Kinetics of NiO reduction by H2 and Ni oxidation at conditions relevant to chemical-looping combustion and reforming

https://doi.org/10.1016/j.ijhydene.2014.03.161Get rights and content

Highlights

  • Holistic analysis of NiO reduction and Ni oxidation at chemical-looping conditions.

  • NiO reduction proceeds via formation and growth of nuclei (Avrami–Erofe'ev model).

  • Ni oxidation follows solid-state kinetics dominated by geometrical contraction.

  • All the available data can be represented by the same type of solid-state kinetics.

Abstract

Twenty solid-state kinetic models, including mechanisms dominated by reaction order, geometrical contraction, diffusion, nucleation, random pore growth, as well as the more flexible Šesták–Berggren and Prout–Tompkins models, are reviewed and applied to describe the reduction of unsupported and supported NiO by H2 and the oxidation of supported Ni by air. In an effort to address the inconsistencies in the literature regarding the suitability of each mechanism to describe the kinetics of Ni-based oxygen carriers in chemical-looping, all the models are compared against experimental data from the literature and with in-house experiments at conditions relevant to chemical-looping combustion and reforming with Ni-based oxygen carriers. A statistical approach to compare models of varying fidelity is employed, involving the Akaike Information Criterion and the F-test. The effect of temperature on the selection of the best-suited model is investigated, supplemented by experimental evidence from X-ray diffraction and scanning electron microscopy analyses. This work reveals that unsupported and supported NiO reduction can be described by nucleation and nuclei growth models. The oxidation kinetics of supported Ni is well-predicted by geometrical contraction models. Calcination and high-temperature treatment during oxygen carrier synthesis are shown to significantly affect the reaction kinetics of the oxygen carrier.

Introduction

Research has shown that nickel oxide (NiO) is a promising oxygen carrier (OC) for chemical-looping combustion (CLC) [1]. Therefore, understanding of the intrinsic kinetics of NiO reduction is crucial for designing efficient CLC systems. Although a breadth of literature exists on the gas–solid kinetics of NiO reduction and Ni oxidation, most studies are not applicable to CLC conditions. The published kinetics is limited to the use of fresh or pre-treated particles, while in the CLC environment the solid particles undergo repeated redox cycles. Moreover, the experimental conditions, e.g., temperature, particle size and gas composition, are usually different than those required for CLC [2]. Finally, the kinetics of NiO reduction and Ni oxidation is often derived from free bulk NiO; whereas, in CLC NiO is dispersed on a support, which increases the active surface area and mechanical strength of the oxygen carrier. Thus, limited information can be extracted from the available reactivity data to determine NiO/Ni reactivity with certainty at CLC-relevant conditions, which is a necessary prerequisite for design purposes.

Generally, the detailed kinetic mechanisms involve a group of individual chemical steps, by which a reactant is converted into a product. However, the information on chemical steps is usually difficult to obtain. Therefore, in solid-state kinetics, the kinetic mechanism study usually involves application and verification of a reasonable model to the experimental data. Ideally, the selection of a best-suited solid-state reaction model is supported by other complementary techniques, such as microscopy, X-ray diffraction (XRD), scanning electron microscopy (SEM), etc. [3]. Mechanisms for solid-state reactions include reaction-order models (F), geometrical contraction models (R), diffusion-limited models (D), nucleation models (Avrami–Erofe'ev, AE) [4], [5], [6], [7], the random pore model (RPM) [8], and the Šesták–Berggren (SB) [9] and Prout–Tompkins (PT) [10] models. Table 1 presents all these models, as they have been adopted in this work, along with their naming and formulas. The principal assumptions and physical meaning of these models are discussed briefly in Section ‘Solid-state kinetic model and mechanisms’.

Table 2 summarizes the relevant kinetic studies on NiO reduction and Ni oxidation in the literature. The most significant observations are discussed here to illustrate the lack of consistency in previous analyses. Janković et al. [11], [12], [13], [14] used the SB model to describe the reduction of NiO in a thermogravimetric analyzer (TGA). Richardson et al. [15], [16] used in-situ hot-stage XRD and observed that the R3 model cannot describe the reduction of NiO. By examining NiO powders in an environmental transmission electron microscope (ETEM) in the presence of H2, Jeangros et al. [17] proposed that the AE model with the rate of nuclei growth as the rate determining step describes the experimental data accurately. Plascencia and Utigard [18] concluded that reduction kinetics of Sinter 75 and Tokyo NiO highly depends on the morphological features and the R3 model is not suitable, based on SEM observations. Instead, they proposed a mixed controlling mechanism with the R and D models as more suitable. Optical microscopy observations led to the conclusion by Utigard et al. [19] that the R3 model represents the kinetics of the reduction of Goro NiO. By comparing the differences in the reduction behavior of bulk NiO and silica-supported NiO, Syed-Hassan and Li [20], [21] related the incubation period of the Ni reduction with the formation of radical species. Lee and Kim [22] studied the kinetics of H2 reduction of nanocrystalline NiO, concluding to the dominance of reaction control at lower temperatures, whereas diffusion control dominates at higher temperatures, which was postulated to be the result of structural changes with temperature [22]. Erri and Varma [23] published a kinetic study of NiO and NiO/NiAl2O4 and concluded that the inconsistencies in the literature value of the activation energy are due to a lack of understanding and consideration of diffusional effects.

