Energy & Fuels 2008, 22, 1485–1492
1485
Lignin Gasification over Supported Ruthenium Trivalent Salts in Supercritical Water Aritomo Yamaguchi,† Norihito Hiyoshi,† Osamu Sato,† Mitsumasa Osada,‡ and Masayuki Shirai*,† Research Center for Compact Chemical Process, National Institute of AdVanced Industrial Science and Technology (AIST), 4-2-1 Nigatake, Miyaginoku, Sendai 983-8551, Japan, and Department of Chemical Engineering, Ichinoseki National College of Technology, Takanashi, Hagisho, Ichinoseki, Iwate 021-8511, Japan ReceiVed February 19, 2008. ReVised Manuscript ReceiVed March 14, 2008
Lignin gasification behavior over unreduced ruthenium trivalent-salts (ruthenium (III) chloride or ruthenium (III) nitrosyl nitrate) supported on titanium oxide and charcoal (RuCl3/TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C) in supercritical water at 673 K was studied in a batch reactor. The Ru(NO)(NO3)3/C and Ru(NO)(NO3)3/TiO2 catalysts showed lignin gasification activities as high as a charcoal supported ruthenium catalyst (Ru/C). Characterization by X-ray diffraction (XRD), X-ray absorption near-edge spectroscopy (XANES), extended X-ray absorption fine structure (EXAFS) measurements, and transmission electron microscopy (TEM) revealed that reduced ruthenium metal particles formed during the lignin gasification were smaller in the case of the Ru(NO)(NO3)3/C and Ru(NO)(NO3)3/TiO2 catalysts, compared with those for the RuCl3/C and RuCl3/TiO2 catalysts. The size of ruthenium metal particles in the case of the Ru/C catalyst did not change by the addition of hydrochloric acid; however, its activity decreased. The RuCl3/C and RuCl3/TiO2 catalysts were less active for the gasification than the Ru(NO)(NO3)3/C and Ru(NO)(NO3)3/TiO2 catalysts because (i) a large ruthenium metal particles were formed and the number of active metal sites was small and (ii) chloride ions poisoned the active sites of ruthenium metal.
1. Introduction Biomass and organic wastes have attracted much attention as renewable energy source because the green house effect of carbon dioxide from the combustion of fossil fuel has to be reduced in view of an environmental problem of global warming.1 The gasification of lignin, which is a major fraction of woody biomass, is needed for its efficient use as a high quality energy source, which can be provided in the form of electric energy using fuel cell or as liquid fuel by Fischer-Tropsch reaction. Supercritical water (Tc ) 647.3 K, Pc ) 22.1 MPa) gasification are efficient because of elimination of drying the biomass; however, high temperature (1073-1273 K) is required for the steam reforming of lignin.2 Low temperature methods are more desirable because waste heat from industrial processes, such as iron or cement production, can be utilized. Supercritical water gasification is a promising technique to reduce the lignin gasification temperature. Several research groups claimed that the lignin gasification in supercritical water was catalyzed by metal catalysts.3–14 We have already reported that charcoal-supported ruthenium (III) salts were active for the lignin gasification in supercritical water at 673 K in which reduced ruthenium metal particles were formed after the lignin gasification.15 As a continuation of this * Corresponding author. E-mail:
[email protected]. Fax: +81-22-2375224. † National Institute of Advanced Industrial Science and Technology (AIST). ‡ Ichinoseki National College of Technology. (1) Huber, G. W.; Iborra, S.; Corma, A. Chem. ReV. 2006, 106, 4044– 4098. (2) Klass, D. L. Biomass for renewable energy, fuels, and chemicals; Academic Press: San Diego, 1998.
