Effect of the deposition process on the stability of Ti₃C₂T_x MXene films for bioelectronics

Abstract

Ti₃C₂T_x MXene is emerging as the enabling material in a broad range of wearable and implantable medical technologies, thanks to its outstanding electrical, electrochemical, and optoelectronic properties, and its compatibility with high-throughput solution-based processing. While the prevalence of Ti₃C₂T_x MXene in biomedical research, and in particular bioelectronics, has steadily increased, the long-term stability and degradation of Ti₃C₂T_x MXene films have not yet been thoroughly investigated, limiting its use for chronic applications. Here, we investigate the stability of Ti₃C₂T_x films and electrodes under environmental conditions that are relevant to medical and bioelectronic technologies: storage in ambient atmosphere (shelf-life), submersion in saline (akin to the in vivo environment), and storage in a desiccator (low-humidity). Furthermore, to evaluate the effect of the MXene deposition method and thickness on the film stability in the different conditions, we compare thin (25 nm), and thick (1.0 μm) films and electrodes fabricated via spray-coating and blade-coating. Our findings indicate that film processing method and thickness play a significant role in determining the long-term performance of Ti₃C₂T_x films and electrodes, with highly aligned, thick films from blade coating remarkably retaining their conductivity, electrochemical impedance, and morphological integrity even after 30 days in saline. Our extensive spectroscopic analysis reveals that the degradation of Ti₃C₂T_x films in high-humidity environments is primarily driven by moisture intercalation, ingress, and film delamination, with evidence of only minimal to moderate oxidation.

1. Introduction

Ti₃C₂T_x MXene, which was reported in 2011 [1], has been extensively investigated for a wide range of bioelectronic applications, including wearable and implantable sensors, as well therapeutic stimulation in recent years [2–4]. The motivation for the wide and growing interest in Ti₃C₂T_x MXene as a versatile bioelectronic material is due to the remarkable combination of high electrical conductivity, volumetric capacitance, and surface-to-volume ratio, in addition to hydrophilic surfaces and scalable liquid-phase processability form colloidal dispersions in pure water [5]. While there is promising potential for Ti₃C₂T_x for chronic use in vivo - for example in implantable electrodes for recording and neuromodulation of brain activity - the limited stability of Ti₃C₂T_x in high-humidity and wet environments makes its suitability for chronic applications still uncertain. Preserving the conductivity and integrity of Ti₃C₂T_x films long term is crucial to ensure the functionality and safety of MXene-based bioelectronic technologies.

The mechanisms of chemical degradation of Ti₃C₂T_x MXene flakes in aqueous dispersions have been extensively studied with the goal of improving the storage life of Ti₃C₂T_x suspensions [6]. Ti₃C₂T_x flake degradation is believed to be driven by oxidation and hydrolysis of Ti and formation of TiO₂ [6]. This phenomenon has been observed in ambient conditions (i.e., at room temperature) near atomic defects in both basal and edge planes of individual flakes [7]. Other works have shown that Ti₃C₂T_x degrades more severely in water and high-humidity conditions, driven largely by hydrolysis reactions resulting in the formation of TiO₂, alongside CO₂, CH₄, and Ti(OH)₄ [8–10]. Harsh oxidizing agents, such as H₂O₂, cause similar effect on Ti₃C₂T_x MXene nanosheets, forming TiO₂ on the flake surface [11]. TiO₂ formation rates are also known to increase with temperature and air humidity levels [12]. Finally, Ti₃C₂T_x is susceptible to deterioration under exposure to ultraviolet (UV) light, likely as a result of Ti₃C₂T_x intrinsically strong absorbance in the 250–300 nm range [13], and the strong photocatalytic properties of TiO₂ once it has formed [8],

A primary strategy to prevent Ti₃C₂T_x MXene chemical degradation is the optimization of the synthesis process. For example, inclusion of excess aluminum (Al) during the MAX precursor phase synthesis results in higher carbon stoichiometry closer to the ideal composition of Ti₃AlC₂ [14]. Thus, both the precursor and the corresponding MXene have fewer defects, which leads to significant improvements in the shelf-life of Ti₃C₂T_x flakes [15]. Beyond synthesis, storage conditions appear to be key determinants of Ti₃C₂T_x chemical stability: argon-saturation [16] and low temperatures [17], even down to freezing [18], have been shown to extend the lifetime of Ti₃C₂T_x. Furthermore, organic solvents such as ethanol [17], isopropanol [9], or acetone and acetonitrile [13] have been shown to slow down degradation of Ti₃C₂T_x suspensions. Several studies have also demonstrated the improved oxidation resistance and stable conductivity of Ti₃C₂T_x dispersions in presence of antioxidant and capping agents, such as sodium ascorbate [19] and L-ascorbic acid [19–21], tannic acid [19]; sodium phosphate, oxalate, and citrate [22]; and sodium polyphosphates, borates, and silicates [23]. The main effect of antioxidants is to protect Ti on the Ti₃C₂T_x edges from oxide formation.

