International Journal of Biological Macromolecules 256 (2024) 128486 Available online 1 December 2023 0141-8130/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Biodegradable and biocompatible collagen-based hybrid materials for force sensing applications Mireia Andonegi a , b , Rafaela M. Meira b , Daniela M. Correia c , Nelson Pereira b , Carlos M. Costa b , e , Senentxu Lanceros-Mendez b , d , f , * , Koro de la Caba a , d , * , Pedro Guerrero a , d , g a BIOMAT Research Group, University of the Basque Country (UPV/EHU), Escuela de Ingeniería de Gipuzkoa, Plaza de Europa 1, 20018 Donostia-San Sebasti´ an, Spain b Physics Centre of Minho and Porto Universities (CF-UM-UP) and Laboratory of Physics for Materials and Emergent Technologies, LapMET, University of Minho 4710057, Braga, Portugal c Center of Chemistry, University of Minho, 4710-057 Braga, Portugal d BCMaterials, Basque Center for Materials, Applications and Nanostructures, UPV/EHU Science Park, 48940 Leioa, Spain e Institute of Science and Innovation for Bio-Sustainability (IB-S), University of Minho, 4710-053 Braga, Portugal f Ikerbasque, Basque Foundation for Science, 48009 Bilbao, Spain g Proteinmat Materials SL, Avenida de Tolosa 72, 20018 Donostia-San Sebasti´ an, Spain ARTICLE INFO Keywords: Collagen Electrical response Pressure sensor ABSTRACT With the aim of replacing synthetic macromolecules by biological macromolecules for advanced applications, collagen films were produced with two different ionic liquids (ILs), choline dihydrogen phosphate ([Ch][DHP]) and choline serinate ([Ch][Seri]), added in order to modulate the electrical responses. The films were prepared by casting, varying IL content between 0 and 6 wt%. The morphology and thermal properties of the resulting films were found to be independent of both IL type and content. However, the highest direct curret (d.c.) electrical conductivity (1.4 ×10 −8 S⋅cm −1 ) was achieved for collagen films containing 3 wt% [Ch][DHP]. Furthermore, it was demonstrated that IL/collagen films were non-cytotoxic, with cell activity values exceeding 70 %. These collagen films were proven to be suitable for force sensing applications, displaying excellent sensitivity and stability upon repeated testing. 1. Introduction Collagen represents the most abundant protein within the extracellular matrix (ECM) of vertebrates, constituting roughly 30 % of the total protein mass in mammals. Specifically, type I collagen, the primary focus of this study, constitutes about 90 % of the collagen found in skin and bones [1,2]. Thus, the main role attributed to collagen is to provide structural support, contributing to the mechanical properties of tissues, such as strength and toughness. Beyond its structural role, collagen has been demonstrated to play a significant part in various functional roles [3]. It plays an important role in tissue healing, providing the biological microenvironment for cell growth, aiding cell attachment, migration, and proliferation [4]. Moreover, collagen exhibits low immunogenicity and antigenicity, rendering it biocompatible [5,6]. Although collagen is resistant to common proteases showing a long lifetime, undergoes biodegradation within the body through certain enzymes from the matrix metalloproteinase (MMP) family, a crucial factor for designing biodegradable collagen-based materials [7]. Consequently, modification or cross-linking is necessary to control the degradation rate for tissue engineering applications [8,9]. The physical and biological properties of collagen, coupled with its versatility in forming gels, films, meshes, scaffolds and fibers make collagen an attractive candidate for various applications. Currently, the majority of industrial-grade collagen type I is sourced from cattle. Cattle skin collagen is used to fortify tendons and promote wound healing, while neonatal bovine dermis is employed in hernia repair, plastic and reconstructive surgery, among other applications [10]. Nevertheless, the utilization of this collagen is limited because of its high manufacturing costs. Fortunately, an untapped and more costeffective source of collagen exists: waste generated by the leather industry. Approximately 80 % of this waste comprises non-tanned residues, which contain about 70–80 % of fibrous collagen [11]. Thus, recovering collagen-containing waste has the potential to be a profitable and environmentally sustainable approach, contributing to the United * Corresponding authors at: BCMaterials, Basque Center for Materials, Applications and Nanostructures, UPV/EHU Science Park, 48940 Leioa, Spain. E-mail address: [email protected] (K. de la Caba). Contents lists available at ScienceDirect International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac https://doi.org/10.1016/j.ijbiomac.2023.128486 Received 3 October 2023; Received in revised form 13 November 2023; Accepted 27 November 2023

