Applying Dynamic Strain on Thin Oxide Films Immobilized on a Pseudoelastic Nickel-Titanium Alloy

Summary

Dynamic, tensile strain is applied on TiO₂ thin films to study the effects of strain on electrocatalysis, specifically proton reduction and water oxidation. TiO₂ films are prepared by thermal treatment of the pseudo-elastic NiTi alloy (Nitinol).

Abstract

Direct alteration of material structure/function through strain is a growing area of research that has allowed for novel properties of materials to emerge. Tuning material structure can be achieved by controlling an external force imposed on materials and inducing stress-strain responses (i.e., applying dynamic strain). Electroactive thin films are typically deposited on shape or volume tunable elastic substrates, where mechanical loading (i.e., compression or tension) can affect film structure and function through imposed strain. Here, we summarize methods for straining n-type doped titanium dioxide (TiO₂) films prepared by a thermal treatment of a pseudo-elastic nickel-titanium alloy (Nitinol). The main purpose of the described methods is to study how strain affects electrocatalytic activities of metal oxide, specifically hydrogen evolution and oxygen evolution reactions. The same system can be adapted to study the effect of strain more broadly. Strain engineering can be applied for optimization of a material function, as well as for design of adjustable, multifunctional (photo)electrocatalytic materials under external stress control.

Introduction

The ability to alter the surface reactivity of catalytic materials by introducing strain has been widely recognized^1,2,3. Effects of strain in crystalline materials can be introduced either by adjusting material architecture (static strain) or by applying a variable external force (dynamic strain). In crystalline materials, static strain can be introduced by doping⁴, de-alloying^5,6, annealing⁷, epitaxial growth on a mismatched crystal lattice² or size confinement^2,3. In polycrystalline materials, strain can occur within grain boundaries due to crystal twinning⁸. Determining the optimal degree of static strain with material architectures requires designing a new sample for each discrete level of strain, which can be time consuming and expensive. Furthermore, introducing static strain often introduces chemical or ligand effects^9,10, making it difficult to isolate the strain contribution. Applying a dynamic strain precisely controlled by an external force allows systematic tuning of a material’s structure/function relationship in order to explore a dynamic range over the strain space without introducing other effects.

To study the effects of dynamic strain on electrocatalysis, metals or metal oxides are deposited on elastic shape or volume tunable substrates, such as organic polymers^{11,12,13,14,15} or alloys^16,17. Applications of mechanical, thermal or electrical loading results in bending, compression, elongation or expansion of an elastic substrate, further inducing a stress-strain response on the deposited catalytic material. So far, catalyst engineering through dynamic strain has been exploited to tune electrocatalytic activities of various metallic and semiconducting materials. Examples include i) the hydrogen evolution reaction (HER) on MoS₂, Au, Pt, Ni, Cu, WC^11,12,13,14, ii) the oxygen evolution reaction (OER) on NiO_x¹⁶, nickel-iron alloys¹⁸ and iii) the oxygen reduction reaction (ORR) on Pt, Pd^12,15,19,20. In most of these reports, organic polymers, such as polymethyl methacrylate (PMMA), were used as elastic substrates. We previously demonstrated the application of elastic metallic substrates, such as stainless steel¹⁶ and a superelastic/shape-memory NiTi alloy (Nitinol^17,21) for strain studies. Nitinol has also been used as an elastic substrate for deposition of platinum films for ORR¹⁹ and deposition of battery cathode materials for energy storage^22,23. Due to its shape memory and pseudoelastic properties, NiTi alloys can be deformed by applying moderate heat¹⁹ or mechanical strain¹⁷, respectively. In contrast to organic elastic substrates, metallic substrates typically do not require deposition of adhesion promoters, are highly conductive and can easily be functionalized. Nitinol is used as a more elastic alternative to stainless steel (SS). While SS can be reversibly strained up to 0.2%, nitinol can be reversibly strained up to 7%. Nitinol owes its unique properties to a martensitic solid state crystal transformation that allows for large elastic deformations^24,25. Both materials are commercially available in different geometries (e.g., foils, wires, and springs). When shaped into elastic springs, metallic substrates can be used to study effects of dynamic strain on electrocatalysis without the need for expensive instrumentation¹⁶; however, defining the stress-strain response is more challenging than for other geometries.