MXene assemblies, such thin films, also suffer from environmental instability as their properties change under specific environmental conditions. In humid environments, water molecules infiltrate the inter-flake space of Ti₃C₂T_x MXene films and increase the interlayer spacing, resulting in an overall decrease in electronic conductivity [24,25]. The strategies that may improve then stability of Ti₃C₂T_x dispersions may not be applicable to or effective when Ti₃C₂T_x is assembled into films. Instead, the proposed strategies to extend their lifetime include binding MXene layers and preventing intercalation of water molecules. It can be achieved, for example, by intercalation of Al³⁺ ions into pre-fabricated Ti₃C₂T_x membranes, which act to strengthen bonds between the oxygen functional groups on the surface of neighboring flakes and restrict moisture ingress [25]. Polymers [26], linker molecules [27], and high temperature treatment [24,28] were shown to be effective as well. Finally, annealing of Ti₃C₂T_x films in an H₂ environment at 900 °C has been shown to recover conductivity after degradation, while also preventing further oxidation [29]. However, while additional chemical or thermal treatments may be useful for free-standing Ti₃C₂T_x films, the solvents and high temperatures typically used are incompatible with the encapsulation and substrate materials commonly used in medical devices — such as elastomers, polymers, and epoxies — which typically have glass transition temperatures T_g < 200 °C [30,31].

In the present work, we systematically investigate the effects of processing and storage conditions on the stability of Ti₃C₂T_x films over time. Specifically, we investigate the stability of films fabricated via scalable manufacturing techniques, such as spray- and blade-coating. For spray-coated films, we also compare the stability of thicker opaque Vs. transparent conductive films. To evaluate the effects of the humidity levels, we investigate three different aging conditions: 1) low humidity (10–18%, desiccator), 2) air (atmosphere), and 3) phosphate buffered saline (PBS, saline) at 37 °C to mimic physiological environments. We use X-ray diffraction (XRD), Raman spectroscopy, scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), and DC conductivity (σ_DC) of films fabricated to monitor changes over time. Finally, we fabricate electrodes featuring the same Ti₃C₂T_x film types and monitor the electrochemical impedance under the same storage conditions. Our findings show that the film morphology and processing methods, as well as the storage conditions, are key determinants of the stability of Ti₃C₂T_x, once assembled into films [32]. Specifically, we show that thicker films made by blade-coating retain their electrical and mechanical properties in all the investigated conditions, including saline. Spray-coated films, especially thin spray-coated films, degrade faster in wet environments than under dry storage, with desiccator storage providing the most stable storage condition regardless of the film deposition method. Finally, when Ti₃C₂T_x films are integrated in functional electrodes, the impedance behavior and overall stability of the electrodes are found to be largely dependent on film thickness, with blade-coated and opaque spray-coated films outlasting transparent spray-coated electrodes.

Materials and Methods

1.1. Preparation of Ti₃C₂T_x Films

Ti₃C₂T_x MXene was synthesized following the modified MAX phase synthesis process developed by Mathis et al [15]. Subsequently, solutions for spray-coating were made by redispersing ~ 12 mg mL⁻¹ Ti₃C₂T_x in DI water to a final concentration of ~1 mg mL⁻¹. Spray-coated films were made using a commercial airbrush (Gocheer) connected to a TC-20 air compressor. The spraying pressure was 52 psi (0.36 MPa), and all substrates were 25 mm × 25 mm glass slides (Fisher Scientific) pre-treated with an air plasma (Harrick Plasma Cleaner: 18 W, 350 sccm air, 2 min. exposure time). Thickness of the spray-coated films was varied to obtain either thin transparent (τ ~ 0.25–0.45 μm) or thick opaque (τ ~ 1.0 μm) films. For blade-coated films, liquid crystalline Ti₃C₂T_x dispersion at a concentration of ~60 mg mL⁻¹ was blade-coated onto glass substrates using an automated blade-coater. Blade coating is only feasible with Ti₃C₂T_x slurries, since more dilute, less viscous dispersions cannot be evenly deposited. The height of the blade from the glass substrate (15 μm) and the traverse speed (40 mm s⁻¹) were kept constants for all samples. These blade-coating parameters were chosen to match the thickness of the spray-coated thick films (τ ~ 1.0 μm), which was measured using a KLA Tencor P7 2D profilometer.