International Journal of Biological Macromolecules 256 (2024) 128486 2 Nations Sustainable Development Goals [12]. Collagen is a promising material for electronic devices, including electrochemical applications [13], and biomemristive memory devices, which exhibit a high-resistance state/low-resistance state ratio of ~100 [14]. In this context, a flexible strain sensor based on collagen fibers was developed, boasting a tensile strength of 59.9 MPa, electrical conductivity of 6.5 S⋅m −1 and super-amphiphobic properties, making it suitable for wearable electronic devices [15]. Furthermore, polymer-based hybrid materials featuring collagen, polyaniline as conductive matrix, and iron oxide nanoparticles as magnetic fillers were developed to achieve tailored dielectric, magnetic and conducting properties. This composite material exhibited an electrical conductivity of 2.44 ×10 −2 S⋅cm −1 and a saturation magnetization of 8.32 emu/g [16]. Within this context, polymer-based hybrid materials incorporating ionic liquids (ILs) offer a novel platform for creating multifunctional hybrid materials with reduced environmental footprint when compared to nanoparticle-containing composites [17]. Building upon previous work involving collagen with various ILs, including choline dihydrogen phosphate ([Ch][DHP]), choline derinate ([Ch][Seri]), choline bis(trifluoromethylsulfonyl)imide ([Ch][TFSI]), and 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide ([EMIM][TFSI]), which demonstrated their suitability for resistive touch sensors [18], this study aims to develop sustainable blends of collagen with low IL content (up to 6 wt%) utilizing two IL types: [Ch][DHP] and [Ch][Seri]. These ILs were selected for their biocompatibility and water miscibility. The evaluation encompasses an assessment of morphology, thermal, mechanical, electrical and biological properties, in addition to the functional sensing response. 2. Materials and methods 2.1. Materials Bovine collagen was supplied by Proteinmat S.L. (Donostia, Spain), choline dihydrogen phosphate [Ch][DHP] (>98 %) and choline derinate [Ch][Seri] (>95 %, 60 % in H 2 O) by Ionic Liquids Technologies GmbH (Germany), and acetic acid by Panreac Quimica S.L.U. (Barcelona, Spain). 2.2. Film preparation Casting was employed in the preparation of collagen films with varying contents (0, 1, 3, and 6 wt%) of [Ch][DHP] and [Ch][Seri]. First, bovine collagen and the corresponding amount of IL for each formulation were mixed in a solution of 0.5 M acetic acid (1:50 collagen/acetic acid). To obtain the films, the mixtures were left at room temperature for 2 h under continuous stirring at 400 rpm before being poured into Petri dishes and allowed to dry at room temperature. Films with an average thickness of 125 μ m were obtained, irrespective of the filler content. These films were identified as 1[Ch][DHP], 1[Ch][Seri], 3[Ch][DHP], 3 [Ch][Seri], 6[Ch][DHP], and 6[Ch][Seri], as a function of IL type and content. Control films were those without ILs. Before undergoing film characterization, all films were conditioned for 48 h in an ACS Sunrise 700 V biochamber (Alava Ingenieros, Madrid, Spain) at 25 ◦C and 50 % relative humidity. 2.3. Film characterization To assess the thermal transitions of the samples, differential scanning calorimetry (DSC) measurements were performed with a Mettler Toledo DSC 822 (Madrid, Spain) equipment. About 3.0 ±0.2 mg of each sample were enclosed in 50 μ L aluminum pans, to avoid mass loss during the experiment, and subjected to a heating ramp from 25 to 300 ◦C at a rate of 10 ◦C/min under an inert atmosphere (10 mL N 2 /min) to avoid thermo-oxidative reactions. Fourier Transformed Infrared (FTIR) measurements were performed using an Alpha II Compact FTIR spectrometer coupled with an attenuated total reflectance (ATR) crystal (ZnSe). The FTIR spectra of the composites were recorded at room temperature from 4000 to 800 cm −1 and collected after 32 scans with a spectral resolution of 4 cm −1 . X-ray diffraction (XRD) was performed at 40 kV and 40 mA with a PANalytic Xpert Pro (PANalytical, Almelo, The Netherlands) apparatus using Cu – K (λ =1.5418) as the radiation source. The data were collected between 2 and 50◦(step size =0.026, time per step =118 s). The morphology of the films was examined by SEM, using an Hitachi S-4800 scanning electron microscope (Hitachi, Madrid, Spain) at 15 kV accelerating voltage. Prior to SEM measurements, films were placed on a metal stub and coated with gold using a JEOL fine-coat ion sputter JFC1100 and argon atmosphere. Mechanical tensile tests of the samples were performed using an Instron 5967 mechanical testing system (Instron, Barcelona, Spain). Tensile tests were carried out at a rate of 1 mm/min on bone-shaped samples (4.75 mm ×22.25 mm) in accordance with the ASTM D 638–03 standard. To determine significant differences between samples, an analysis of variance (ANOVA) was performed using SPSS software (SPSS Statistic 25). For comparison across various systems, Tukey's test with statistical significance at the P <0.05 level was considered. The d.c. electrical conductivity ( σ , S/cm) of the films was obtained at room temperature through a Keithley 287 picoammeter/voltage source, applying voltage between ±10 V. Measurements were performed in the parallel plate configuration and the electrical conductivity was obtained from the I-V curves through Eq. (1): σ =d R.A(1) where d (cm) is the sample thickness, R (Ω) is the resistance value, and A (cm 2 ) is the electrode area. 2.4. Film degradation and cytotoxicity For the degradation assay, circular samples with a 13 mm diameter were cut from different collagen films, weighed (wi) and placed in a 24well culture plate with 500 μ L of phosphate-buffered saline solution (PBS, pH 7.4) and cell culture medium (DMEM). The samples were subsequently incubated in an oven at 37 ◦C for 1, 2, 4 and 7 days. At each time point, the samples were removed from the culture plate and left to dry at room temperature before being weighed (wt). The weight loss for three specimens of each sample was calculated using Eq. (2): Weight loss (%) = wi −wt wi ⨉100 (2) To assess the cytotoxicity of the samples, MC3T3-E1 pre-osteoblast cells were cultivated in Dulbecco's modified Eagle's medium (DMEM, Gibco) containing 1 g⋅L −1 glucose, 10 % fetal bovine serum (FBS, Biochrom) and 1 % penicilin/streptomycin (P/S, Biochrom) under standard culture conditions (at 37 ◦C in a 95 % humidified air containing 5 % CO 2 ). The culture medium was refresehed every 2 days and the cells were harvested before reaching ~70 % confluence. The citotoxicity of the IL/collagen composite films was evaluated using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Promega) assay. Circular samples with a 13 mm diameter were cut from collagen films with varying [Ch][DHP] and [Ch][Seri] contents (1, 3 and 6 wt%). First, the samples were exposed to ultraviolet light (UV) for 40 min (20 min on each side), washed twice with sterile phosphate-buffered saline solution (PBS, pH 7.4) for 30 min per wash on an orbital shaker to remove any residual solvent. Subsequently, they were exposed again to UV light for an additional hour (30 min on each side) to ensure complete sterilization. Following sterilization, the samples were placed in a 24well tissue culture polystyrene plate containing cell culture medium and incubated at 37 ◦C in 95 % humidified atmosphere containing 5 % CO 2 for 72 h. MC3T3-E1 cells (cell density =3 ×10 4 cells⋅mL −1 ) were cultivated in 96-well tissue culture polystyrene plates for 24 h to M. Andonegi et al.

International Journal of Biological Macromolecules 256 (2024) 128486 3 facilitate cell attachment. After this period, the cell culture medium in the 96-well plate was replaced and the MC3T3-E1 cells were exposed to a culture medium that had been in contact with various samples. After 72 h, the medium from each well was removed and fresh medium containing MTT solution in a 1:10 ratio was added. After 2 h of incubation, 100 μ L from each well was transferred in quadruplicate to a 96well plate and the optical density was measured at 490 nm. MC3T3-E1 cell viability was calculated using Eq. (3): Cell viability (%) = absorbance of sample absorbance of negative control ⨉100 (3) 2.5. Force sensor development An interdigit matrix of 9 circular digits (digit width and spacing of 500 μ m and a channel length of 95 mm) was fabricated with conductive silver ink from Novacentrix (Metalon HPS-021LV). A manual screenprinting machine was used with a mesh of 100 threads by centimeter and the conductive patterns were deposited on top of a PET substrate (Melinex 506, 100 μ m thick PET film), as shown in Fig. 1a. For the sensing assays, the films were cut in circles of 12 mm in diameter and placed on top of a paper separator with a hole of 10 mm in diameter, allowing an air gap between the interdigit and the collagen film (Fig. 1b), and the matrix was encapsulated with another PET film. The electrical connection of the matrix with the electronic system was achieved with a flat flexible cable with a pitch of 1 mm, glued with Z-axis conductive tape from 3 M (9703). A teensy 4.0 microcontroller was used to acquire the resistance variation of the matrix through a voltage divider with a resistance (R) of 10 MΩ. The microcontroller analog-to-digital converter (ADC) acquires the voltage value for the nine sensors and sends the converted data to a Graphical user interface (GUI) via USB, as shown in Fig. 2a). Fig. 2b shows the developed tic-tac-toe game, taking advantage of the ninesensor matrix. 