Infrared spectroscopy has a rich heritage in heterogeneous catalysis, where it has frequently been used to deduce the structure of chemisorbed species present on high surface area catalysts [i] . CO oxidation and the water gas shift reaction are two mechanisms, which are of great interest and have therefore been subjected to mechanistic and kinetic study using in situ infrared spectroscopy. This is commonly considered to be operando methodology.
Operando methodology allows the observation of catalytic parameters as well as surface structure. This methodology combines in situ spectroscopic techniques with simultaneous monitoring of catalytic performance. It is possible to investigate both characteristics in the same experiment on the same sample. Operando methodology is considered the best way to produce accurate representations of reactions [ii] .
Get your grade
or your money back
using our Essay Writing Service!
The techniques used to analyse surface species include diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) as well as Steady-state isotopic-transient kinetic analysis (SSITKA)
SSITKA involves replacing reactants with an alternative corresponding isotope. The isotope is used as a label which can be used to follow the reaction through adsorption onto the surface of the catalyst, conversion and then desorption. A mass spectrometer is used to monitor the isotope. This provides key information such as the coverage of any surface intermediates as well as the average residence time of these surface intermediates.
SSITKA is often coupled with IR analysis such as DRIFTS, which can be used to determine the nature of any surface intermediates.
Simple IR analysis of reactions often does not provide a clear representation of all the species in the reaction. The IR bands can overlap masking vital information needed to deduce reaction pathways and kinetics. However inlet concentration modulation can be adopted to overcome this particular obstacle. This involves periodic variation of reactant concentrations.
The number of workers active in this particular area of research (IR assisted mechanistic investigations of heterogeneously catalysed gas phase reactions) is surprisingly small. In order to provide a broad perspective of the discipline, the work of three different groups will be described in this review of. Firstly, McDougall et al. study spectroscopic identification of the active site for CO oxidation on Rh/Al2O3 by concentration modulation in situ DRIFTSiii. Secondly, Meunier et al. outline quantitative analysis of the reactivity of formate species seen by DRIFTS over a Au/Ce(La)O2 waster-gas shift catalyst. Finally, Chuang and Guzman describe their mechanistic investigation of heterogeneous catalysis by transient infrared methodsviii. This includes the NO-CO reaction, ethylene hydroformylation and photocatalysed reactionsviii.
McDougall et al. used diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) to identify the active site for CO oxidation on Rh/Al2O3 [iii] . The group also used this technique to determine the surface species present on the catalyst as well as identifying the active species and those playing a spectator role.
Determining whether or not a surface species is active or not requires the use of transient techniques, which may include flow switching or pulse adsorption. The inlet composition is altered periodically and the corresponding DRIFTS data, as well as mass spectrometry data, are compared allowing further insight into the nature of particular surface species.
Figure 1: DRIFTS spectrum after equilibration of the sample under the reaction mixture (5%CO, 3%O2 in He at a total flow rate of 65 sccm) at 567K.
The broad peak displayed between 1700 and 1400 cm-1 of figure 1 is often associated with carbonate species. This is the most substantive feature of figure 1 and prompted the group to conclude that there was considerable build up of carbonates, primarily of the Al2O3 sites.
The peaks at 2081 and 2010 cm-1 can be assigned to geminal dicarbonyl speciesiii. Adsorbed CO in bridged sites is indicated by the weak feature at 1880 cm-1 however there is no obvious band suggesting linear CO on Rh metal.
Figure 2: Collected MS and DRIFTS trend analysis during modulation of the inlet reactant concentration at 0.11Hz
Always on Time
Marked to Standard
The gaseous concentrations of CO, O2 and CO2 were followed by a mass spectrometer (MS). This was coupled with the integration of peak areas collected by DRIFTS spectra and allowed insight into the detection of a number of surface species.
The MS and DRIFTS data were collected at various wavenumbers in order to observe the changes in the asymmetric and symmetric vco of the geminal dicarbonyl species, bridged CO and the surface carbonate. These wavenumbers included 2123 to 2047 cm-1, 2047 to 1945 cm-1, 1915 to 1820 cm-1 and 1772 to 1580 cm-1 The traces were collected over 10 modulation cyclesiii.
