Using probe molecule FTIR spectroscopy to identify and characterize Pt group metal based single atom catalysts

Chinese Journal of Catalysis 38 (2017) 1473 1480 催化学报 2017 年第 38 卷第 9 期 www.cjcatal.org available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/chnjc Perspective (Special Issue of the International Symposium on Single Atom Catalysis (ISSAC 2016)) Using probe molecule FTIR spectroscopy to identify and characterize Pt group metal based single atom catalysts Chithra Asokan a,, Leo DeRita a,, Phillip Christopher a,b,c, * 1. Introduction Optimizing the efficiency of platinum group metal (PGM) utilization in supported catalysts is critical due to the naturally low abundance, high costs, current demands, and expanding proposed applications of these metals. An emerging class of supported PGM catalysts that provides the potential for perfect metal utilization efficiency is single atom, or isolated site, PGMiso, catalysts where single PGM atoms are dispersed on a support [1 10]. The potential for perfect metal utilization requires that all PGMiso sites exist in similar local bonding environments creating a constant reactivity across all sites. While many reports have demonstrated the utility of PGMiso catalysts, the limited space for reactant adsorption at such sites suggests that the range of potential applications may ultimately be dictated by an ability to introduce additional functionality around PGMiso species. In addition to the design of synthesis methods for producing these atomically precise supported catalysts, it is critical to develop site specific characterization approaches that identify the existence of PGMiso species on supports, provide insight into their local bonding environment, and assess their reactivity. Aberration corrected scanning transmission electron microscopy (STEM) is an essential tool to provide proof of the existence of PGMiso on supports, although STEM analysis cannot offer details regarding the local environment of single metal atoms unless the support is very well defined [11]. In addition, a significant number of images are needed per sample to develop quantitative data regarding the concentration or fraction of PGMiso species in a sample. Thus, coupling STEM with other characterization approaches is useful for analyzing the uniformity of single atom existence throughout large samples and providing insights into the PGMiso environment. X ray absorption spectroscopy (XAS) is often coupled with STEM to characterize supported PGMiso species and is particularly useful when studying homogeneous single site species [11]. However, XAS analysis of supported PGMiso species can be difficult to interpret when there is coexistence of multiple single site species (ie. species located on different sites on the support) or the co existence of small clusters and PGMiso species. In this perspective, we describe how probe molecule Fourier transform infrared spectroscopy (FTIR) is a powerful site specific characterization technique that can be used to identify and quantify the concentration of PGMiso species in a catalyst sample, and provide insights into local geometry, stability, reactivity and homogeneity of supported PGMiso species. Probe molecule FTIR is particularly well suited for characterizing supported PGMiso species because bonds formed between probe molecules and PGMiso species are sensitive to local geometric and electronic environments, which induces changes in band width [1], frequency [12] and other characteristics of the adsorbed molecule s IR spectrum [13,14]. Furthermore, temperature programed FTIR experiments following coverage as a function of temperature can provide site specific adsorption energies [13,15,16]. As compared to STEM or XAS, probe molecule FTIR is economical in that it can detect low concentrations of PGMiso, is inexpensive, and only requires short times. We start this perspective by describing the use of CO probe molecule IR to identify supported single Pt atoms and then single Rh, Ir and Os atoms, with the defining difference between these systems being the number of CO molecules that stably binds per PGM atom, and end with outlooks on how probe molecule IR can be used to provide insights into reactivity, local structure, and homogeneity of PGMiso species. 2. Identifying PGMiso species The choice of probe molecule in FTIR characterization of catalysts is important to ensure that the frequency or band appearances of the adsorbed molecule will respond to various characteristics of the adsorption site, allowing for site specific analysis. For the characterization of supported PGMiso species, CO provides several beneficial attributes, such as: (1) each adsorption site is comparable to the size of one CO; (2) CO generally binds strongly enough to supported PGM structures to enable characterization by room temperature IR; (3) the polarization of CO is responsive to changes in metal charge, dipole dipole coupling with nearby CO, and the coordination number of the PGM site [17,18]. Other probe molecules such as