3. Results and discussion 3.1. Thermal and physicochemical properties DSC analysis was used to determine the thermal transitions and stability of collagen films containing ILs. The influence of IL inclusion on the glass transition temperature (T g ), denaturation temperature (T d ), and enthalpy (ΔH) of the films is shown in Table 1. All samples showed an endothermic peak between 45 ◦C and 185 ◦C, associated with the dehydration and thermal denaturation of the amorphous region of collagen [19]. It was observed that T d increased from 84.1 ◦C to 91.3 ◦C for 6[Ch][DHP] and 98.4 ◦C for 6[Ch][Seri] samples, indicating that a slight improvement in the thermal stability of collagen upon the inclusion of both ILs, particularly in the case of [Ch][Seri]-containing samples [20]. ΔH values, which represent the energy required to release free and bound water as well as for collagen denaturation [21], decreased from 324.5 J/g in control films to 223.5 J/g for 6[Ch][DHP] and 193.6 J/g for 6[Ch][Seri], in accordance with the decrease in the structural order observed in the XRD analysis (Fig. 5). Regarding the glass transition temperature, T g decreased from 38.5 ◦C in control films to 36.2 ◦C for 6[Ch][DHP] and 32.53 ◦C for 6 [Ch][Seri]. This reduction in Tg can be attributed to the interactions between collagen and ILs that disrupt hydrogen bond in collagen, resulting in changes in the collagen structure [22]. The influence of incorporating different IL types and contents was studied by ATR-FTIR measurements. As can be seen in Fig. 3a), the main absorption bands of collagen are observed in the FTIR spectra of all samples, irrespective of IL concentration or type. The absorption band at 3500–3000 cm −1 is attributed to N – H stretching vibration of amide A and the band at 1632 cm −1 is associated with the C – – O stretching vibration of amide I. The band at 1547 cm −1 corresponds to the N – H bending vibration of amide II, and the band at 1238 cm −1 represents the C – N stretching vibration of amide III [23,24]. Furthermore, the absorption bands at 1080 and 1032 cm −1 are associated to C – O and C-O-C stretching vibrations, respectively [23,25]. Regarding ILs, the slight absorption band detected at 960 cm −1 in [Ch][DHP]/collagen and [Ch] [Seri]/collagen films is attributed to the C – N stretching of the [Ch] + cation [18,26,27]. Additionally, for the 6[Ch][DHP] sample, two shoulders at 946 cm −1 and 1080 cm −1 are noticeable, corresponding to the P-OH group and P – – O groups of [DHP] − , respectively [18,28,29]. When ILs were added, slight changes in the relative intensity between the amide I and amide II bands were observed (Fig. 3b). These changes may be attributed to conformational changes in the collagen structure induced by the presence of ILs. Specifically, the intensity of the amide I band was lower than that of the amide II band for the control films. However, for the 6[Ch][Seri] sample, the intensity of amide I band became smaller than that of amide II. This shift is attributed to the physical interactions, specifically ionic interactions between collagen chains and ILs. It is worth noting that these interactions can change the native structure of collagen [20]. Still, at low concentrations, ionic liquids primarily affect the supramolecular structure of collagen without significantly impacting the secondary structure of collagen fibers at molecular level [30]. This observation aligns with the SEM images of the film cross-section (Fig. 4). In fact, [Ch][DHP] was found to promote the formation of intraand intermolecular forces within the helices of collagen due to electrostatic interactions between collagen and [Ch] [DHP] [31]. 3.2. Morphological analyses SEM and XRD tests were conducted to stablish a connection between the previously mentioned properties and the structural characteristics of the collagen film. This investigation is essential because ILs have the potential to disturb the triple helix structure of collagen. The crossFig. 1. Schematic representation of the developed pressure sensor system. M. Andonegi et al.

International Journal of Biological Macromolecules 256 (2024) 128486 4 section morphology of the IL/collagen composites was evaluated by SEM and representative images are shown in Fig. 4, which includes pristine collagen (control), 6[Ch][DHP] and 6[Ch][Seri]. It is noteworthy that all films showed the compact and uniformly dense fibrillar structure consistent with collagen. These findings confirm that the fibrillar structure of the samples remained intact irrespective of the addition of ILs [32,33]. Additionally, all samples showed the semicrystalline XRD pattern characteristic of collagen, featuring diffraction peaks at 2ɵ =7.80◦and 20.50◦(Fig. 5). On the one hand, the distinct peak denoted as A corresponds to the crystalline structure of collagen and represents the distance between the collagen chains. On the other hand, the broader band denoted as B is associated with the scattering caused by components Fig. 2. Schematic representation of the electronic circuit for data acquisition (a) and image of the application developed (b). Table 1 Glass transition temperature (T g ), denaturation temperature (T d ), and enthalpy (ΔH) values obtained by DSC analysis of collagen films containing [Ch][DHP] and [Ch][Seri]. Films T g (◦C) ±1 ◦C T d (◦C) ±1 ◦C ΔH (J/g) ±1 % Control 38.5 84.1 324.5 1[Ch][DHP] 36.9 86.4 275.5 3[Ch][DHP] 36.3 88.6 266.3 6[Ch][DHP] 36.2 91.3 223.5 1[Ch][Seri] 34.9 91.1 213.7 3[Ch][Seri] 32.8 92.4 210.5 6[Ch][Seri] 32.5 98.4 193.6 Fig. 3. – FTIR-ATR spectra of collagen films with different contents of [Ch][DHP] and [Ch][Seri]: a) from 4000 to 800 cm −1 and b) from 1800 to 800 cm −1 . Fig. 4. SEM cross-section images of a) pristine collagen (control), b) 6[Ch][DHP], and c) 6[Ch][Seri] films. M. Andonegi et al.

