts. The carbons of the phenolate arm on the side of the 16-H hydrogen are numbered as C1–C7 whereas those on the 16-Me side are numbered as C9–C15. The coordination geometry at the central Mo atom as well as the preferred solution conformation of cis-1 and trans-1 were deduced mainly from. 1) the observed 1 H– 1 H NOE correlations (indicating spatial proximity between the correlated protons) 2) the chemical shift values of the 7-H and 9-H protons (as protons close to the Cl ligand become deshielded) 3) the three-bond 1 H– 13 C correlations seen in the HMBC spectra optimized for 10 Hz long-range couplings (the strongest correlations are expected for such H–X–Y–C coupling pathways where the H–X and Y–C bonds are anti-periplanar about the X–Y bond). Fig. 2. Solution structure of trans-1 in CDCl 3 as deduced by NMR spectroscopy. Table 1 The chemical shifts δ (ppm) of selected protons of cis-1, trans-1, and 2, and their relevant NOE and HMBC correlations. The atom numbering and the syn/anti descriptors are explained in the text and in Figs. 1 and 2. Complex proton δ/ppm NOE correlations HMBC correlations [a] cis-1 7-H syn 4.79 16-H, 7-H anti C9 16-H 4.01 7-H syn , 16-Me C9 7-H anti 3.73 5-H, 7-H syn , 9-H syn , 16Me – 9-H syn 3.57 11-H, 7-H anti , 9-H anti , 16-Me – 9-H anti 3.42 9-H syn C16 16Me [b] 1.51 11-H, 16-H, 7-H anti , 9H syn C16 trans-1 7-H anti 4.82 9-H anti , 7-H syn C16 9-H anti 4.22 7-H anti , 9-H syn C16 7H syn[b] 3.745 5-H, 7-H anti , 16-H, 16Me – 9H syn[b] 3.739 11-H, 9-H anti , 16-Me – 16-H 3.57 7-H syn , 16-Me C9 16-Me 1.26 11-H, 7-H syn , 9-H syn C16 (trans-) 2 7-H anti 4.77 7-H syn C16 7-H syn 3.81 7-H anti , 16-H(w), 16Me C9, C16(w) 9-H anti 3.63 (br) n.d. [c] n.d. [c] 9-H syn [b] 3.48 C7/C16 16-H [b] 3.29 C9 16-Me 1.13 11-H, 7-H syn , 9-H syn , 16-H C16 [a] HMBC (opt. for n J CH =10 Hz) correlations from the specified proton to the C7, C9 and C16 carbons. [b] Could not be individually selected in a 1D NOESY experiment due to signal overlap; the NOE correlations could be detected and assigned by selecting the correlation partner instead. [c] Not detected due to exchange broadening of the proton peak. E E Fig. 3. The predominant solution conformers A (major) and B (minor) of the complex anion (trans-2) as obtained from MM conformational search followed by DFT geometry optimization. Fig. 4. The 5.0–3.1 ppm region of the 1 H NMR spectrum of complex 2 in DMSO‑d 6 solution at 303 K. Black: observed, red: simulated with dynamic lineshape fitting, assuming fast-intermediate exchange between the conformers A and B with pop(A) =90%, k(forward) =1250 s −1 and Δδ sim -values between the two limiting conformers as given in Table 2. The overlapping 16-H & H 2 O peak was not included in the simulation and was modelled with a single broad line. A. Peuronen et al. Inorganica Chimica Acta 553 (2023) 121519 5 A selected set of these data are summarized in Table 1. The solution structures shown in Figs. 1 and 2 for cis-1 and trans-1, respectively, were obtained from molecular-mechanics (MM2) geometry optimizations of starting structures that were hand-built to satisfy the above-mentioned NMR spectroscopic geometry constraints. The optimized structures agree well with all the observed NMR parameters. For example, in cis-1 the chemical shift of 7-H syn (4.79 ppm) is strongly deshielded by the Cl ligand compared with 7-H anti (3.73 ppm), and a strong NOE correlation between 7-H syn and 16-H can be seen, as well as a HMBC correlation between 7-H syn and C9 (indicating their anti-periplanar positions about the C7–N8 bond). It is thus concluded that in cis-1 and trans-1 the phenolate oxygens are cis and trans coordinated, respectively, to the central Mo atom and that in CDCl 3 solution these complexes adopt a single predominant conformation