1.2. Aging of Ti₃C₂T_x Films

Ti₃C₂T_x films on glass substrates were stored in a desiccator cabinet and kept at stabilized humidity levels (relative humidity, RH 10–18%) at 25 °C. Other samples were submerged in 0.01 M PBS (pH 7.4) held at 37 °C in an air oven, to simulate in vivo aging conditions in physiological environments. PBS levels were monitored over the course of the study and refreshed as needed. Finally, a third set of samples was left under ambient atmosphere at 25 °C, stored in a cabinet out of direct sunlight with no humidity control. All samples were aged for 30 days and tested at 3 time points: immediately after fabrication (t₀), then after 15 (t_mid), and 30 days (t_f). All samples were made in the same initial batch to ensure uniformity. We tested three films per film fabrication method and storage condition at each time point, with an additional three films tested for each sheet resistance measure (n = 6). No individual sample was re-used for any subsequent time point.

Sheet resistance was measured for all film types using a Loresta handheld AX-MCP-T370 resistivity meter (Nittoseiko Analytech Co., Ltd). The UV/Vis spectra of thin transparent films were also collected at the same time points as sheet resistance measurements, using a Spectronic 200 spectrophotometer (ThermoScientific) in the 300–1000 nm wavelength range. In this window, the transparency of Ti₃C₂T_x MXene has a characteristic local minima at ~ 780 nm, which was used to monitor film changes in each aging condition.

Raman and XRD spectra were collected at t₀ and t_f. The Renishaw InVia confocal Raman microscope was used with λ = 785 nm, 5% power (~0.15 mW), and a 20x objective. Spectra were measured with a 1200 lines/mm grating and 30 s of exposure. XRD patterns were acquired using a Rigaku MiniFlex benchtop X-ray diffractometer (Rigaku Co. Ltd.) with a Cu Kα (λ = 0.1542 nm) source at 40 kV voltage and 15 mA current, for 2θ = 3°–60°, at a rate of 7° min⁻¹, with a step of 0.02°. The d-spacing value of the (002) peaks in the Ti₃C₂T_x diffraction patterns was calculated using Bragg’s law. XPS was also taken at t₀ and t_f, targeting the Ti 2p core-level region. Ti₃C₂T_x films were mounted on conductive carbon tape with a copper tape bridge, using a PHI VersaProbe 5000 instrument (Physical Electronics, USA) with a 200 μm, 50 W monochromatic Al Kα X-ray source. Samples were not sputtered, pass energy and step size were set at 23.5 eV and 0.05 eV, respectively, and subsequent quantification and peak fitting were conducted using CasaXPS V2.3.19 software. Finally, scanning electron microscopy (SEM) images were also acquired for the films using a Zeiss Supra 50VP Field Emission SEM (Zeiss Group) with a 3 kV accelerating voltage. Additionally, the water contact angle of films was measured using a Keyence VHX6000 digital microscope. The advancing angle of a water droplet placed on the samples was calculated from the images (n = 1 for each film type).

1.3. Preparation and aging of Ti₃C₂T_x electrodes

Ti₃C₂T_x electrodes were prepared by spray-coating Ti₃C₂T_x MXene colloidal dispersions onto glass slides pre-patterned with gold (Au) lines at the top for electrical connections. An insulated wire was then soldered to the Au pads for connection to the data acquisition system. Samples were encapsulated with polyimide tape (Kapton) pre-patterned with a CO₂ laser to create the electrode vias. Two sets of electrodes were aged for 30 days in the desiccator and air, respectively, while a separate set of electrodes was aged in saline for 7 days, the maximum duration for which all samples remained measurable and did not experience encapsulation failures. PBS-aged electrodes were placed in containers such that the connection wires were not exposed to the solution, and only the electrode area was submerged in solution. Electrochemical impedance spectroscopy (EIS) was acquired for all electrodes at the initial and final timepoints, using a Gamry Reference 600 Potentiostat/Galvanostat/ZRA (Gamry Instruments, Inc.). A three-electrode configuration was used with a graphite rod counter electrode, aqueous Ag/AgCl reference electrode (1M KCl), and with an input voltage of 10 mV_rms, sweeping the frequencies from 1–10⁵ Hz.