International Journal of Biological Macromolecules 256 (2024) 128486 5 within collagen fibers, representing the amorphous phase [31,34]. In the case of samples containing [Ch][DHP] (Fig. 5a), a slight decrease in the intensity of peak A was observed upon the incorporation of 1 wt% IL. This reduction became more pronounced in samples with 3 and 6 wt% [Ch][DHP], indicating a decrease of the structural order in collagen as IL contents increased. This observation may be attributed to the ability of ILs to disrupt hydrogen bonds in collagen and modify its structure [22,35,36]. The shift in this peak position to 7.35◦for 6[Ch][DHP] samples indicates an increase in the distance between molecular chains from 1.14 nm to 1.20 nm. Additionally, an increase in the intensity of the band B was observed with the addition of [Ch][DHP], regardless of the IL content, and a shoulder at approximately 40◦appeared in the 6[Ch] [DHP]samples. A similar effect was observed with the incorporation of [Ch][Seri] (Fig. 5b), resulting in a more significant decrease in the intensity of peak A for the samples with 3 and 6 wt% [Ch][Seri] than for the samples with the same [Ch][DHP] contents. This observation points to a more substantial reduction in the collagen structural order for [Ch][Seri] containing samples, although no peak shift was observed. It must be noted that the absence of the peak at around 30◦, characteristic of the triple helical structure of collagen [37], suggests that the triple helical structure of collagen is partially disrupted in all samples [38], an effect attributed to the processing conditions. 3.3. Mechanical and d.c. electrical properties Tensile tests were used to assess the effect of ILs on the mechanical properties of the films, and the results are shown in Fig. 6a) and summarized in Table 2. It was observed that there was no significant difference in tensile strength but a slight increase in elongation at break was noticed for [Ch][DHP] containing samples as the IL content increased. This phenomenon can be attributed to the hygroscopic nature of this IL, which enhanced the film water retention capacity [39,40]. In contrast, for samples containing [Ch][Seri], a significant decrease in tensile strength and an increase in elongation at break were observed in the case of 3[Ch][DHP] and 6[Ch][DHP] films. This behavior is ascribed to the plasticizing effect of the IL. Furthermore, the decrease in the structural order in the samples containing 3 and 6 wt% [Ch][Seri], as observed by XRD analysis (Fig. 5b), could also contribute to the decrease in tensile strength. The d.c. electrical conductivity value of collagen films as a function of IL content is shown in Fig. 6b. The inset shows the characteristic I-V curves for the films containing 6 wt% IL, demonstrating nearly linear Ohmic behavior, with slight non-linear contributions. Regardless of IL type, the electric conductivity ( σ ) increased with rising IL content up to 6 wt%, when a slight decrease is observed. This improvement in electrical response can be attributed to the increased presence of mobile anions and cations facilitated by ILs [41]. At higher IL concentrations (6 wt%) IL-IL interactions intensify, limiting ion mobility and leading to a Fig. 5. XRD patterns of a) [Ch][DHP] and b) [Ch][Seri] containing collagen films. Fig. 6. a) Stress/strain curves obtained by tensile tests and b) d.c. electrical conductivity values of collagen films containing [Ch][DHP] and [Ch][Seri] ILs. Inset: I-V curves for 6[Ch][DHP] and 6[Ch][Seri] collagen films. Table 2 Tensile strength (TS) and elongation at break (EB) of collagen films prepared with different [Ch][DPH] and [Ch][Seri] contents. Films TS (MPa) EAB (%) Control 25.16 ±4.29 b,c 10.41 ±0.66 a 1[Ch][DHP] 26.92 ±1.20 c 16.59 ±1.45 b 3[Ch][DHP] 27.08 ±1.97 c 17.71 ±1.45 b 6[Ch][DHP] 26.03 ±2.59 b,c 16.88 ±0.83 b 1[Ch][Seri] 27.40 ±3.78 c 12.97 ±0.54 a 3[Ch][Seri] 21.05 ±3.94 a,b 15.70 ±2.43 b 6[Ch][Seri] 18.09 ±2.57 a 16.09 ±1.56 b a-c: Two means followed by the same letter in the same column are not significantly (P >0.05) different through the Tukey's multiple range test. M. Andonegi et al.

