Investigation of Electrochemical Characteristics and Interfacial Contact Resistance of TiN-Coated Titanium as Bipolar Plate in Polymer Electrolyte Membrane Fuel Cell

Abstract

In this research, titanium nitride (TiN) was applied to grade 1 titanium as a bipolar plate for a proton exchange membrane fuel cell (PEMFC). The TiN was deposited by the arc ion plating method (AIP) to investigate the electrochemical characteristics of the anode and cathode environments in the PEMFC. The corrosion experiments were conducted in an aqueous solution of pH 3 (H₂SO₄ + 0.1 ppm HF, 80 °C) determined by the Department of Energy (DoE). The hydrogen gas and air were bubbled to simulate the anode and cathode environments. The potentiodynamic polarization experiment showed that there was no active peak. The potentiostatic experiment showed that the current densities of the TiN-coated specimens were less than 1 μA/cm² in both the anode and cathode. As a result of observing the surface with an SEM before and after the potentiostatic experiment, only pinholes generated during the coating process were observed, and no corrosion damage was observed. Furthermore, electrochemical impedance spectroscopy (EIS) analysis showed that the coated specimens had a higher charge transfer resistance than the titanium substrate. In the case of interfacial contact resistance (ICR), the TiN-coated specimen displayed lower resistance than the titanium substrate and satisfied the DoE technical target of less than 10 mΩ·cm² at 140 N/cm².

1. Introduction

As regulations on environmental pollution are strengthened due to global warming, interest in eco-friendly and highly efficient power generation systems is increasing. The proton exchange membrane fuel cell (PEMFC) uses hydrogen as a fuel and is highly efficient because it converts chemical energy directly into electrical energy through an electrochemical reaction. In addition, PEMFCs are applicable to various industrial fields, such as power generation facilities and transportation, because they do not generate pollutants, have the advantage of a relatively low operating temperature, and can start quickly [1,2,3,4]. The PEMFC includes bipolar plates and membrane electrode assembly (MEA) composed of a proton exchange membrane, a catalyst layer, and a gas diffusion layer. In particular, the bipolar plate is one of the essential components of a PEMFC, and it uniformly supplies hydrogen, oxygen, and cooling water to the MEA. It also collects and transmits the electric current generated by the electrochemical reaction. In addition, stacking multiple cells supports non-rigid components and discharges water resulting from an electrochemical reaction [5,6]. The properties required for a bipolar plate include excellent electrical conductivity, corrosion resistance, low gas permeability, and rigidity. In the early days, graphite or graphite composite materials with excellent corrosion resistance and conductivity were used as bipolar plates for PEMFCs [7]. The metal bipolar plate has excellent mechanical strength, so volume and weight can be reduced due to thinning. However, compared to graphite, it has disadvantages in terms of corrosion resistance and poor electrical conductivity. When 140 N/cm² compressive stress was applied, the interfacial contact resistance of AA 5052 was 61.58 mΩ·cm², while that of the 904L stainless steel was 60.0 mΩ·cm², indicating a lower electrical conductivity compared with graphite, which was 6.39 mΩ·cm² [8,9]. Therefore, research on applying metals, such as stainless steel, aluminum alloy, and titanium, to bipolar plates has been actively conducted [8,9,10]. Using the ion plating method, Kang et al. improved corrosion and ICR with a TiN coating on SUS 316L. Hsu et al. researched TiN, TiCN, TiC, and other coatings of 5051 aluminum alloys using the PVD method for bipolar plates [11,12]. In particular, titanium can produce high-power stacks due to its low density of 4.50 g/cm³ and excellent electrical conductivity of 2.34 × 10⁶·S/m [13]. However, in the fuel cell operating environment, a non-conductive oxide TiO₂ is formed on the surface of titanium, increasing corrosion resistance but reducing electrical conductivity due to increased interfacial resistance [14]. Therefore, in this study, titanium specimens were coated with TiN using the PVD method (cathodic arc deposition) to improve electrical conductivity and corrosion resistance. In addition, the electrochemical characteristics and ICR were investigated in an aqueous solution that simulated a PEMFC.

