Temperature programmed reduction is a very simple experiment which is common-place in almost all catalytic and chemical laboratories. The setup for the reactor is very simple (figure 1.3) and consists of flowing a reducing gas (typically H2) over the surface of the catalyst (typically a metal oxide), at a set flow rate, and mea-suring the uptake of the gas as a function of time by recording the entrance and exit readings of the reactor. The reactor is heated at a constant rate, and as such TPR is considered a non-isothermal technique. Due to this linear heating,instead of plotting the thermogram as a function of time, it is often plotted as a function of temperature instead.
A TPR experiment can be expressed using the chemical equation 1.1:
n(MOx) + H2 −−→ (n-1) (MOx) + MOx−1+ H2O (1.1)
Where n is the number of moles of the metal oxide (MOx) present in the catalyst sample. Throughout the TPR thermogram the adsorption of H2 will increase as the temperature is increased. As the reduction process is able to overcome its activation energy barrier, the adsorption of H2 will rise, until it reaches a peak and then
Figure 1.3: Basic schematic of a TPR experiment.
decays back to the baseline as the material converts from the metal oxide (MOx) to it’s metallic or reduced state (MOx-1). A standard TPR thermogram can be seen in figure 1.4 which shows this rise and fall as a function of temperature.
As the TPR thermogram rises and falls back to the baseline we can assign a variable α using equation 1.2 which we define as the degree of reduction. Where Tm is the temperature at which the degree of reduction (α) is being measured, T0 and Tf
being the start and final temperatures of the thermogram respectively, and finally cH2 being the concentration of hydrogen being absorbed.
α(Tm) =
Tm
T0
cH2 dT
Tf T0
cH2 dT
(1.2)
Using equation 1.2 to calculate α for the TPR thermogram one can start applying current kinetic analysis.
Chapter 1 Kinetic Analysis and Modelling in Heterogeneous Catalysis
Figure 1.4: Thermogram for the reduction of CeO2 as a function of temperature at three heating rates: 5 K min-1, 10 K min-1, 15 K min-1
The idea for non-isothermal kinetic analysis first originated in the late 1950’s as a result of the development of various thermal based analytical methods. It was only then that technology had reached a point where accurate thermal and adsorption readings could be taken. After the results of these experiments were studied in depth it was theorised that these temperature programmed thermograms, during which some process is occurring (be it oxidation, reduction or desorption), contain kinetic information, and mathematical models were developed to explain these processes.[7–
10] Over the last 40 to 50 years these models have barely changed, with the majority still in use to this day,[11] but unfortunately their application has been sparse at best.
The majority of TPR analysis, particularly in catalysis, is rudimentary at best.
If one was to perform a sampling of the current literature they would see that for the majority of papers which contain temperature programmed experiments they simply study the location of the peaks and comment on any shifts that occur,[12]
or they look at quantitative data e.g. the quantity of H2 absorbed during a TPR thermogram.[13] When the kinetic analysis is seen in the literature it is often in a paper which in which the kinetic modelling, rather than the actual catalytic system,
is the main focus of the paper.[14, 15]The few papers in which the kinetic modelling has been applied to describe a system can be used to gain deep insight into the behaviour of a catalyst. One of the model papers for this would be the 2004 paper by Munteanu et al.[16] In this paper they use the kinetic modelling - in a similar method as is used in this thesis - in order to describe the influence gold has on the reduction behaviour of a complex catalyst (Au−V2O5/CeO2).
In the paper they took varying forms of the Au−V2O5/CeO2) catalyst and per-formed a TPR on each sample. They then used the kinetic constants calculated from the regression algorithm to understand the structure of the surface, and how the gold influences the reduction behaviour.
Figure 1.5: Parameters calculated from regression algorithm taken from [16]
It was found that when gold was included in the sample, the activation energy was found to be highly dependent on the degree of reduction, this, along with other parameters, showed that the gold was highly involved in the reduction process, and that the reduction process starts initially on the grain boundary with the gold particles, and then spreads out over the catalyst surface. It was also found that the gold changed the bulk characteristics of the material, facilitating the diffusion of O atoms though the lattice. The paper shows that the kinetic theory can be used
Chapter 1 Kinetic Analysis and Modelling in Heterogeneous Catalysis
to gain deep understanding of the catalyst material, which is invaluable in catalyst design.
The work outlined in this thesis is involved in attempting to automate the kinetic analysis of TPR experiments using the methods outlined in the literature. The kinetic analysis has been programmed into a "front-end" focused software, which is based around a central graphical user interface named CCI-TPR shown in figure 1.6.
Alongside automating the current analysis of TPR thermograms a new more robust methodology for calculating the various kinetic parameters from experimental TPR data has been developed. Using this new method the accuracy of the calculated kinetic parameters is increased greatly, meaning that the user can gain a deeper insight into the surface reduction of their catalytic material. Users can apply this new information to their catalytic system helping them create more efficient and better catalysts. With TPR being such a commonplace experiment to perform it
Figure 1.6: The CCI-TPR main graphical user interface with overlaying activation energy subplot
would be in the best interest of catalytic chemists to use these newly developed tools in order to greatly increase the quality and quantity of their data.