Effect of TiN Particle Size and Impurities on the Contact Resistance of TiN-Powder-Decorated Stainless-Steel Separator Electrodes for Fuel Cells

Masahiko Hatakeyama; Koichiro Nagae; Masaki Naganawa; Ichiro Yoshino; Satoshi Sunada

doi:10.2320/matertrans.MT-MA2022004

Abstract

In recent years, stainless steel has been used as a separator in polymer electrolyte fuel cells (PEFCs). An electrophoretic deposition (EPD) method has been developed as a surface treatment, in which titanium nitride (TiN) is coated on stainless steel to impart conductivity. Using the EPD method, the surface of SUS316L steel was coated with three types of TiN–styrene-butadiene rubber (SBR) with different TiN particle diameters and impurity contents to improve the contact resistance. A smaller impurity content in the TiN resulted in a lower contact resistance. Furthermore, when the TiN particle size was between 50 nm and 1.5 µm, a larger particle size provided a lower contact resistance. In the polarization curve measurements, no deterioration of the corrosion resistance was observed for TiN–SBR. At the potential of an actual PEFC environment, low current densities were maintained in both the anodic and cathodic environments. Furthermore, scanning electron microscopy and electron probe microanalysis before and after the polarization test confirmed that there was no TiN detachment in the PEFC environment. The above results suggest that TiN-coated SUS316L steel is a promising separator for PEFCs.

Fig. 4 Contact resistance between the GDL and SUS316L stainless steels coated with TiN with different impurity contents.

1. Introduction

In recent years, alternatives to fossil fuels such as gasoline have attracted public attention with increasing interest in environmental issues, such as energy affairs and global warming. To solve these problems, fuel cells have received considerable attention as clean and environmentally friendly next-generation energy sources owing to their excellent energy conversion efficiency and low environmental impact.

Among fuel cells, polymer electrolyte fuel cells (PEFCs) are expected to be widely used because they have high power densities and small sizes and weights. Their broad use in fuel-cell electric vehicles or next-generation home generators is expected because of their low operating temperature and short start-up time.

PEFCs have polymer electrolyte membranes (PEMs), which conduct hydrogen ions and convert the chemical energy generated by the chemical reaction between hydrogen and oxygen into electrical energy. The PEM is held between two platinum/carbon catalyst layers (anode/cathode) and the gas diffusion layer (GDL) to form a membrane electrode assembly (MEA). The MEA is held between two separators (generally referred to as bipolar plates) to form a unit cell. The separator has various shapes in practical applications. Because the separator contributes significantly to the total weight, volume, and cost of a stacked form, the reduction of the wall thickness as well as the cost are major challenges.¹⁾

Owing to their chemical stability, thermal stability, and high conductivity, graphite and its composites have been developed as candidate materials for separators.²^,³⁾ However, graphite has various shortcomings, such as low mechanical strength and high production cost, when applied to separators intended for use in fuel-cell vehicles.⁴^,⁵⁾ Accordingly, low-cost metallic materials have been investigated for use as separators in fuel-cell vehicles. Recently, stainless steels have attracted significant attention. In particular, there are many studies on SUS316 stainless steel as a separator material.⁶^–⁹⁾

On the surface of stainless steel, there is a passivation film with a thickness of approximately 3 nm, which primarily contains ferric oxyhydroxide and chromium oxyhydroxide.¹⁰⁾ Electricity generation tests using unprocessed stainless steel as a base material showed an increase in the contact resistance between the stainless steel and GDL owing to the rapid growth of the passive film during electricity generation, which could reduce the electricity generation performance. In unprocessed stainless steel, dissolved metallic ions (iron, chromium, etc.) are captured by the MEA and degrade generation performance.¹¹^,¹²⁾ Numerous surface treatments to increase the conductivity and corrosion resistance of stainless steel surfaces have been attempted.¹³^–¹⁹⁾ However, these treatments require expensive procedures, such as noble metal plating or physical/chemical vapor deposition in a vacuum environment, and the cost is problematic in practical applications.