To further investigate the changes at the Ti₃C₂T_x-electrolyte interface over the course of the study, we also fit the impedance spectra of the electrodes at t₀ and t_f using previously established equivalent circuit models for Ti₃C₂T_x MXene (detailed in Supplementary Information) [4,33]. All circuit models were built using Gamry’s EChem Analyst software package and featured a charge-transfer resistance (R_ct), Warburg impedance (W), and constant phase element (CPE) to describe the Ti₃C₂T_x electrodes, while saline was described by its spread resistance (R_s). Double-layer capacitance (C_dl) was derived from the CPE and R_s parameters, according to previously published literature [34]. The initial value of R_s was estimated to be 0.36 ± 0.07 kΩ cm² for blade coated films, 59.34 ± 25.93 kΩ cm² for spray coated thick films, and 1.31 ± 0.89 kΩ cm² for spray coated thin films based on geometric considerations [34,35], while the Ti₃C₂T_x-related terms were initialized from literature values [4]. Spectra for three separate electrodes for each film type and aging condition were averaged to obtain representative results at each timepoint.

Electromyography (EMG) recordings

To demonstrate the use of Ti₃C₂T_x films for bioelectronic applications, EMG recordings were taken from the forearm (Flexor digitorum) to measure muscle contractions using an Intan RHS2000 Stimulation/Recording Controller (Intan Technologies) at a 20 kHz sampling rate. Reference and ground were placed on the inner and outer elbow using Natus disposable adhesive electrodes, with the Ti₃C₂T_x film Kapton electrodes placed on the middle of the forearm. To visualize the muscle contraction a 60 Hz notch filter was applied to the raw signal data, followed by a 10 – 150 Hz bandpass filter. The amplitude envelope of the raw signal data was overlaid to show an isolated contraction for each film type incorporated into an electrode.

2. Results and Discussion

2.1. Stability of film conductivity

To investigate the contributions of different processing methods on the stability of Ti₃C₂T_x films, we fabricated Ti₃C₂T_x films via spray- and blade-coating (Figure 1). Transparent thin films were formed by spray-coating on glass substrates until a transmittance of ~60% at 550 nm was achieved (Supplementary Figure S1 C). This transmittance correlates with a thickness of approximately 25 nm, as obtained from previously established relations between film thickness and absorbance [36]. We deposited both thick spray-coated and blade-coated films to a thickness of ~1.0 μm (Supplementary Figure S1 A, B). The choice of these deposition methods allowed us to explore the influence of (i) thickness (thick Vs. transparent thin spray-coated), and (ii) flake alignment (spray-coated Vs. blade-coated) on the stability of the films. In blade-coated films, Ti₃C₂T_x nanosheets have a higher degree of alignment due to the liquid crystalline nature of the Ti₃C₂T_x slurries required for this process [37]. In addition to films on glass slides, we also fabricated 3 mm diameter electrodes using the same types of films (Figure 1(a)). We aged films and electrodes in three different environments that are relevant for applications in bioelectronic interfacing: (i) atmosphere at room temperature (shelf-life), (ii) desiccator storage (low-humidity packaging) (ii) saline at 37 °C (in vivo physiological environments) (Figure 1(b)).

Figure 1. Sample fabrication and storage conditions.

(a) Schematic of the film and electrode fabrication process. (b) Storage conditions for 30-day aging in i) atmosphere at 25 °C, ii) desiccator storage, and iii) 1X PBS (saline) at 37 °C.

At the initial timepoint, thicker films had lower sheet resistance, with blade-coated thick films being the most conductive (0.37 ± 0.12 Ω sq⁻¹; n = 6 samples) followed by spray-coated thick films (2.21 ± 0.19 Ω sq⁻¹; n = 6 samples), and spray-coated thin films (58.1 ± 6.45 Ω sq⁻¹; n = 6 samples). All film types maintained their initial sheet resistance values over one month in ambient conditions (blade-coated: 0.52 ± 0.09 Ω sq⁻1, spray-coated thick: 2.84 ± 0.28 Ω sq⁻¹, spray-coated thin: 302 ± 9.69 Ω sq⁻¹, Figure 2(a)). It should be noted, however, that all film types showed minor fluctuations in sheet resistance over the course of one month in air, likely due to natural temperature and/or humidity fluctuations in the environment.

Figure 2. Electronic and morphological changes of the Ti₃C₂T_x films in different aging conditions.

(a) Sheet resistance. (b) SEM images of films at the initial and final time points. Scale bars: 2μm.