as shown in Figs. 1 and 2. These conformers are almost identical with the corresponding solid-state (Xray) structures discussed later. The dioxo product 2 was poorly soluble in chloroform and was dissolved in DMSO‑d 6 for NMR measurements. In the 1 H and 13 C NMR spectra, the peaks of both the diamagnetic, monomeric complex anion 2 and a triethylammonium cation were seen, with ca. 1:1 M ratio of the ions. Thus, the synthesis of complex 2 as described above yields 2 as its triethylammonium salt (Et 3 NH)[MoO 2 (L ala )]. The full list of assigned chemical shifts is given in the Experimental section. Exchange broadening (denoted here with “br”) was clearly visible in many of the 1 H and 13 C NMR peaks of anion 2 at 30 ◦C (cf. Figs. 4 and 5), i.e. the anion in solution undergoes some dynamic process(es) whose rate is within the “fast–intermediate” exchange regime on the NMR time scale. In order to elucidate the nature of the exchange process of anion 2, computational conformational search followed by geometry optimization and 1 H/ 13 C NMR chemical shift calculation was employed (cf. Experimental for details). Both cisand trans-geometries of the complex anion were modelled, but only the trans-geometry produced computational results consistent with the observed NMR spectra. Thus, it is concluded that the dioxo product 2 is the trans-2 isomer, as shown in Scheme 1. Two low-energy conformers were found for trans-2 by molecular-mechanics conformational search, and their geometries were further optimized at B3LYP/def2-TZVP level of theory to yield conformer A (ref. 0 kJ/mol) and B (+4.6 kJ/mol in gas-phase and +5.5 kJ/mol in DMSO, using PCM solvent model). The conformers trans-2-A and trans-2-B are shown in Fig. 3. The structure of the major conformer A closely resembles the structure of trans-1. Trans-2-B, the higher-energy conformer, has a computational Boltzmann factor of ca. 0.11 relative to A at 30 ◦C in DMSO solution. Its C9-carrying ring adopts a conformation clearly different from that in conformer A. As a result, e.g. the 9-H syn bond in B adopts a syn-clinal position relative to the Mo–N8 bond, in contrast with the anti-periplanar position in A. This conformational change is accompanied with the benzene ring attached to C9 turning away from the L-alanine arm, so that the overall ligand geometry becomes more bent than in A. We postulate the observed exchange broadening in some of the 1 H and 13 C NMR spectral peaks of 2 to be caused by the interconversion of conformers A and B. This was supported by chemical shift calculations for these conformers together with dynamic NMR lineshape analyses of the experimental NMR spectra, as described below. The 1 H and 13 C isotropic shielding constants σ iso were calculated for conformers A and B of trans-2 by using the GIAO method (cf. Experimental). Since the oxochloro complexes cis-1 and trans-1 displayed nearuniform solution conformations they were convenient reference compounds for calibration of the computational method. Thus, the shielding constants were calculated also for the DFT optimized structures of cis-1 and trans-1 and their observed chemical shifts were fitted to the calculated shielding constants, yielding the following linear regression equations: δH= − 0.95293⋅ σ iso +30.44383 (N=16,r=0.998,RSS =0.16) δC= − 1.03768⋅ σ iso +196.4006 (N=34,r=0.9996,RSS =49.8) These equations were then used to convert the calculated σ iso values for trans-2-A and -B to chemical shifts δ calc (Table 2). According to the calculated 1 H chemical shifts, there is a large difference between the 9-H anti proton shifts in the limiting conformers A and B (Δδ calc =0.84 ppm). This can be explained by deshielding effect of the adjacent Mo =O bond on 9-H anti in conformer A. This is exactly the proton