In previous experimental studies with transition metal catalysts, changes in activities of catalytic surfaces under strain have been attributed to changes in the energetics of the d orbitals colloquially known as d-band theory²⁶. In contrast, the effects of strain on metal oxides is significantly more complex, as it can effect bandgap, carrier mobility, diffusion and distribution of defects and even direct/indirect transitions^{21,27,28,29,30,31}. Herein we provide detailed protocols for the preparation and characterization of n-type doped TiO₂ thin films, as well as protocols to study electrocatalytic activities of these films under tunable, tensile strain. The equivalent system can be applied to study electrocatalytic activities of different materials as a function of dynamic strain.

Protocol

1. Preparation of NiTi/TiO₂ electrodes

1. Chemical and mechanical polishing of NiTi substrates
   1. Cut the superelastic NiTi foil (0.05-mm thickness) into 1 cm x 5 cm strips.
   2. Polish sample using 320-, 600- and 1200-grit sandpaper, and then rinse with ultrapure water (18.2 MΩ).
   3. Polish sample with 1 μm diamond, 0.25 μm diamond, and 0.05 μm alumina polish.
   4. After polishing, sonicate for 5 min in sequential baths of ultrapure water (18.2 MΩ), isopropanol, ethanol, ultrapure water (18.2 MΩ), and then dry under nitrogen (used organic solvents were reagent grade).
      CAUTION: Organic solvents are flammable, can irritate skin and eyes, poisonous if ingested. Use with caution in well ventilated areas.
      NOTE: Foils should be treated gently. Repeated bending or twisting can result in nano-to-micro sized fissures, which will affect its elastic properties decreasing the effects of strain on the electrocatalytic activities.
2. Preparation of TiO₂ films
   1. Oxidize NiTi foils by placing foils in a 500 °C oven under aerobic conditions (Figure 1).
   2. For preparation of 50 nm thick rutile TiO₂ films, heat NiTi foils for 30 min at 500 °C. Longer heating will result in thicker TiO₂ films. Heating will cause a change in the surface color from gray to blue/purple (Figure 2).
3. Applying tensile stress on NiTi/TiO₂
   1. Gently clamp foil (1 cm x 5 cm strip) in a mechanical tester (Table of Materials) with 1 cm of foil exposed at each end.
   2. Strain the NiTi/TiO₂ samples at a rate of 2 mm/min. Keep the strain at desired level (0-3%).
      NOTE: Extension of the available 3 cm NiTi/TiO₂ lengthwise from 0.0 to 2.1 mm is considered straining from 0 to 7%, which can be calculated by simple equation strain=(l-l₀)/l₀ , where l₀ is initial and l final length of foil exposed to tensile strain. Typical stress-strain curve is shown in Figure 3.
4. To start electrochemical measurements, pre-strain the foil to 5 N (taken as 0% strain).
   NOTE: The slight pre-straining of the foil leads to more reproducible results.