work, we further investigated the effects of the supports, ruthenium precursors, and counteranions of ruthenium salts on the lignin gasification over ruthenium trivalent salts supported on activated charcoal carbon and titanium oxide. The mechanism of ruthenium reduction during the lignin gasification is also discussed. 2. Experimental Section 2.1. Catalyst Preparation. The various catalysts used in this work were prepared by an impregnation method using activated charcoal powder (Wako Pure Chemical Industries) or titanium oxide, titania (Degussa, P-25) and an aqueous solution of ruthenium (III) chloride hydrate (Wako Pure Chemical Industries) or ruthenium (III) nitrosyl nitrate solution in dilute nitric acid (Aldrich) as follows. The aqueous solution of ruthenium precursor and charcoal or titanium oxide powder were stirred for 1 h at ambient temperature and evaporated to dryness at 323 K under reduced pressure by a rotary evaporator. Then the samples were dried for 10 h at 373 K in an oven, which are represented hereinafter as RuCl3/C, Ru(NO)(NO3)3/C, RuCl3/TiO2, and Ru(NO)(NO3)3/TiO2, based on the corresponding ruthenium precursors and supports. The amount of ruthenium in all the catalysts was regulated to be 5 wt %. A charcoal-supported ruthenium (Ru/C, 5 wt % of ruthenium) catalyst was purchased from Wako Pure Chemical Industries and used as a reference catalyst.15 2.2. Lignin Gasification Reaction. Lignin (organosolv-lignin powder) was purchased from Aldrich and used without further purification.4 Gasification of lignin was carried out in a SUS 316 tube, of which inner volume was 6.0 cm3. The catalyst (0.15 g), lignin (0.10 g), and water (3.0 g) were loaded in the tube, and the reactor was purged with argon gas. The reactor was submerged into a molten-salt bath (KNO3-NaNO3) at 673 K for a given reaction time. The partial pressure of water at 673 K and 0.5 g
10.1021/ef8001263 CCC: $40.75 2008 American Chemical Society Published on Web 05/03/2008
1486 Energy & Fuels, Vol. 22, No. 3, 2008
Yamaguchi et al.
Table 1. Product Yield and Gas Composition of Catalytic Lignin Gasification in Supercritical Watera gas composition (%) catalyst
gas yield (C%)
H2
CO
CH4
CO2
C2-C4 gases
water soluble (C%)
THF solubleb (C%)
THF insoluble (C%)
Charcoal RuCl3/TiO2 Ru(NO)(NO3)3/TiO2 RuCl3/C Ru(NO)(NO3)3/C Ru/C
8.4 20.2 61.8 24.3 75.4 87.5
6.9 19.7 6.2 14.6 3.1 3.1
23.6 0.0 0.0 0.0 2.8 0.7
15.4 21.4 43.8 17.1 41.9 47.4
52.8 57.0 48.5 65.7 51.4 48.2
1.4 1.9 1.5 2.6 0.9 0.6
10.9 13.5 2.7 7.1 1.0 0.1
42.3 44.7 34.0 39.2 19.4 9.4
38.4 21.6 1.5 29.4 4.2 3.0
a Reaction conditions; lignin 0.10 g, catalyst 0.15 g, water density 0.50 g cm-3, 673 K, 1 h. b THF soluble (C%) was calculated as 100 - (gas yield (C%)) - (water soluble (C%)) - (THF insoluble (C%)).
cm-3 of water density was 37.1 MPa in the supercritical phase.16 After the reaction, the tube was submerged into a water bath for cooling to ambient temperature. Gaseous products were analyzed by a gas chromatography (Shimadzu, GC-8A) using a Shincarbon ST column and a thermal conductivity detector. Liquid and solid products in the tube were recovered with water and filtered to separate water-insoluble fraction from the water-soluble fraction. Amounts of organic carbon and ruthenium species in the watersoluble fraction were evaluated using a total organic carbon analyzer (Shimadzu, TOC-VCSN) and inductively coupled plasma emission spectrometry (Seiko, ICP-SPS 1500R), respectively. The waterinsoluble fraction was washed with tetrahydrofuran (THF, Wako Pure Chemical Industries) and filtered to separate THF-insoluble solid fraction from THF-soluble fraction. Amount of THF-insoluble product (char) was estimated by subtracting the weight of the catalyst loaded from the amount of THF-insoluble solid fraction. A product yield based on carbon and gas composition are defined as given below, product yield based on carbon (C%) ) (mol of carbon atom in product)/ (mol of carbon atom in lignin loaded) × 100 (1)
gas composition (%) ) (mol of gas product)/ (sum of mol of gas product) × 100 (2) 2.3. Characterization. X-ray diffraction (XRD) patterns of the catalysts before and after the gasification were recorded using a Rigaku RINT 2200VK/PC with Cu KR radiation. X-ray absorption near edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) measurements were performed using a synchrotron radiation ring at AR-NW10A, Photon Factory, KEK with a Si (311) double-crystal monochromator in transmission mode. The catalyst samples were recovered from the SUS reactors quickly and filled into EXAFS cells as water suspensions. The XANES and EXAFS spectra of the catalysts, which were exposed to air, were measured at ambient temperature. The XANES spectra were normalized by their edge jumps to be unity. The EXAFS spectra were analyzed by the UWXAFS package.17 After background subtraction, a k3-weighted EXAFS function in the k range of 30-140 nm-1 was Fourier transformed into an R-space. The spectrum was fitted in the corresponding k-space of the R range 0.10-0.30 nm. The backscattering amplitudes and phase shifts were calculated by the FEFF8 code.18 The coefficient of effective amplitude reduction factor (s02) of Ru-Ru bond was estimated from
Figure 1. Carbon yield and gas composition for lignin gasification in supercritical water at 673 K and 0.5 g cm-3 of water density: (a) gas yield (catalyst RuCl3/TiO2 (0), Ru(NO)(NO3)3/TiO2 (4), RuCl3/C (9), Ru(NO)(NO3)3/C (2), and Ru/C (b)); gas composition over RuCl3/TiO2 (b) and Ru(NO)(NO3)3/TiO2 (c) (H2 (O), CH4 (4), CO ( × ), CO2 (0), and C2-C4 gases (3)).