In saline, the sheet resistance of all films increased over time (Figure 2(a)). For thick spray- and blade-coated films, the sheet resistance at t_f increased by ~6x and ~5x (t_f, 13.7 ± 2.80 and 1.88 ± 0.39 Ω sq⁻¹), respectively, while for the thin spray-coated films, the sheet resistance at t_f was ~32x higher than the initial value (t_f, 1.86 ± 0.30 kΩ sq⁻¹). Sheet resistance at t_mid for blade-coated samples was ~1.5x higher than the final values (t_mid, 2.71 ± 0.23 Ω sq⁻¹). The observed minor decrease in sheet resistance in saline aged blade coated samples from 14 days to 28 days may be attributable to the difference in structure of blade coated films compared to spray coated films and merits further investigation in future studies to understand the underlying phenomena. We also saw visible changes in the films, such as salt crystal formation and tearing of the sheet layer, in the surface morphology of the films aged over the course of a month in saline (Figure 2(b)). Most importantly, both thin and thick spray-coated films lost their integrity, while blade-coated films remained intact even after delaminating from the glass substrate (Supplementary Figure S2).

2.2. Effect of aging on structure and chemical properties

To investigate whether film degradation is driven by oxidation of the Ti₃C₂T_x flakes, by moisture ingress and increased interlayer spacing between the Ti₃C₂T_x layers — or a combination of both — we acquired Raman, XPS, and XRD data on the films at the initial and final study timepoints.

Raman spectra of all samples indicate that they are all Ti₃C₂T_x MXene (Figure 3). The in-situ studies used the out-of-plane carbon and surface termination vibrations to track oxidation. At t₀, blade- and thick spray-coated films had the same A_1g(C) peak position—roughly 725 cm⁻¹—which suggests no compositional differences [38]. At t_f, all aged samples showed only a minor blue shift of the A_1g(C) peak (centered around ~720 cm⁻¹), with higher shifts indicating more oxidation in samples (Supplementary Table S1)[39]. Though all samples had minor shifts, within acceptable error, blade-coated films are expected to have more aligned Ti₃C₂T_x MXene flakes compared to spray-coated films [37,38], which should enhance the film stability. This hypothesis was confirmed by the smaller shift in the A_1g(C) peak position of blade coated films following air and desiccator storage compared to thick spray-coated films in the same aging conditions (Supplementary Table S1). This same trend did not hold true for thicker films aged in saline in which blade coated films had a larger shift in the A_1g(C) peak, while spray coated thick films had a smaller shift. This may be caused by a difference in moisture intercalation in blade coated films that can be attributed to the resulting flake stacking from this processing method [38]. Still, we observed shifts in the Ti₃C₂T_x MXene resonance peaks at ~120 cm⁻¹ for all samples. These peaks are related to Ti₃C₂T_x MXene vibronic properties [38], and changes in their location may be correlated with the observed decreases in DC conductivity of the Ti₃C₂T_x films. Finally, we observed that at t_f none of the samples showed the D or G bands of amorphous graphitic carbon (expected around 1350–2700 cm⁻¹), suggesting that the Ti₃C₂T_x in all samples was not fully degraded, and that Ti–C bonds were still present (Supplementary Figure S3). There were also no TiO₂ peaks or photoluminescence from any of the samples, so while the A_1g peak shifts may suggest a higher oxygen content, the films were not fully oxidized [38,40,41].

Figure 3. Raman spectra of pristine vs. aged Ti₃C₂T_x films over 30 days.

To probe the film surface oxidation state, we acquired XPS on air and desiccator-aged samples (Table 1 and Supplementary Figure S4). Saline-aged spray-coated samples were fragmented and covered by a deposit from PBS (salt crystals). XPS measurements were only able to detect phosphorus and miscellaneous carbon rather than the underlying Ti₃C₂T_x in these samples. XPS spectra from all pristine Ti₃C₂T_x MXene films were similar, confirming minimal oxidation as evidenced by the small (~10%) contributions from TiO₂ components near 459 eV and 465 eV to the overall Ti 2p core-level spectra. This small amount of “intrinsic” oxide is attributed to oxidation of the edges of the Ti₃C₂T_x flakes during synthesis or after film deposition [15,42]. The long-term stability of the Ti₃C₂T_x films was most influenced by their thickness: while blade-coated and thick spray-coated films showed virtually no change in the degree of oxidation after 30 days, regardless of storage conditions (desiccator: ~10% TiO₂, air: ~12–15% TiO₂), the thin spray-coated films showed a slight increase in oxidation after storage in desiccator (~19% TiO₂ at t_f) but significantly higher oxidation after storage in air (~53% TiO₂ at t_f). The combination of Raman spectroscopy and XPS-derived oxidation assays indicate that - although moderate oxidation is present - oxidation alone may not be the primary driving mechanism of degradation of Ti₃C₂T_x films at varying humidity levels. This is especially true considering that XPS is sensitive to only the top ~1–10 nm of the surface. Therefore, even for the most damaged sample as in the thin spray-coated samples aged in air, while the surface may be oxidized, the bulk is still pristine MXene, as indicated by the Raman spectra coming from a depth of a few tens of nanometers.