2. Experimental Procedure

2.1. Sample Preparation

Grade 1 titanium was used as the substrate for the bipolar plate of the PEMFC; its chemical composition is listed in

Table 1. Chemical compositions of grade 1 titanium (wt%).

Fe	C	N	H	O	Ti
≤0.20	≤0.08	≤0.03	≤0.015	≤0.18	Bal

. The substrate was etched to remove impurities and oxides from the surface through an ion beam, and a Cr buffer layer was deposited to enhance the adhesion of TiN. TiN was then deposited using cathode arc deposition to form a coating layer. The detailed process is shown in

Table 2. Deposition conditions of TiN coating with PVD.

Process	Source	Ampere (A)	Voltage (V)	Gas Flow (sccm)	Time (min)	Bias (-V)	Temp (°C)
Etching	Ion beam	0.1~0.5	800~2000	50~80 (Ar)	100~150	50~200	250
Cr Buffer layer	Sputter	4~8	300~500	80~130 (Ar)	30~80	50~200
TiN	Arc	16~18	30~90	300~500 (N2)	70~100	50~200

.

2.2. Characterization of TiN Films

To analyze the coating thickness and corrosion damage, cross-sections and surfaces were observed with a scanning electron microscope (FE-SEM, JEOL, JSM-6700F, Tokyo, Japan). In addition, the elements of the coating layer were quantitatively analyzed using energy dispersive X-ray spectroscopy (EDS, OXFORD, AZTec Energy, Oxford, UK). The microscope operated under the following conditions: accelerating voltage 15 kV; probe current 12 nA; emission current 56.4 µA; decreased vacuum in the chamber with the pressure of 50 Pa. Phase analysis was performed using XRD with a Dmax-2500/PC (RIGAKU, The Woodlands, TX, USA). For XRD analysis, Cu Kα with λ = 1.546 Å was used under the conditions of a step size of 0.02°, a scan time of 1.0 s/step, and 2Ө of 30°~90°.

2.3. Electrochemical Experiment

In this research, a potentiostat (Gamry, US/PCI4/750, Warminster, PA, USA) was used to evaluate the electrochemical characteristics of TiN-coated titanium in an aqueous solution that simulates the operating environment of a PEMFC. The electrochemical cell was composed of a three-electrode corrosion cell consisting of a working electrode, a counter electrode, and a reference electrode. A holder was used for the working electrode with an exposed area of 1 cm². A silver/silver chloride electrode (Ag/AgCl/sat. KCl) was used as the reference electrode, and a platinum net of 2 cm × 2 cm was used as the counter electrode. The operating environment of the PEMFC was simulated using a pH 3 (H₂SO₄ + 0.1 ppm HF, 80 °C) solution recommended by the DoE [15]. Oxygen and hydrogen were continuously supplied to simulate the operating environments of the cathode and anode, respectively. After 60 min of stabilization, potentiodynamic polarization tests were performed in the potential range of −0.25 V to +1.6 V with respect to open circuit potential (OCP) with a scan rate of 1.0 mV/s. The corrosion potential (E_corr) and corrosion current density (I_corr) were calculated using the Tafel extrapolation method in the range of ±0.25 V based on the OCP of the potentiodynamic polarization curves. The potentiostatic experiment was conducted at potentials of +0.6 V_SSCE and −0.1 V_SSCE for 24 h. EIS was carried out at a frequency of 100 kHz to 10 MHz, with an amplitude of 10 mV. To measure the change in the corrosion resistance of the coating layer in the operating environment, the experiment was conducted after immersion for 24 h in the cathode (+0.6 V_SSCE) and anode (−0.1 V_SSCE) environments.

2.4. Measurement of Interfacial Contact Resistance (ICR)

Wang’s method, illustrated in Figure 1, was used to evaluate the ICR of titanium specimens coated with TiN [16]. A TiN coating was applied on both sides to accurately measure the ICR. After placing the specimen between the gold-coated copper plate and the conductive carbon paper (Toray TGPH-090) as a gas diffusion layer (GDL), the compressive stress was increased by 0.2 MPa, from 0.2 MPa to 1.8 MPa (R & B, VDS-1200). Next, the value of R₁ and R₂ were measured after a stabilization time of 10 s (Agilent, 34461 A, Santa Clara, CA, USA).

