�� 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 3294 www.advmat.de www.MaterialsViews.com COMMUNICATION wileyonlinelibrary.com Adv. Mater. 2011, 23, 3294���3297 George Akiyama , Ryotaro Matsuda , * Hiroshi Sato , Masaki Takata , and Susumu Kitagawa* Cellulose Hydrolysis by a New Porous Coordination Polymer Decorated with Sulfonic Acid Functional Groups G. Akiyama, Dr. R. Matsuda , Dr. H. Sato , Prof. S. Kitagawa Exploratory Research for Advanced Technology (ERATO) Kitagawa Integrated Pores Project Japan Science and Technology Agency (JST) Kyoto Research Park Bldg#3���405 Shimogyo-ku, Kyoto 600���8815, Japan E-mail: ryotaro.matsuda@kip.jst.go.jp kitagawa@icems.kyoto-u.ac.jp Dr. R. Matsuda, Dr. H. Sato, Prof. S. Kitagawa Institute for Integrated Cell-Material Sciences (iCeMS) Kyoto University Katsura, Nishikyo-ku, Kyoto 615���8510, Japan Dr. R. Matsuda, Dr. H. Sato, Prof. M. Takata , Prof. S. Kitagawa RIKEN SPring-8 Center Sayo-gun, Hyogo 679���5148, Japan Prof. M. Takata Japan Synchrotron Radiation Research Institute/SPring-8 and RIKEN SPring-8 Center Sayo-gun, Hyogo 679���5148, Japan DOI: 10.1002/adma.201101356 Biofuels have attracted attention in this decade as an alternative energy resource to limited fossil fuels. The world production of bioethanol considerably increased from 17 billion liters in 2000 to more than 46 billion liters in 2007. [ 1 ] Most bioethanol is currently produced from food crops, such as corn and sug- arcane, as biomass resources because the food crops derive water-soluble starch that can be easily converted to bioethanol by yeast bacteria or acid catalysts. However, this ethanol pro- duction process consumes huge amounts of food crops and competes with the supply of food. To overcome this problem, starch-derived ethanol is anticipated to be replaced by cellu- lose-derived ethanol because the cellulose can be conveniently obtained from wooden or grain plant fi bers that do not com- pete with the food supply. Recently, several solvent media and catalysts for the hydrolysis of cellulose have been investigated, including supercritical water, ionic liquids, mesoporous carbon functionalized with ruthenium metal or SO 3 H groups, and sul- fonated ion exchange resins. [ 2���9 ] Nevertheless, further develop- ment of more effective and ���green��� materials for hydrolysis of cellulose is required. Porous coordination polymers (PCPs) or metal-organic frameworks (MOFs), composed of metal ions and various organic ligands, have been extensively studied as a new class of porous solids. One of the advantages of PCPs is their highly designable framework, which provides a large variety of pore surfaces and pore structures. [ 10���16 ] Here we show the synthesis of a new PCP (compound 1 ) that has the sulfonic acid groups as a very strong acid on its pore surface and dem- onstrate its catalytic performance for cellulose hydrolysis. It is essential to decorate the pore surface with a Br��nsted acid for the fabrication of a solid acid catalyst. Whereas many types of functional groups can be introduced into PCP frame- works, it is generally very diffi cult to retain a strong Br��nsted acid site such as sulfonic acid in the framework. This is because even if a ligand with a Br��nsted acid is used to construct a PCP framework, deprotonation takes place in the synthesis solution resulting in the formation of a PCP framework composed of a ligand with the conjugate Br��nsted base form. To avoid depro- tonation of the acid group, it is necessary to use a strong acid solution as the synthesis solvent. Although most of the reported PCPs or MOFs are decomposed in such strong acid solutions, among the many PCPs synthesized to date there are several robust and water durable PCPs. In particular, MIL-101, which is composed of a chromium oxide cluster and terephthalate ligands, is highly stable even in strong aqueous acid solution because it is synthesized in hydrofl uoric acid media. [ 17 ] From this point of view, we adopted the MIL-101 framework as a plat- form to create a new solid acid catalyst with highly acidic func- tional groups in the porous framework. We synthesized MIL-101 [ 17 ] and a new PCP as a crystalline powder composed of a chromium ion and 2-sulfoterephthalate instead of the unsubstituted terephthalate in MIL-101. Com- pound 1 was prepared by solvothermal reaction of chromium(VI) oxide, monosodium 2-sulfoterephthalic acid, and hydrochloric acid in water. To confi rm the structure of compound 1 , an X-ray powder diffraction pattern of 1 was measured at the BL44B2 beam line at SPring-8 using synchrotron radiation with a wave- length of 0.8 ��. [ 18 ] The pattern is in very good agreement with a simulated pattern of reported MIL-101, indicating that 1 has the same structure as MIL-101 except for the sulfonic acid group (Figure S1, Supporting Information). A model structure of 1 derived from MIL-101 