Table 1.

Summary of the XPS results for different film types at the initial (t₀) and final (t_f, one month) timepoints of the study. (Values for the PBS aging condition are not reported, as at t_f spray-casted films were too fragmented to be analyzed). The third column shows the ratio of the titanium oxide component to the sum of the titanium carbide with varying surface termination composition in MXene derived from the Ti 2p spectra. The fourth column shows the ratio of the titanium oxide component to the titanium bonded to oxygen surface terminations component in MXene derived from the O 1s spectra.

Film Type	Aging Condition	[TiO2]:[Ti-Tix]	[TiOx oxide]:[Ti-O MXene]
Thin spray-cast	t 0	0.12	1.08
	tf in Air	1.11	1.07
	tf in Desiccator	0.23	0.87
Thick spray-cast	t 0	0.11	0.46
	tf in Air	0.17	0.63
	tf in Desiccator	0.11	0.48
Blade-coated	t 0	0.11	0.65
	tf in Air	0.14	0.57
	tf in Desiccator	0.11	0.67

With Raman and XPS analysis indicating that oxidation cannot alone explain the decrease in conductivity of Ti₃C₂T_x in humid and wet environments, we then acquired XRD patterns to track changes in the d-spacing values of the (002) peak (Figure 4). Table 2 provides the d-spacing values for all film types and aging conditions. At t₀, the d-spacing of blade- and spray-coated thick films was 12.64 ± 0.03 Å and 12.61 ± 0.07 Å, respectively (n = 3 samples), in line with previous reports [40]. After aging in saline for 30 days, moisture ingress caused an increase in d-spacing of ~2–3 Å for these films, which is consistent with the average size of a water molecule [43]. Thus, as an additional layer of water separated MXene flakes, the conductivity decreased. Comparatively, in low humidity and atmosphere aging conditions, the d-spacing of the thick spray-coated and blade-coated films remained largely unchanged (Figure 4(b), (c)). The initial d-spacing of transparent thin films was higher than that of the thicker films (14.80 ± 0.25 Å), suggesting that the Ti₃C₂T_x flakes were not as uniformly aligned in the thin films compared to thicker films. In saline, the final d-spacing of thin films slightly decreased (13.82 ± 0.24 Å, Figure 4(a)). Since the thin films consist of less uniformly aligned flakes to begin, moisture infiltration might occur more easily, resulting in this apparent equilibrium in d-spacing. In air, however, there was still a minor increase in the d-spacing for the thin films, likely due to greater interlaying spacing that allows for ambient moisture ingress between the flakes. Higher order peaks in the XRD spectrum correspond to the (006) and (008) planes. These peaks can be assigned to the out-of-plane 00l reflections typical of MXene nanosheets stacked uniformly layer-by-layer with respect to the surface of the substrate [44]. The difference in appearance of these peaks for both types of thick films aged under the desiccator condition is likely a result improved layer-by-layer stacking of flakes, as moisture was driven from the films (Figure 4(b), (c)). Altogether, XRD analysis revealed that the interlayer spacing of the films is strongly dependent on the overall film thickness rather than flake alignment. Furthermore, moisture ingress seems to contribute to degradation of the film properties due to an increase in the interlayer spacing of the Ti₃C₂T_x films. To further support this, we measured film wettability over time. Spray coated thin films initially have a lower contact angle (20.13 degrees) in comparison to spray coated thick films (51.27 degrees) and blade coated thick films (56.18 degrees), indicating thinner films with less uniformly aligned flakes are more wettable and hydrophilic, thus more prone to moisture ingress. In dry storage conditions spray coated thin films do not show much variation in contact angle over time. However, for both spray coated, and blade coated thick samples, the water contact angle does decrease over time, with a larger observable decrease in air (t_f spray-thick: 28.37 degrees, t_f blade-thick: 26.97 degrees) rather than desiccator storage (t_f spray-thick: 47.34 degrees, t_f blade-thick: 38.11 degrees) indicating that films become more wettable when aged in air as the film surface becomes more hydrophilic. We found all film types showed a notable decrease (~3x) in water contact angle when aged in saline, thus samples became more hydrophilic and wettable (Supplementary Table S3). These results, coupled with very moderate surface oxidation evidenced by Raman spectroscopy and XPS, highlight water intercalation as the driving mechanisms for the degradation of Ti₃C₂T_x MXene transparent thin films in ambient and high-humidity environments.