1474 Chithra Asokan et al. / Chinese Journal of Catalysis 38 (2017) 1473 1480 NO may also be useful, although NO tends to dimerize, complicating site specific analysis, and NO is more likely to dissociate on metal surfaces compared to CO [19]. While CO is an ideal probe molecule in many regards, it is also important to note that CO can induce reconstruction of supported PGM species, which must be considered and is discussed further below [20 23]. In this section we start by discussing isolated Pt atoms, Ptiso, on supports. The first consideration in using CO probe molecule FTIR to differentiate the existence of Ptiso from other Pt structures, such as metallic Pt nanoparticles, Ptmetal, on a support is the expected stretching frequency of CO when adsorbed to each site. CO adsorbed to Ptmetal sites has been extensively studied, where it is known that the stretching frequency for linearly adsorbed CO is between 2030 2100 cm 1, with variations derived from the coordination number of the Pt adsorption site and charge transfer between Ptmetal and the support. In addition to the linear adsorption geometry of CO on metal surfaces, CO can also bind in a bridge configuration between two metal atoms. The stretching frequency of CO when bound to bridge sites is typically redshifted to a lower frequency (~1750 1950 cm 1 ) compared to the stretching frequency of linearly bound CO. Due to the bridging adsorption geometry requiring two metal atoms, the presence of this vibrational mode is indicative of metal nanoparticles existing. The absence of CO bound in the bridge geometry, while insufficient evidence alone, is a practical starting point to identify the presence of single atom adsorption sites, particularly for Pd atoms as CO binds preferentially to Pd nanoparticle surfaces in a bridge configuration [24]. By contrast, due to coordination with O atoms on the surface of metal oxide supports, or in the lattice of zeolites, Ptiso species are expected to exist in a cationic charge state, which is manifested spectroscopically in a blueshifted (higher wavenumber) band position of CO adsorbed to Ptiso compared to Ptmetal sites (either in the linear or bridge adsorption geometry), with reported stretching frequencies ranging from 2080 2170 cm 1 [17,18,25]. In general it has been reported that the stretching frequency for CO on Ptiso is shifted to 40 50 cm 1 higher frequencies compared to CO adsorbed in a linear geometry on Ptmetal clusters on the same support. The exact positions of both stretching frequencies are highly dependent on the direction and magnitude of charge transfer between Pt and the support. An example of this differentiation is shown in Fig. 1(a), where the IR spectra of CO adsorbed to Pt on H ZSM5 at room temperature for a range of catalysts with varying Pt loading are shown [2]. In these systems an organometallic Pt precursor was used to specifically promote Ptiso formation through interaction between the precursor and Al framework bonding sites and the catalyst was not reduced prior to characterization to minimize Ptmetal formation (see discussion further below on this). At the lowest weight loading, saturation of the catalyst by CO at room temperature produced a CO stretching band at 2115 cm 1, while at higher weight loadings an increasing intensity of a band at 2070 2090 cm 1 was observed. Through correlation with STEM imaging of similar low loading catalysts, the 2115 cm 1 band was assigned to CO adsorbed on Ptiso species, while the band at 2070 2090 cm 1 was assigned to CO adsorbed in a linear geometry to Ptmetal sites. In this case, as expected, CO on Ptiso exhibited a blueshifted (higher frequency) stretching frequency compared to CO on Ptmetal and furthermore it was demonstrated that complimentary techniques of probe molecule IR and STEM imaging provide an excellent way to identify Ptiso species. Apart from the frequency of the CO stretch, another critical spectroscopic signature for differentiating CO on Ptiso and Ptmetal sites (and this is general for all PGMs) is that the CO stretching frequency is independent of coverage when adsorbed to Ptiso, while the CO frequency redshifts with decreasing coverage on Ptmetal. Fig. 1(b) shows an example of this, where IR spectra were collected from a reduced 0.5% Pt/H mordenite (Hmor) catalyst that contained both Ptiso and Ptmetal species following CO adsorption at room temperature and purging of the system at increasing temperature. On Ptmetal sites the CO stretching frequency redshifts (from 2099 to 2073 cm 1 ) as coverage decreased due to increasing temperature. Conversely, as