Bifunctional Characteristics of Al 2 O 3 supported Ni in the HI Decomposition of Sulfur-Iodine Process

The Sulfur-Iodine process is in need of catalytic reactor for HI decomposition because the HI decomposition reaction rate is very slow. Nickel as an alternative catalyst for platinum was investigated in this study. Al2O3 supported Ni catalysts were prepared by impregnation method. Ni amounts loaded over Al2O3 were in the range of 0.1~20 wt. %. HI decompositions were carried out in the temperature range of 573 ~ 773 K using the fixed-bed quartz reactor. The difference of catalysts before and after the reaction was analyzed using BET, CO/H2 chemisorption, XRD, XRF and SEM. It was confirmed by XRD and SEM-EDX analysis that Ni was converted to NiI2 during the HI decomposition. Catalyst deactivation due to the formation of NiI2 leads to a reduction of HI conversion. Although Ni of catalyst converted to NiI2, HI decomposition with low loading (up to 3 wt. %) catalyst showed a little decrease of HI conversion. However, with more than 5 wt. % catalyst, the initial HI conversion was considerably decreased. In the particular case of 20 wt. %, the initial conversion was increased close to 60 %, which is higher than 20 % as an equilibrium conversion at 723 K. These results showed that Ni had not only a catalytic function for HI decomposition, but also function as a sorbent to absorb I2 produced from HI.


Introduction
Sulfur-Iodine cycle is the promising thermochemical water-splitting process to produce large amounts of hydrogen [1].Sulfur-Iodine hydrogen production process consists of the following three chemical reactions: HI decomposition: 2HI ֖ H 2 + I 2 (573 ~ 773 K) ( Bunsen reaction (Eq.( 2)) is a water splitting reaction of producing H 2 SO 4 and HI.HI and H 2 SO 4 can be separated by specific gravity difference.In the H 2 SO 4 decomposition process (Eq.( 3)), H 2 SO 4 is decomposed into H 2 O, SO 2 and O 2 .Products of H 2 SO 4 decomposer are then fed back to Bunsen process.H 2 and I 2 are produced by thermal decomposition of HI (Eq.( 1)).I 2 is returned to Bunsen process as complexes of HIx•yH 2 O [2].In the temperature range of 600 ~ 800 K, HI conversion is around 20 % in the equilibrium state.In addition, the rate of HI decomposition reaction is very slow without a catalyst [3].The platinum as a catalyst for the HI decomposition has been mainly studied [4].Ni catalysts supported on various supports have been investigated for the cost savings resulting from alternating of noble metal.However, the life time of Ni supported catalyst was not good while Pt supported catalysts showed very good performance.Favuzza et al. [5] reported that deactivation of Ni-based catalyst was due to NiI 2 formation.In the present work, Ni catalysts supported on Ȗ-Al 2 O 3 were investigated.To identify the optimum loading amounts of Ni, Ni catalysts were prepared in the range of 0.1~20 wt.% by impregnation method.To compare the activities of catalysts, HI decompositions were conducted in the controlled conditions to be below theoretical equilibrium conversion.

Preparation of Catalysts.
The Ni supported catalysts on Ȗ-Al 2 O 3 were prepared by impregnation method (Fig. 1).As a support, Ȗ-Al 2 O 3 (Alfa-aesar, 99 %) was used.Al 2 O 3 particles in the range of 300 ~ 450 ȝm were obtained through pulverizing-sieving.The Al 2 O 3 particles were thermally treated to remove impurities at 973 K in the air atmosphere.Nickel(II) nitrate hexahydrate (Ni(NO 3 ) 2 ā 6H 2 O, Sigma-Aldrich, 98.5%) as Ni precursor was used.The Al 2 O 3 particles were wetted in Nickel nitrate solution using a rotary vacuum evaporator (BUCHI, R-210).The wetted particles were dried at 373 K for 24 h and then calcined at 973 K for 6 h in 5 vol.% H 2 /Ar.

Characterization and Activity Test of Prepared Catalysts.
Surface area and pore size distribution were measured by BEL-sorp mini.Metal dispersion and specific metal surface area were measured by BELCAT-B.10 vol.% CO/He and 5 vol.% H 2 /He were used as absorbing gas.The XRD (DMAX-2500, Rigaku) and SEM (S-4800, HITACHI) analysis were carried out to identify the physical and chemical change of catalysts before and after the activity test.
Activity test apparatus (Fig. 2) consisted of the evaporator of HI acid solution, the quartz tubular reactor with catalyst, traps and gas chromatograph.Activity tests were carried out with the vertical fixed-bed quartz reactor (9 mm inside and 13 mm outside diameter).Catalysts were used about 0.4 g (catalyst bed height: about 11 mm).The reaction temperature was kept at 723 K using heating jacket.HI aqueous solution (56 wt.%, Kanto) was fed into the evaporator in 0.2 mL/min using a peristaltic pump (Cole-Parmer, Masterflex L/S).Ar (99.999 %) as the carrier gas was fed in 50 mL/min into the quartz reactor.
The reacted gas was cooled to room temperature to trap I 2 , H 2 O and unreacted HI as soon as it was passed through the reactor.The resulting gas was then passed through the scrubber with the 2 M KI solution, in order to identify H 2 amount using the gas chromatograph (7890A, Agilent Technologies).