Gasification oVer Supported Ru TriValent Salts
Energy & Fuels, Vol. 22, No. 3, 2008 1487 dav )
∑ d n /∑ d n 3
i
i
2 i i
i
(di: particle size;ni: number of particles)
i
(3)
3. Results and Discussion
Figure 2. XRD patterns for supported ruthenium trivalent salts catalysts before and after lignin gasification in supercritical water at 673 K for 1 h: TiO2 (a), RuCl3/TiO2 before (b) and after (c) the lignin gasification, Ru(NO)(NO3)3/TiO2 before (d) and after (e) the lignin gasification, charcoal (f), RuCl3/C before (g) and after (h) the lignin gasification, and Ru(NO)(NO3)3/C before (i) and after (j) the lignin gasification. The closed circles indicate ruthenium metal.
the fitting result of a standard sample of ruthenium metal powder. The fitting parameters were coordination number (CN), interatomic distance (R), Debye-Waller factor (σ2), and a correction of the threshold energy (∆E0). Transmission electron microscopy (TEM) images were measured using an FEI TECNAI-G20 electron microscope operated at 200 kV of an accelerating voltage. The catalyst powders were dispersed in methanol, and then loaded on a grid coated with carbon films for TEM measurement. The average sizes (dav) of ruthenium metal particles from the TEM images were determined by the following equation:
3.1. Gasification of Lignin over Supported Ruthenium Trivalent Species. Table 1 shows the product yields and gas composition of lignin gasification at 673 K for 1 h in supercritical water over the ruthenium-based catalysts and the charcoal support without ruthenium species. The gas yields over the supported ruthenium trivalent catalysts, such as RuCl3/TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C, as well as the reduced Ru/C catalyst were much higher than those over the charcoal support without ruthenium species, leading to the fact that all supported ruthenium trivalent catalysts possessed lignin gasification activities. The activities of lignin gasification were influenced significantly by the precursor of ruthenium species. The gas yields over supported ruthenium (III) nitrosyl nitrate catalysts showed about three times higher activities than those over supported ruthenium (III) chloride catalysts. Figure 1 shows the gasification profiles over the supported ruthenium catalysts in supercritical water over the RuCl3/TiO2 and Ru(NO)(NO3)3/TiO2 catalysts as a function of reaction time. Lignin in the reactor was almost completely gasified over the Ru(NO)(NO3)3/TiO2 catalyst in supercritical water at 673 K for 3 h; however, lignin gasification over RuCl3/TiO2 reached a plateau at about 30 C% lignin conversion after 1 h (Figure 1a). We have reported that complete gasification and high methane selectivity were observed over the Ru(NO)(NO3)3/C catalyst.15
Figure 3. Ru K-edge XANES spectra for supported ruthenium catalysts before and after lignin gasification in supercritical water at 673 K. (1) Ruthenium metal powder (a), RuCl3 (b), and Ru(NO)(NO3)3 (c). (2) RuCl3/C (d), RuCl3/TiO2 (e), Ru(NO)(NO3)3/C (f), and Ru(NO)(NO3)3/TiO2 (g) before the gasification. (3) RuCl3/C (reaction time, 5 min (h) and 1 h (i)), RuCl3/TiO2 (reaction time, 1 h (j)), Ru(NO)(NO3)3/C (reaction time, 5 min (k) and 1 h (l)), and Ru(NO)(NO3)3/TiO2 (reaction time, 1 h (m)) after the lignin gasification.