Figure 2 indicates that the carbonate species are unique in showing a clear modulation which is almost completely in phase with the concentrations of CO, O2 and CO2. The higher frequency symmetric dicarbonyl mode is the only other species which shows any sort of variation with concentration. This variation however is only very slight. This therefore suggests that the kinetics of the elementary reaction steps are fast relative to the input modulation frequency under the investigated conditions of the reaction.
The traces associated with both dicarbonyl species display a gradual drop in absorbance over time. There is also a shift in phase between all the traces, however, this shift is negligible. This negligible shift indicates that the kinetics of the elementary reactions steps are very quick with respect to the input modulation.
Figure 3 is produced by the collection and coaddition of data from four modulation cycles. The spectra detect all species identified in figure 1. The most significant aspect is the peak positioned at 2100 cm-1 which is well resolved. This is attributed to CO adsorbed on oxidised Rh(I) sites. The difference spectrum shows that the intensity of the peak is related to the differential conversion.
The information extracted from the difference spectrum allows the conclusion that the geminal carbonyl is a spectator whilst the CO adsorbed on oxidised Rh(I) is the active species in the oxidation of CO. Although it was previously thought that geminal carbonyl species were inactive [iv] , the active nature of CO on oxidised Rh is contrary to previous claim [v] .
Linear CO was also observed during the reaction however it was not observed by the subtraction technique (figure 3) and therefore is regarded as inactive in the reaction mechanism. This is also the case for bridged CO as well as CO on more strongly oxidised Rh(II-III) sites which were not detected within any of the data collected from this study.
Figure 3: DRIFTS spectra, each produced from coaddition of 40 individual spectra collected during maxima and minima in the modulation cycle. The inset shows the result of subtraction of the two spectra in the region for forms of adsorbed CO. The vertical dotted lines are at the maximum of the band in the inset (2100 cm-1)
From the results presented previously, the group were able to draw a number of conclusions. Due to the use of in situ DRIFTS during concentration modulation, active species for CO oxidation on Rh/Al2O3 could be identified. Without the use of concentration modulation there is a possibility of bands overlapping due to the presence of coadsorbed inactive 'spectator' species. A spectator species is a moiety that is detectable via spectroscopy but one that plays no major role in the mechanism, i.e. it is kinetically irrelevant to the main process chemistry. This issue was therefore overcome.
CO adsorbed on oxidised Rh sites was shown to be the active species in the oxidation of CO whereas geminal carbonyls and linear CO on Rh metal were identified as inactive during the reaction.
This technique of concentration modulation coupled with DRIFT analysis could be used to further analyse the kinetics of the reaction as proposed by the group. The adsorption, desorption and surface reaction kinetics could be explored further by the use of this frequency response technique.
Meunier et al. used DRIFTS to examine the water gas shift reaction over a Au/Ce(La)O2 catalyst [vi] . DRIFTS was used initially to measure the rate of the reaction.
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
Figure 4: Arrhenius plot relating to the rate of CO2 formation measured over the AuCL sample in a quartz plug-flow reactor in Medford and in the modified DRIFT cell in Belfast. The feed was 2% 12CO + 7% H2O in Ar or He
Figure 4 shows the reaction rates at various temperatures and hence an activation energy was calculated. The data was collected using DRIFTS which monitored the formation of CO2.
DRIFTS was also used to identify surface species. IR spectra of the Au catalyst were observed and the IR bands were assigned to various intermediate species.
Figure 5: In situ DRIFT spectra obtained over the 0.6 AuCL at steady-state conditions at 493K under (a) 7% H20 in Ar and (b) the WGS feed: 2% CO +7% H2O in Ar. A mirror signal was used as background
Figure 5 displays two DRIFT spectra, the IR bands of which have been assigned to various surface species.
The lower spectra, (a), contains a large band at 3700 - 2500 cm-1 which corresponds to the stretching mode of hydrogen-bonded hydroxyls. The band at 3658 cm-1 indicates the presence of hydroxyl groups doubly coordinated to the catalyst. Between 1800 and 1100 cm-1 there appears a broad band which is commonly associated with OCO-containing compounds observed over basic solids, including ceria. These compounds include carbonates and carboxylates. The presence of carbonate species is further justified by the 855 cm-1 peak which corresponds to the out-of-plane bending vibration of a carbonate.
The upper spectrum, (b), contained new bands present at 2946 and 2830 cm-1. These were associated with the combination (C-H) + vs(OCO) and v(C-H) of a bidentate formate species respectively. OCO groups of formate and carbonate species contributed to additional bands observed in the region 1800-1200 cm-1.