3. Results and Discussion

3.1. Characterization of TiN Films

A cross-section image of the TiN-coated specimen is shown in Figure 2. The thickness of the coating layer ranged from 2.5 to 3 μm. It was confirmed that the Ti, Cr buffer layer, and TiN coating layer were composed of dense microstructures. The Cr buffer layer was deposited using sputtering to improve the adhesion of TiN. EDS analysis showed that Ti = 81.94 wt% and O = 13.17 wt% for the titanium substrate, and Ti = 64.52 wt% and N = 29.54 wt% for the TiN coating layer. Figure 3 shows the results of the XRD diffraction patterns for the TiN coating layer. In the case of the Ti substrate in Figure 3a, diffraction peaks of α-titanium were observed on the (100), (002), (102), (110), (103), and (112) planes [17,18]. As a result of the TiN coating layer analysis shown in Figure 3b, the Ti, Cr, and TiN diffraction peaks were measured. TiN presented strong diffraction peaks on the (111), (200), (220), (311), and (222) planes, indicating that a stable phase was formed [19]. In the face-centered cubic (FCC) structure, the (200) and (111) planes have the lowest surface energy and strain energy, respectively. It can be interpreted that the total free energy decreases as the coating layer grows along the (111) plane when the strain energy is dominant [20].

3.2. Electrochemical Experiments

Figure 4 shows the results of the potentiodynamic polarization experiment performed by supplying hydrogen gas and oxygen in a solution of pH 3 (H₂SO₄ + 0.1 ppm HF, 80 °C) to the titanium substrate and TiN-coated specimen. The pH 3 solution is used because the Nafion membrane of the PEMFC is a perfluorosulfonic acid ionomer [21]. The Nafion membrane has a sulfonic acid group (−SO₃H) and a small amount of F^- ions caused by the deterioration of the Nafion membrane, creating a corrosive environment [22]. In the cathode environment supplied with oxygen, as shown in Figure 4a, the corrosion potential (E_corr) and the corrosion current density (I_corr) of the titanium substrate were 82.7 mV and 0.180 μA/cm², respectively. The TiN-coated specimens were 138.28 mV and 0.070 μA/cm², respectively. In the anode environment supplied with hydrogen gas, depicted in Figure 4b, the E_corr and I_corr of the titanium substrate were 124.9 mV and 0.111 μA/cm², respectively. The TiN-coated specimens were 125.1 mV and 0.035 μA/cm², respectively.

In both the anode and cathode environments, the TiN-coated specimen showed a lower corrosion current density value than the titanium substrate, did not cause an active peak due to the anodic dissolution, and met the DoE technical target presented in

Table 3. DoE technical targets for bipolar plate [20].

Characteristic	Units	2020 Targets
Corrosion, anode	μA/cm2	<1 and no active peak
Corrosion, cathode	μA/cm2	<1
Areal specific resistance	mΩ·cm2	<10

[23]. The porosity of the coating layer generated in the PVD process is listed in

Table 4. Electrochemical parameters deduced from potentiodynamic polarization curves in Figure 4 of titanium substrate and TiN-coated specimen.

-	Specimen	Icorr (μA/cm2)	Ecorr (mV)	βa (V/dec)	βc (V/dec)	Rp (× 103 Ω cm2)	Porosity (%)
Cathode	Ti	0.180	82.764	0.3251	0.1147	204.53	-
	TiN-coated	0.070	138.280	0.2039	0.1126	449.976	1.5
Anode	Ti	0.111	124.9	0.2974	0.1575	402.799	-
	TiN-coated	0.035	125.1	0.4867	0.0695	754.489	0.5

. E_corr, I_corr, and Tafel slope (β_a, β_c) values calculated using the Tafel extrapolation method from the potentiodynamic polarization curve were used for calculating the porosity. This result was calculated using Equations (1) and (2) and represents the polarization resistance (R_p) and the porosity (P), respectively [24].