1488 Energy & Fuels, Vol. 22, No. 3, 2008
Yamaguchi et al.
Table 2. Structural Parameters for RuCl3/TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, Ru(NO)(NO3)3/C, and Ru/C Catalysts after the Lignin Gasification in Supercritical Water at 673 K for 1 h and Ru Metal Powder, Determined by a Curve Fitting Analysis of the EXAFS Fourier Transforms
a
sample
CNRu-Ru
R (nm)
σ2 (10-5 nm2)
∆E0 (eV)
Rf (%)
Ru metal powder RuCl3/TiO2 Ru(NO)(NO3)3/TiO2 RuCl3/Cb RuCl3/C Ru(NO)(NO3)3/Cb Ru(NO)(NO3)3/C Ru/C
12.0a 10.1 ( 3.1 6.9 ( 2.8 10.2 ( 2.3 11.2 ( 2.3 5.9 ( 1.4 6.1 ( 1.2 6.7 ( 1.1
0.267 ( 0.001 0.267 ( 0.001 0.266 ( 0.002 0.267 ( 0.001 0.267 ( 0.002 0.265 ( 0.001 0.266 ( 0.002 0.266 ( 0.001
5.7 ( 0.5 4.7 ( 1.5 6.5 ( 1.8 4.7 ( 1.0 5.3 ( 0.2 7.3 ( 1.3 6.5 ( 0.4 6.6 ( 0.8
-0.6 ( 2.9 0.6 ( 3.7 -2.5 ( 4.0 0.5 ( 2.5 -3.7 ( 5.1 -6.3 ( 2.7 -5.1 ( 8.2 -5.4 ( 1.8
2.5 0.6 1.4 0.6 1.0 3.2 4.0 2.3
The coordination number was fixed at 12.0. b The reaction time was 5 min.
Figure 4. Fourier transforms of k3-weighted EXAFS spectra at Ru K-edge for supported ruthenium trivalent-salt catalysts after the lignin gasification in supercritical water at 673 K. (1) RuCl3/C (reaction time, 5 min (a) and 1 h (b)) and RuCl3/TiO2 (reaction time, 1 h (c)). (2) Ru(NO)(NO3)3/C (reaction time, 5 min (d) and 1 h (e)) and Ru(NO)(NO3)3/TiO2 (reaction time, 1 h (f)). The line and circle represent measured Fourier transforms and calculated Fourier transforms, respectively.
We also claimed that carbon monoxide and hydrogen were produced from the lignin gasification and then carbon dioxide and methane were produced via water-gas shift reaction and methanation reaction.4 The higher methane selectivity obtained over the Ru(NO)(NO3)3/TiO2 catalyst than that over RuCl3/TiO2 shows that methanation activities over the Ru(NO)(NO3)3/TiO2 catalysts would have higher number of methanation sites, which need ensembles of surface ruthenium atoms. We have discussed the methane selectivity along with structural characterization results in section 3.4. 3.2. Structure of Supported Ruthenium Trivalent Species during Gasification. Figure 2 shows XRD patterns of the RuCl3/TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C samples before and after the lignin gasification in supercritical water for 1 h at 673 K. Three diffraction peaks at
38.8, 42.4, and 44.4° attributed to ruthenium metal were observed in the XRD patterns of the RuCl3/TiO2 and RuCl3/C catalysts after the lignin gasification (closed circles in Figure 2c and h). On the other hand, these peaks were not observed clearly in the Ru(NO)(NO3)3/TiO2 and Ru(NO)(NO3)3/C catalysts after the gasification (Figure 2e and j). The metal particle size of the RuCl3/TiO2 and RuCl3/C catalysts after the gasification could not be estimated using Scherrer formula because of poor peak separation. Figure 3 shows Ru K-edge XANES spectra of the RuCl3/ TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C catalysts before and after the lignin gasification in supercritical water at 673 K and the reference samples of ruthenium metal powder, ruthenium (III) chloride hydrate powder, and ruthenium (III) nitrosyl nitrate solution. The XANES spectra of the RuCl3/ TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C catalysts before the lignin gasification were similar to those of the corresponding precursors, ruthenium (III) chloride and ruthenium (III) nitrosyl nitrate, respectively as shown in Figure 3b-g, implying that ruthenium species of the RuCl3/TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C catalysts before the lignin