CO desorbs from Ptiso sites at increasing temperature there is no change in the stretching frequency. The coverage dependent shift on Ptmetal sites is due to dipole dipole coupling of adsorbed CO, where neighboring CO molecules vibrate in unison leading to a higher vibrational energy at higher coverage. Due to spatial separation of Ptiso species, there are no adjacent CO molecules in close enough proximity to induce coverage dependent frequency shifts due to dipole dipole coupling [3,25]. The lack of frequency shift with variation in coverage is an important signature of CO adsorption on PGMiso sites. While distinct identification of the existence of Ptiso and Ptmetal sites can be achieved by realizing that the frequencies and the coverage dependent frequency shifts for adsorbed CO are different, it is also important to differentiate Ptiso species from oxidized Pt clusters, Ptox, where Pt is in a cationic state for both structures. Differentiation of Ptiso and Ptox by CO IR requires analysis of the IR spectra of adsorbed CO following different catalyst pre treatments, as the band position for CO on these cationic sites has been reported to be similar (2100 2130 cm 1 ). For example, Fig. 1(c) shows spectra of CO adsorbed to a 5% Pt/Hmor catalyst at room temperature, which was previously reduced at increasing temperature in hydrogen. With a mild reduction below 200 C where Ptox is known to only partially reduce the IR spectrum of adsorbed CO shows a broad band centered at ~2120 cm 1, a similar stretching frequency as was seen for CO on Ptiso in Fig. 1(b) [25]. When the 5% Pt catalyst was fully reduced and probed with CO, the band intensity associated with CO on cationic Pt sites completely disappeared. When comparing Fig. 1(b) and (c), the signature of CO bound to Ptiso remained after complete reduction (Fig. 1(b)), while the signature of CO bound to Ptox clusters disappeared after complete reduction (Fig. 1(c)). This example demonstrates that analyzing the stability of cationic Pt sites through increasing reduction temperature is critical for separating Ptiso and Ptox particles active sites when using IR analysis, as they otherwise exhibit similar stretching frequencies of adsorbed CO. Thus, combining variations in metal weight loading (where lower weight loading should promote Ptiso formation,

Chithra Asokan et al. / Chinese Journal of Catalysis 38 (2017) 1473 1480 1475 Fig. 1. Unique identification of Ptiso species. (a) IR spectra of CO adsorbed at saturation coverage and room temperature to different Pt/HZSM 5 catalysts containing varying Pt weight loading. Reproduced with permission from Ref. [2]. Copyright 2013 American Association for the Advancement of Science. (b) IR spectra of CO adsorbed on 0.5% Pt/Hmor catalyst at room temperature following a pretreatment of calcination at 500 C and reduction at 350 C, after purging with inert for 30 min at (1) 20, (2) 100, (3) 250 and (4) 350 C. (c) IR spectra of CO adsorbed on 5% Pt/Hmor catalyst at room temperature following pretreatment by calcination at 500 C and reduction in H2 at (1) 50, (2) 100, (3) 200 and (4) 500 C. (b) and (c) reproduced with permission from Ref. [25]. Copyright 1994 Royal Society of Chemistry. while higher loading should promote cluster formation) and catalyst pretreatment prior to CO IR analysis provide a good platform for differentiating the existence of Ptiso and Ptox species. However, it is worth mentioning that reduction of the catalyst could induce mobility of Ptiso species and Pt cluster formation and as a result, corollary characterization of the catalyst as a function of reduction temperature by STEM imaging is useful for ensuring the pre treatment used to differentiate Ptiso and Ptox species does not cause structural changes to the catalyst. We note that although Pt was the primary focus in this section, due to the similar electronic structure supported Pd species should exhibit similar spectroscopic trends to those described for Pt, with a primary difference being the preferred bridge adsorption site of CO on Pd metal clusters, as compared to atop adsorption site on Ptmetal clusters. 3. Evaluating PGMiso reactivity The utility of IR to study PGMiso extends beyond identification of the sites. Once the sites have been unambiguously identified, an in situ IR setup allows one to probe reactivity, compare adsorption strength of probe molecules on different sites, and explore possible reaction mechanisms by identifying the active site and quantifying changes in the spectra over the course of a reaction. There is strong evidence from surface science studies that CO adsorbs more strongly to Ptox sites compared to Ptmetal sites [15,16]. In characterizing the catalytic re

