Exploring substrate/ionomer interaction under oxidizing and reducing environments

Local gas transport limitation attributed to the ionomer thin-film in the catalyst layer is a major deterrent to widespread commercialization of polymer-electrolyte fuel cells. So far functionality and limitations of these thin-films have been assumed identical in the anode and cathode. In this study, Nafion thin-films on platinum(Pt) support were exposed to H 2 and air as model schemes, mimicking anode and cathode catalyst layers. Findings indicate decreased swelling, increased densification of ionomer matrix, and increased humidity-induced aging rates in reducing environment, compared to oxidizing and inert environments. Observed phenomenon could be related to underlying Pt-gas interaction dictating Pt-ionomer behavior. Presented results could have significant implications about the disparate behavior of ionomer thin-film in anode and cathode catalyst layers.


Introduction
As polymer-electrolyte fuel cells (PEFCs) gain traction in the energy-device landscape, they face a major hurdle from significant mass-transport losses associated with the ionomer/catalyst interface [1], [2]. Sources of mass-transport losses include: confinement driven gas transport losses in ionomer thin-film coating carbon-supported platinum, interfacial resistances caused by structural changes at local ionomer-platinum boundary, and partial electrochemical deactivation of platinum surfaces [3]- [6]. The latter can impact overall kinetics on platinum(Pt) surfaces [7], [8], however such effects on ionomer mass-transport and the interplay with reducing atmospheres are unknown. As a result, explicit understanding of losses at the ionomer/Pt interface is required for optimal electrode-ionomer design and accelerating market penetration of PEFCs.
Ionomer thin-films cast onto a Pt surface can serve as model systems providing a focused glimpse into the catalyst layer. Although bulk, continuous polycrystalline Pt does not fully describe Pt nanoparticle phenomenon present in real catalyst layers, it can still elucidate surface specific interactions that impact ionomer properties and morphology [9], [10]. While impact of Pt substrate on ionomer performance have been shown [8], [11], efforts to clarify the source of this impact have been contradictory, especially in elucidating the role of water on oxidized and unoxidized Pt surfaces [12], [13]. Additionally, the extent of Pt surface influence on ionomer during exposure to oxidative/reductive environments remains unexplored. In this study, watervapor-sorption dynamics of dispersion-cast Nafion thin-films under reducing (H 2 ), oxidizing (Air), and inert (Ar, N 2 ) environments are investigated in order to understand the Pt/ionomer interaction in anode and cathode catalyst layers.

Thin-film Preparation
Nafion dispersions (5 wt%, 1100 g/mol SO 3equivalent-weight, Sigma Aldrich) were diluted in isopropanol, spin cast onto Pt-coated Si, and Si/SiO 2 wafers to form ~50 nm films. Pt substrates were prepared via e-beam evaporation of 5nm Ti adhesion layer followed by 60nm of Pt. Pt substrates were cleaned with benchtop Ar plasma for 6 minutes prior to casting. Thin-films were annealed at 150°C under vacuum for 1 hr before measurement.

Water-Uptake Measurement
Thickness change of Nafion films was monitored using in-situ spectroscopic ellipsometry (J.A. Woollam) as detailed in Ref [14]. Measurements shown are the average of at least two separate samples measured <15 minutes after annealing. To create a consistent water history, all measurements were preceded with an hour exposure to dry (0%) and saturated (96%) relative humidity (RH) (See Fig 1a for hydration protocol). Humidity-dependent thickness (L(t,RH)) was an average of the last 10 min of set humidity. The % change from dry ( L o ) is given by:

Grazing Incidence Small Angle Scattering (GISAXS) Measurements
Pt-coated Nafion films were placed into an in-house built environmental chamber with X-ray transparent Kapton windows as in Ref [6]. The sample was equilibrated in dry H 2 and N 2 gas at room temperature and GISAXS patterns were collected after multiple purges for 5 to 10 minutes in each gas, at varying incidence angles (α i ).

Mechanical-Property Measurement
100 nm Nafion films were prepared on Pt-coated thin Si cantilever wafers (105μm thickness by approximately 0.5cm x 4cm). Sample was clamped in an environmental cell with humidified gas feeds. Constrained swelling due to the substrate results in a compressive force, which bends the Si cantilever. Using a laser array reflected off the backside of sample, change in curvature of the cantilever was measured and related to stress-thickness via Stoney's equation, see Ref [15].
Humidity-induced stress-strain curves were generated by combining stress and strain (from ellipsometry, see Equation 1) under the same humidity conditions, and the deformation energy density was calculated by integrating the area under the curve.  The reversibility and persisting impact of the gaseous environment on ionomer swelling was explored using humidity-cycling by alternating inert and reducing gas exposure. In-situ ionomer thickness change on Pt was monitored over three hydration cycles: first, a single step of dry to 96% RH gas exposure (Cycle 0, gas 1); second, humidity was stepped down to 0% RH prior to hydration cycling (Cycle 1, gas 1); finally, gas 1 was switched and stepped RH was applied (Cycle 2, gas 2). Here, the dry reference thickness was set to the thickness from Cycle 0. conditions at a rate that is between that of H 2 -and Ar-only environments, confirming the impact of the H 2 environment. Thickness change in (a) Dry and (b) Saturated (96% RH) relative to dry and saturated thickness in Cycle 0 exposed to gas 1, respectively.

