**1. Introduction**

Methane steam reforming (SR) is a leading reaction in syngas production with a wide employment on an industrial scale [1–6]. However, in order to allow high conversions of methane, very high temperatures (900 ◦C) are employed, and this requires feeding the reactor with a large amount of heat, which is provided by burning part of the methane reagen<sup>t</sup> in an external furnace. Therefore, in some conditions, lower operative temperatures (400–500 ◦C) may be advisable, for instance when very high H2/CO ratios are required, in what is called Low Temperature Steam Reforming (LTSR) [7–9]. In cases of excess of steam in the process feed, with the right catalyst the two SR and WGS (Water Gas Shift) reactions take place consequently in the same reactor [10,11]. This allows for a drastic decrease in CO

content with respect to the classical steam reforming reaction, providing an outlet gas with a high H2/CO ratio. Moreover, the temperatures employed are compatible with special applications such as use of microreactors, CO2 capture enhanced reforming, and Pd membrane reactors for pure hydrogen production [12–15]. In the latter case, the steam reforming occurs inside a tubular membrane from which hydrogen is separated. The employment of a membrane requires high operative pressures that reduce the methane conversion. However, the hydrogen removal from the retentate allows to increase the hydrogen yield and methane conversion with regard to a classical reactor, allowing to realize an effective process [11,12]. For these reasons, very active, selective, and stable catalysts are required to efficiently perform LTSR in a membrane reactor. In fact, these must be able to activate methane at low temperatures, thus improving the performances with respect to high active SR and CPO (catalytic partial oxidation) catalysts [16]. Moreover, active-phase stability is important for the application cited above, in order to minimize the operations related to catalyst replacement. For this reason, Rh was selected as active phase rather than the mostly used Ni. In fact, Rh was reported to have both higher activity and stability in steam reforming than other active phases [17–21].

The Rh4(CO)12 cluster [22] had been studied almost 30 years ago for catalyzing di fferent reactions, such as the hydrosilylation of isoprene, cyclohexanone, and cyclohexenone [23], or, more recently, the hydroformylation of cyclopentene co-promoted by HMn(CO)5 [24], all in homogeneous catalysis. Rh4(CO)12-derived catalysts, supported on Al2O3, MgO, and CeO2, had also been tested in 2001 in the CPO process for syngas production [25–27]. However, the CPO conditions and the related generated hot spot do not allow a simple comparison among catalysts. Furthermore, no tests on heterogeneous supported catalysts based on the Rh4(CO)12 cluster have been reported yet.

This work describes the preparation of Rh-based oxide-supported catalysts by exploiting the Rh4(CO)12 carbonyl cluster as source of active phase, as well as their performances on low-temperature methane steam reforming reactions, with the aim of developing a suitable catalyst for future membrane reactor operations. The Rh4(CO)12 cluster was deposited on two di fferent supports, namely Ce0.5Zr0.5 O2 and ZrO2, which were obtained by microemulsion synthesis. Notably, this method gave rise to an improved Rh-supported CeZr-oxide catalyst with respect to classical CeZr ones, in terms of activity and stability in the oxy-reforming reaction [28–31].

The e ffect of the presence of Ce in the support and its influence on the catalyst deactivation were also investigated. Noteworthy, no deactivation studies of Rh4(CO)12-derived catalysts for syngas production have been reported thus far.
