Pulse response analysis using the Y-procedure: A data science approach
Introduction
Catalysts are viewed as a fundamental solution to unlocking new energy sources that are sustainable, affordable, and resilient and are used in 80–90% of all industrial processes, critical to the production of fuels, plastics, agrochemicals, and pharmaceutics (Lim, 2016, Morgan et al., 2017). Historically, heterogeneous catalyst development has been dominated by trial-and-error methods and generally focuses on optimizing the chemical and physical characteristics of multi-component solids, which often requires years of research and development (Gleaves et al., 2010, Chorkendorff and Niemantsverdriet, 2005). Recent advances in surface science experimental methods and theoretical computational chemistry are providing greater molecular scale insights for catalyst development but the materials-pressure gap is still a concern for translating research knowledge to industry-relevant solutions under industrial operating conditions (Friend and Xu, 2017, Morgan et al., 2017, Vojvodic and Nørskov, 2015, Medford et al., 2015, Pérez-Ramírez and Kondratenko, 2007, Gleaves et al., 2010).
Current industrial catalysts are conventionally tested under steady-state conditions, which can only provide insights about global behavior, as low time resolution limits these approaches to the slow steps of a complex reaction mechanism (Morgan et al., 2017, Kalz et al., 2017). Using the continuously stirred-tank reactor (CSTR), the rate of substance change can be obtained with no a priori assumptions about the kinetic model, i.e., a model-free procedure (Temkin, 1979, Denbigh, 1952). This rate is determined as a difference between “in” and “out” advection fluxes per unit of the active catalyst surface under complete mixing conditions (Berty, 1984). However, transient kinetic methods offer additional and unique insights into how catalytic materials influence reaction rates and pathways (Gleaves et al., 2010, Redekop et al., 2011, Redekop et al., 2014). While the advantages of transient experiments are well accepted by the research community, the experimental sophistication and complex analysis techniques have prevented broader adoption.
The Temporal Analysis of Products, or TAP reactor system is a fully automated research instrument for precise time-resolved kinetic characterization of materials ranging from single metal particles to multi-component solids and is designed to inject a small gas pulse (combination of reactant and inert) into a microreactor containing a solid catalyst sample. The three-zone reactor configuration is most commonly used for TAP experiments (Constales et al., 2001a, Constales et al., 2001b, Constales et al., 2004, Constales et al., 2006). There are two principle operating modes – state-defining (a single-pulse does not alter the catalyst state) and state-altering (a sequence of pulses is used to gradually change the catalyst in a controlled manner). Observable quantities are the exit flux of gas species (reactants, inert, products) via mass spectrometer voltage readings.
A limiting case of the three-zone configuration is the thin-zone TAP reactor (TZTR) (Shekhtman et al., 1999) and has the following advantages: (Constales et al., 2017).
- •
In the Knudsen diffusion regime, intermolecular collisions can be neglected and gas transport is well-defined
- •
There is an insignificant change of solid active material during a single experiment
- •
The chemical composition of the active material is highly uniform in space
In a TZTR, gas concentration gradients across the catalyst zone can be neglected; further, diffusion and chemical reaction can be separated (Shekhtman et al., 1999). Under these conditions, non-steady-state kinetically model-free information can be extracted directly from exit flow data by utilizing of the Y-Procedure (Constales et al., 2017).
Currently, there are three methods to analyze thin-zone pulse response data:
- 1.
Model fitting (Kondratenko and Pérez-Ramírez, 2006)
- 2.
Method of moments (Shekhtman et al., 2003)
- 3.
The Y-Procedure (Yablonsky et al., 2007)
Model fitting requires a prioriassumed kinetic models and the method of moments has the disadvantage of averaging kinetic information over the entire flux. The Y-Procedure is a model-free approach that preserves transient intrinsic information across each pulse while expanding the theory beyond steady-state experiments.
Theoretical details of the Y-procedure have been described elsewhere (Yablonsky et al., 2007, Redekop et al., 2011). Our focus is to provide practical knowledge of the Y-procedure data science work flow including all of the pre-processing steps that are required such that it might be more broadly utilized.
Section snippets
Experimental data
The mechanism of adsorption was investigated over platinum (Good Fellow, purity) with a maximum particle size of 150 m. Data was collected using a commercial TAP-3 reactor system manufactured by Mithra Technologies, Inc. via a state-defining, consecutive pulse experiment (Gleaves et al., 1997). This work was designed to reproduce the experiment described by Redekop et al. (2011) in order to evaluate the veracity of data collected using different TAP installations. The data collected
Data science work flow
The data science work flow is shown in Fig. 2 with each activity explained in the following sections.
General methodology
In the previous section the data science work flow was described in detail for the following steps: data collection, data pre-processing and correction, diffusion coefficient estimation and finally Y-Procedure analysis. In general, any interpretation of results begins following data pre-processing and correction with simple qualitative flux analysis. In fact, many TAP experimental investigations have used such a logic based approach for comparison of pulse response shape and timing to garner a
Conclusion
The TAP reactor pulse response experiment and analysis aides the development of catalysts through efficient kinetic characterization. More specifically, the Y-Procedure is a revolutionary technique applied to TAP pulse response experiments, which enables experimentalists to derive information about the gas concentration, surface concentration, reaction rates, and kinetic information as a function of time within each pulse. This technique transforms the TAP reactor into a high throughput device,
Acknowledgements
Major support for this work was provided by the U.S. Department of Energy (USDOE), Office of Energy Efficiency and Renewable Energy (EERE), Advanced Manufacturing Office Next Generation RD Projects under contract No. DE-AC07-05ID14517. Additional support was provided through the INL Laboratory Directed Research Development (LDRD) Program under DOE Idaho Operations Office Contract DE-AC07-05ID14517. Accordingly, the U.S. Government retains a nonexclusive, royalty-free license to publish or
References (45)
- et al.
Role of intermediates in reaction pathways from ethene to hydrocarbons over h-zsm-5
Appl. Catal. A: General
(2017) - et al.
Oxidative coupling of methane–a complex surface/gas phase mechanism with strong impact on the reaction engineering
Catal. Today
(2014) - et al.
Multi-zone tap-reactors theory and application, I: the global transfer matrix equation
Chem. Eng. Sci.
(2001) - et al.
Multi-zone tap-reactors theory and application: Ii. the three-dimensional theory
Chem. Eng. Sci.
(2001) - et al.
Multi-zone tap-reactors theory and application, iii: multi-response theory and criteria of instantaneousness
Chem. Eng. Sci.
(2004) - et al.
Multi-zone tap-reactors theory and application, iv: ideal and non-ideal boundary conditions
Chem. Eng. Sci.
(2006) - et al.
Precise non-steady-state characterization of solid active materials with no preliminary mechanistic assumptions
Catal. Today
(2017) - et al.
Transient kinetics from the tap reactor system: application to the oxidation of propylene to acrolein
J. Catal.
(1995) - et al.
Tap-2: an interrogative kinetics approach
Appl. Catal. A: General
(1997) - et al.
Temporal analysis of products (tap)–recent advances in technology for kinetic analysis of multi-component catalysts
J. Mol. Catal. A: Chem.
(2010)