Figure 4.

XRD patterns and d-spacing values of aged films at initial and final time points across all aging conditions.

Table 2.

X-ray Diffraction (XRD) d-spacing values for all samples at the initial and final timepoints. Values reported as avg. ± std. dev. across n = 3 samples.

		(002) d-spacing (Å)
Sample	Aging Condition	t 0	tf (1 month)
Blade-coated	Saline	12.64 ± 0.03	14.21 ± 0.81
	Air		12.45 ± 1.24
	Desiccator		11.21 ± 0.08
Thick, spray-coated	Saline	12.61 ± 0.07	14.44 ± 0.44
	Air		14.40 ± 0.10
	Desiccator		12.38 ± 0.85
Thin, spray-coated	Saline	14.80 ± 0.25	13.82 ± 0.24
	Air		15.27 ± 0.59
	Desiccator		14.96 ± 0.59

2.3. Effect of aging on the electrochemical behavior of Ti₃C₂T_x electrodes

In addition to the electronic, chemical, and structural changes observed in the Ti₃C₂T_x films, we also investigated the effects of aging on the frequency response of the Ti₃C₂T_x electrode impedance (Figure 5 and Supplementary Figure S6). In bioelectronic applications, maintaining low interface impedance is crucial to achieving efficient ionic to electronic current transduction, with minimal loss and distortion for biosensing, as well as high charge storage and injection capacity for stimulation [45]. Certain bioelectrical signals in areas such as the brain, heart, and muscles can exhibit characteristic oscillations within the 10 Hz frequency range [46]. Looking at the signal behavior collected from electrodes at a frequency close to or at this range, allows for the specific signal activity correlated to physiological events to be captures and analyzed. At 10 Hz, the initial impedance modulus for electrodes fabricated from thicker films was comparable between the different coating methods (spray-coated thick: 0.32 ± 0.03 kΩ, blade-coated thick: 0.32 ± 0.03 kΩ) and much lower than thin-film electrodes (spray-coated thin: 0.73 ± 0.06 kΩ). The 10 Hz impedance did not vary (< 110 ± 80 Ω change) for any of the electrodes stored in dry conditions for 30 days (Figure 6 (a–c)). In contrast, after 7 days in saline, the 10 Hz modulus of spray-coated thick and thin-film electrodes increased by ~8x and ~6x, respectively (spray-coated thick: 2.40 ± 1.07 kΩ, spray-coated thin: 4.07 ± 3.21 kΩ, Figure 6(d)). Electrodes fabricated from blade-coated thick films remarkably showed only a < 2x increase in impedance modulus at 10 Hz when stored in saline (0.55 ± 0.001 kΩ at t_f.). The phase behavior of the electrodes also changed over the 7-day period (Figure 5). After 7 days in saline, the impedance of the electrodes was no longer measurable due to significant fragmentation and/or detachment of all Ti₃C₂T_x films from the substrates, and the saline tests were interrupted.

Figure 5.

Electrochemical Impedance Spectroscopy (EIS) results for all electrode types and aging conditions. (a,c,e) Impedance modulus and (b,d,f) phase for electrodes aged over 30 days month in the desiccator and ambient conditions. (g,i,k) Impedance modulus and (h,j,l) phase for electrodes aged over 7 days in 37 °C PBS. (a,b,g,h) Thin spray-coated films; (c,d,i,j) thick spray-coated films; (e,f,k,l) thick blade-coated films. Points represent means, shaded regions are std. dev. across n = 3 samples for all film types and timepoints.

Figure 6. 10Hz Impedance of Ti₃C₂T_x electrodes.

(a-c). Impedance at 10 Hz after 28 days in low-humidity and air storage for (a) thin spray-coated, (b) thick spray-coated, (c) blade-coated films. (d) Impedance of all film types after 7 days in PBS (saline).

To characterize the changes in the electrochemical response of the electrodes, we fit the total electrochemical impedance at t₀ and at t_f with previously established equivalent circuit models for Ti₃C₂T_x electrodes [4,33]. From the fitting, we extracted the equivalent circuit parameters representing charge-transfer resistance (R_ct), double-layer capacitance (C_dl), and the Warburg impedance (W, Supplementary Figure S5 and Table 3), and compared their values before and after aging. Both R_ct and C_dl decreased for all electrode types and in all aging environments. These findings are indicative of a decrease in the Ti₃C₂T_x double layer capacitance and increase in the direct electron transfer across the interface, due to widening of the interlayer spacing. The Warburg impedance was highest for all electrodes after 7 days in saline (Supplementary Figure S5 C, F, I and Table 3). In comparison, the Warburg element for electrodes aged in dry environments was smaller and comparable across all electrode types (Supplementary Table 3). We also observed a thickness dependence in the Warburg impedance magnitude, as the thin spray-coated films had higher diffusion occurring than both types of thicker films. Since the Warburg element models mass-transport and diffusion reactions [33], these changes could be indicative of moisture ingress in the films.