Figure 6: In situ DRIFT spectra obtained over the 0.6AuCL at steady-state conditions at 493K under WGS feed (a) 2% 12CO + 7% H2O in Ar and (b) 2% 13CO + 7% H2O in Ar. The measured CO conversion was 26%. A spectrum of the same sample under 7% H2O /Ar at the same temperature was used as background.
The effect of switching the feed from 12CO to 13CO is shown in figure 6. Both the formate and carbonate bands shift to lower wavelengths as a result of the 13C containing reactant. The exchange between the 12C species and 13C species was completed after 30 mins. However the band at 2132 cm-1 remained unchanged following the introduction of 13CO. This is contrary to the expected red shift down ca. 50 cm-1 which suggests that this band is not associated with any carbonyl species. It was proposed that this band is due to an electronic transition occurring over reduced ceria [vii] .
Figure 7: Details of the formate and carbonate regions of the in situ DRIFT spectra obtained over of the 0.6 AuCL under 2% 12CO + 7% H2O in Ar (bottom spectrum) and 30 mins under 2% 13CO + 7% H2O in Ar at various temperatures (three top spectra). A spectrum of each sample under 7% H2O /Ar at the same temperature was used as background.
Figure 7 revealed that the exchange of carbonates and formates was only partially completed after 30 minutes. Exchanged varied between 48% and 95% within the temperature range investigated. Despite this the exchange of CO2 was completed within 20 minutes at all temperatures investigated.
Figures 8 and 9: (8) Relative evolution of the intensity of 12C-containing carbonate band, 12C-containing formate band and 13CO2 signals with time on stream at 428K under 2% 13CO + 7% H2O, following steady-state under 2% 12CO + 7% H2O.
(9) Relative evolution of the intensity of 12C-containing carbonate band, 12C-containing formate band and 13CO2 signals with time on stream at 493K under 2% 13CO + 7% H2O, following steady-state under 2% 12CO + 7% H2O
Figures 8 and 9 display the failure of formates and carbonates to exchange within the first 30 minutes of the reactions. The rates of exchange for both species are similar and appear to correspond to the rate of exchange of CO2 initially before subsequently exchanging much more slowly. As is expected, the carbonates and formates exchange much quicker in the reaction at 493 K. However the intensity of both species fails to reach zero within the time taken for CO2 to completely exchange between 12C and 13C.
The rate of exchange of CO2 can be approximated by a single exponential function contrary to that of both the formates and carbonates due to their nonuniform exchange. This led to the formation of a two species model of both slow and fast exchanging formates and carbonates.
Figure 10: Arrhenius plots of rate constant related to the exchange of the fast formates, the slow formates, the fast carbonates and the slow carbonates. The corresponding apparent activation energies are also reported.
The fast and slow decomposition of formates and carbonates were characterised by determination of their rates and associated activation energies. The fast formates were exchanged between 10 and 20 times faster than the slow formates depending on the reaction temperature. These slow formates exhibited a significantly higher activation energy than all the other surface species displayed in figure 9.
The carbonates showed similar variation between the rates of slow and fast species however their respective activation energies appeared similar.
The activation energies observed for these surface species do not correlate with the formation of CO2 calculated as 40.0 kJmol-1. This is significantly different and therefore suggests that the formation of CO2 is not the direct result of the decomposition of formates or carbonates.
Figure 11: (a) Upper spectrum: DRIFT spectrum observed over a Au-free Ce(La)O2 material at 373K in Ar impregnated with 0.125 wt% of formate deposited from sodium formate. Lower spectrum: in situ DRIFT spectrum observed over the 0.6 AuCL under 2% 13CO + 7% H2O in Ar. T = 458K. (b) Calibration plot relating the formate CH stretching band area of the Ce(La)O2 sample impregnated with various loading of sodium formates. The dotted line is associated with the in situ spectrum shown in (a).
The formate concentration was investigated by impregnating the Ce(La)O2 catalyst with sodium formate at various concentrations. The concentration of added formate was plotted against the corresponding area of the DRIFTS band. This was then used to determine the in situ formate concentration of a WGS reaction. The information obtained was then used to draw conclusions about the adsorptivity of the formate species. This information is displayed above in figure 6.
Table 1: Parameters representing the two-term exponential model of the formate and carbonate exchange curves.