Results and Discussion
Specific surface area (BET) was measured before and after loading Ni on the support.BET of Ni loaded catalyst was lower than that of support.BET of Al 2 O 3 as purchased was 218 m 2 /g.BET of Ni loaded catalysts was decreased owing to clogging of pores of support with increasing of loading amount of Ni.BET of Ni loaded catalyst decreased down to 128 m 2 /g when the Ni loading amount was 20 wt.%.
Metal dispersions were measured by following CO and H 2 chemisorption, respectively.Metal dispersion values detected by CO chemisorption showed fluctuating results.CO adsorption on supported Ni is considerably more complex, the stoichiometry of which varies with equilibrium pressure, temperature, metal crystallite size, and metal loading [6].Therefore, CO adsorptions were not accepted as results in the present work.The table shows results measured by H 2 chemisorption.Metal dispersions were not measured when the Ni loading amount was less than 3 wt.%.The Ni dispersion was decreased with increasing the Ni contents of the catalyst.The evaluation of catalytic activity was performed in the temperature range of 573 ~ 773 K.The HI conversions were increased with increasing the reaction temperature [7].Fig. 3 represents the results at 723 K graphically with Favuzza's result at 773 K [5].The gas hourly space velocity was controlled to 34750 h -1 in order to be below than theoretical equilibrium conversion.The HI conversions were linearly increased with increasing Ni contents of catalyst up to 3 wt.%.However, it showed the similar results between 3 and 10 wt.% of Ni/Al 2 O 3 catalysts.The decomposition results were less than Favuzza's because of the high GHSV and the low reaction temperature.It was necessary to note the variation in the onset of decomposition.It was thought that the high conversion in the onset of the reaction was occurred since Ni catalyst had the bifunctional characteristics of the sorbent of I 2 produced from HI decomposition, as well as the catalyst.These effects took place in the case of the high Ni contents of the catalyst.The another concern was that NiI 2 had the catalytic activity in the HI decomposition.
Fig. 4 shows the XRD patterns of 10 wt.% Ni/Al 2 O 3 before and after the HI decomposition.The Ni and Al 2 O 3 peaks were detected in the catalyst before the reaction.In the catalyst after the HI decomposition, it was found that most of the Ni was converted to NiI 2 .Therefore, the HI decomposition in the end of the reaction could be thought as results due to the catalytic properties of NiI 2 .In order to determine the characteristics of Ni as a sorbent and NiI 2 as a catalyst, HI decomposition was conducted with the high Ni contents (20 wt.% Ni/Al 2 O 3 ) in the condition of low GHSV of about 4,630 h -1 .HI conversion was increased close to 60 % in the onset of the reaction and then decreased to 20 %.HI decomposition was maintained in the equilibrium state of 20 % after 50 minutes.When H 2 or I 2 is removed, the conversion rate exceeding 20 % that is the theoretical equilibrium conversion value occurs.Hence, the conversion rate of 60 % was caused by removing I 2 from the product gas.As a result, the Al 2 O 3 supported Ni metal in the HI decomposition showed the characteristics of a catalyst as well as the sorbent.It was also revealed that the NiI 2 formed from HI decomposition had the catalytic activity.

Conclusion
In this study, HI decompositions were performed to identify characteristics of Al
2 O 3 supported Ni catalyst.The HI conversions were increased with increasing the reaction temperature in the temperature range of 573 ~ 773 K.It was confirmed from XRD analysis that Ni was converted to NiI 2 during the HI decomposition.The results exceeding the theoretical equilibrium took place in the case of the high Ni contents of the catalyst.In the case of high Ni contents and low GHSV, HI conversion was increased close to 60 % in the onset of the reaction.It was reconfirmed that Ni catalyst had the bifunctional characteristics of the sorbent of I 2 produced from HI decomposition, as well as the catalyst.It was also revealed that the NiI 2 formed from HI decomposition had the catalytic activity.financial resource from the Ministry of Trade, Industry & Energy, Republic of Korea.(No.20153030040910) 1) ĹĻ Bunsen reaction: SO 2 + I 2 + 2H 2 O ֖ H 2 SO 4 + 2HI (293 ~ 393 K) (2) ĹĻ H 2 SO 4 decomposition: H 2 SO 4 ֖ H 2 O + SO 2 + 0.5O 2 (1073 ~ 1173 K)

Table 1 .
Surface Characteristics Of Support And Prepared Catalysts