1476 Chithra Asokan et al. / Chinese Journal of Catalysis 38 (2017) 1473 1480 activity of Ptiso species (and PGMiso species in general) a first step is to compare the adsorption energy of reactive species to Ptiso versus Ptmetal and Ptox sites. Initial work has suggested that the support can significantly influence the adsorption strength of CO on Ptiso (and Pdiso) species, where on FeOx and Al2O3 CO adsorbs weakly to the Ptiso and Pdiso sites, while on H ZSM5 CO adsorbs quite strongly to Ptiso [2 4]. This suggests the support plays a very important role, acting as a ligand to modify the reactivity of PGMiso species [9]. In addition to comparing the adsorption energy of probe molecules on each type of site, the reactivity in temperature programmed experiments can also be analyzed. An example of this type of experiment, where IR spectra of a CO saturated 2.6% Pt/HZSM 5 catalyst containing both Ptmetal and Ptiso species (confirmed by STEM imaging) during exposure to O2 at room temperature and at elevated temperatures is shown in Fig. 2(a). It was observed that CO molecules on Ptmetal sites (red area) were highly reactive in O2 (by the disappearance of the band), while CO on Ptiso (green) was stable in the O2 environment. The results suggest that Ptiso on HZSM 5 is much less reactive for CO oxidation compared to Ptmetal on HZSM 5. Clearly, the use of in situ temperature programmed IR experiments provide advantages over a traditional experiment using a thermal conductivity detector (TCD) or mass spectrometry signal, because this provides site specific information about the stability and reactivity of probe molecules. While differentiation of Ptiso (and likely Pdiso) from Ptmetal and Ptoxide sites using CO probe molecule IR requires consideration of band positions, coverage dependent frequency shifts, and pretreatment dependent signatures, other PGMiso species are easier to identify due to distinct probe molecule adsorption characteristics. In the case of Rh, Os, and Ir, it is energetically favorable to form metal gem dicarbonyl, M(CO)2, structures when site isolated species are exposed to sufficient CO pressure [1,9,13,14,26]. The M(CO)2 structures exhibit two characteristic CO stretches associated with the symmetric and asymmetric stretch of the 2 CO molecules. Bands associated with the symmetric and asymmetric stretches of the M(CO)2 structures typically occur at frequencies that are distinct from either linear or bridge bound CO to nanoparticles or clusters of the same metal, in either the oxidized or metallic state. These structures and their associated spectra when exposed to CO have been studied in great detail and the assignments relating structures to spectra have been rigorously substantiated [13,14,26]. An example of this is shown for a 4% Rh on TiO2 catalyst that was saturated by CO at room temperature followed by IR analysis (Fig. 2(b)). The modes at 2097 and 2028 cm 1 are assigned to the symmetric and asymmetric stretches of the Rh(CO)2 species, respectively, while the modes at 2068 and 1860 cm 1 are assigned to CO adsorbed in linear and bridge bound geometries, respectively, at Rhmetal cluster surfaces. We note that there are a few reports of gem dicarbonyls forming at defect sites on certain metal and metal oxide particles following harsh treatments [27]. Thus, conclusive relationships between the existence of the gem dicarbonyl bands in IR spectra and the existence of single metal atoms should be corroborated by coverage dependent measurements that show constant CO band position, or STEM measurements. In addition to rigorous assignment of CO adsorbed to various geometries of Rh sites, extinction coefficients for these stretches have been measured, enabling the use of CO IR for quantification of the relative or absolute number of each type of adsorption site [5,26]. For example, Rh/TiO2 catalysts with varying Rh weight loading were prepared and the IR spectra at Fig. 2. Probing the reactivity of PGMiso. (a) IR spectra of CO adsorbed on 2.6 wt% Pt/HZSM 5 catalyst in flowing O2 at varying exposure time and temperature. The band at >2100 cm 1 in green was assigned to CO on Ptiso and the band at <2100 cm 1 in red was assigned to CO on Ptmetal clusters. Reproduced with permission from Ref. [2]. Copyright 2013 American Association for the Advancement of Science. (b) IR spectrum of CO adsorbed at room temperature and saturation coverage on 4% Rh/TiO2. Ball and stick models of the assigned vibrational modes are also shown. (c) IR spectra of CO adsorbed to Rh/TiO2 catalysts at varying weight loadings is shown. (d) The fraction of exposed Rh atoms existing as Rhiso derived from quantitative IR is shown in a black line and the reverse water gas shift (r WGS) turn over frequency (TOF) is shown in red dots as a function of Rh weight loading. (b d) reproduced with permission from Ref. [5]. Copyright 2015 American Chemical Society.