Figure 12: Rate of CO2 production and rate of formate decomposition over the 0.6AuCL at three different temperatures under 2% 12CO + 7% H2O.
Figure 12 importantly illustrates the large difference between the rate of formate decomposition and rate of CO2 formation. The difference is almost an order of magnitude whereby CO2 formation is significantly faster.
Using this kinetic data, it is possible to conclude that formate decomposition does not directly lead to CO2 formation. The reaction must proceed through another step.
Figure 13: Schematic representation of the species populating the surface of the 0.6 AuCL catalyst during the water-gas shift reaction. It is suggested that the surface species close to the Au centres are the most reactive, while those further away are essentially spectators. The origin of the difference of reactivity is unclear.
The schematic above is used to describe a suggested explanation for the difference in decomposition rates of both formates and carbonates leading to the concept of slow and fast formates and carbonates. The fraction of the fast and slow species varies with temperature as indicated in table 1.
This was not investigated in this particular study by Meunier et al. however certain information was used to form a proposed explanation for this unusual occurrence. A separation of reactivity zones could be present near the active Au sites. The higher reactivity zone leads to faster decomposition whilst the lower reactivity zone causes slower decomposition.
Figure 14: Suggested reaction scheme taking place over the 0.6 AuCL catalysts during the water-gas shift reaction. The intermediate X could possibly be a species formed by reaction between CO adsorbed on Au and an oxide ion, as in a redox mechanism, or an adduct species, possibly a formate not resolved on the DRIFT spectrum.
The scheme above summarises a number of different events which occurred during the reaction. Clearly it was felt that the formation of CO2 occurred by an unknown intermediate with an activation energy corresponding to 40.0 kJmol-1. There was also evidence that formates were converted into CO2 however this was considered a minor pathway.
Another suggestion was that a significant amount of CO2(g) readsorbed onto the active sites as carbonates. The exchange of CO2 and both slow and fast carbonates had very similar activation energies.
Despite these findings this reaction process was considered highly complex as shown by the significant fraction of carbonates which were found not to exchange in the time taken to complete the formation of CO2. SSITKA was not used in this study due to the high basicity of the ceria. It was deemed that this technique would not provide unambiguous results.
The following conclusions were drawn by Meunier et al. as a result of the data expressed previously. The rate of exchange of the surface species as well as their respective absolute amounts was vital in forming conclusions regarding reaction intermediates. The lack of this information often leads to incorrect reaction pathways.
The rate of CO2 formation in the WGS reaction was found to be approximately 60 times that of both fast and slow formate decomposition. This suggests that the formate surface species were not significant within the main reaction mechanism.
Nonuniform reactivity of the surface formates and carbonates was observed. These species appeared to decompose both quickly and slowly in different instances. This characteristic could be fitted to a bimodal distribution containing both fast and slow exchange species.
The exchange curves and number of precursors during the reaction was partially disguised by a significant readsorption of the product CO2. SSITKA was therefore required to determine both concealed aspects of the reaction.
Chuang and Guzman used transient methods to perform a mechanistic investigation of heterogeneous catalysis [viii] . Their work includes a number of reactions which include the NO-CO reaction, ethylene hydroformylation and a brief study of generic photocatalytic reactions.was based around the NO-CO reaction which is a major R&D goal due to the stringent standard for NO emissions.
Figure 15: SSITKA coupled with infrared methods: (a) a differential reactor packed with catalyst particles, (b) a simplified reaction pathway with the transient responses of gaseous reactants, adsorbed species, and gaseous products.
A differential reactor, outlined in figure 15(a), operates at a constant temperature and pressure by feeding in the reactant fluid, R, which causes the production of the product, P, under steady state conditions. This apparatus coupled with the use of gas chromatography allows the production of rate data as well as determining the product distribution.
Transient methods can also be adopted while using a differential reactor. The inlet concentrations of the components of the reactant fluid can be altered either in a step change manner or by the use of pulse changes. These concentration variations can be monitored by mass spectrometry. This inlet concentration modulation was used by McDougall et al. as explained previously.
The transient methods, outlined in the introduction to this review, as well as the equipment and kinetic models, outlined in figure 15, were carried out on three reactions. Firstly the NO-CO reaction on Rh and Pd catalysts, a heterogeneous hydroformylation reaction and photocatalytic reactions were all studied.