The electrophoretic deposition (EPD) method is promising to reduce the cost of the separator because it can modify the surface properties under ambient temperature and pressure.¹⁸^,¹⁹⁾ Kumagai et al. conducted electricity generation tests using the EPD method with cells employing titanium nitride–styrene-butadiene rubber (TiN–SBR)-coated SUS304 and SUS310S steels¹⁸^,¹⁹⁾ as separators. They reported that TiN particles, which are a hard material with high conductivity, penetrate the passive film and directly contact the base material. The SBR used as a binder maintained the stability of the TiN particles and supported them on the surface of the steel to impart conductivity and corrosion resistance to the surface. This is a promising surface modification method for stainless steel. However, the influence of impurities contained in TiN particles, such as oxygen, on cell performance was not mentioned in previous studies.

In this study, SUS316L stainless steel, which is a candidate for the separator base material, was selected, and its applicability as a PEFC separator was evaluated. The impurity content in TiN was varied, and a TiN–latex layer was clad onto the surface of the stainless steel. The contact resistance between the specimen and GDL, as well as the polarization curves, were measured.

2. Experimental Method

2.1 Specimens

Table 1 lists the chemical composition of the SUS316L stainless steel used in this study.

Table 1 Chemical composition of the SUS316L stainless steel (mass%).

The specimens were preprocessed by wet polishing using water-resistant emery papers up to #2000. After removal of the grime on the surface of the specimens, acid washing was applied to activate the surface. Three types of TiN powder with different impurity contents (prepared by Wako Pure Chemical Corporation) were selected as conductive particles, as shown in Table 2. The concentration was fixed at 0.1 mass% in a dispersion bath. A 40% latex-containing compound was diluted to 6.7% and adjusted to 1.5 mass% in a dispersion bath as a binder. The powders were then dispersed in the bath by ultrasonication for 10 min. The EPD treatment was conducted for 45 s under a constant voltage of 500 V using a direct current power supply (GPR-100H05D, INSTEK Japan Corporation). The SUS316L steels were dried for 10 min at 353 and 453 K using vacuum dryers (AVO-200NS-D and AVO-310NB, respectively, AS ONE Corporation).

Table 2 Chemical compositions of the TiN particles (mass%).

2.2 XRD and EPMA

The crystallographic characteristics of the TiN particles with different impurity contents were confirmed by X-ray diffraction (XRD) (D8 DISCOVER, BRUKER). Cu Kα X-rays were applied with a measurement range of 30° to 145° (2θ). The surfaces of the SUS316L steels coated with TiN–latex layers with different impurity contents were analyzed by scanning electron microscopy (SEM) and electron probe microanalysis (EPMA) (JXA-8230, JEOL Ltd.).

2.3 Contact resistance measurements

Based on the method reported by Davis et al., the contact resistance between the stainless steel and GDL was measured under arbitrary stress.²⁰⁾ Commercially available carbon paper (TGP-H-060, Toray Industries, Inc.) was used as the GDL. Arbitrary stress was applied using a load measuring instrument (LTS-1kNB-S20, Minebea Inc.), and the resistance was measured using a digital multimeter (Battery High Tester, Hioki E.E. Corporation). The resistance values are denoted R₁, R₂, and R₃; the gold-deposited copper electrode is denoted Au; and the measurement sample is denoted Sample.

\begin{equation} R_{1} = \text{Au/GDL/Sample/GDL/Au} \end{equation}

(1)

\begin{equation} R_{2} = \text{Au/GDL/GDL/GDL/Au} \end{equation}

(2)

\begin{equation} R_{3} = \text{Au/GDL/GDL/Au}. \end{equation}

(3)

After R₁, R₂, and R₃ were measured, the contact resistance between the sample and GDL (R_Sample) was calculated using the following equation:

\begin{align*} R_{\text{Sample}} & = (R_{1} + R_{2} - 2R_{3})/2\\ & = 0.5 (R_{1} + R_{2} - 2R_{3}). \end{align*}

2.4 Polarization measurements

Potentiodynamic and potentiostatic polarization curves were measured to confirm the corrosion resistance of the SUS316L steels coated with TiN–latex using the simulated PEFC environment proposed by Wang et al.²¹⁾ The three-electrode method was used in this study, where SUS316L steels with and without a TiN–latex coating were used as working electrodes, a Pt electrode was used as the counter electrode, and a saturated calomel electrode (SCE) was used as the reference electrode. The test solution comprised 0.05 M SO₄²⁻ + 2 ppm F⁻ (pH 3.3) heated to 353 K. The SUS316L steels coated with TiN–latex were immersed in an Ar-saturated test solution. Cathodic treatment was conducted by sweeping the potential from the immersion potential to −500 mV at 1 mV·s⁻¹, followed by a potentiodynamic polarization test up to +1500 mV.