$${\mathrm{R}}_{\mathrm{p}}=\frac{{\mathsf{\beta }}_{\mathrm{a}}\times {\mathsf{\beta }}_{\mathrm{c}}}{2.303\times {\mathrm{I}}_{\mathrm{corr}}\times \left({\mathsf{\beta }}_{\mathrm{a}}+{\mathsf{\beta }}_{\mathrm{c}}\right)}$$

(1)

$$\mathrm{P}=\left({\mathrm{R}}_{\mathrm{ps}}/{\mathrm{R}}_{\mathrm{p}}\right)\times {10}^{-\mid \Delta {\mathrm{E}}_{\mathrm{corr}}\mid /{\mathsf{\beta }}_{\mathrm{a}}}$$

(2)

In Equation (2), R_ps and R_p are the polarization resistance of the titanium substrate and the polarization resistance of the TiN-coated specimen, respectively. ΔE_corr represents the difference in corrosion potential between the titanium substrate and the TiN-coated specimen. β_a represents the positive Tafel slope of the titanium substrate. The TiN-coated specimen showed 1.5% and 0.5% porosity in the cathode and anode, respectively. When the electrolyte and the substrate react through the pores formed on the surface of the coating layer to form a corrosion product, the corrosion current density is increased and the polarization resistance is reduced. Therefore, when the coating is deposited using the PVD method, defects due to pinholes and inclusions may occur and cause corrosion; therefore, reducing porosity is essential [25].

Figure 5 presents the results of the potentiostatic corrosion experiment of the titanium specimens coated with TiN and the titanium substrate in a pH 3 solution (H₂SO₄ + 0.1 ppm HF, 80 °C). The conventional potentiostatic corrosion experiment for evaluating the durability of a PEMFC bipolar plate has different and conflicting results due to the different concentrations of sulfuric acid and hydrofluoric acid [26,27,28,29]. S. H. Lee et al. investigated the excellent electrochemical properties of CrN by studying 316L stainless steel coated with CrN and TiN in a solution of 0.1 N H₂SO₄ and 2 ppm F^- [26]. D. Zhang et al. coated TiN on 304 stainless steel and observed the corrosion of TiN using 0.5 M H₂SO₄ and 2 ppm F^- aqueous solution [27]. Based on R. Tian et al.’s research, TiN was coated on 316L stainless steel, and the experiment was conducted in 0.05 M H₂SO₄ and 2 ppm F^- aqueous solution. No damage by corrosion was found due to the excellent corrosion resistance of TiN [28]. Meanwhile, W. Meng et al. coated titanium with a C/TiC nanocomposite and conducted research using 0.5 M H₂SO₄ and 5 ppm F^- aqueous solution and obtained results that satisfied the DoE technical target [29]. In the PEMFC, the temperature, relative humidity, and operating temperature of the fuel (hydrogen and oxygen) affected the pH inside the cell. As a result of conducting experiments under various conditions, the pH range of the water discharged from the cell ranged from 3 to 7 [30]. Therefore, in this study, an experiment was conducted based on the solution (pH 3) and time (24 h) recommended by the DoE.

Figure 5a shows the i–t curves of the potentiostatic corrosion experiment of a titanium substrate and a TiN-coated specimen in a cathode environment (+0.6 V_SSCE) with an oxygen supply. After 24 h of the experiment, the current densities of the titanium substrate and the TiN-coated specimen were 0.19 μA/cm² and 0.57 μA/cm², respectively, satisfying the DoE technical target. Figure 5b shows the i–t curves of the potentiostatic corrosion experiment of a titanium substrate and a TiN-coated specimen in the anode environment (−0.1 V_SSCE) with hydrogen gas supply. After 24 h of the experiment, the current densities of the titanium substrate and the TiN-coated specimen were −2.87 μA/cm² and −0.28 μA/cm², respectively; only the TiN-coated specimen satisfied the DoE technical target. However, the current density of the titanium substrate increased gradually, so it did not satisfy the DoE target value (less than 1 μA/cm²). This was likely caused by the reductive dissolution of TiO₂, a passivation film, in the acidic solution, which led to the formation of Ti³⁺ ions [31]. Moreover, F^- reacts with TiO₂ and causes deterioration of the oxide film [32]. The deterioration of the oxide film due to the acidic environment and fluorine ions is shown in Equations (3) and (4).