gasification retained the structure of precursors. The XANES spectra of the RuCl3/TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C catalysts after the lignin gasification were similar to those of the ruthenium metal (Figure 3), indicating that ruthenium species of these catalysts were reduced to ruthenium metal during the lignin gasification. It should be noted that the reduction of trivalent ruthenium species in the supported ruthenium trivalent catalysts to ruthenium metal occurred within 5 min of the lignin gasification, as proved by the XANES spectra of the RuCl3/C and Ru(NO)(NO3)3/C catalysts after 5 min of the gasification (Figure 3h and k). Inductive coupled plasma emission spectrometric analysis showed that no ruthenium species were present in the aqueous phase after the lignin gasification for all of the catalysts, RuCl3/ TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C, implying that all the ruthenium species were reduced on the supports. Figure 4 shows Fourier transforms of k3-weighted EXAFS spectra at Ru K-edge of the RuCl3/TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C catalysts after the lignin gasification in supercritical water at 673 K. The peaks attributed to Ru-Ru metal bond were observed at 0.23 nm (phase shift uncorrected) in all the ruthenium catalysts after the lignin gasification, indicating that ruthenium metal particles were also formed in the Ru(NO)(NO3)3/TiO2 and Ru(NO)(NO3)3/C catalysts similar to those in the RuCl3/TiO2 and RuCl3/C catalysts. Table 2 shows structural parameters determined by a curve fitting analysis of the EXAFS Fourier transforms for the RuCl3/ TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C catalysts after the lignin gasification along with the ruthenium metal powder as a reference sample. The coordination numbers
Gasification oVer Supported Ru TriValent Salts
Energy & Fuels, Vol. 22, No. 3, 2008 1489
Table 3. Structural Parameters for RuCl3/C and Ru(NO)(NO3)3/C Catalysts after the Treatment in Supercritical Water without Lignin at 673 K for 1 h, Determined by a Curve Fitting Analysis of the EXAFS Fourier Transforms sample
CNRu-Ru
R (nm)
σ2 (10-5 nm2)
∆E0 (eV)
Rf (%)
RuCl3/TiO2 RuCl3/C Ru(NO)(NO3)3/C
9.4 ( 2.5 11.0 ( 1.1 6.2 ( 0.6
0.267 ( 0.001 0.267 ( 0.001 0.265 ( 0.001
4.7 ( 1.3 4.7 ( 0.4 7.2 ( 0.4
2.5 ( 4.2 3.0 ( 1.2 -6.3 ( 0.9
1.1 1.2 1.3
(CN) of Ru-Ru bonds for the Ru(NO)(NO3)3/TiO2 and Ru(NO)(NO3)3/C catalysts after the gasification were 6.1 and 6.9, which correspond19 to 1.0 and 1.1 nm of the ruthenium metal particle size, respectively. The CN of the Ru(NO)(NO3)3/C catalyst after 5 min of the gasification was 5.9, corresponding to the particle size of 1.0 nm, and increased slightly to 6.1 after 1 h of the gasification, indicating that the ruthenium species of the Ru(NO)(NO3)3/C catalyst were reduced to small ruthenium
metal particles in 5 min of the gasification and the sizes of these particles remained the same after 1 h of gasification. On the other hand, the CNs of Ru-Ru bonds for the RuCl3/TiO2 and RuCl3/C catalysts after the gasification were 10.1 and 11.2, respectively, indicating that large Ru metal particles were formed in these catalysts. The CN of the RuCl3/C catalyst after 5 min of the gasification was 10.2 and increased slightly to 11.2 after 1 h of the gasification, indicating that the ruthenium species of
Figure 5. TEM images of RuCl3/TiO2 (a), Ru(NO)(NO3)3/TiO2 (b), RuCl3/C (c), Ru(NO)(NO3)3/C (d), and Ru/C (e) after lignin gasification in supercritical water at 673 K for 1 h. The arrows indicate ruthenium metal particles.