is shown in Figure 1 . According to the reported structure of MIL-101, compound 1 is expected to have giant cages with an inner diameter of ��� 3 nm and an extremely large specifi c surface area. MIL-101 contains fl uoride anions that come from the hydrofl uoric acid synthesis solution to compen- sate the positive charges of the skeleton. However, an elemental analysis of compound 1 shows that no halide ions are occluded in the solid. It is probable that one of the three SO 3 H groups has an anionic form instead of the inclusion of halide ions and the others have the acid form to achieve electroneutrality. Ther- mogravimetric analysis shows that 1 releases adsorbed water (41 molecules per three chromium cations) below ca. 380 K, is stable up to 600 K, and then follows thermal decomposi- tion (Figure S2, Supporting Information). Consequently, the chemical formula of 1 is determined as {Cr 3 (H 2 O) 3 O[(O 2 C)��� C 6 H 3 (SO 3 H)���(CO 2 )] 2 [(O 2 C)���C 6 H 3 (SO 3 )���(CO 2 )]} �� n H 2 O (where n is ��� 38). The nitrogen (N 2 ) adsorption isotherm of desolvated 1 was measured at 77 K ( Figure 2 ). The isotherm shows a sudden N 2 uptake in the low relative pressure ( P / P 0 , where P

3295 www.advmat.de www.MaterialsViews.com �� 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim COMMUNICATION wileyonlinelibrary.com Adv. Mater. 2011, 23, 3294���3297 is the vapor pressure and P 0 saturated vapor pressure) region, then a gradual increase in the amount adsorbed up to P / P 0 = 0.3, indicative of the presence of micropores with different sizes in the structure. The Brunauer���Emmett���Teller (BET) surface area was estimated to be 1915 m 2 g ��� 1 , which is lower than that of MIL-101 (3124 m 2 g ��� 1 ) because of the bulky sulfonic groups on the pore surface. As is expected from the chemical for- mula, when 1 was immersed in water, it produced an acidic solution and the concen- tration of surface sulfonic acid groups was estimated to be 1.8 mmol g ��� 1 by acid���base titration using 0.1 M sodium hydroxide solu- tion. It is worth noting that compound 1 is highly stable even in boiling water for more than 24 h, and no change in the powder X-ray pattern or the nitrogen sorp- tion behavior was found after this treat- ment (Figure S3, Supporting Information). The high thermal stability in water is one of the most important charcteristics for het- erogeneous catalysts of cellulose hydrolysis and as a result, 1 is expected to show a high performance for cellulose hydrolysis as a new solid acid catalyst. We demonstrated cellulose hydrolysis at 393 K with catalysts (compound 1 or MIL- 101) and without a catalyst as control experi- ments. The same amount of MIL-101 or compound 1 (0.2 g) was used for the cata- lytic reaction. The resulting solution was analyzed by high-performance liquid chro- motography (HPLC) and NMR. The results of the reaction are summarized in Table 1 . The cellulose was not hydrolyzed in the absence of catalyst and was hydrolyzed only a small amount with MIL-101. On the other hand, the reaction of 1 for 3 h pro- duced hydrolysis of cellulose with yields of 2.6, 1.4, and 1.2% of xylose, glucose, and cellobiose, respectively therefore, 5.3% of the cellulose in total was converted to mono- or disaccha- rides. The yields of xylose, glucose, and cellobiose continually increased when the reaction time was increased from 1 to 5 h (Figure S4, Supporting Information). It is well known that 5-hydroxymethylfurfural, levulinic acid, and formic acid are normally produced as unfavorable side products after the reaction of cellulose hydrolysis under strongly acidic conditions. [ 6 , 19 ] We found these side products after cellulose hydrolysis using sulfuric acid aqueous solution while they showed high catalytic activity (glucose: 9.0%, xylose: 12%). On the other hand, when we use comound 1 , these side products were not detected after the reaction, as confi rmed by 1 H NMR spectroscopy, indicating that this catalytic reac- tion is clean and effective. It should be hard for cellulose to diffuse deep inside the porous solid, even though the chain diameter of cellulose is smaller than the size of the pores, because most of the cellulose is crystallized and insoluble in Figure 1 . Schematic representation of the structure of compound 1 . Compound 1 is composed of Cr 3 + ions and 2-sulfoterephthalates. Chromium and oxygen atoms of metal clusters are drawn in the stick model with light pink and green colors, respectively. Sulfonic groups are drawn using a ball and stick model as bluish colors. Figure 2 . Nitrogen gas sorption isotherm at 77 K for 1 in the relative pressure ( P / P 0 ) range from 0 to 0.94. P 0 is the saturated vapor pressure equal to 101 kPa for nitrogen at 77 K. Stp stands for standard temperature and pressure. 0.0 0.2 0.4 0.6 0.8 1.0 200 400 600 Relative pressure P/P0 Amount adsorbed / ml(stp) g -1