that shows considerable exchange broadening in the experimental 1 H NMR spectrum of complex 2 (cf. Fig. 4). For the carbon nuclei, the largest Δδ calc were calculated for C7 (3.85 ppm), C14 (4.47 ppm), C15 (–3.88 ppm), C16 (–2.80 ppm), and C17 (–2.28 ppm). Consistently, all these carbons displayed broad lineshapes in the experimental 13 C NMR spectrum of 2 (cf. Fig. 5). While in DMSO‑d 6 solution (m.p. 19 ◦C) it was not possible to lower the temperature sufficiently to freeze out the individual NMR spectra of the limiting conformers, and only their dynamical average was experimentally observed, the good correlation Fig. 5. The aromatic region of the 13 C NMR spectrum of complex 2 in DMSO‑d 6 solution at 303 K, showing exchange broadening of many of the carbon peaks. The peak assignment follows our numbering for the carbons (cf. e.g. Fig. 3). Table 2 Selected calculated 1 H and 13 C chemical shifts δ calc (ppm) of the limiting conformers A and B, Δδ calc =δ calc (A) – δ calc (B) and the corresponding Δδ sim from dynamic lineshape fitting to observed NMR spectra, and the experimental dynamically averaged chemical shifts δ obs for (trans-) 2. nucleus δ calc (A)/ppm δ calc (B)/ppm Δδ calc Δδ sim δ obs /ppm 7-H anti 5.35 5.26 0.09 0.45 4.77 7-H syn 3.88 3.82 0.06 0.27 3.81 9-H anti 4.20 3.36 0.84 1.00 3.63 (br) 9-H syn 3.66 3.68 −0.02 −0.33 3.48 16-H 3.49 3.74 −0.25 – 3.29 C1 160.61 161.93 −1.32 160.3 C2 135.90 136.82 −0.92 136.5 C3 122.38 122.18 0.20 120.6 C4 139.45 140.25 −0.80 139.7 C5 122.41 121.67 0.74 123.2 C6 122.76 123.13 −0.37 122.1 C10 124.70 125.95 −1.25 124.0 (br) C11 122.32 123.52 −1.20 124.8 C12 140.02 140.31 −0.29 139.4 (br) C13 123.88 123.70 0.18 122.6 C14 138.25 133.78 4.47 135.4 (br) C15 161.23 165.11 −3.88 162.2 (br) C17 175.46 177.74 −2.28 174.9 (br) A. Peuronen et al. Inorganica Chimica Acta 553 (2023) 121519 6 between Δδ calc and the observed exchange broadening of 1 H and 13 C NMR peaks strongly supports the postulated conformational equilibrium for complex 2 in solution. Some observed HMBC correlations are also nicely in agreement with the presence of minor conformer B (cf. Table 1). For example, a HMBC correlation peak between 9-H syn and C7 is expected to be weak or missing in a structure like conformer trans-2-A due to unfavorable H syn –C9–N8–C7 dihedral angle: this correlation was, indeed, not observed for the analogous structure trans-1. Yet, this correlation peak was observed in the HMBC spectrum of complex 2, and can be explained with the favorable anti-periplanar relation between the 9H syn and N8–C7 bonds in the minor conformer B. Finally, dynamic lineshape fitting and simulation was performed for selected regions of the exchange-broadened 1 H and 13 C NMR spectra of complex 2. Based on the calculated Boltzmann factor of 0.11 for B, the population of conformer A was set at 90% and that of B at 10% in this simulation. The calculated chemical shift differences guided the initial guess of these values in the simulation. An example of a dynamic 1 H lineshape simulation is shown in Fig. 4 for the geminal protons at C7 and C9. From the lineshape analysis of NMR spectra in a fast-intermediate exchange regime (i.e. when a single, averaged spectrum is observed) it is very difficult or impossible to obtain precise rate constants k due to interdependence of the various fitting parameters (populations, chemical shifts of the limiting forms, rate constants). An approximate value k =1250 s −1 (303 K) was obtained from this example simulation for the forward rate constant of conformational change A → B, when using the aforementioned populations and reasonable Δδ sim values (cf. Table 2) in the simulation. This should be regarded as an order-of-magnitude estimate; the corresponding 13 C lineshape analyses tended to converge to kvalues approximately one half of the above value. In summary, the observed exchange broadening in the 1 H and 13 C NMR solution spectra of 2 can be explained very well by the postulated fast-intermediate conformational equilibrium process between conformers A and B. The dynamic NMR simulations based on this hypothesis and computational data produce results that are in good agreement with the observed NMR spectra. 