Influence of Supports on Chemical Looping Methane Reforming Activity of NiO-Based Oxygen Transfer Materials

Influence of supports on chemical looping methane reforming activity of NiO-based oxygen transfer materials

 

Introduction

Hydrogen is an energy carrier that can be used to efficiently store regenerative energy, and it fulfills the requirements of being the most environmentally friendly fuel for several applications. Although the steam methane reforming process is currently the major source of hydrogen production, its high energy demands  along with  increased CO₂ production make the overall process undesirable both from economic and environmental considerations. The Sorption Enhanced Chemical Looping Steam Methane Reforming (SE-CL-SMR) is an emerging technology that can simultaneously minimize the aforementioned drawbacks and produce pure hydrogen. The stable and active oxygen transfer materials (OTMs) and CO₂ sorbents are the major requirements for a viable SE-CL-SMR process. NiO supported OTMs have been promising oxygen carriers for chemical looping methane reforming reactions due to their superior redox reactivity, catalytic performance, stability and negligible volatility. In this study we investigated the influence of supports on redox activity of OTMs in the absence of steam in a thermo-gravimetric analyzer (TGA) unit, and in conventional and chemical looping steam methane reforming reactions (CLR) in a fixed bed flow unit.

Experimental

Four NiO-based OTMs with 40wt% NiO loading were synthesized by wet impregnation using commercial supports: (SiO₂, TiO₂, Al₂O₃, ZrO₂). Both pre- and post- characterizations of the OTMs such as BET, XRD, H₂-TPR, TPO and TEM were performed. Twenty redox cycles, methane reduction at 650°C/air oxidation at 800°C, were conducted in a TGA unit connected to a mass spectrometer (MS). Conventional steam reforming activity of the pre-reduced OTMs was tested at 650°C in a bench scale fixed bed flow unit with a steam/methane (S/C) ratio of 3 and GHSV=10⁵h^-1. In a second step, NiO/ZrO₂ and NiO/Al₂O₃, which exhibited satisfactory reforming activity and stability, were tested under CLR conditions. The materials were initially exposed to CH₄/steam in their oxidized state at 650°C for 1h (S/C=3). At the end of reforming stage, metallic Ni was re-oxidized with air at 850°C for 15min and the cycle was then repeated.

Results and discussion

BET surface area of the 40%NiO-OTMs decreased after loading NiO compared to the pure supports. XRD confirmed the presence of NiO phase in all OTMs, and strong metal-support interaction compounds (NiAl₂O₄ and NiTiO₃) were identified in Al₂O₃- and TiO₂- supported OTMs. During the 20 redox cycles in TGA, all OTMs except 40%NiO/g-Al₂O₃ attained very high reduction degrees (94% - 100%) (Fig. 1a). Degree of reduction (RD, %) of NiO/Al₂O₃ slowly increased up to cycle 11 reaching a maximum and steady value of 86%. This relatively low reducibility is mainly attributed to the strong metal-support interaction, as identified by XRD. In order to investigate the unusual trend in the reduction degree of NiO/Al₂O₃, temperature programed reduction (TPR) was performed for fresh and used (after 10^th redox cycles) OTM in the TGA (not shown). Results confirmed the decomposition of NiAl₂O₄ and increase in the NiO/NiAl₂O₄ ratio with cycles, indicating that re-crystallization of metallic Ni after reduction in methane weakens the alumina-nickel oxide interaction. All four OTMs underwent complete re-oxidation during the oxidation step of the cycling experiments, reaching the same oxidation degree level of 100% in all cycles. The H₂, CO, CO₂ and H₂O were the products detected during CH₄ reduction, whereas CO₂ was the only oxidation product detected by MS.

Figure 1 The variation in (a) the reduction degree and (b) the amount of carbon with cycles

The amount of carbon deposited during reduction steps, calculated from the observed weight gain in TGA, is shown in Fig. 1b. NiO/Al₂O₃ OTM exhibited a volcano shape curve which is possibly caused by changes in catalyst structure (gradual NiAl₂O₄ reduction) with cycles (Fig. 1b). For all OTMs trends in hydrogen formation (not shown) followed the same trend as C-deposition, indicating that there is a relationship between deposited carbon and H₂ formation during the reduction step. Data analysis showed that hydrogen production occurs primarily via CH₄ split. The amount of carbon deposition and hydrogen production was higher during initial cycles on Al₂O₃ and ZrO₂ OTMs, followed by decrease in subsequent cycles. This indicates deactivation of these OTMs towards the CH₄ decomposition reaction.