1490 Energy & Fuels, Vol. 22, No. 3, 2008
Yamaguchi et al.
Table 4. Product Yield and Gas Composition of Catalytic Lignin Gasification in Supercritical Water in the Presence of Inorganic Acidsa gas composition (%) catalyst RuCl3/C Ru(NO)(NO3)3/C Ru/C
inorganic
acidb
HNO3 HCl HCl
gas yield (C%) 22.9 42.4 30.2
H2
CO CH4 CO2 C2-C4 gases water soluble (C%) THF solublec (C%) THF insoluble (C%)
15.5 0.9 18.2 63.5 8.3 0.5 27.2 62.1 11.6 2.0 30.4 54.6
2.0 2.0 1.5
9.1 2.7 9.1
41.6 23.6 19.1
26.4 31.3 41.6
a Reaction conditions; lignin 0.10 g, catalyst 0.15 g, water density 0.50 g cm-3, 673 K, 1 h. b Inorganic acids are added in the same moles of anion species of ruthenium salts. c THF soluble (C%) was calculated to be 100 - (gas yield (C%)) - (water soluble (C%)) - (THF insoluble (C%)).
Figure 6. Size distribution of ruthenium metal particles of RuCl3/TiO2 (a), Ru(NO)(NO3)3/TiO2 (b), RuCl3/C (c), and Ru(NO)(NO3)3/C (d) after the lignin gasification in supercritical water at 673 K for 1 h, obtained from their sizes of more than 200 particles in the TEM images.
this catalyst were reduced to form large ruthenium metal particles within 5 min. The ruthenium metal particle sizes as well as the lignin gasification activities were influenced significantly by the precursor of ruthenium species. Figure 5 shows TEM images of the RuCl3/TiO2, Ru(NO)(NO3)3/ TiO2, RuCl3/C, Ru(NO)(NO3)3/C, and Ru/C catalysts after the lignin gasification. The ruthenium metal particles (indicated by arrows) were observed in all the images. Size distributions of ruthenium metal particles of supported ruthenium catalysts after the lignin gasification were obtained from the sizes of more than 200 particles observed in TEM (Figure 6). The sizes of ruthenium metal particles of the RuCl3/TiO2 and RuCl3/C catalysts ranged mainly from 4 to 18 nm and their averaged sizes (dav) were 11.4 and 11.9 nm, respectively. On the other hand, those of the Ru(NO)(NO3)3/TiO2 and Ru(NO)(NO3)3/C catalysts ranged mainly from 1 to 3 nm and their dav values were 1.7 and 2.3 nm, respectively. The particle size values estimated by TEM are larger than those from EXAFS analysis (ca. 1.0 nm), which might be due to the difficulty to observe the ruthenium metal particles of less than 1 nm in TEM. 3.3. Structural Change of Supported Ruthenium Trivalent Species in Supercritical Water. We investigated the structure of supported ruthenium trivalent species after super-
critical water treatment without lignin at 673 K to understand the mechanism of structural changes that supported ruthenium trivalent species undergo to form reduced metal particles. Figure 7 shows Ru K-edge XANES spectra and k3-weighted EXAFS Fourier transforms of the RuCl3/TiO2, Ru(NO)(NO3)3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C samples after the treatment in supercritical water without lignin at 673 K. The ruthenium species of the RuCl3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C catalysts were reduced to ruthenium metal without lignin; however, the Ru(NO)(NO3)3/TiO2 catalyst was not reduced after the treatment in supercritical water (Figure 7). Structural parameters determined by a curve fitting analysis of their EXAFS spectra of the RuCl3/TiO2, RuCl3/C, and Ru(NO)(NO3)3/C catalyst after the supercritical water treatment are shown in Table 3. The CNs of Ru-Ru bonds in the RuCl3/ TiO2, RuCl3/C, and Ru(NO)(NO3)3/C samples after the supercritical water treatment were almost the same as those after the lignin gasification. A small amount of gas (ca. 4 cm3; gas composition hydrogen 10%, methane 5%, carbon dioxide 85%) was formed during the supercritical water treatment of the Ru(NO)(NO3)3/C catalyst at 673 K for 1 h; however, gas evolution did not occur during the supercritical water treatment of Ru(NO)(NO3)3/TiO2 catalyst. We had also observed that
Gasification oVer Supported Ru TriValent Salts
Figure 7. Ru K-edge XANES spectra (1) and k3-weighted EXAFS Fourier transforms for supported ruthenium trivalent species after the treatment in supercritical water without lignin at 673 K for 1 h (catalyst; RuCl3/C (a), RuCl3/TiO2 (b), Ru(NO)(NO3)3/C (c), and Ru(NO)(NO3)3/ TiO2 (d)). The line and circle represent measured Fourier transforms and calculated Fourier transforms in part 2, respectively.