Furthermore, a potentiostatic polarization test was conducted to confirm the stability of the TiN–latex layer. The SUS316L steels coated with TiN–latex were immersed in the test solution, and cathodic treatment was applied by sweeping the potential from the immersion potential to −500 mV. After the cathodic treatment, the potential was swept to −100 or +600 mV, which is referred to as the PEFC environment, where it was maintained for 8 h.

3. Results and Discussions

3.1 Surface observations of the TiN–latex-coated SUS316L stainless steels

The XRD results for TiN A, TiN B, and TiN C with different impurity contents are shown in Fig. 1. According to the International Center for Diffraction Data (ICDD), the crystallographic orientations of the specimens were identical. Regarding the peaks around 143°, those for TiN A and TiN B exhibited smooth shapes. The particle size of TiN C was small, and the disorder of the crystal lattice on the surface was easily affected, which appeared in the measurement. This phenomenon can occur when the particle size is on the nanoscale. However, this did not appear to have a significant effect on the experiment.

Fig. 1

XRD patterns of the TiN particles.

The SEM observations of the stainless steels coated with TiN–latex layers using the EPD method are shown in Fig. 2. Pinhole-like defects were not observed in the TiN–latex layers. Regarding the aggregation of TiN A–latex, particles 3–15 µm in size along with particles on the order of 100 nm were characteristic. In contrast, the particles in TiN B–latex were almost all on the order of 100 nm, and these particles clustered to form aggregates approximately 10 µm in size. The particles approximately 3–15 µm in size observed in TiN A–latex were not observed in TiN B–latex, and aggregates on the order of 100 nm were dominant. TiN C–latex consisted of finer particles, less than 100 nm in size, which were smaller than those in TiN B–latex. Aggregates approximately 10 µm in size were observed in some areas. Among the coatings, only TiN A–latex exhibited particles 3–15 µm in size. This was thought to contribute to a lower contact resistance in conjunction with a lower impurity content in TiN A, as discussed later.

Fig. 2

SEM images of the SUS316L stainless steels coated with (a) TiN A–latex, (b) TiN B–latex, and (c) TiN C–latex.

The results of the quantitative EPMA for the TiN–latex-coated SUS316L steels are shown in Fig. 3. Ti, O, N, and C were detected in all TiN–latex-coated specimens. Fe, Cr, and Ni, which are the primary components of SUS316L steel, were not detected, confirming that the TiN–latex layer thoroughly covered the surface of the stainless steel.

Fig. 3

Comparison of the qualitative analysis patterns of the SUS316L stainless steels coated with (a) TiN A-latex, (b) TiN B-latex, and (c) TiN C-latex. This data shows the elements detected when three samples are measured by EPMA. The X-axis is the position of the spectroscope when each element is detected.

3.2 Evaluation of the contact resistance between the TiN–latex-coated SUS316L steels and GDL

The contact resistances between the GDL and graphite, untreated SUS316L steel, and three different TiN–latex-coated SUS316L steels under arbitrary stress are shown in Fig. 4. The resistances were measured in a similar manner as in other reports.¹⁸^,¹⁹⁾ The contact resistance of each specimen decreased monotonically with increasing applied stress. According to Kumagai et al., this is because the GDL and latex deformed, and the contact area increased owing to the increase in stress. In addition, because TiN is a hard material and has high electrical conductivity, it is thought that the contact resistance was reduced because the TiN particles became embedded in the stainless steel surface and formed a circuit directly with the base material. Therefore, it was confirmed that the contact resistance of the TiN–latex-coated steels was lower than that of untreated SUS316L steel. In addition, TiN A–latex exhibited a low contact resistance similar to that of graphite. It is thought that the larger TiN particles penetrated the stainless steel more deeply and provided a larger contact area, resulting in a decrease in the contact resistance. Further, examples of impurities contained in TiN particles include titanium carbide (TiC) and titanium oxide (TiO). The electrical resistivity of TiN and TiC is 25 to 52 µΩ·cm, and the electrical resistivity of TiO is 281 µΩ·cm.²²^,²³⁾ It is considered that the lower the content of impurities such as TiC and TiO, the closer to the value of TiN, which has a low electrical resistivity.