In order to investigate the nature and types of carbon deposited on OTMs by methane decomposition during the reduction cycles, TPO and TEM analysis were performed. TPO characterization of all OTMs after the first methane reduction step in TGA showed a single CO₂ peak centered at ~500⁰C – 550⁰C indicating the presence of one type of carbon on the surface. However the TPO profile of NiO/Al₂O₃ after the methane reduction step of the fourth cycle indicated the presence of two carbon species, oxidized to CO₂ at 575⁰C and 700⁰C respectively. In order to obtain further insights into the nature of the deposited carbon species, the samples were examined with transmission electron microscopy. Both graphitic and filamentous carbon were observed on Al₂O₃ and ZrO₂ supported OTMs. TEM image of Al₂O₃ supported OTM after the first cycle is shown in Fig. 2 for illustration. On the other hand, only graphitic carbon was observed on the SiO₂ and TiO₂ supported OTMs.

Experiments under conventional reforming conditions (with pre-reduced OTMs) showed a satisfactory performance of NiO/Al₂O₃ and NiO/ZrO₂ at 650°C, with a high methane conversion and less than 8% deactivation after 10 h on stream. On the other hand, NiO/SiO₂ and NiO/TiO₂ were found to be inactive, with initial CH₄ conversion less than 12%. The two most promising materials were then tested under chemical looping reforming conditions for 20 consecutive steam methane reforming/air oxidation cycles. Methane conversion as a function of cycles is shown in Figure 3. NiO/ZrO₂ exhibited a satisfactory activity with initial CH₄ conversion around 80% and less than 2% deactivation after 20 cycles, corresponding to 20h exposure to CH₄/steam. Initial CH₄ conversion on NiO/Al₂O₃ OTM was similar (~79%), but it deactivated more rapidly, leading to a CH₄ conversion of 59% in the 20^th cycle (Fig. 3). This loss of activity can be attributed to the lower hydrothermal stability of alumina compared to zirconia under the steam reforming conditions.

Figure 2 TEM image of Al2O3 OTM after 1st reduction cycle

Figure 3 CH4 conversion in chemical looping reforming experiments for NiO/ZrO2 and NiO/Al2O3  (Reduction-Reforming: 650⁰C, 1h, S/C = 3,  GHSV =100000h-1; Oxidation: 800⁰C, 15min, Air)

Conclusions

Support was found to have a significant effect on performance of NiO oxygen carriers in chemical looping steam methane reforming. All investigated OTMs supported on silica, alumina, zirconia and titania displayed satisfactory activity and high stability during 20 methane reduction/air oxidation cycles in TGA experiments. The 40%NiO/Al₂O₃ OTM showed unusual behavior compared to the other OTMs due to gradual reduction of metal-support compounds to more easily reducible species with cycles. Carbon deposition in the absence of steam in the TGA occurs primarily via CH₄ decomposition on reduced Ni. Only graphitic carbon was observed on SiO₂- and TiO₂- OTMs, whereas both graphitic and filamentous carbon were observed on Al₂O₃- and ZrO₂- OTMs. Testing of all OTMs as reforming catalysts in conventional SMR showed that NiO on titania and silica is inactive, while alumina and zirconia supported OTMs had high activity and stability characteristics. The latter two materials were tested in the chemical looping methane reforming process, where they exhibited promising performance. NiO/ZrO₂ showed good activity with initial CH₄ conversion around 80% and less than 2% deactivation after 20 cycles. NiO/Al₂O₃ had also a high initial CH₄ conversion, but had higher deactivation rate with cycles, probably due to a lower hydrothermal stability of alumina.

ACKNOWLEDGMENTS

This work was made possible by NPRP grant 5–420–2–166 from QNRF (member of Qatar Foundation). The statements made herein are solely the responsibility of the authors.