3.3. Solid-state structures. Crystals of cis-1 and trans-1 were obtained from acetonitrile and their solid-state structures were verified by single crystal X-ray diffraction (Table 3). Cis-1 crystallised with two molecules of acetonitrile in the asymmetric unit whereas there are two complex units in the asymmetric unit of trans-1, respectively. As suggested by the NMR data (see above), both isomers form monomeric molecules in which the L-alanine bisphenolate group has coordinated as a tetradentate trianionic ligand through three oxygen donors and one nitrogen donor (Figs. 6 and 7, Table 4). In both compounds, the nitrogen donor is located trans to the terminal oxido ligand while the Mo–N bonds are relatively long due to the strong structural trans effect of multiple bonded oxido ligands [29]. As expected, the coordination sphere around the Mo(VI) is very similar to those found earlier for trans-MoOCl(L gly ) (L gly =tetradentate glycine Table 3 Summary of crystallographic data for cis-1 and trans-1. cis-1 trans-1 Formula C 37 H 54 ClMoN 3 O 5 C 33 H 48 ClMoNO 5 CCDC number 2,241,830 2,241,831 Crystal system Monoclinic Monoclinic Space group P2 1 P2 1 a/Å 11.8644(3) 13.8290(3) b/Å 11.1198(2) 16.6221(4) c/Å 14.5986(3) 14.7935(3) β/◦97.253(2) 90.9678(19) V/Å 3 1910.60(7) 3400.06(14) Z 2 4 μ (Mo-K α )/cm −1 3.794 4.180 ρ calc g/cm 3 1.308 1.309 2Θ range/◦7.512 to 153.908 8.002 to 139.914 Independent reflections 7932 12,183 R int 0. 0345 0.0297 Parameters 468 796 R 1 0.0223 a (0.0230) b 0.0281 a (0.0303) b wR 2 0.0548 a (0.0553) b 0.0702 a (0.0719) b GoF (F 2 ) 1.035 1.042 Peak, hole/e Å −3 0.23/-0.55 0.63/-0.61 Flack parameter −0.022(3) −0.025(7) a I ≥2 σ (I). b For all data. Fig. 6. The molecular structure of cis-1. Solvent molecules and hydrogen atoms are omitted for clarity. Displacement ellipsoids are presented at the 50% probability level. Fig. 7. The molecular structure of trans-1. Only one of two distinct molecules in the asymmetric unit is shown. Hydrogen atoms are omitted for clarity. Displacement ellipsoids are presented at the 50% probability level. Table 4 Selected bond lengths [Å] and angles [◦] for cis1, trans-1 and trans-MoOCl(L gly ). cis-1 trans-1 a trans-MoOCl(L gly ) b Mo1-O1 1.8719(19) 1.878(4) 1.8724(14) Mo1-O2 1.910(2) 1.875(3) 1.8811(14) Mo1-O3 1.682(2) 1.677(3) 1.6766(15) Mo1-O18 1.985(2) 2.012(2) 2.0149(14) Mo1-N8 2.398(2) 2.428(3) 2.4266(17) Mo1-Cl1 2.3696(7) 2.3446(8) 2.3513(5) O3-Mo1-N8 175.00(9) 168.29(14) 168.80(7) O1-Mo1-O2 93.93(8) 162.38(13) 162.11(6) O1-Mo1-O18 155.69(8) 90.53(15) 86.58(6) O18-Mo1-Cl1 85.89(6) 162.56(8) 162.15(4) Mo1-O1-C1 142.47(17) 146.5(3) 141.50(13) Mo1-O2-C15 134.94(17) 143.4(3) 142.33(13) Mo1-O18-C17 125.38(19) 127.2(2) 128.17(13) N8-C16-C17-O18 32.9(3) 25.9(6) 2.3(3) a For one of two similar molecules in the asymmetric unit. b From reference [14]. A. Peuronen et al. Inorganica Chimica Acta 553 (2023) 121519 7 bisphenol), cis/trans-MoCl(L ae ) (L ae =tetradentate aminoethanol bisphenol) and MoOCl(L tris ) (L tris =amine trisphenol) [13,15,30,31]. In both cis and trans isomers, the bonding parameters between the metal centre and the trianionic amino carboxylate bisphenolate ligand are in accordance with the previous studies on oxidomolybdenum(VI) complexes with tripodal aminoalcohol bisphenolates and amine trisphenolates. Unlike the two isomers of 1, we were not