Fig. 4

Contact resistance between the GDL and SUS316L stainless steels coated with TiN with different impurity contents.

The time dependence of the contact resistance was evaluated. Constant potentials of −100 and +600 mV were applied as the anodic and cathodic environments, respectively, in a 0.05 M H₂SO₄ solution for 8 h. Figure 4 shows the results obtained at a load of 1.5 MPa, which are summarized in Table 3. Considering the nature of graphite, its contact resistance is assumed to be a constant value of 2.06 mΩ·cm². The contact resistances before polarization were 60.03, 2.10, 5.16, and 3.12 mΩ·cm² for the untreated, TiN A–latex-coated, TiN B–latex-coated, and TiN C–latex-coated SUS316L steels, respectively. The contact resistances after 8 h of polarization increased remarkably for the untreated SUS316L steel in both the cathodic and anodic environments. In contrast, the contact resistances of the three TiN–latex-coated steels increased only slightly (less than 1.7 times higher than those before polarization). These values were below the target value of 10 mΩ·cm² under a stress of 1.38 MPa, as proposed by the United States Department of Energy (DOE). In particular, the TiN A–latex-coated SUS316L steel showed approximately the same resistance as that of graphite before polarization, and the increase in the contact resistance after polarization was remarkably small compared to that of the other specimens. Comparing the TiN A–latex-coated and TiN B–latex-coated SUS316L steels, the TiN particle size increased, and the contact resistance due to the GDL decreased as the impurity content decreased. Comparing the TiN B–latex-coated and TiN C–latex-coated steels, the TiN particle size decreased, and the contact resistance between the TiN–latex-coated steel and GDL decreased as the impurity content decreased. Between the TiN C–latex-coated and TiN A–latex-coated SUS316L steels, the TiN A–latex-coated SUS316L steel with a larger TiN particle size exhibited a lower contact resistance.

Table 3 Comparison of the contact resistance at 1.5 MPa between SUS316L stainless steel and the GDL before and after polarization for 8 h in a 0.05 M SO₄²⁻ (pH 3.3) + 2 ppm F⁻ solution at 353 K.

From these results, it is shown that the contact resistance between the TiN–latex-coated SUS316L steels and the GDL decreased as the TiN particle size increased and the impurity content decreased. This is because the electrical resistance of the titanium oxide and titanium carbide impurities is higher than that of TiN. Therefore, it is thought that with fewer impurities, the electrical resistance is closer to the value of TiN. In addition, from the results of TiN A and TiN C, the particle size is thought to have a greater effect on the contact resistance than that of the impurity content. This suggests the possibility of improving the power generation performance.

3.3 Corrosion resistance evaluated in simulated PEFC environments

The potentiodynamic polarization curves of untreated SUS316L steel and the three types of TiN–latex-coated SUS316L steel are shown in Fig. 5. The simulated PEFC environment was prepared using a solution containing 0.05 M SO₄²⁻ + 2 ppm F⁻ (pH 3.3). The current density at 0.1 V higher than the minimum value in each polarization curve is considered to be the current density of the active state. The current densities of the TiN–latex-coated SUS316L steels were lower than that of the untreated steel. In the passive range of the untreated steel (from −200 to +700 mV), the current densities of the TiN–latex-coated SUS316L steels were lower than that of the untreated steel. In the PEFC environment proposed by Wang et al.²⁴⁾ (from −100 to +600 mV), the TiN–latex-coated SUS316L steels exhibited lower current densities than that of the untreated steel. This is because the reaction area was limited by the TiN–latex coating. Therefore, it was confirmed that coating with TiN–latex did not deteriorate the activation–passivation properties of SUS316L steel.