Table 3.

Summary of the EIS fitting results for each film type at the initial and final timepoints, R_ct and C_dl normalized by the electrode geometric surface area (GSA). Values are provided as avg. ± std. dev. across n = 3 electrodes for each condition. BC — Blade-coated; Thick — Thick spray-cast; Thin — Thin spray-cast. Des — Desiccator (25 C); Air — Atmosphere (25 C); PBS — Saline (37 C)

	Rct (kΩ cm2)			Cdi (mF cm−2)			Warburg coefficient, W (kΩ s−1/2)
	BC	Thick	Thin	BC	Thick	Thin	BC	Thick	Thin
to	1.22 ± 0.21	286.89 ± 122.51	413.95 ± 206.44	(3.82 ± 0.97) 103	234.41 ± 63.28	152.66 ± 46.97	—	—	—
Des	0.68 ± 0.39	4.71 ± 0.99	18.76 ± 25.61	(1.96 ± 0.44) 103	54.15 ± 35.84	24.83 ± 3.70	2.99 ± 0.17	2.68 ± 0.76	1.69 ± 0.38
Air	0.39 ± 0.07	40.58 ± 7.46	0.44 ± 0.40	855.79 ± 797.71	15.55 ± 1.69	48.37 ± 11.62	3.53 ± 0.28	1.89 ± 1.00	4.59 ± 0.29
PBS	0.36 ± 0.21	59.34 ± 25.93	1.31 ± 0.89	536.81 ± 181.15	2.60 ± 1.50	0.38 ± 0.36	4.24 ± 0.28	3.22 ± 0.25	28.67 ± 1.78

Thus, for the Ti₃C₂T_x electrodes, the electrolytic environment accelerated film deterioration in all film types, as the devices loss integrity and delaminated after only 7 days in saline, becoming nonfunctional. To ensure adhesion of Ti₃C₂T_x films to biomedical-grade polymeric substrates used in medical devices, additional strategies such as adhesion promoters and additives will have to be considered in the future. However, for non-invasive dry epidermal biosensing Ti₃C₂T_x films without additional skin adhesion promoters seem sufficient to detect electrophysiological activity generated by muscle contractions (Figure 7).

Figure 7.

Electromyography (EMG) recording of forearm (flexor digitorum) muscle contractions for each electrode type. MXene electrodes (1) on the forearm with reference and ground disposable electrodes (2 and 3) on the elbow.

3. Conclusions

This study explored aging of Ti₃C₂T_x MXene films in environments relevant for applications in bioelectronics, and demonstrated that MXene structural, electronic, electrochemical, and chemical properties are not affected by aging in low-humidity and ambient conditions, regardless of processing methods and film thickness. As those environments mimic conditions at which bioelectronic devices might be stored before use, they represent a shelf-life benchmark. While Ti₃C₂T_x films are stable in ambient conditions, we observed significant degradation in the films after aging in a physiologically relevant electrolyte (i.e., saline), leading to the increase in sheet resistance and impedance modulus over time, as well as visible changes in the film morphology and integrity. However, blade coated films were notably more robust in saline environment. Blade coating is known to produce films with highly aligned flakes, allowing the resulting films to maintain their integrity and electronic properties even after submersion in saline for 30 days. Spray coated films had lower stability in saline and higher susceptibility to damage and degradation over time, regardless of the film thickness.

Using a combination of Raman spectroscopy, XPS, and XRD, we detected only minor surface oxidation after a month of aging in saline. The observed increase in interlayer spacing from XRD suggests that the driving force of Ti₃C₂T_x degradation in saline appears to be moisture ingress between individual Ti₃C₂T_x layers. The increase in the Warburg impedance element found from equivalent circuit modeling of the electrolytic impedance also supports the hypothesis that diffusion of water and other ions through Ti₃C₂T_x films is a major cause of the overall observed impedance increase and film deterioration. Blade-coated films with smaller inter-flake spacing and better alignment withstand various environmental storage conditions without significantly degrading. This suggests that, for future development of implantable bioelectronic and other medical devices based on Ti₃C₂T_x films, blade coating and other processing methods that yield thicker films with densely packed and highly aligned flakes might ensure long-term stability in aqueous environments.

Data Accessibility

All data and codes will be made available upon reasonable request.