Chithra Asokan et al. / Chinese Journal of Catalysis 38 (2017) 1473 1480 1477 saturation CO coverage and room temperature were collected (Fig. 2(c)). With increasing Rh weight loading the spectra shifted from being dominated by the modes associated with CO on Rhiso to the modes associated with CO on Rhmetal. Using the spectra in Fig. 2(c), the fraction of Rh sites existing at the surface of Rhmetal nanoparticles and as Rhiso species was quantified by deconvoluting the spectra, integrating the area under the peaks associated with the symmetric Rh(CO)2, linear CO on Rhmetal, and bridge CO on Rhmetal stretches, and normalizing the area by the site specific CO adsorption stoichiometry and extinction coefficients [5,26]. As shown by the black line in Fig. 2(d), the fraction of exposed Rh sites that exist as Rhiso species decreases from ~60% to ~8% as Rh weight loading increased from 0.5% to 6%. By relating the measured quantity of exposed Rhiso and Rhmetal sites on this series of catalysts to turn over frequencies (TOF) it was shown that selectivity in the reaction of CO2 reduction by H2 was controlled by the population of each type of site, where CO production through the reverse water gas reaction occurred only on Rhiso sites, while CH4 formation through CO2 methanation occurred on Rhmetal particles. This can be seen as the quantitative correlation between the fraction of sites existing as Rhiso and TOF for CO formation as a function of Rh weight loading in Fig. 2(d). This example shows conclusively how powerful the CO probe molecule IR technique can be, where the relative fraction of Rhiso species was quantified by CO IR and related to a unique catalytic characteristic of this site. These mechanistic insights were recently exploited to demonstrate that catalysts containing exclusively Rhiso can drive the water gas shift reaction with >95% CO conversion without forming CH4, which is important for industrial application [6]. Using extinction coefficients and FTIR band intensities to quantify single atom sites is one way to assign reactivity, but measuring changes in the gem dicarbonyl peak positions in situ can also provide insight about reaction intermediates and mechanisms. For example, recently the interaction of CO with Rhiso species and mechanism of CO oxidation on Rhiso species supported on phosphotungstic acid (NPTA) clusters was analyzed by following shifts in the band positions of Rh(CO)2 species and relating this to changing charge states of the complex [7]. When exposed to CO the catalyst consisting of solely Rhiso species on NPTA clusters exhibited 2 sets of dicarbonyl bands at 2108 & 2048 cm 1 and 2093 & 2034 cm 1, which were assigned to Rh(CO)2 3+ and Rh(CO)2 +, respectively. By following the loss of bands associated with Rh(CO)2 3+ species as a function of temperature and reactant partial pressure it was demonstrated that reduction of Rh(CO)2 3+ to Rh(CO)2 + could occur by CO abstraction of oxygen bound to Rh(CO)2 3+, which is a step involved in the CO oxidation catalytic cycle. 4. Characterizing local environment of PGMiso species We have discussed how CO probe molecule IR can be used to identify the existence of PGMiso sites, assign their reactivity in catalytic processes, and help elucidate insights into catalytic mechanisms. Another interesting area in utilizing probe molecule IR for characterizing PGMiso is in exploiting spectral characteristics to infer details associated with the local geometry of these species. For example, it is interesting to consider if PGMiso species reside in a single adsorption site on a support, or if there are multiple sites that exhibit degenerate adsorption energies for the single metal atom, thus meaning that the various PGMiso species on a sample are not homogeneous in their local environments. It was recently proposed that the full width at half max (FWHM) of modes assigned to CO on PGMiso species could be used as a qualitative indicator of the homogeneity of PGMiso CO complex adsorption sites on a support [1]. An example of this analysis is shown in Fig. 3(a), where the spectra of Ir(CO)2 species in solution and on various supports is compared. For Ir(CO)2 complexes synthesized on zeolite supports under optimized