Transient Surface Oxides Form on Pt(111) - But Vanish During Ammonia Oxidation

TL;DR

The study addresses whether surface oxides form on Pt(111) during ammonia oxidation and how such phases relate to activity and selectivity. Using operando SXRD, CTR analysis, and NAP-XPS across varying $p_{O_2}/p_{NH_3}$ and temperatures near 450–500 K, the authors show that Pt(111) does not sustain a stable surface oxide under ammonia-rich conditions; transient hexagonal monolayers and a Pt(111)-(8×8) oxide-related phase emerge under oxygen-rich feeds above light-off but are removed when ammonia is introduced. The reaction follows a Langmuir–Hinshelwood mechanism, with NO formation favored by available atomic oxygen and N2 production dominating under ammonia-rich conditions, underscoring the critical role of surface oxygen availability. These results reveal a facet-dependent oxide stability on platinum and highlight oxide-free Pt(111) operation under ammonia oxidation, offering structure–chemistry insights to guide design of catalytic interfaces for improved stability and NO selectivity in industrial environments.

Abstract

Ammonia oxidation on platinum catalysts is pivotal for industrial nitric acid production and environmental abatement, yet the role of surface oxides in this process remains debated. Using operando surface X-ray diffraction (SXRD), crystal truncation rod (CTR) analysis, and near-ambient pressure X-ray photoelectron spectroscopy (NAP-XPS), we reveal that Pt(111) does not form stable surface oxides under ammonia oxidation conditions. Instead, transient hexagonal monolayers and a Pt(111)-(8x8) superstructure emerge under oxygen-rich atmospheres and above the catalyst light-off temperature, but vanish upon ammonia exposure. Real-time mass spectrometry and NAP-XPS demonstrate that the reaction proceeds via a Langmuir-Hinshelwood mechanism, where adsorbed NHx and O species availability dictate selectivity toward NO or N2. Reducing the oxygen pressure by an order of magnitude slows the kinetics of oxide growth, only detected after 24 hr, and facilitated by transient and precursor structures.

Figures (20)

• Figure 1: Reciprocal space in-plane maps collected under different atmospheres, computed using the hexagonal lattice of Pt(111). The measurement of the large reciprocal space in-plane map takes about 330, and 115 for the small map. Both measurements under oxygen share the same colormap.
• Figure 2: Super-structure rods (SSRs) measurements for three different [H, K] positions at $p_{Ar} = \qty{420}{\milli\bar{}}$ and $p_{O_2} = \qty{80}{\milli\bar{}}$ (a) $\delta t$ designates the elapsed time since the introduction of oxygen, until the end of the measurement. The empty circles correspond to the Pt(111)-($8\times8$) structure. X-ray reflectivity curves at $p_{Ar} = \qty{420}{\milli\bar{}}$ and $p_{O_2} = \qty{80}{\milli\bar{}}$ for different exposure times (b) Curves fitted using GenX are shown as lines. $\delta t$ designates the time elapsed since the introduction of oxygen. $\rho_{ox}$, $t_{ox}$, and $\sigma_{ox}$ are the oxide density, thickness, and root mean square roughness. $\sigma_{sub}$ is the substrate root mean square roughness.
• Figure 3: Reciprocal space in-plane maps collected under different atmospheres, computed using the hexagonal lattice of Pt(111).
• Figure 4: X-ray reflectivity curves under different atmospheres during ammonia oxidation. Curves fitted using GenX are shown as lines. $\sigma_{sub}$ is the substrate root mean square roughness.
• Figure 5: Large reciprocal space in-plane maps collected at $p_{Ar} = \qty{495}{\milli\bar{}}$ and $p_{O_2} = \qty{5}{\milli\bar{}}$ for different exposure times.