2. Conducting electrochemical measurements under strain

1. Applying tensile stress on working electrode
   1. To conduct electrochemical experiments under applied strain, assemble the custom-made electrochemical cell (Figure 4 and Figure 5) loosely around the NiTi/TiO₂ foil. Ensure that the center of the NiTi/TiO₂ foil is exposed by carefully positioning the cell in the middle (Figure 5).
   2. Tighten the cell gently onto the sample to create a solution-tight cell for the electrochemical measurements.
   3. Fill up with an electrolyte and purge the solution gently with nitrogen.
   4. Increase strain to specific levels, typically 0 to 3% in 0.5% increments and conduct electrochemical experiments for each discrete strain value.
   5. Before each strain adjustment, loosen the electrochemical cell around NiTi/TiO₂ foil, so that the sample can move freely. Then realign the cell by gently tightening back onto the sample and refill the electrolyte for the next electrochemical measurements.
      NOTE: Tightening and untightening the cell around the NiTi/TiO₂ foil is obviously more laborious and time consuming than working with a continuously tightened cell through the experiments. Nevertheless, this approach minimizes possible wrinkling of NiTi/TiO2 foil leading to the most reproducible results and the highest effects of strain.
2. Electrochemical characterization of strained working electrode
   1. As an initial experiment, conduct cyclic voltammetry (CV) or linear sweep voltammetry (LSV) measurements (Figure 6A). Further characterization could include impedance, electrolysis, chronoamperometry, etc.
   2. Collect electrochemical measurements with samples exposed to discrete, increasing levels of strain (e.g., from 0 to 3% in 0.5% increments), followed by gradual decreasing of applied strain (e.g., from 3 to 0% in 0.5% increments).
   3. Collect data for multiple experimental cycles (0%→3%→0%) to test the system mechanical stability and data reproducibility.
   4. Alternatively, keep the foil strained at a discrete amount of strain for prolonged time periods (e.g., hours or days) and conduct electrochemical experiments periodically (e.g., voltammetry) or continuously (e.g. electrolysis).
3. HER experiments
   1. Use 0.5 M sulfuric acid as the electrolyte, Ag/AgCl (1 M NaCl) as the reference electrode, and a coiled platinum wire (0.5 mm diameter by ~10 cm length) as the counter electrode.
      CAUTION: Sulfuric acid causes severe skin burns and eye damage. Do not breathe mist, vapors, or spray. Wear protective gloves, protective clothing, eye protection, and face protection. Immediately wash exposed skin with copious amounts of water if exposed.
   2. Scan the potentials between the open-circuit voltage (OCV) to -0.8 V vs RHE, starting with the highest potential value with scan rate 5-50 mV/s (Figure 6A).
4. OER experiments
   1. Use 1 M sodium hydroxide as the electrolyte, Hg/HgO (1 M NaOH) as the reference electrode, and a coiled platinum wire (0.5 mm diameter by ~10 cm length) as the counter electrode.
      CAUTION: 1 M sodium hydroxide can cause skin burns and eye damage Do not breathe mist, vapors, or spray. Wear protective gloves, protective clothing, eye protection, and face protection. Immediately wash exposed skin with copious amounts of water if exposed.
   2. For OER experiments, scan the potential between OCV to 2 V vs RHE, starting with the lowest potential value, with scan rate 5-50 mV/s (Figure 6B).
5. Impedance
   1. Carry out electrochemical impedance spectroscopy (EIS) measurements at frequencies ranging from 1 Hz-100 kHz at a potential where no Faradaic process is observed (OCV) (Figure 6C).
6. Analyzing time profile, system stability and products
   1. To test the stability of the system and measure products (e.g., H₂ and O₂), conduct electrolysis experiments.
   2. For amperometric i-t measurements, choose the most suitable potential based on CV or LSV results (e.g., -0.25 V vs RHE for HER).
   3. Alternatively, for chronopotentiometry experiments, choose the most suitable current density based on CV results.
   4. If gas chromatograph is available, measure in-line hydrogen (from HER) or oxygen (from OER) gas produced electrochemically (Figure 4B).
      NOTE: These are examples of electrochemical analyses. Electrochemical characterization can be tailored for a specific study.

