Delicate Distinction between OH Groups on Proton-Exchanged H‑Chabazite and H‑SAPO-34 Molecular Sieves

We observed surprising differences in the FTIR (Fourier transform infrared) hydroxyl spectra of the structurally isomorphous, proton-exchanged H-CHA and H-SAPO-34 molecular sieves when measured by transmission (TR) or diffuse reflectance (DRIFT) techniques. Experimental and density functional theory (DFT) based model evidence is presented in this paper to prove that the essential reason for this spectral difference is that DRIFT emphasizes the vibrations of surface hydroxyl sites. Vibrations of the bulk Brønsted acidic hydroxyls shift to higher frequencies when they become surface species, and the IR beam is reflected from approximately the top ∼15 to 20 Å thick layer of the particles; hence, the proportion of surface related IR bands becomes significant compared to the bulk related ones in the DRIFT spectra while the opposite is valid for the TR spectra. We demonstrate that the surface hydroxyls are Brønsted acidic on both the H-CHA and the H-SAPO-34 particles, and the upshifted vibrations noticed primarily in the DRIFT spectra are Al–OH vibrations on the surface even of H-SAPO-34, not P–OH groups as most researchers believe. We also show that the bulk Brønsted sites might involve HO1, HO2, and HO4 type hydroxyls associated with the known geometrically different oxygen positions on both molecular sieves, but only HO1 surface hydroxyls are associated with the red-shifted vibration intensified in the DRIFT spectra. Moreover, a single surface model cannot account for every vibration observed in DRIFT spectra. From the combination of IR vibrations of three adequate surface models one can as properly match the experimental DRIFT spectra as the TR spectra from the combination of the calculated bulk HO1···HO4 vibrations of these molecular sieve crystals.我们观察到，结构同构的质子交换型H-CHA与H-SAPO-34分子筛的傅里叶变换红外（Fourier transform infrared, FTIR）羟基光谱，在采用透射（transmission, TR）或漫反射（diffuse reflectance, DRIFT）技术测试时，呈现出令人意外的显著差异。本文提供了实验数据与基于密度泛函理论（density functional theory, DFT）的模型佐证，证明该光谱差异的核心成因在于：DRIFT技术更侧重采集表面羟基位点的振动信号。当体相布伦斯台德（Brønsted）酸性羟基转变为表面物种时，其振动峰会向高波数方向偏移；同时红外光束仅在颗粒表层约15至20埃的薄层内发生反射，因此在DRIFT光谱中，与表面相关的红外吸收带占比相较体相相关吸收带会显著提升，而透射（TR）光谱的情况则恰好相反。我们证实，H-CHA与H-SAPO-34颗粒表面的羟基均具备布伦斯台德酸性；且在DRIFT光谱中显著观测到的高波数偏移振动，实则为H-SAPO-34表面的铝羟基（Al–OH）振动，而非多数研究者所认定的磷羟基（P–OH）。此外，我们发现体相布伦斯台德位点可能包含与两种分子筛已知几何构型不同的氧位点相关的HO1、HO2与HO4型羟基，但仅有HO1型表面羟基与DRIFT光谱中增强的红移振动相关。进一步研究表明，单一表面模型无法解释DRIFT光谱中观测到的全部振动信号。通过组合三种适配的表面模型的红外振动信号，可如通过计算这些分子筛晶体的体相HO1···HO4振动来匹配TR光谱一般，精准复现实验得到的DRIFT光谱。