13 Scientific Article The Israel Chemist and Chemical Engineer Issue 8 · November 2021 · Kislev 5782 with adsorbates, depending on the mode of operation, spectroscopic signals are generally of low relative intensity. This holds true both for surface-sensitive spectroscopies, and bulk techniques. Figure 1 shows some applications and limitations of commonly applied spectroscopic techniques, X-ray absorption spectroscopy (XAS) and infrared (IR) spectroscopy, to study catalytic reactions. Firstly so, they are limited because the exposed surface-fractions of the metals that comprise nanoparticles are often in the singledigit percentile, as is the fraction of Brønsted acid sites of a zeolite. For example, an 8 nm Ni nanoparticle consists of approximately 25,000 atoms , of which approximately 0.1 are surface atoms. Secondly, spectral information is often heavily clouded by the least active species  as different types of sites, or surface fractions, have different interactions with adsorbates. Furthermore, of the relatively small fraction of exposed surface, an even smaller fraction does the actual job . In practice, this means that the majority of the signals in a spectroscopic measurement come from the molecules that are the most stable (i.e., the least reactive), instead of the molecules that are undergoing reaction or, in other words, “spectator species” mask active species. Thirdly, surfaces and in particular nanoparticles are structurally highly dynamic under the elevated temperatures and pressures at which they are often applied, and particularly when exposed to adsorbates or reactants . To summarize, our fundamental knowledge of the interaction of adsorbates with nanoparticles is lacking due to: 1. Small spectroscopic signal changes in noisy environments 2. Broad and convoluted signals 3. Adsorbate-sur face systems change dynamica l ly, particularly so as a result of the processes under study. 1.2 Past and present approaches For these reasons, an overwhelming majority of studies in the literature have oversimplified the systems under study, for example, by using single crystal facets, studied under ultra-high vacuum conditions (an approach that is termed surface science) [3, 7-11]. These approaches have yielded many of the important insights on which we currently build our understanding. However, in recent years it has become increasingly apparent that adsorbates and surfaces have completely different physical parameters, such as surface stability, mobility of species, surface coverages, and surface energies, at relevant conditions of pressure and temperature [3, 12, 13]. Several surface-science groups around the world are attempting to close this gap by performing surface-science experiments at less high-vacuum conditions, approaching “ambient” pressures [13–15]. However, many of these approaches still make use of model surfaces rather than the supported nanoparticles that are used in applications, and these measurements are generally performed in a static manner. When a static measurement is performed on a surface that is in chemical and physical equilibrium, much information is lost as to how this equilibrium was reached. That is, for a given reaction intermediate to cover a surface, often several sequential elementary reaction steps have occurred on the time scale of milliseconds to seconds, each of which affect the surface and subsequent steps [1, 16]. Notable approaches to studymodel surfaces at conditionsmore relevant to applications are, for example, high-pressure scanning tunneling microscopy (STM) of model hydrodesulfurization catalysts working together with an industrial partner [17, 18]. The same group is working on integrating STM and Figure 1. Two ensemble (bulk) spectroscopy types (X-ray absorption spectroscopy, and infrared absorption spectroscopy) which can be applied at high pressures and temperatures, and at high temporal resolution and their limitations in the field of study.