Fig. 5

Anodic polarization curves of as-polished and TiN–latex-coated SUS316L stainless steels in a 0.05 M SO₄²⁻ (pH 3.3) + 2 ppm F⁻ solution at 353 K.

Figure 6 shows the evolution of the current density during the 8 h potentiostatic polarization tests in a simulated PEFC environment. Figure 6(a) shows the evolution of the current density in the anodic environment (−100 mV). The rapid decay in the current density for the untreated and TiN–latex-coated SUS316L steels immediately after the start of the experiment is thought to be caused by passive film formation. Subsequently, a lower current density was maintained, and no increment in the current density was observed. This suggests a lower possibility of rusting for TiN–latex-coated SUS316L steel in the actual PEFC environment. A similar tendency can also be confirmed in Fig. 6(b) and in other studies.¹⁰^,¹⁸^,¹⁹⁾ Therefore, it is suggested that using SUS316L steel as a separator has very little effect on the MEA performance due to melting.

Fig. 6

Potentiostatic polarization test results for the as-polished and TiN–latex-coated SUS316L stainless steels in a 0.05 M SO₄²⁻ (pH 3.3) + 2 ppm F⁻ solution at 353 K: (a) −100 mV and (b) +600 mV.

3.4 Evaluation of the contact resistance between the steel and GDL before and after the polarization test

The contact resistances between the GDL and the untreated and three types of TiN–latex-coated SUS316L steel are shown in Fig. 7. Similar to other reports,¹⁰^,¹⁹⁾ the contact resistance increased after the 8 h potentiostatic polarization test for the untreated SUS316L steel. In particular, it increased by 10 times when the cathodic potential was +600 mV. This appears to have been caused by the growth of the passive film. In contrast, the current densities of the three types of TiN–latex-coated SUS316L steel only increased slightly after the 8 h potentiostatic polarization test. Specifically, the contact resistance of the TiN A–latex-coated SUS316L steel was comparable to that of graphite, even after the 8 h potentiostatic polarization test, which suggests that it can stably maintain a small contact resistance.

Fig. 7

Contact resistance between (a) untreated, (b) TiN A–latex-coated, (c) TiN B–latex-coated, and (d) TiN C–latex-coated SUS316L stainless steels and the GDL before and after polarization for 8 h in a 0.05 M SO₄²⁻ (pH 3.3) + 2 ppm F⁻ solution at 353 K.

Because the tightening torque applied to a cell during power generation is typically 1.5 MPa, the contact resistances between the GDL and the untreated and TiN–latex-coated SUS316L steels under a load of 1.5 MPa before and after the polarization test were obtained, as shown in Fig. 8. For comparison, the contact resistance between the GDL and a reference graphite specimen was 2.06 mΩ·cm². The contact resistances between the steel and GDL after the potentiostatic test in the simulated anodic PEFC environment were 252.50, 2.14, 5.14, and 5.28 mΩ·cm² for the untreated, TiN A–latex-coated, TiN B–latex-coated, and TiN C–latex-coated SUS316L steels, respectively, while they were 576.50, 2.34, 5.82, and 5.30 mΩ·cm², respectively, after the potentiostatic test in the simulated cathodic PEFC environment. The untreated steel showed an increased contact resistance in the simulated PEFC environment. In particular, an increase of 10 times was observed in the simulated cathodic PEFC environment. In contrast, the contact resistance between the GDL and steel did not increase for the TiN–latex-coated SUS316L steels, and the contact resistance was less than 10 mΩ·cm², which is the target proposed by the DOE. In particular, SUS316L steel coated with TiN A–latex containing minimal impurities showed a low contact resistance comparable to that of graphite.

Fig. 8

Contact resistance at 1.5 MPa between stainless steel and the GDL before and after polarization for 8 h in a 0.05 M SO₄²⁻ (pH 3.3) + 2 ppm F⁻ solution at 353 K.