International Journal of Biological Macromolecules 256 (2024) 128486 6 slight decrease of the conductivity [42]. Furthermore, the type of IL slightly affected electric conductivity due to differences in size and interactions stablished within the collagen matrix, affecting ion mobility. Collagen films containing 3 wt% of [Ch][DHP] exhibited the highest electrical conductivity with a value of 1.4 ×10 −8 S⋅cm −1 . 3.4. Degradation and cytotoxicity Degradation assays were performed to evaluate the behavior of collagen films containing IL when exposed to PBS and DMEM. As can be seen in Fig. 7a, all samples completely degraded after 4 days in PBS despite their different profile. Collagen and [Ch][DHP]/collagen composites displayed a similar degradation profile. After 1 day, collagen films revealed a weight loss of ~25 %, which increased to ~30 % on the second day, and reached 100 % by the fourth day. In particular, the weight loss increased from 19 % on the first day to 28 % on the second day for1[Ch][DHP] films, from 18 % on the first day to 24 % on the second day for 3[Ch][DHP]films, and from 17 % on the first day to 27 % on the second day for 6[Ch][DHP] films. These results indicated that the incorporation of [Ch][DHP] did not affect the degradation profile of the polymer. In contrast, the addition of IL [Ch][Seri] induced a faster weight loss when compared to pristine collagen. An increase in [Ch] [Seri] content led to a significant increase in the weight loss of the composite films. In relation to the degradation assays in DMEM (Fig. 7b), it was observed that the degradation profiles of the samples were quite similar, with the exception of 6[Ch][Seri] films, which presented a faster degradation rate. After 1 day in DMEM, 6[Ch][Seri] films revealed a weight loss of 25 %, which increased approximately to 43 % on the second day, 83 % on the fourth day, and 88 % on the seventh day. Additionally, the cytotoxicity of the samples was evaluated and the obtained results are presented in Fig. 8. None of the IL/collagen films exhibited citotoxixiy, as all samples presented MC3T3-E1 cell viabilities higher than 70 %. This demonstrates the suitability of both [Ch][DHP]/ collagen and [Ch][Seri]/collagen composites for biomedical applications. 3.5. Sensor response Taking the electrical properties into account, 3[Ch][DHP] films were selected to develop a force sensor. In Fig. 9a, the resistance variation during100 compression cycles over time is depicted using a Shimadzu AG-IS with a 500 N load cell. A more detailed representation of these variations over 3 cycles is presented in Fig. 9b, illustrating the correlation between the resistance variation and the force on the sensor. The resistance variation increased proportionally with the force applied to the film and vice versa. Variations in the air gap between the interdigit and the collagen film allowed variations in the contact area with the interdigit, changing the electrical behavior and resulting in the decrease of the resistance value (increase of ΔR) with the applied force (Fig. 9b), measured as an electrical signal. Fig. 9c) shows the resistance variation as a function of the applied force for a specific sensor, which is representative of the remaining sensors. The decrease in resistance was only detected when the applied force reached approximately 1.2 N, the minimum force required to place the collagen film in contact with the interdigit. The sensor exhibited a sensitivity of approximately 8 MΩ/N, and its response remained stable throughout cycling with a good repeatability and no hysteresis. Fig. 9 d) shows the variation received by the application when the different sensors were pressed sequentially with a finger, from sensor S1 to S9, showing an identical variation in each sensor and allowing a reliable detection of the pressure. The GUI built in Qt Modeling Language (QML) received the value of the 9 sensors and detected if the resistance value was above a threshold, based on the applied pressure. If the resistance value was less than the threshold, the algorithm assumed that the button was pressed, changing the state of each square. Fig. 7. Weight loss relative to the original mass after degradation for 7 days in a) PBS and b) DMEM media for IL/collagen films. Fig. 8. Cytotoxicity assays with MC3T3-E1 cells in contact with the as-prepared extraction media exposed to the IL/collagen composites with different [Ch] [DHP] and [Ch][Seri] contents for 72 h (relative cell viability was presented as the percentage of the negative control (DMEM, n =3 ±standard error of mean (SEM)). M. Andonegi et al.