Extensive studies on the reduction kinetics of supported NiO have also been published (Table 2), among which studies on Al2O3 (α- and γ-Al2O3) are prominent. Dueso et al. [24], [25] selected the R3 and D4 models for Ni reduction and oxidation kinetics of NiO and NiAl2O4 supported on Al2O3. Ryu et al. [26], [27], [28] investigated Ni/bentonite oxidization and NiO/bentonite reduction kinetics, concluding to the superiority of R3 and D4 models for reduction and oxidation, respectively. The R3 model was also found satisfactory in studies by Ishida et al. [29], [30], [31], Mattisson et al. [32], [33], Moghtaderi and Song [34], Chen and Shiue [35] and Abad et al. [36] for the reduction of NiO on various supports. However, it was suggested by Adánez [2] that the R3 model is not suitable for describing the reduction of oxygen carriers of small size and high porosity. Also, Hossain et al. [37], [38], [39], [40] found that their experimental data can be better matched by the nucleation and nuclei growth model with a random nucleation mechanism (the AE1 model). The AE1 model was also selected by Son and Kim [41] to fit their NiO reduction data. Readman et al. [42] used in-situ powder XRD to study NiO supported on NiAl2O4, showing that the NiAl2O4 support remained inert during the redox process.

In summary, the lack of agreement between the various analyses is striking. The variability of the rate mechanisms adopted could be a consequence of variations in the experimental conditions or in the morphology of the Ni/NiO samples. Kruggel-Emden et al. proposed new empirical correlations for the reduction reactions of CaSO4 and Mn3O4 and oxidation reactions of Cu, Ni, and MnO, which achieved better fits compared to the common kinetic models [43]. The present work aims to analyze statistically and experimentally solid-state models that describe the reduction and oxidation of unsupported and supported NiO and Ni, which was not studied as comprehensively by Kruggel-Emden et al. [43]. In Sections ‘Solid-state kinetic model and mechanisms’ and ‘Model discrimination and identification methods’, the methodologies used for modeling and statistical discrimination are briefly discussed. Section ‘Reduction of unsupported NiO’ statistically analyzes solid-state models applied to describe H2 reduction of unsupported NiO of different origins and morphology from the literature. In Section  ‘Reduction of supported NiO’, reduction experiments of supported NiO by H2 from the literature and in-house TGA tests are examined. The best solid-state reaction model is selected for each case through statistical comparison of the capability of each model to match the experimental data. The effect of temperature in the range of 600–950 °C on selected kinetic models is investigated, supported experimentally by scanning electron microscopy (SEM) and in-situ hot-stage XRD. Analysis of the oxidation kinetics of supported Ni is presented in Section ‘Oxidation of supported NiO’.

Section snippets

Solid-state kinetic model and mechanisms

Fig. 1 presents the schematic description of the models of Table 1; commonly categorized as reaction-order (F), geometrical contraction (R), diffusion (D), and nucleation (Avrami–Erofe'ev, AE). The same ordering and categorization was followed in Table 1. In summary, the reaction order-based models (F) assume a homogeneous reaction process. Mechanisms R2 and R3 of Table 1 are forms of the classic shrinking core model with phase-boundary reaction control of different geometries [44]. The R2

Model discrimination and identification methods

Common approaches of evaluating the kinetic parameters of a reaction include isoconversion and isothermal reactivity analysis methods [14]. The former encompass the stationary point method [48], Kissinger's method [49], Friedman method [50], Flynn–Wall–Ozawa method [51], [52], Kissinger's–Akahira–Sunose (KAS) method [53] and Málek's method [54]. In this study, the isothermal method was chosen because CLC is typically operated within a rather narrow temperature range [2]. Additionally, as shown

Results and discussion

This study uses 20 different solid-state reaction models (Table 1) to match experimental data obtained from the literature and this work. The models in Table 1 include 17 one-parameter models, 2 two-parameter models (RPM and AEn) and 1 three-parameter model (SB). Reduction data of 10 NiO materials (unsupported and supported) by H2 and oxidization data of 4 Ni based oxygen carriers by air are analyzed. The quality of fit (RSS) of all the models against the experimental data and their

Conclusions

The summary of the selected solid-state kinetic models for the reduction of supported and unsupported NiO by H2 and the oxidation of supported NiO by air is shown in Table 19. On the basis of the Hancock and Sharp analysis, as well as model fitting methods the two-parameter Avrami–Erofe'ev model with n ranging from 0.9 to 1.6 is shown to be the most successful in representing reduction kinetics of supported and unsupported NiO. Despite the remarkable inconsistency in previous analyses, the

Acknowledgments

This material is based upon work supported by the National Science Foundation under Grant No. 1054718. Support by W.R. Grace & Co. by providing the Al2O3/SiO2 matrices is gratefully acknowledged.