conditions the FWHM of the bands associated with the symmetric and asymmetric stretch are ~5 cm 1, which approaches the value for Ir(CO)2 complexes in solution, ~4 cm 1, suggesting that Ir(CO)2 species were adsorbed at a single site, or very few but similar sites, on the zeolite [1]. For the analogous case of Rh(CO)2, this is confirmed by theoretical calculations that identify a single most stable adsorption site for Rh(CO)2 a on zeolite support, see Fig. 3(b) [28]. In the Ir case, by varying synthetic conditions or zeolite characteristics the FWHM of the CO stretches increased, suggesting careful synthetic procedures are required for producing homogeneous PGMiso species. In contrast to zeolite supported Ir(CO)2 species, when using MgO or Al2O3 as a support, the FWHM of the CO stretches increases significantly, suggesting a plethora of Ir(CO)2 adsorption sites on the support. This can be explained by the many adsorption sites for PGMiso species that exist on metal oxides, which exhibit similar binding energies, as shown for the case of Rh(CO)2 on Al2O3 in Fig. 3(c) [29]. This is an important concept because when the catalytic reactivity of PGMiso species is examined at a mechanistic level it is critical to consider the local environment. When the local environment surrounding PGMiso species exist in many forms (as is the case shown for Ir(CO)2 on MgO in Fig. 3(a)), it is more difficult to understand PGMiso reactivity as compared to well defined systems, such as on zeolite supports. Interestingly, it can also be seen here that for acidic supports (zeolites) the CO stretching frequency of Ir(CO)2 is blueshifted (higher frequency) compared to the frequency in solution, while on basic supports the CO stretching frequency redshifts (lower frequency), demonstrating that the support directly acts as a ligand for PGMiso species [7]. It has previously been proposed that the spectral signature of the gem dicarbonyl species can indicate the local bonding geometry of Rhiso on a specific support [13,14]. Based on the ratio of the asymmetric and symmetric peak intensities, the bond angle between the two Rh CO bonds, θ, can be estimated using the following equation, Iasym 2 tan Isym where Iasym, and Isym are the symmetric and asymmetric peak intensities, respectively. The calculated bond angle is dependent on both the precious metal and the support. While angle is not necessarily an indicator of adsorption strength, certain sites may maintain stronger bonds when undergoing tempera

1478 Chithra Asokan et al. / Chinese Journal of Catalysis 38 (2017) 1473 1480 Fig. 3. PGMiso adsorption site homogeneity. (a) IR spectra of Ir(CO)2 complexes on various supports and in solution. Reproduced with permission from Ref. [1]. Copyright 2016 American Chemical Society. (b) Schematic of the location and structure of Rh(CO)2 in a zeolite support from theoretical calculations. Reproduced with permission from Ref. [28]. Copyright 2000 American Chemical Society. (c) Schematics of the structure of various Rh(CO)2 adsorption sites on hydroxylated (001) and (100) γ Al2O3 surfaces. Reproduced with permission from Ref. [29]. Copyright 2013 American Chemical Society. ture programmed desorption and thus the overall average angle may change as a function of temperature. However, it is critical to consider that unless the sample contains only a single M(CO)2 adsorption site on the support, the resulting IR spectrum is a composite of the many M(CO)2 species that exist, complicating the analysis. 5. Outlook and conclusions As research expands into PGMiso species of various metals on various supports, it has become apparent that the support influences the reactivity of these catalysts as significantly as the metal. This is because in these catalysts, the metal atom and local environment of the support make up the active catalytic site and commonly the support plays an active role either by supplying OH or labile O species to the catalytic cycle [9,30]. The support is a ligand that sterically and electronically modifies the PGMiso species, but also can be actively involved in catalysis. Given the sensitivity of CO probe molecule IR to the local environment of PGMiso species, it is expected that this technique will play