International Journal of Biological Macromolecules 256 (2024) 128486 7 4. Conclusions Biocompatible ionic liquid (IL)/collagen films were developed for force sensor applications. These collagen films were prepared using the casting method incorporating two different ILs, choline dihydrogen phosphate ([Ch][DHP]) and choline derinate ([Ch][Seri]), with IL content variations (1, 3, 6 wt%). These films were characterized by a compact and homogeneous fibrillar structure, which provided suitable thermal and mechanical properties. Additionally, IL/collagen films were proved to be non-cytotoxic with a cell viability >70 %. Notably, the d.c. electrical conductivity of the composites increased particularly for collagen films with 3 wt% of [Ch][DHP], reaching 1.4 ×10 −8 S.cm −1 . Consequently, this particular formulation was selected for the development of force resistive sensors, which showed excellent response and sustained stability over time. Author statement We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements Grant PID2021-124294OB-C22 funded by MCI/AEI10.13039/ 501100011033 and by “ERDF A way of making Europe”. This work was also supported by the Basque Government (IT1658-22) and the Portuguese Foundation for Science and Technology (FCT) under strategic funding UIDB/04650/2020, UID/FIS/04650/2021, project PTDC/FISMAC/28157/2017, 2022.05932.PTDC and Investigator FCT Contract 2020.02915.CEECIND (D.M.C) and 2020.04028.CEECIND (C.M.C.) funded by national funds through FCT and by the ERDF through the COMPETE2020-Programa Operacional Competitividade e Internacionalizaç˜ ao (POCI). The authors also acknowledge funding from the Basque Government Industry and Education Department under the ELKARTEK program. M.A thanks the Basque Government1 for her fellowship (POS_2022_1_0007). References [1] C. Ding, et al., The response of collagen molecules in acid solution to temperature, Polymer 55 (22) (2014) 5751–5759. [2] P.C. Balaure, et al., In vitro and in vivo studies of novel fabricated bioactive dressings based on collagen and zinc oxide 3D scaffolds, Int. J. Pharm. 557 (2019) 199–207. [3] S.A. Ghodbane, M.G. Dunn, Physical and mechanical properties of cross-linked type I collagen scaffolds derived from bovine, porcine, and ovine tendons, J. Biomed. Mater. Res. A 104 (11) (2016) 2685–2692. [4] A. Irastorza, et al., The versatility of collagen and chitosan: from food to biomedical applications, Food Hydrocoll. 116 (2021), 106633. [5] A.K. Lynn, I.V. Yannas, W. Bonfield, Antigenicity and immunogenicity of collagen, J. Biomed. Mater. Res. B Appl. Biomater. 71B (2) (2004) 343–354. Fig. 9. a) 100 cycles of loading and unloading force applied to the sensor and the corresponding resistance variation over time. b) Magnification of some of the force application cycles and c) resistance variation correlation with the force applied to the sensor. d) Digital output received in the software for the 9 sensors pressed sequentially. M. Andonegi et al.

International Journal of Biological Macromolecules 256 (2024) 128486 8 [6] M. Andonegi, et al., Physicochemical and biological performance of aloe Veraincorporated native collagen films, Pharmaceutics 12 (2020), https://doi.org/ 10.3390/pharmaceutics12121173. [7] J.L. Lauer-Fields, D. Juska, G.B. Fields, Matrix metalloproteinases and collagen catabolism, Pept. Sci. 66 (1) (2002) 19–32. [8] D. Benayahu, et al., Unique collagen fibers for biomedical applications, Mar. Drugs 16 (2018), https://doi.org/10.3390/md16040102. [9] M. Andonegi, et al., Structure-properties relationship of chitosan/collagen films with potential for biomedical applications, Carbohydr. Polym. 237 (2020), 116159. [10] A. Le´ on-L´ opez, et al., Hydrolyzed collagen-sources and applications, Molecules 24 (2019) 22. [11] L. Maistrenko, et al., Collagen obtained from leather production waste provides suitable gels for biomedical applications, Polymers 14 (2022), https://doi.org/ 10.3390/polym14214749. [12] J. Bebbington, J. Unerman, Achieving the United Nations sustainable development goals, Account. Audit. Account. J. 31 (1) (2018) 2–24. [13] X. Zhang, et al., Collagen-based flexible electronic devices for electrochemical energy storage and sensing, Macromol. Rapid Commun. 44 (10) (2023), p. 2200977. [14] Y. Zeng, et al., A sustainable biomemristive memory device based on natural collagen, Materials Today Chemistry 13 (2019) 18–24. [15] Z. Bai, et al., Versatile nano–micro collagen fiber-based wearable electronics for health monitoring and thermal management, J. Mater. Chem. A 11 (2) (2023) 726–741. [16] B.T. Mekonnen, M. Ragothaman, T. Palanisamy, Bifunctional hybrid composites from collagen biowastes for heterogeneous applications, ACS Omega 2 (8) (2017) 5260–5270. [17] D.M. Correia, et al., Ionic liquid–polymer composites: a new platform for multifunctional applications, Adv. Funct. Mater. 