3. Controls

1. Capacitance measurements
   1. To determine if increases in HER activities are simply due to increases in electroactive surface, conduct capacitance measurements at different strain values.
   2. Run CV experiments at different scan rates (e.g., 1 and 500 mV/s) at a potential range where Faradic currents are negligible, so that currents represent only the charge/discharge of the electric double layer (e.g., 0 to 0.1 V vs RHE).
   3. Plot scan rates versus currents (Figure 7A).
   4. Compare increases in capacitance with strain with increases in electrocatalytic activities (e.g., HER or OER) with strain (Figure 7A).
      NOTE: If increases in electrocatalytic activities are higher than increases in capacitance, it can be concluded that simple increase in grain separation and electroactive surface is not the only contributor to the increase in electrocatalytic activities.
2. Characterization of cracked films
   1. Purposely crack NiTi/TiO₂ foil by keeping the foil strained at 7% for 30 min or longer for 50 nm TiO₂ films (Figure 8). Thicker TiO₂ films (100 nm) can be cracked at lower strains (3% strain).
   2. Analyze the surface for cracking by scanning electrochemical microscopy (SEM), or other surface analysis methods, as described below.
   3. Conduct electrochemical measurements as described above with pristine and purposely cracked TiO₂ films at different incrementally increased and then decreased strain values from 0%→3%→0% (Figure 6D). NiTi/TiO₂ foils with 50 nm thick TiO₂ films that were never strained pass 3% are considered pristine, elastic.
      NOTE: Determine the specific “elastic limit”: the maximum stress that can be applied on a material before the onset of an irreversible deformation (e.g., grain rearrangement or even film cracking). Elastic range depends on film type, thickness and deposition method. For example, we show that 100 nm thick TiO₂ films crack at lower strains than 50 nm thick TiO₂ films.
3. Characterization of NiTi foils (i.e., unoxidized foils)
   1. Polish NiTi folis as described in step 1.1, but do not thermally treat them.
   2. Run all the electrochemical experiments, as described above, with NiTi foils that were not thermally treated as a control.

4. Surface characterization

1. Sample preparation
   1. Cut and pretreat NiTi/TiO₂ as described in steps 1.1 and 1.2.
      NOTE: The size of the sample foil depends on the size of the sample holder, which depends on a specific instrumentation used for the surface characterization.
   2. Wash samples with water to remove any residual salt if used in electrochemical experiments before the characterization.
   3. Assemble NiTi/TiO₂ foil in the tensile stretcher and strain to a desired level as described in section 1.3.
   4. Assemble the custom-made sample holders around the strained sample and gently tighten the screws (Figure 9).
2. Surface characterization
   1. To check film quality and changes in film topology with strain, collect scanning electrochemical microscopy (SEM) images.
   2. Use other available surface analysis methods to monitor changes in surface chemical composition, grain rearrangements and exposed crystal lattices (e.g., Raman spectroscopy, XPS or XRD experiments) (Figure 10).
   3. To check if a sample holder kept constant strain during the surface characterization experiments untighten the sample from the sample holder and look for any curl in the sample between the strained portion under the clamp and the unrestrained portion that was previously in the tensile tester.

Results

Pre-treated NiTi foils are oxidized at 500 °C under aerobic conditions (Figure 1). Due to the oxophilic nature of titanium, calcination at elevated temperatures results in a surface layer of rutile TiO₂. The thickness of the layer and degree of n-type doping are affected by annealing time and temperature, which is reflected in color change from gray (untreated sample) to uniform blue/purple after 20 min heating (Figure 2). Longer heating time res...

Discussion

Nitinol is a suitable elastic substrate for applying mechanical stress on thin films. It is commercially available, highly conductive and can be easily functionalized. Preparation of rutile TiO₂ thin films by thermal treatment of nitinol, results in highly n-type doped TiO₂. It is important to emphasize that NiTi/TiO₂ is a unique system where TiO₂ films are prepared by thermal treatment of NiTi rather than a deposition method. Our previous publications have shown that strain ap...

Materials

Name	Company	Catalog Number	Comments
2-Propanol	Sigma Aldrich	109634
Ag/AgCl (3M NaCl) Reference Electrode	BASi	MF-2052
Alkaline Reference Electrode	Basi	EF-1369
Ethyl alcohol, Pure, 200 proof, anhydrous, =99.5%	Sigma Aldrich	459836
MT I I / F u l l am SEMTester Series	MTI Instruments
Nitinol foil, 0.05mm (0.002in) thick, superelastic, flat annealed, pickled surface	Alfa Aesar	45492
PK-4 Electrode Polishing Kit	BASi	MF-2060
Potentiostat 600D	CHI instruments	600D
Pt wire	Sigma Aldrich	267228-1G
Sodium hydroxide	Sigma Aldrich	221465
Sulfuric acid	Sigma Aldrich	30743