Heterogeneous Catalysis Hot Paper DOI: 10.1002/anie.201306899 Size-Dependent Phase Transformation of Catalytically Active Nanoparticles Captured In Situ** Nico Fischer, Brett Clapham, Theresa Feltes, Eric van Steen, and Michael Claeys* Abstract: The utilization of metal nanoparticles traverses across disciplines and we continue to explore the intrinsic size- dependent properties that make them so unique. Ideal nano- particle formulation to improve a processs efficiency is classically presented as exposing a greater surface area to volume ratio through decreasing the nanoparticle size. Although, the physiochemical characteristics of the nano- particles, such as phase, structure, or behavior, may be influenced by the nature of the environment in which the nanoparticles are subjected [1, 2] and, in some cases, could potentially lead to unwanted side effects. The degree of this influence on the particle properties can be size-dependent, which is seldom highlighted in research. Herein we reveal such an effect in an industrially valuable cobalt Fischer–Tropsch synthesis (FTS) catalyst using novel in situ characterization. We expose a direct correlation that exists between the cobalt nanoparticles size and a phase transformation, which ulti- mately leads to catalyst deactivation. Fischer–Tropsch synthesis (FTS) is a surface polymerization reaction producing a range of valuable hydrocarbon products from synthesis gas (syngas ; a mix of H 2 and CO) and with the industrial output of FTS increasing, so is the necessary research to improve the process. [3, 4] In case of supported cobalt nanoparticles, a major focus of current research is relating ideal activation to a distinct nanoparticle size. [5–7] Owing to the presence of both oxidizing and reducing compounds during FTS, similar to many other commercial reactions utilizing nanocrystallite catalysts, [8] an important alternative in studying nanoparticle size effects is exploring certain forms of deactivation that can also be crystallite size- dependent. For FTS, it is argued that kinetically water, the potentially oxidizing species, does not play a role, [9, 10] how- ever, numerous reports hypothesize it does promote deacti- vation. This includes a rather broad range of mechanisms including re-oxidation to CoO, the promotion of cobalt- support interactions such as the formation of cobalt alumi- nates, H 2 O aided sintering, or surface reconstruction. [11–14] It is through re-oxidation that we can begin to focus on crystallite size-dependence. Yet, this conditional phase transformation has only been studied theoretically [12] or only been proposed using indirect methods which do not monitor the oxidation of the cobalt directly, that is, with ex situ characterization techniques [14] with no solid conclusion to date. Cobalts sensitivity to re-oxidation by air is the leading obstacle to this research. To overcome this, we introduce two complementary in situ techniques developed in our labora- tory to study the catalyst nanoparticles. An advantage in studying ferromagnetic materials is that it is possible to distinguish the intrinsic properties, which occur in the nano- particles as their crystallite size and phase changes. Herein, we report a novel in situ magnetometer that is capable of monitoring the magnetic cobalt catalyst at high temperatures and pressures during activation (e.g. the reduction of Co 3 O 4 to the active metallic Co 0 phase) and reaction, thus, giving insight into the catalysts composition and crystallite size. [15] X-ray diffraction (XRD) is a common tool for materials science, but by supplying a new capillary cell reactor [16] we can monitor the crucial phase and crystallite size changes, which can occur rapidly, in operando and importantly in the presence of high water concentrations. Model-like cobalt catalysts supported on an industrially relevant alumina support material were studied with these techniques. The Co 3 O 4 nanoparticles were synthesized by the reverse micelle technique, which yields crystallites with a narrow size distribution as we previously reported. [17] The sample names reflect their reduced metallic cobalt crystallite sizes from XRD measurements (additional characterization results can be found in the Supporting Information, Table S1). Our initial experiments focused on in situ magnetic measurements (Figure 1 a) of the two catalysts, CAT 9.5 (5.0 wt.% Co) and CAT 4.5 (1.8 wt. % Co). Full details on the background and experimental details of the technique can be found in the Supporting Information. Briefly, owing to the magnetic properties of the cobalt species, at temperatures higher than 13 8C, the only magnetic cobalt species present is metallic cobalt, Co 0 (see Table S2). This situation allowed us to measure and monitor the degree of reduction (DOR) throughout high-temperature activation in H 2 and during the reaction (conditions listed in Table S3). This monitoring was mainly done by taking magnetic measurements at an external field strength of 20 kOe then at 0 kOe to assess the amount of metallic cobalt and the remnant magnetization, respectively. Only metallic crystallites with a size larger than a critical diameter (depending on conditions and crystallite phase 7– 20 nm) [4, 18] will retain a level of magnetization after removal [*] Dr. N. Fischer, B. Clapham, Dr. T. Feltes, Prof.Dr. E. vanSteen, Prof. Dr. M. Claeys Centre for Catalysis Research and c*change (DST-NRF Centre of Excellence in Catalysis), Department of Chemical Engineering University of Cape Town Cape Town (South Africa) E-mail: michael.claeys@uct.ac.za [**] This work is supported by the Centre for Catalysis Research at the University of Cape Town and c*change (DST-NRF Centre of Excellence in Catalysis). We also thank J. Macke, M. Wüst, and C. de Vries for their contributions to the work and Sasol R&D for support of the development of the in situ magnetometer. Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/anie.201306899. . Angewandte Communications 1342  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2014, 53, 1342 –1345