Figure 16: The responses of CO and CO2 to a step change in the inlet concentration of 13CO during the NO-CO reaction on Rh/SiO2. The inset is a result of switching from 13CO back to 12CO.
Figure 17: Infrared spectra of adsorbed species during switching of the inlet flow from 12CO to 13CO in the NO-CO reaction on Rh/SiO2 catalysts.
The NO-CO reaction is of great economic importance due to the scarcity of Rh and Pd for use as catalysts and the environmental importance of limiting NO emissions. Therefore, it is vital that this reaction is studied and consequently improved.
The mechanism of this reaction has been previously determined (1451 reference) however very little information has been collected regarding the active sites and spectator species present in the reaction. This information is much more useful in designing selective and active catalysts.
Figure 17 displays a number of species. These include CO2 bands at 2358 cm-1, a Rh-NCO band at 2189 cm-1, geminal dicarbonyl bands at 2098 and 2035 cm-1, a NO+ band at 1913 cm-1 and a NO- band at 1763 and 1691 cm-1.
The presence of geminal dicarbonyl and NO- suggests that both Rh+ and Rh0 sites are present during reaction. Bands ca. 1831 cm-1 can be attributed to gaseous and neutral NO. There is also a band at 2298 cm-1 associated with Si-NCO suggesting that NCO migrates from Rh, where it is produced, to the SiO2 support.
Switching from 12CO to 13CO indicated that the geminal dicarbonyl containing 12C exchanged rapidly to 13C. Figure 3 elucidates that the time taken for the isotopic change of the geminal carbonyl is completed in 33s. Unfortunately this rapid exchange therefore does not allow the determination of adsorbed CO in this reaction. This particular problem is evident in a number of reactions including NO in the NO-CO reaction and CO in CO hydrogenation. The gradual exchange in intensity of the Rh-NCO band suggests that it is not involved in the formation of CO2.
Figure 17 also outlines the rapid exchange of gaseous CO with geminal carbonyl species on the surface. This was suggested as an adsorption-assisted desorption process as the rate of desorption of CO showed a strong dependence on the partial pressure of CO.
Figure 18: Selective enhancing and poisoning. (a) Addition of H2 into the NO-CO, (b) Addition of O2 into the NO-CO
Figure 18 illustrates the process of selective enhancing and poisoning of the catalyst. The addition of hydrogen enhances the catalyst whilst the oxygen addition causes poisoning. This technique is used to investigate the reactivity of surface species. In this case adsorbed NO was studied.
The added hydrogen reacts with adsorbed nitrogen and oxygen which leads to the depletion of the adsorbed linear NO on Pd0. Alternatively figure 18(b) outlines that the addition of oxygen caused accumulation of linear NO.
This evidence, coupled with the lack of exchange of linear CO, bent NO and bridged NO as illustrated by the selective enhancing and poisoning technique, leads to the conclusion that linear NO is an active adsorbed species. It is also clear that the reduced Pd0 is an active site for NO dissociation. Bent NO and bridged NO can be regarded as spectators.
The potential for improving the NO-CO reaction has therefore increased as result of the identification of the active catalytic site. Maintaining the Pd0 it's reduced state ensures the maximisation of potential active sites for the reaction to take place.
Figure 19: (a) Ethylene hydroformylation pathway (b) the reaction of adsorbed CO with ethylene/hydrogen on Rh/SiO2 catalyst.
The reaction indicated in figure 5(a) known as heterogeneous hydorformylation involves the reaction of an olefin with syngas to produce an aldehyde. It has been previously established that this reaction contains the following key steps. Firstly an adsorbed ethylene is hydrogenated to an adsorbed ethyl. An adsorbed CO is then inserted into the metal-carbon bond of the adsorbed ethyl species which forms an adsorbed acyl species. This acyl species is then hydrogenated forming propionaldehyde.
The insertion step of an adsorbed CO species was investigated by determining the nature of active sites involved in this step. Figure 5(b) shows the presence of linear CO on Rh0, bridged CO on two Rh sites and geminal dicarbonyl species. The addition of the olefin results in the reduction of the linear CO band demonstrating that this is the species involved in the insertion step which forms the aldehyde. This also suggests that the Rh0 sites are active in catalysing the reaction. This conclusion led to the use of sulphur and Ag in the hydroformylation as they 'break up' two fold adsorption sites therefore increasing the concentration of active Rh0 sites for the reaction to occur. This in turn increases the rate of reaction.