3.5 SEM observations after the 8 h potentiostatic polarization test

Thus far, the three types of TiN–latex-coated SUS316L steel have shown lower contact resistances with the GDL than that of untreated SUS316L steel. Furthermore, the contact resistances of the TiN–latex-coated steels were nearly identical to that of graphite. However, the polarization curves of the three types of TiN–latex-coated SUS316L steel indicated the existence of a stable passive film. Stable conductivity was also measured after the potentiostatic polarization test. Therefore, SEM observations and EPMA analyses (which are shown in the next section) were performed for the three types of TiN–latex-coated SUS316L steel after the 8 h potentiostatic test to confirm their stability.

SEM observations of the three types of TiN–latex-coated SUS316L steel after the 8 h potentiostatic test in the anodic environment (−100 mV) are shown in Fig. 9. Compared to the SEM images of the specimens before the test (Fig. 2), the three types of TiN–latex-coated SUS316L steel showed similar morphologies, except for differences caused by different fields of view. The SEM observations after the potentiostatic test in the cathodic environment (+600 mV) are shown in Fig. 10. Compared to the SEM images of the specimens before the test (Fig. 2), the three types of TiN–latex-coated SUS316L steel also showed similar morphologies, except for some differences caused by different fields of view.

Fig. 9

SEM images of the (a) TiN A–latex-coated, (b) TiN B–latex-coated, and (c) TiN C–latex-coated SUS316L stainless steels after polarization for 8 h in a simulated anodic environment.

Fig. 10

SEM images of the (a) TiN A–latex-coated, (b) TiN B–latex-coated, and (c) TiN C–latex-coated SUS316L stainless steels after polarization for 8 h in a simulated cathodic environment.

3.6 EPMA analyses after the 8 h potentiostatic polarization test

The results of the EPMA analyses for the three types of TiN–latex-coated SUS316L steel after the 8 h potentiostatic test in the anodic environment (−100 mV) are shown in Fig. 11. After the polarization test in the anodic environment, Ti, O, N, and C, which are the main components of TiN–latex, were detected in all three samples. The main components of SUS316L steel (Fe, Cr, and Ni) were not detected, confirming that there was no exfoliation of TiN–latex during the polarization test.

Fig. 11

Comparison of the qualitative analysis patterns of the (a) TiN A–latex-coated, (b) TiN B–latex-coated, and (c) TiN C–latex-coated SUS316L stainless steels after polarization for 8 h in a simulated anodic environment.

The results of the EPMA analyses for the three types of TiN–latex-coated SUS316L steel after the 8 h potentiostatic test in the cathodic environment (+600 mV) are shown in Fig. 12. After the polarization test, Ti, O, N, and C were detected in all three samples. In addition, Fe, Cr, and Ni, which are the main components of SUS316L steel, were not detected, confirming that there was no exfoliation of TiN–latex, even in the cathodic environment.

Fig. 12

The above results confirm that the TiN–latex coating was stable on both the anode and cathode sides in the simulated PEFC environment. This suggests that SUS316L steel coated with TiN–latex would be an effective PEFC separator in an actual PEFC.

4. Conclusion

SUS316L steel was selected as a candidate base material for PEFC separators, and its applicability as an PEFC separator was evaluated by measuring the contact resistances between the GDL and stainless steels coated with TiN–latex layers with various impurity contents. The following conclusions were drawn.

(1) The contact resistances between the GDL and SUS316L steels coated with TiN–latex layers with various impurity contents were measured. The steel coated with TiN–latex with a low impurity content showed the lowest contact resistance, suggesting that the impurity content of the TiN particles affects the contact resistance with the GDL.
(2) The TiN–latex-coated SUS316L steels showed no increase in the current density during the potentiostatic test, confirming a high corrosion resistance in the simulated PEFC environment.
(3) The contact resistances between the TiN–latex-coated SUS316L steels and the GDL were measured before and after the polarization test in a simulated PEFC environment to confirm their stability. The contact resistances between the GDL and TiN–latex-coated SUS316L steels after the polarization tests showed no increase in current density compared to those obtained before the polarization test. The SUS316L steel coated with TiN–latex with a low impurity content exhibited a low contact resistance comparable to that of graphite, and this low contact resistance was maintained even after polarization.

Accordingly, it is suggested that SUS316L stainless steel coated with a TiN–latex layer with minimal impurities using the EPD method is a promising separator for PEFCs.

REFERENCES

Corresponding author

Register with J-STAGE for free!