$${\mathrm{TiO}}_2+4{\mathrm{H}}^{+}+\mathrm{e}\to {\mathrm{Ti}}^{3+}+{\mathrm{H}}_2\mathrm{O}$$

(3)

$${\mathrm{TiO}}_2+4{\mathrm{H}}^{+}+6{\mathrm{F}}^{-}\to \mathrm{TiF}{6}^{2-}+{\mathrm{H}}_2\mathrm{O}$$

(4)

In Figure 5b, the electrochemical behavior of the titanium substrate is thought to be affected by the acidic environment and fluoride ions. However, in the case of the TiN-coated specimen, both the anode and cathode environments displayed a current density of 1 μA/cm² or less, confirming that it can be applied to a PEMFC bipolar plate.

Figure 6 shows a scanning electron microscope observation of the surface of the TiN-coated specimen before and after the potentiostatic corrosion experiment. Pinholes with a diameter of 0.5 to 1 μm were observed in all specimens during the PVD coating process, but the size and depth were very small, and no corrosion damage occurred after the experiment.

Figure 7 shows the Nyquist plot, bode plot (impedance and phase angle vs. frequency), and the equivalent circuit used for data fitting, which was obtained after the TiN-coated specimens and the titanium substrate were immersed in a pH 3 solution (H₂SO₄ + 0.1 ppm HF, 80 °C) for 24 h. The titanium substrate was immersed for 24 h, and the TiN-coated specimen was immersed for 24 h in a cathode environment of +0.6 V_SSCE supplied with oxygen and an anode environment of −0.1 V_SSCE supplied with hydrogen gas. In the equivalent circuit illustrated in Figure 7b, R_s is the resistance between the aqueous solution and the working electrode, R_p is the resistance due to the pores of the coating layer, and R_ct is the charge transfer resistance between the substrate and the coating interface. Furthermore, C_f is the capacitance of the coating layer, and C_dl is the electric double-layer capacitance at the interface between the substrate and the coating. Considering the surface unevenness and defects of the oxide film and coating layer during the fitting process, the capacitance was substituted for the constant-phase element (CPE) and expressed as Equation (5) [33,34].

$${\mathrm{Z}}_{\mathrm{CPE}}={[\mathrm{C}{(\mathrm{j}\mathsf{\omega })}^{\mathrm{n}}]}^{-1}$$

(5)

Here, ω represents the angular frequency, j is √−1, and C is the CPE constant. In addition, n is a value representing the deviation from purely capacitive behavior. The value of n can vary from 1 for a perfect capacitor to −1 for a perfect inductor. The fitted values are listed in

Table 5. Fitting parameters of titanium substrate and TiN-coated specimen obtained from Nyquist plot.

-	Rs (× 103 Ω cm2)	CPEf (× 10−3 Ω−1 cm−2 sn)	nf	Rp (× 106 Ω cm2)	Rct (× 103 Ω cm2)	CPEdl (× 10−3 Ω−1 cm−2 sn)	ndl
Ti	1.188	0.173	0.86	0.480	41.022	0.863	0.71
Ti + TiN (Cathode)	1.652	0.049	0.92	2.893	443.09	437.3	0.94
Ti + TiN (Anode)	1.491	0.098	0.88	2.986	122.79	301.7	0.80

. The charge transfer resistance of the TiN-coated specimen in the cathode and anode environments was 443.09 kΩ·cm² and 122.79 kΩ·cm², respectively, indicating higher values than the 41.022 kΩ·cm² of the titanium substrate. This means that a TiN coating can improve the corrosion resistance of titanium in a PEMFC environment. In addition, the bode plots in Figure 7c,d presented the same trend as the fitted data. In the low-frequency region, the impedance coefficient of the TiN-coated specimen was greater than that of the titanium substrate. The peak of the phase angle was larger and wider, confirming that corrosion resistance improved [35,36].