successful in obtaining a well-resolved crystal structure of the anionic complex 2 (see ESI for further information). We were, however, able to acquire a preliminary structure (ESI, Figure S1), which matches well with the NMR results demonstrating clearly the formation of the expected trans isomer of 2 in which the two phenolato moieties are in a trans arrangement and the two oxido O atoms at cis positions relative to each other. The complex is monoanionic as also suggested by the 1:1 ratio of complex anions and triethylammonium cations present in the crystal structure. 3.4. Catalytic oxidation of alkenes Trans-1 and 2 were studied in the catalytic epoxidation of alkenes, namely styrene, cyclohexene and their naturally occurring derivatives, i. e. S(–)-limonene and (–)- α -pinene (Fig. 8), using 30% aqueous H 2 O 2 or tert-butyl hydroperoxide (tBuOOH) as oxidants in 1,2-dichloroethane (DCE) solutions. To find appropriate reaction conditions for a maximum conversion of alkenes to epoxides, different reaction parameters were considered, that is the temperature, reaction time, amount of catalyst (0.5, 1, 2 and 3 mol% loadings) and oxidant (1:1, 2:1, 3:1 and 4:1 M ratios to substrate). When the reaction conditions were optimized, 2 mol% of catalysts were considered sufficient to run the oxidations at 80 ◦C in the 1 h of the reaction time with 3:1 M ratio of oxidant to substrate. Styrene is generally oxidized by H 2 O 2 or tBuOOH as oxidants and metal complexes as catalysts to yield five oxidation products, specifically styrene oxide, benzaldehyde, benzoic acid, phenylacetaldehyde and 1-phenylethane-1,2-diol (Fig. 9). Styrene oxide can be formed in the first step, but very fast further reaction converts the product into benzaldehyde via nucleophilic attack of the oxidant to styrene oxide followed by the cleavage of the intermediate hydroperoxystyrene, which can be also further oxidized to benzoic acid [32]. Besides, the formation of benzaldehyde may also result from a radical mechanism by the direct oxidative cleavage of the styrene side-chain double bond. The partial decomposition of the catalyst and thus the very low conversion of styrene, in case of H 2 O 2 , can be attributed to the presence of water. Similarly, the presence of water can also bring on the formation of 1phenylethane-1,2-diol by the hydrolysis of styrene oxide. Finally, phenylacetaldehyde can be formed during the isomerisation of styrene oxide. In our case, the oxidation of styrene leads to the formation of Fig. 8. Substrates used for the catalytic oxidation studies. Fig. 9. Various products of catalytic oxidation of styrene. Table 5 Oxidation of styrene. Catalyst (amount) Oxidant Conversion (%) Epoxide (%) a Benzaldehyde (%) Benzoic acid (%) trans-1 (1 mol%) 30% H 2 O 2 12 31 69 – trans-1 (2 mol%) 30% H 2 O 2 16 41 59 – trans-1 (1 mol%) tBuOOH 65 25 53 22 trans-1 (2 mol%) tBuOOH 94 16 51 33 2 (1 mol %) 30% H 2 O 2 11 10 83 7 2 (2 mol %) 30% H 2 O 2 28 17 68 15 2 (1 mol %) tBuOOH 60 30 62 8 2 (2 mol %) tBuOOH 91 19 71 10 a The absolute calibration curve method was used for selectivity determination. A. Peuronen et al. Inorganica Chimica Acta 553 (2023) 121519 8 benzaldehyde as a major component along with smaller amounts of styrene oxide and benzoic acid (Table 5). The yields of all products were rather low when H 2 O 2 was used as an oxidant, but use of tBuOOH lead to high conversions, whereas both catalysts have a rather similar reaction outcome. Formation of benzaldehyde by the oxidative cleavage of the C – – C double bond in styrene is frequently seen in epoxidation experiments, thus only a few selective epoxidation catalysts have been reported. [33–35]. Oxidation of cyclohexene generally results in epoxidation products, i.e. cyclohexene oxide, and after its hydrolysis