References (82)

  • T.A. Utigard et al.

    Reduction kinetics of Goro nickel oxide using hydrogen

    Chem Eng Sci

    (2005)
  • S.S.A. Syed-Hassan et al.

    NiO reduction with hydrogen and light hydrocarbons: contrast between SiO2-supported and unsupported NiO nanoparticles

    Appl Catal A Gen

    (2011)
  • C. Dueso et al.

    Reactivity of a NiO/Al2O3 oxygen carrier prepared by impregnation for chemical-looping combustion

    Fuel

    (2010)
  • C. Dueso et al.

    Reduction and oxidation kinetics of nickel-based oxygen-carriers for chemical-looping combustion and chemical-looping reforming

    Chem Eng J

    (2012)
  • T. Mattisson et al.

    Chemical-looping combustion using syngas as fuel

    Int J Greenh Gas Control

    (2007)
  • M.M. Hossain et al.

    Nickel on lanthanum-modified γ-Al2O3 oxygen carrier for CLC: reactivity and stability

    Catal Today

    (2009)
  • M.M. Hossain et al.

    Chemical-looping combustion (CLC) for inherent CO2 separations — a review

    Chem Eng Sci

    (2008)
  • K.E. Sedor et al.

    Reactivity and stability of Ni/Al2O3 oxygen carrier for chemical-looping combustion (CLC)

    Chem Eng Sci

    (2008)
  • H. Kruggel-Emden et al.

    A comparative study of reaction models applied for chemical looping combustion

    Chem Eng Res Des

    (2011)
  • M.E. Brown

    The Prout-Tompkins rate equation in solid-state kinetics

    Thermochim Acta

    (1997)
  • J. Málek

    The kinetic analysis of non-isothermal data

    Thermochim Acta

    (1992)
  • L. Han et al.

    Heterogeneous modeling of chemical-looping combustion. Part 1: reactor model

    Chem Eng Sci

    (2013)
  • A. Abad et al.

    Mapping of the range of operational conditions for Cu-, Fe-, and Ni-based oxygen carriers in chemical-looping combustion

    Chem Eng Sci

    (2007)
  • P. Cho et al.

    Comparison of iron-, nickel-, copper- and manganese-based oxygen carriers for chemical-looping combustion

    Fuel

    (2004)
  • J. Málek

    The applicability of Johnson-Mehl-Avrami model in the thermal analysis of the crystallization kinetics of glasses

    Thermochim Acta

    (1995)
  • J.W. Christian

    Chapter 12 – formal theory of transformation kinetics

  • Z.J. Yan et al.

    On the crystallization kinetics of Zr60Al15Ni25 amorphous alloy

    J Alloys Compd

    (2004)
  • Z. Zhou et al.

    Model-based analysis of bench-scale fixed-bed units for chemical-looping combustion

    Chem Eng J

    (2013)
  • P. Gayán et al.

    NiO/Al2O3 oxygen carriers for chemical-looping combustion prepared by impregnation and deposition–precipitation methods

    Fuel

    (2009)
  • S. Vogel et al.

    In-situ investigation of the reduction of NiO by a neutron transmission method

    Mater Sci Eng A

    (2002)
  • H. Rashidi et al.

    Application of random pore model for synthesis gas production by nickel oxide reduction with methane

    Energy Convers Manag

    (2013)
  • A. Khawam et al.

    Solid-state kinetic models: basics and mathematical fundamentals

    J Phys Chem B

    (2006)
  • M. Avrami

    Kinetics of phase change. I general theory

    J Chem Phys

    (1939)
  • M. Avrami

    Kinetics of phase change. II transformation-time relations for random distribution of nuclei

    J Chem Phys

    (1940)
  • M. Avrami

    Granulation, phase change, and microstructure kinetics of phase change. III

    J Chem Phys

    (1941)
  • B.N. Erofe'ev

    Generalized chemical kinetics equation and its application to reaction with solid participants

    Dokl AN SSSR

    (1946)
  • S.K. Bhatia et al.

    A random pore model for fluid-solid reactions: I. Isothermal, kinetic control

    AIChE J

    (1980)
  • E.G. Prout et al.

    The thermal decomposition of potassium permanganate

    Trans Faraday Soc

    (1944)
  • Q. Jeangros et al.

    Reduction of nickel oxide particles by hydrogen studied in an environmental TEM

    J Mater Sci

    (2012)
  • S.S.A. Syed-Hassan et al.

    Effects of crystallite size on the kinetics and mechanism of NiO reduction with H2

    Int J Chem Kinet

    (2011)
  • J. Lee et al.

    Synthesis and related kinetics of nanocrystalline Ni by hydrogen reduction of NiO

    Mater Trans

    (2001)
  • Cited by (89)

    • Economic model predictive control for packed bed chemical looping combustion

      2024, Chemical Engineering and Processing - Process Intensification
    View all citing articles on Scopus
    View full text