a critical role in developing insights into broad trends that exist and define how the support activates or inactivates PGMiso species for catalysis. By tailoring the microenvironment of the support PGMiso complex, and exploring this local environment with IR approaches, it is expected that predictive insights into site specific reactivity will be developed that enable enhanced catalytic activity, and thus better metal utilization efficiency, and unique reactivity that could bridge the gap between the specificity of heterogeneous and homogeneous catalysts. In addition to identifying the existence and characterizing the local environmental of single precious metal atoms on oxide supports, there is significant potential for the same approach to be useful in the characterization of single atom alloy catalysts [31,32]. However, the local coordination of the single atom dopants in metal surfaces is likely to significantly modify the behavior of the single atom compared to when it is supported on an oxide. There are significant challenges with using probe molecule FTIR to characterize PGMiso species. The first challenge, as

Chithra Asokan et al. / Chinese Journal of Catalysis 38 (2017) 1473 1480 1479 mentioned above, is the homogeneity of the PGMiso local environments on the support. The most useful insights into local environment from IR spectroscopy will be obtained for samples where the PGMiso species exist in a single location. Although, using the FWHM of probe molecule bands to identify samples with heterogeneous local environments may help to identify the particular locations on the support with desired reactivity. Furthermore, calcination, reduction, and exposure of probe molecules to catalysts may induce reconstruction, causing additional challenges for characterization [20 23]. Monitoring these effects during or after various treatments via corollary in situ or ex situ STEM and XAS characterization is critical for developing definitive probe molecule IR assignments to different adsorption sites. Given the well known ability of CO to induce mobility of PGMiso species of the support, causing particle formation or fragmentation depending on the metal and conditions, IR analysis at sub ambient temperature is expected to be useful for identifying the as synthetized structure of PGMiso based catalysts. In conclusion, it has become apparent that catalysts consisting of isolated or single atoms of PGMs on oxide support are introducing novel catalytic functionality and the possibility for enhanced metal utilization efficiency. The further development of these concepts requires the ability to characterize such dispersed metal species in terms of identifying their existence, analyzing their reactivity and providing insights into the local environment of these species. Probe molecule IR spectroscopy using CO is emerging as a very powerful approach to achieve these insights and push our understanding of PGMiso based catalysts forward. It is expected that with carefully continued development of this technique and corollary characterization by STEM imaging and XAS, we will begin to be able to tailor microenvironments surrounding PGMiso species to rationally design their reactivity. Acknowledgments P.C. acknowledges funding from the NSF CAREER (1554112). These authors contributed equally to this work. Phillip Christopher a Department of Chemical and Environmental Engineering, University of California, Riverside, Riverside, CA 92521, USA b Program in Materials Science, University of California, Riverside, Riverside, CA 92521, USA c UCR Center for Catalysis, University of California, Riverside, Riverside, CA 92521, USA Tel: +1 951 8277959 Fax: +1 951 8273188 E mail: Christopher@engr.ucr.edu Received 9 June 2017 Published 5 September 2017 DOI: 10.1016/S1872 2067(17)62882 1 References [1] A. S. Hoffman, C. Y. Fang, B. C. Gates, J. Phys. Chem. Lett., 2016, 7, 3854 3860. [2] K. Ding, A. Gulec, A. M. Johnson, N. M. Schweitzer, G. D. Stucky, L. D. Marks, P. C. Stair, Science, 2015, 350, 189 192. [3] B. T. Qiao, A. Q. Wang, X. F. Yang, L. F. Allard, Z. Jiang, Y. T. Cui, J. Y. Liu, J. Li, T. Zhang, Nat. Chem., 2011, 3, 634 