3.3. ICR Measurement

The interfacial contact resistance of the TiN-coated specimen and the titanium substrate were measured by calculating R₁ and R₂, as shown in Figure 1. Using R₁, which is the total resistance in Figure 1a, and R₂, which is the contact resistance between the carbon paper and the gold-coated copper plate in Figure 1b, the interfacial contact resistance of the bipolar plate was calculated according to Equation (6).

$${\mathrm{R}}_1-{\mathrm{R}}_2={\mathrm{R}}_{\mathrm{ss}}+2{\mathrm{R}}_{\mathrm{cp}}+2{\mathrm{R}}_{\mathrm{cp}/\mathrm{coating}}\approx 2{\mathrm{R}}_{\mathrm{cp}/\mathrm{coating}}$$

(6)

The resistance of the specimen (R_ss) and the resistance of the carbon paper (R_cp) are neglected.

Figure 8 compares the ICR values of the titanium substrate and the TiN-coated specimen. The TiN-coated specimen exhibited superior electrical conductivity compared with the titanium substrate in all ranges. The details are presented in

Table 6. Interfacial contact resistance measured at various compressive stress.

Specimen		ICR(mΩ·cm2)
		20 N/cm2	40 N/cm2	60 N/cm2	80 N/cm2	100 N/cm2	120 N/cm2	140 N/cm2	160 N/cm2	180 N/cm2
Ti substrate		204.0	146.2	100.0	76.8	60.2	52.0	42.2	36.8	33.0
TiN- coated titanium	Bare	68.2	23.6	15.6	12.8	9.8	8.0	6.8	6.2	5.6
	Anode	67.8	21.4	15.8	12.6	10.2	8.6	7.2	6.4	6.0
	Cathode	68.8	27.4	18.4	13.4	10.6	9.8	8.0	6.8	6.6

. The high ICR of the titanium substrate was a result of the TiO₂ passivation film formed on the surface [37]. In particular, when a compressive force of 140 N/cm² was applied, the ICR values of the TiN-coated specimen and titanium substrate were 6.8 mΩ·cm² and 42.2 mΩ·cm², respectively. After the potentiostatic experiment, the ICR values of the TiN-coated specimen were 8.0 mΩ·cm² and 7.2 mΩ·cm² at the cathode and anode, respectively. This was interpreted as the formation of a passivation film on the TiN coating layer during the potentiostatic corrosion experiment [38]. The ICR increased after the potentiostatic experiment, but all TiN-coated specimens satisfied the technical target of the DoE (10 mΩ·cm² or less at a compressive force of 140 N/cm²).

4. Conclusions

The results of the research on the electrochemical characteristics and ICR of TiN-coated titanium as a PEMFC bipolar plate are as follows:

The titanium substrate and the TiN coating layer were densely deposited due to the formation of the Cr buffer layer, and a coating layer with a thickness of 2.5 to 3 μm was formed. As a result of the XRD analysis, all Ti, Cr, and TiN peaks appeared.

• As the TiN coating layer of the FCC structure was deposited, the peak value of the (111) plane was the largest;
• In the potentiodynamic polarization experiment, an active peak in current density due to corrosion was not observed in the TiN-coated specimen in either the anode or cathode environments. The corrosion current density was less than 1 μA/cm². In addition, the porosity of the TiN coating layer calculated from the data extracted from the potentiodynamic polarization curve ranged from 0.5% to 1.5%;
• The TiN-coated specimen exhibited a current density of 1 μA/cm² or less in both the anode and cathode environments. The surface analysis of the coating layer before and after the potentiostatic corrosion experiment demonstrated that although defects caused by pinholes were observed in the TiN-coated specimen, there was no corrosion damage;
• The EIS analysis showed that the TiN-coated specimen exhibited a higher charge transfer resistance than the titanium substrate, indicating excellent corrosion resistance in a PEMFC environment.

The TiN-coated specimen exhibited superior ICR compared to the titanium substrate. After the potentiostatic experiment in the anode and cathode environments, the ICR increased due to the formation of a passivation film. Still, it satisfied the DoE technology target of less than 10 mΩ·cm² at a compressive force of 140 N/cm².