Results and Discussion
The findings in Fig. 1 and 2 are consequences of changes at the ionomer/Pt interface induced by gas/Pt interaction. Surface oxidation on Pt metal can occur via electrochemical and thermochemical pathways [17]. In a thermochemically oxidized Pt surface, exposure to an oxidative gaseous environment like air will enlarge oxidized metal islands on Pt, while exposure to a reducing environment like H 2 can reduce the unstable passivated surface even under ambient conditions [17]- [20]. Pt substrates in this study are likely to exist with some surface oxidation as they are stored under ambient conditions. This oxide surface continues to grow with continued exposure to an oxidizing environment or, is reduced and saturated with dissociated atomic hydrogen during H 2 exposure; a phenomenon that has been reported experimentally and computationally [10], [21]- [23]. As a result, during exposure to Air and H 2, Pt interface can exist at varied states of oxidation and reduction resulting in sample-to-sample variability. Nonetheless, adsorbed hydrogen reduces the solid-surface free energy [21], resulting in a more hydrophilic but nonpolar Pt/H interface compared with that of oxidized Pt. This phenomenon was verified by using bare Pt-coated crystal in a quartz-crystal microbalance, which exhibited significant adsorption of H 2 on Pt surface when dry, and greater absorption of water when saturated due to greater affinity for water at the Pt/H interface (data not shown). The Pt/H interface lacks strong electrostatic interactions, resulting in possible ionomer restructuring to orient hydrophilic sidechains towards the Pt/H interface, where water molecules are likely to gather, thereby creating a dense region of hydrophobic ionomer away from the interface. In such a scenario, the bulk of the ionomer behaves like a higher equivalent-weight ionomer with lower water uptake. On the other hand, negatively charged oxygen atoms on an oxidized Pt surface, which, while comparatively less hydrophilic, induce a strong polar dipole and enhance electrostatic interactions between hydronium ions and sulfonic-acid moieties. Similar depression in watervapor uptake in thin-films on Si/SiO 2 support under H 2 also point towards impact of oxidized surfaces. Under ambient conditions, growth of native oxide layer of 1 to 2 nm is expected on a Si substrate. Continued layer-by-layer growth of SiO 2 ; however, requires presence of both water and oxygen [24], [25]. Although reduction of the oxide layer is not occurring under H 2 environment on Si/SiO 2 support, oxide formation is actively being facilitated under humidified air. These interactions enhance the overall effective water uptake within the ionomer on oxidized surface, which is consistent with predictions from molecular-dynamics simulations [26]. Figure 3 schematically portrays the balancing impacts of polarity and hydrophilicity in reducing and oxidizing environment. The above hypothesis is supported by morphological changes tracked by GISAXS and mechanical response of Nafion thin-film on Pt exposed to H 2 and N 2 gases. When α i of x-ray beam is below the critical angle of the polymer film, α c ,film , total external reflection occurs with a surface-sensitive scattering [27], whereas above α c ,film , , the x-ray beam penetrates through the entire film and scattering from the paracrystalline Pt surface is observed. As shown in Figure 4a, the paracrystalline peak is present at α i = 0. 16  Despite being the least understood component, the gas/ionomer/Pt interface in the catalyst layer bears the utmost duty for PEFC performance. Thus, there is need for greater understanding of pairwise interaction between gas/ionomer, ionomer/Pt, and gas/Pt interfaces to reduce critical transport losses and improve electrode design. To that effect, this study focused on how gas/Pt interaction impacts Pt surface and ionomer thin-film morphology and properties.
Unexpectedly, a reduced swelling, increased densification, decreased deformation energy density and continual reduction in effective water uptake in the ionomer during cycling were observed under H 2 relative to oxidizing or inert environment. These observations demonstrate the coupled impact of gas/substrate and ionomer/substrate interactions on ionomer thin-film's behavior and ultimately it's transport properties [28]. Therefore, there is a need for increased electrode-specific investigations and separate ionomer design for anode and cathode catalyst layers. The impact of electronic potential going from oxidation to reduction potentials can also affect the surface-state identity and ionomer thin-film morphology, which is a focus of current research. Furthermore, existence of a water-rich phase at the Pt/ionomer interface in a reducing environment can impact surface conductivity significantly, which may not occur in an oxidizing environment. The findings herein also indicate heightened vulnerability to delamination of ultra-thin ionomer films in the anode due to increased water-layer thickness and reduced deformation energy density.