30 (24) (2020) 1909736. [18] M. Andonegi, et al., Sustainable collagen blends with different ionic liquids for resistive touch sensing applications, ACS Sustainable Chemistry & Engineering 11 (15) (2023) 5986–5998. [19] J. Gonz´ alez-Masís, et al., Self-assembly study of type I collagen extracted from male Wistar Hannover rat tail tendons, Biomaterials Research 24 (1) (2020) 19. [20] A. Tarannum, et al., Are aqueous solutions of choline-based ionic liquid biocompatible cross-linkers for collagen? J. Mol. Liq. 303 (2020), 112654. [21] V. Samouillan, et al., The use of thermal techniques for the characterization and selection of natural biomaterials, Journal of Functional Biomaterials 2 (2011) 230–248, https://doi.org/10.3390/jfb2030230. [22] A. Tarannum, J.R. Rao, N.N. Fathima, Choline-based amino acid ILs–collagen interaction: enunciating its role in stabilization/destabilization phenomena, J. Phys. Chem. B. 122 (3) (2018) 1145–1151. [23] K. Belbachir, et al., Collagen types analysis and differentiation by FTIR spectroscopy, Anal. Bioanal. Chem. 395 (3) (2009) 829–837. [24] T. Riaz, et al., FTIR analysis of natural and synthetic collagen, Appl. Spectrosc. Rev. 53 (9) (2018) 703–746. [25] X. Bi, et al., A novel method for determination of collagen orientation in cartilage by Fourier transform infrared imaging spectroscopy (FT-IRIS), Osteoarthr. Cartil. 13 (12) (2005) 1050–1058. [26] R.M. Meira, et al., Ionic liquid-based electroactive materials: a novel approach for cardiac tissue engineering strategies, J. Mater. Chem. B 10 (34) (2022) 6472–6482. [27] V. Rigual, et al., Protic, aprotic, and choline-derived ionic liquids: toward enhancing the accessibility of hardwood and softwood, ACS Sustain. Chem. Eng. 8 (3) (2020) 1362–1370. [28] S.M. Kamal Mohamed, et al., Facile preparation of biocompatible and transparent silica aerogels as Ionogels using choline dihydrogen phosphate ionic liquid, Appl. Sci. 11 (2021), https://doi.org/10.3390/app11010206. [29] A. Reizabal, et al., Silk fibroin bending actuators as an approach toward natural polymer based active materials, ACS Appl. Mater. Interfaces 11 (33) (2019) 30197–30206. [30] A. Mehta, J.R. Rao, N.N. Fathima, Effect of ionic liquids on the different hierarchical order of type I collagen, Colloids Surf. B Biointerfaces 117 (2014) 376–382. [31] S.Y. Bak, et al., Assessment of the influence of acetic acid residue on type I collagen during isolation and characterization, Materials 11 (2018), https://doi.org/ 10.3390/ma11122518. [32] P. Ji, et al., Collagen film with bionic layered structure and high light transmittance for personalized corneal repair fabricated by controlled solvent evaporation technique, Journal of Functional Biomaterials (2022) 13, https://doi. org/10.3390/jfb13020052. [33] Y. Ma, et al., A top-down approach to improve collagen film’s performance: the comparisons of macro, micro and nano sized fibers, Food Chem. 309 (2020), 125624. [34] N. Re´ ategui-Pinedo, et al., Characterization of collagen from three genetic lines (gray, red and F1) of Oreochromis niloticus (Tilapia) skin in young and old adults, Molecules (2022) 27, https://doi.org/10.3390/molecules27031123. [35] Y. Hu, et al., Evaluation of 1-ethyl-3-methylimidazolium acetate based ionic liquid systems as a suitable solvent for collagen, J. Appl. Polym. Sci. 130 (4) (2013) 2245–2256. [36] M.M. Giraud-Guille, et al., Structural aspects of fish skin collagen which forms ordered arrays via liquid crystalline states, Biomaterials 21 (9) (2000) 899–906. [37] C. Li, et al., Structural properties of pepsin-solubilized collagen acylated by lauroyl chloride along with succinic anhydride, Mater. Sci. Eng. C 55 (2015) 327–334. [38] Z. Meng, et al., Dissolution and regeneration of collagen fibers using ionic liquid, Int. J. Biol. Macromol. 51 (4) (2012) 440–448. [39] M. Rahman, C.S. Brazel, Ionic liquids: new generation stable plasticizers for poly (vinyl chloride), Polym. Degrad. Stab. 91 (12) (2006) 3371–3382. [40] E. Veroutis, et al., Intraand inter-molecular interactions in choline-based ionic liquids studied by 1D and 2D NMR, J. Mol. Liq. 322 (2021), 114934. [41] M. Qu, et al., Ion transport in ionic liquid/poly(vinylidene fluoride) system under electric fields: a molecular dynamics simulation, Colloids Surf. A Physicochem. Eng. Asp. 642 (2022), 128328. [42] F.C.A. Silva, et al., Polymer-ion interactions in PVDF@ionic liquid polymer electrolytes: a combined experimental and computational study, Electrochim. Acta 427 (2022), 140831. M. Andonegi et al.