Figure 20: The transient response of 13CO and C2H513CHO to a pulse of 13CO in the 12CO feed during ethylene hydroformulation for varying C2H4 and H2 partial pressure.
Figure 21: (a) Illustration of the formation of photogenerated electrons which absorb infrared radiation and (b) difference infrared spectra of adsorbed ethanol with TiO2 catalyst during photocatalytic oxidation of ethanol.
Transient techniques employed to study surface adsorbed species can also be adopted to investigate photocatalysed reactions. These reactions are initiated by photogenerated electrons which can be followed in order to develop an understanding of their transient nature.
TiO2 catalysed oxidation of ethanol was used to study the role photogenerated electrons play in reactions. On H2O deficient TiO2, ethanol is oxidised to form acetic acid. Photogenerated electrons are also present and build up. This buildup of electrons, indicated in the IR between 1300 and 2500 cm-1, is unique to H2O deficient TiO2 therefore it is suggested that H2O accelerates the returning of photogenerated electrons to their respective positive 'holes'.
These positive holes are also suggested to be consumed by ethanol and contribute to the reaction pathway. These holes are unable to diffuse away from the catalyst surface and therefore are prone to consumption by surface species. This is often either ethanol or ethoxy species in this instance. The consumption of these holes leads to a buildup of photogenerated electrons where there is a lack of electron acceptors such as H2O.
This study is useful in outlining the usefulness of in situ IR to monitor photocatalytic reactions. However it is also clear that more work is required to understand it's full potential.
Chuang and Guzman express the value in the use of transient techniques in determining the nature of surface species and active sites of a heterogeneous catalyst. These transient IR methods were shown to identify the spectator species and active surface moieties as well as investigate the kinetic aspects of the reaction. Reaction rate and coverage data allowed verification of kinetic models and the respective parameters.
The monitoring of photogenerated electrons gives arise to the potential for studying photocatalysis. The use of these methods in such a manner is currently under investigation by the group.
The three groups using transient techniques in situ in order to monitor various reactions provide information which highlights the usefulness of this approach. These techniques allow a number of aspects of reactions to be known. Namely the kinetics of the reaction, including reaction rates and associated activation energies, as well as better understanding of the reaction pathway by identifying spectator and active surfaces species and catalytic sites.
The use of modulation of inlet concentration is key to the usefulness of using operando techniques. IR bands of various species often overlap which distorts the information leading the inconclusive results. However, monitoring species individually as the reactant concentrations are altered avoids this possibility leading to higher clarity of results.
SSITKA is an alternative technique which avoids the overlapping of IR bands. The use of different isotopes allows the reactants, products and intermediates of a reaction to be labelled and therefore followed by the use of mass spectrometry. Different intermediate species can then be determined which also leads to potentially developing a reaction pathway.
SSITKA also provides proof of mechanisms which is where kinetic data falls short. Despite the usefulness of kinetic data is developing reaction models, it cannot be used to prove mechanisms.
Whilst the techniques highlighted previously provide very valuable information, there are a number of limitations. This is illustrated by the work done by Meunier et al. where an activation energy of 40.0 kJmol-1 is found however this is not assigned to any particular intermediate species. Identifying species on a DRIFT spectrum is limited to only those which are resolved on DRIFT spectra.
As well as this, the exchanging between carbon containing compounds with different isotopes in SSITKA cannot always be used to determine the role of certain species. This is the case outlined in the Chuang and Guzman report whereby 12CO is replaced by 13CO rapidly. This disallows the possibility of investigating whether CO is a spectator or an active species.
Despite the limitations with the techniques used, the use of in situ operando and transient techniques are extremely useful. The limitations of these techniques are small relative to the amount of conclusive evidence of reaction kinetics and mechanisms provided.
In summary, the experiments described involve an extra level of instrumental complexity compared to that normally associated with conventional reaction testing of heterogeneous catalysts. This observation can in some way explain the limited number of examples of the use of infrared spectroscopic operando techniques to probe heterogeneously catalysed reaction systems. However, as demonstrated here, the information gleaned by the operando methodology is significant and can provide insight to the elementary surface processes not necessarily accessible via conventional micro-reactor studies. As a consequence, it is anticipated that the application of methods outlined above, as represented by the work of 3 leading groups, will proliferate more in contemporary heterogeneous catalysis research.