to cyclohexene-1,2-diol, or in allylic oxidation products, i.e. 2-cyclohexen-1-ol and 2-cyclohexen-1-one (Fig. 10). In this study, the monoand bicyclic monoterpenes possessing cyclohexene ring, i.e. S(–)-limonene and (–)- α -pinene also give the corresponding oxidation products. Cyclohexene, S(–)-limonene and (–)- α -pinene were oxidized more selectively than styrene with both catalysts while generally the epoxide was the major product (Table 6). Complex trans-1 gave 25% yield but 100% selectivity whereas for complex 2 the conversion of cyclohexene is only slightly higher (29%) giving epoxide in a high selectivity with cyclohex2-en-1-one as a minor product when H 2 O 2 was used as an oxidant. In case of monoterpenes, the yields are much lower but selectivity is still very high. On the other hand, the use of tBuOOH provided epoxide practically quantitatively with excellent conversions using both catalysts, but including the formation of diepoxide when trans-1 was a catalyst in oxidation of S(–)-limonene possessing additionally exocyclic isopropenyl moiety. Cisand trans-1,2-limonene oxides are formed almost in equal proportions, but when H 2 O 2 was used as oxidant the ratio between the cis/trans isomers is 1:2. It is also noteworthy that in the oxidation of cyclohexene, S(–)-limonene and (–)- α -pinene, in comparison to other molybdenum(VI) complexes (e.g. with Schiff base ligands) [36–38], trans-1 gave in each case much higher conversions and selectivities towards epoxides with 2 mol% loading, excluding oxidation of S (–)-limonene. Complex 2 show similar conversions and selectivities for monoterpenes, but in contrast to cyclohexene when much higher 99% conversion was observed. In general, trans-1 seems to be slightly more active and selective as an epoxidation catalyst than dioxido complex 2. Studies on monomeric and polymeric dioxidomolybdenum(VI) complexes with pyridoxal thiosemicarbazone ligands also revealed different reactivity between the neutral and charged cationic compounds in the oxidation of cyclooctene by aqueous tBuOOH [39]. Under the solvent-free conditions and with a low catalyst:substrate ratio (1:2000), the monomeric neutral MoO 2 L(MeOH) complex is the most efficient catalyst with very good conversion and selectivity (both 97%) after 6 h of reaction time. While similar charged complex [MoO 2 LH (MeOH)]Cl, with protonated pyridoxal nitrogen, is much less active with 48% conversion and 74% selectivity towards cyclooctene oxide. On the other hand, corresponding charged polymer {[MoO 2 LH]Cl} n showed Fig. 10. Various products of catalytic oxidation of cyclohexene. Table 6 Oxidation of cyclohexene, S(–)-limonene and (–)- α -pinene. Catalyst (amount) Substrate Oxidant Conversion (%) Epoxide (%) a trans-1 (2 mol%) cyclohexene 30% H 2 O 2 25 100 trans-1 (1 mol%) cyclohexene tBuOOH 72 94 trans-1 (2 mol%) cyclohexene tBuOOH 99 100 trans-1 (2 mol%) S (–)-limonene 30% H 2 O 2 6 86 trans-1 (2 mol%) S (–)-limonene tBuOOH 99 27 b trans-1 (2 mol%) (–)- α -pinene 30% H 2 O 2 5 81 trans-1 (2 mol%) (–)- α -pinene tBuOOH 99 100 2 (2 mol%) cyclohexene 30% H 2 O 2 29 93 c 2 (1 mol%) cyclohexene tBuOOH 63 96 2 (2 mol%) cyclohexene tBuOOH 99 99 2 (2 mol%) S (–)-limonene 30% H 2 O 2 18 100 2 (2 mol%) S (–)-limonene tBuOOH 87 100 2 (2 mol%) (–)- α -pinene 30% H 2 O 2 4 76 2 (2 mol%) (–)- α -pinene tBuOOH 71 100 a The absolute calibration curve method was used for selectivity determination. b Diepoxide was found with selectivity of 65%. c Formation of cyclohex-2-en-1-one as by-product. Fig. 11. Spectral changes during titration of (top) 2 ml of a 4.0 ⋅ 10 -5 M solution of trans-1 and (bottom) 2 ml of a 4.5 ⋅ 10 -5 M solution of 2 catalysts in DCE after successive addition of one-drop portions of aqueous 30% H 2 O 2 . A. Peuronen et al.