641. [4] E. J. Peterson, A. T. DeLaRiva, S. Lin, R. S. Johnson, H. Guo, J. T. Miller, J. H. Kwak, C. H. F. Peden, B. Kiefer, L. F. Allard, F. H. Ribeiro, A. K. Datye, Nat. Commun., 2014, 5, 4885. [5] J. C. Matsubu, V. N. Yang, P. Christopher, J. Am. Chem. Soc., 2015, 137, 3076 3084. [6] H. L. Guan, J. Lin, B. T. Qiao, S. Miao, A. Q. Wang, X. D. Wang, T. Zhang, AIChE J., 2017, 63, 2081 2088. [7] B. Zhang, H. Asakura, N. Yan, Ind. Eng. Chem. Res., 2017, 56, 3578 3587. [8] M. Yang, J. L. Liu, S. Lee, B. Zugic, J. Huang, L. F. Allard, M. Flytzani Stephanopoulos, J. Am. Chem. Soc., 2015, 137, 3470 3473. [9] J. Guzman, B. C. Gates, Dalton Trans., 2003, 3303 3318. [10] J. Y. Liu, ACS Catal., 2017, 7, 34 59. [11] J. D. Kistler, N. Chotigkrai, P. Xu, B. Enderle, P. Praserthdam, C. Y. Chen, N. D. Browning, B. C. Gates, Angew. Chem. Int. Ed., 2014, 53, Graphical Abstract Chin. J. Catal., 2017, 38: 1473 1480 doi: 10.1016/S1872 2067(17)62882 1 Using probe molecule FTIR spectroscopy to identify and characterize Pt group metal based single atom catalysts Chithra Asokan, Leo DeRita, Phillip Christopher * University of California, USA The utility, applications, and future directions of using probe molecule IR spectroscopy for identifying and characterizing catalysts consisting of supported Pt group metal single atom catalysts is presented in this perspective. Absorbance (a.u.) Probe molecule IR 2150 2100 2050 2000 Wavenumber (cm 1 ) Identification, characteristics and reactive properties = Pt = Ti = O = C

1480 Chithra Asokan et al. / Chinese Journal of Catalysis 38 (2017) 1473 1480 8904 8907. [12] J. Ryczkowski, Catal. Today, 2001, 68, 263 381. [13] H. Miessner, I. Burkhardt, D. Gutschick, A. Zecchina, C. Morterra, G. Spoto, J. Chem. Soc., Faraday Trans. 1, 1989, 85, 2113 2126. [14] J. T. Yates, T. M. Duncan, S. D. Worley, R. W. Vaughan, J. Chem. Phys., 1979, 70, 1219 1224. [15] R. W. McCabe, L. D. Schmidt, Surf. Sci., 1977, 65, 189 209. [16] R. W. McCabe, L. D. Schmidt, Surf. Sci., 1976, 60, 85 98. [17] K. Chakarova, M. Mihaylov, K. Hadjiivanov, Microporous Mesoporous Mater., 2005, 81, 305 312. [18] A. Y. Stakheev, E. S. Shpiro, O. P. Tkachenko, N. I. Jaeger, G. Schulz Ekloff, J. Catal., 1997, 169, 382 388. [19] C. Lamberti, A. Zecchina, E. Groppo, S. Bordiga, Chem. Soc. Rev., 2010, 39, 4951 5001. [20] M. J. Kale, P. Christopher, ACS Catal., 2016, 6, 5599 5609. [21] T. Avanesian, S. Dai, M. J. Kale, G. W. Graham, X. Pan, P. Christopher, J. Am. Chem. Soc., 2017, 139, 4551 4558. [22] F. Solymosi, M. Pasztor, J. Phys. Chem., 1985, 89, 4789 4793. [23] B. R. Goldsmith, E. D. Sanderson, R. Ouyang, W. X. Li, J. Phys. Chem. C, 2014, 118, 9588 9597. [24] H. Unterhalt, G. Rupprechter, H. J. Freund, J. Phys. Chem. B, 2002, 106, 356 367. [25] V. L. Zholobenko, G. Lei, B. T. Carvill, B. A. Lerner, W. M. H. Sachtler, J. Chem. Soc., 1994, 90, 233 238. [26] T. M. Duncan, J. T. Yates, R. W. Vaughan, J. Chem. Phys., 1980, 73, 975 985. [27] W. F. Lin, S. G. Sun, Electrochim. Acta, 1996, 41, 803 809. [28] J. F. Goellner, B. C. Gates, G. N. Vayssilov, N. Rösch, J. Am. Chem. Soc., 2000, 122, 8056 8066. [29] O. M. Roscioni, J. M. Dyke, J. Evans, J. Phys. Chem. C, 2013, 117, 19464 19470. [30] E. W. McFarland, H. Metiu, Chem. Rev., 2013, 113, 4391 4427. [31] G. X. Pei, X. Y. Liu, A. Q. Wang, A. F. Lee, M. A. Isaacs, L. Li, X. L. Pan, X. F. Yang, X. D. Wang, Z. J. Tai, K. Wilson, T. Zhang, ACS Catal., 2015, 5, 3717 3725. [32] G. Kyriakou, M. B. Boucher, A. D. Jewell, E. A. Lewis, T. J. Lawton, A. E. Baber, H. L. Tierney, M. Flytzani Stephanopoulos, E. H. Sykes, Science, 2012, 335, 1209 1212. 利用探针分子红外光谱识别和表征铂族金属基单原子催化剂 Chithra Asokan a,, Leo DeRita a,, Phillip Christopher a,b,c,* a 加利福尼亚大学河滨分校化工与环境工程系, 加利福尼亚 92521, 美国 b 加利福尼亚大学河滨分校材料科学项目, 加利福尼亚 92521, 美国 c 加利福尼亚大学河滨分校催化 UCR 中心, 加利福尼亚 92521, 美国 摘要 : 单原子催化是提高贵金属利用率的有效手段, 而表征单原子催化剂是理解单原子催化的基础. 探针分子红外光谱可 用于识别和定量催化剂样品中孤立的 Pt 族金属物种的浓度, 从而得到负载的孤立的 Pt 族金属物种的局部几何形状 稳定性 活性及其分散性. 本文讨论了该技术用于识别和表征含负载型孤立的 Pt 族金属原子催化剂的效能 应用 以及未来的发展 方向. 关键词 : 单原子催化剂 ; 铂族金属 ; 红外光谱 ; 表征 ; 载体效应 收稿日期 : 2017-06-09. 接受日期 : 2017-07-01. 出版日期 : 2017-09-05. 共同第一作者. * 通讯联系人. 电话 : +1-951-8277959; 传真 : +1-951-8273188; 电子信箱 : Christopher@engr.ucr.edu 基金来源 : 国家科学基金会 (1554112). 本文的英文电子版由 Elsevier 出版社在 ScienceDirect 上出版 (http://www.sciencedirect.com/science/journal/18722067).