Figure 8. XRD patterns for supported ruthenium trivalent-salt catalysts after the lignin gasification in supercritical water at 673 K for 1 h: RuCl3/C in the absence (a) and presence (b) of nitric acid, Ru/C in the absence (c) and presence (d) of hydrochloric acid, and Ru(NO)(NO3)3/C in the absence (e) and presence (f) of hydrochloric acid. The closed circles indicate ruthenium metal.
lignin was gasified without any catalyst (gas yield 3.7 C%4 for 15 min) in supercritical water. These results show that ruthenium trivalent cations could be reduced to metal particles by hydrogen evolved during the reaction and ruthenium metal particles played a key role as catalysts for lignin gasification. 3.4. Effect of Counter Anion of Ruthenium Trivalent Salts on Lignin Gasification. Ruthenium metal particles were formed in all supported ruthenium trivalent catalysts during the lignin gasification and the order of size for ruthenium metal
Energy & Fuels, Vol. 22, No. 3, 2008 1491
particles was RuCl3/C ≈ RuCl3/TiO2 > Ru(NO)(NO3)3/TiO2 ≈ Ru(NO)(NO3)3/C. The order of gasification activity was Ru(NO)(NO3)3/C ≈ Ru(NO)(NO3)3/TiO2 > RuCl3/TiO2 ≈ RuCl3/C. These two orders indicate that the anionic species of a ruthenium precuor played an important role in bringing about the structural changes of ruthenium species during the gasification and therefore in the lignin gasification activities. We carried out the lignin gasification over supported ruthenium catalysts with an additon of nitric or hydrochloric acid (Table 4). The liginin gasification yield of the RuCl3/C catalyst with nitric acid was 22.9 C%, which is almost of the same value of the RuCl3/C catalyst itself (24.3 C%), indicating that the liginin gasification activity was unaffected by the presence of nitric acid. On the other hand, the liginin gasification yields of the Ru(NO)(NO3)3/C and Ru/C catalysts with an addition of hydrochloric acid were 42.4 and 30.2 C%, respectively, which were much smaller than those without hydrochloric acid (75.4 and 87.5 C%, respectively). Thus, the addition of hydrochloric acid inhibited the lignin gasification activities of the Ru(NO)(NO3)3/C and Ru/C catalysts. The structural change of ruthenium species of RuCl3/ C, Ru(NO)(NO3)3/C, and Ru/C after the addition of nitric or hydrochloric acid in lignin gasification were investigated by XRD (Figure 8). The diffraction peaks ascribed to ruthenium metal particles in the Ru(NO)(NO3)3/C and Ru/C samples were not observed by the addition of hydrochloric acid, indicating that the size of particle did not increase; however, the addition of hydrochloric acid reduced drastically the lignin gasification activities of the Ru(NO)(NO3)3/C and Ru/C catalysts. Therefore, hydrochloric acid would inhibit the gasification because of the adsorption of hydrochloric anions on the ruthenium metal surface. Lignin gasification over the RuCl3/C and RuCl3/TiO2 catalysts reached a plateau after 1 h (Figure 1a). The deactivation of the RuCl3/C and RuCl3/TiO2 catalysts within 1 h could be explained by the decrease in number of active sites by the formation of large metal particles and the adsorption of chloride anions on the catalyst surfaces. We have previously reported that the lignin gasification over Ru/C in supercritical water proceeded via the following steps: (i) lignin decomposition to alkylphenols and formaldehyde by supercritical water, (ii) gasification of alkylphenols and formaldehyde over ruthenium catalysts, and (iii) formation