Analysis of Platinum Distribution within a Nafion 212 Membrane during Electroless Deposition

Abstract

Ionic polymer metal composite (IPMC) is an electrically-responsive material made of perfluorinated sulfonic acid (PFSA) ionomers electroless plated with noble metal, such as platinum. Recently, interest in thin IPMC has been rising because of its potential applications in micro-electromechanical systems. However, little effort has been made so far on studying metal electroless deposition within thin PFSA membranes (thinner than 100 um). In this paper we study the platinum distribution within a thin Nafion membrane (Nafion 212, 50 µm-thick) using standard manufacturing recipe. We compare the obtained distribution with the distribution reported in literature based on thick Nafion membrane (Nafion 117, 180 µm-thick). We show that in Nafion 212 the maximum Pt concentration is situated in the bulk of the membrane, while in Nafion 117 the maximum peak is situated at the liquid-membrane boundary, suggesting a different rate determining deposition mechanism. Electroless deposition of platinum in Nafion 212 Ionic polymer metal composite (IPMC) is rising interest in the microelectromechanical system (MEMS) since smaller and thinner IPMC substrates would expand the applications. IPMCs are mostly made of a perfluorinated sulfonic acid ionomers (PFSA) membrane, such as commercially-available Nafion, coated with platinum. The prevalent way of manufacturing IPMC relies on the chemical reduction of platinum on the surface of the PFSA membrane, i.e. electroless deposition (1). Most of the reports on IPMC manufacturing rely on Nafion 117 (thickness of 0.180 mm, equivalent weight 1100 g/mol) while Nafion 212 has been poorly studied (thickness 0.050 mm, equivalent weight 1100 g/mol). This contribution aims at filling this gap. Materials and methods A PFSA membrane (Nafion 212 membrane) has been coated with platinum using the standard recipe already describe elsewhere (1). An immersion-reduction step was performed in an immersion bath composed of tetraammineplatinum chloride (Pt(NH3)4Cl2, 0.2%), and reduction baths made of sodium borohydride (NaBH4, 5%). After the first immersion (30 min and 2 hours, respectively), the samples were immerged in the reduction bath until the end of the precipitation. Samples were then frozen in liquid nitrogen and cut, and analyzed with electrons dispersive x-ray spectroscopy (EDS) and processing of scanning electron microscopy (SEM) images (Fig 1a). Figure 1. a) EDS detection of platinum (yellow) overlapped with SEM picture of the cross-section of IPMC. Platinum distribution and Rayleigh fit for single reduction step (NaBH4) of platinum in a Nafion 212 membrane, after 30 minutes (b) and 2 hours (c) of immersion. Results and discussion The distribution of platinum within Nafion 117 has already been studied by Millet et al. (2). The Pt distributions thereby reported strongly differs from the profiles we obtained (Fig 2). In the aforementioned profiles, the maximum Pt concentration is located at the liquid-substrate boundary of the Nafion membrane (2), whereas the Pt distribution in Nafion 212 reaches the maximum inside the Nafion membrane, between 0.5 µm and 1 µm below the surface (Fig 1b 1c). As reported by Millet et al., two limiting types of kinetics may be encountered in a dynamic exchange process, such that the deposition rate-determining mechanism is either diffusion in the liquid-film boundary (F-mechanism), or diffusion in the membrane (M-mechanism). The position of the peaks in our experimental data suggests that the mechanism determining the Pt precipitation rate inside Nafion 212 is diffusion inside the membrane while using Nafion 117 with the same reducing agent concentration leads to a precipitation rate determined by diffusion at the liquid/film boundary. It has been proven experimentally that ionic conductivity and water uptake of Nafion 212 is higher than for Nafion 117 (3). Thus we can hypothesize that the different behavior of thinner Nafion could explain the different empirical Pt distribution obtained using same reducing agent concentration. We could obtain a good fit of the Pt distribution (R2=0.919 and R2=0.945) within Nafion 212 with the Rayleigh distribution for immersion time of 30 minutes and 2 hours respectively (Eq 1.): f(x,σ 2) = (x/σ 2) exp(-x/2σ 2) [1] The scale parameter σ was found to be 4.19 and 4.58 for the sample immerged for 30 minutes and 2 hours, respectively. Further investigation of the precipitation mechanism inside Nafion 212 membrane and estimations of the phenomenological equations of mass transport correlated to Rayleigh distribution is reserved for future work. References K. Kim, M. Shahinpoor, Smart Mater. Struct., 12, 1, p. 65 (2003). P. Millet, R. Durand, E. Dartyge, G. Tourillon, and A. Fontaine J. Electrochem. Soc., 140, 5, pp. 1373-1380 (1993). A. Kusoglu, A.Z. Weber Chem. Rev. 117, 3 pp. 987-1104 (2017). Figure 1