of char from alkylphenols and formaldehyde. The cross-linking reaction between alkylphenols and formaldehyde would proceed to form char, leading to the deactivation of the RuCl3/C and RuCl3/TiO2 catalysts. The selectivities of carbon dioxide and hydrogen were higher and that of methane was lower over the RuCl3/TiO2 and RuCl3/C catalysts, compared with those over the Ru(NO)(NO3)3/TiO2 and Ru(NO)(NO3)3/C catalysts. We have reported that carbon monoxide and hydrogen were formed from lignin and that carbon dioxide and methane were produced via water-gas shift reaction and methanation reaction, respectively, during the lignin gasifiaction.4 It is also reported that the water-gas shift reaction does not need ensembles of surface metal atoms; on the other hand, the methanation reaction needs the ensembles of metal atoms.20 Chloride anions adsorbed on the surface might reduce (3) Osada, M.; Sato, T.; Watanabe, M.; Adschiri, T.; Arai, K. Energy Fuels 2004, 18, 327–333. (4) Osada, M.; Sato, O.; Watanabe, M.; Arai, K.; Shirai, M. Energy Fuels 2006, 20, 930–935. (5) Osada, M.; Sato, O.; Arai, K.; Shirai, M. Energy Fuels 2006, 20, 2337–2343. (6) Osada, M.; Sato, T.; Watanabe, M.; Shirai, M.; Arai, K. Combust. Sci. Technol. 2006, 178, 537–552. (7) Osada, M.; Hiyoshi, N.; Sato, O.; Arai, K.; Shirai, M. Energy Fuels 2007, 21, 1854–1858.
1492 Energy & Fuels, Vol. 22, No. 3, 2008
the number of ruthenium ensembles, inducing methanation selectivity over the RuCl3/TiO2 and RuCl3/C catalysts to be much less than those in the case of Ru(NO)(NO3)3/TiO2 and Ru(NO)(NO3)3/C catalysts. Notably, the addition of hydrochloric acid decreased the methane selectivity from 41.9% to 30.4% over Ru(NO)(NO3)3/C and from 47.4% to 27.2% over Ru/C, despite the fact that hydrochloric acid did not affect ruthenium (8) Osada, M.; Hiyoshi, N.; Sato, O.; Arai, K.; Shirai, M. Energy Fuels 2007, 21, 1400–1405. (9) Matsumura, Y.; Minowa, T.; Potic, B.; Kersten, S. R. A.; Prins, W.; van Swaaij, W. P. M.; van de Beld, B.; Elliott, D. C.; Neuenschwander, G. G.; Kruse, A., Jr. Biomass Bioenergy 2005, 29, 269–292. (10) Yoshida, T.; Matsumura, Y. Ind. Eng. Chem. Res. 2001, 40, 5469– 5474. (11) Yoshida, T.; Oshima, Y. Ind. Eng. Chem. Res. 2004, 43, 4097– 4104. (12) Yoshida, T.; Oshima, Y.; Matsumura, Y. Biomass Bioenergy 2004, 26, 71–78. (13) Watanabe, M.; Inomata, H.; Osada, M.; Sato, T.; Adschiri, T.; Arai, K. Fuel 2003, 82, 545–552. (14) Sato, T.; Furusawa, T.; Ishiyama, Y.; Sugito, H.; Miura, Y.; Sato, M.; Suzuki, N.; Itoh, N. Ind. Eng. Chem. Res. 2006, 45, 615–622. (15) Yamaguchi, A.; Hiyoshi, N.; Sato, O.; Osada, M.; Shirai, M. Catal. Lett. 2008, 122, 188–195. (16) Wagner, W.; Prub, A. J. Phys. Chem. Ref. Data 2002, 31, 387– 535. (17) Stern, E. A.; Newville, M.; Ravel, B.; Yacoby, Y.; Haskel, D. Physica B 1995, 208, 117–120. (18) Ankudinov, A. L.; Ravel, B.; Rehr, J. J.; Conradson, S. D. Phys. ReV. B 1998, 58, 7565–7576. (19) Jentys, A. Phys. Chem. Chem. Phys. 1999, 1, 4059–4063. (20) Bond, G. C. Catalysis by metals; Academic Press: London, 1962.
Yamaguchi et al.
metal particle sizes. This result also supports that the addition of hydrochloric acid decreased the number of ensembles of surface metal sites, leading to the suppression of methanation reaction. 4. Conclusions Structure and lignin gasification behavior of ruthenium trivalent catalysts in supercritical water at 673 K were studied in detail. The order of gasification activity was Ru/C ≈ Ru(NO)(NO3)3/C ≈ Ru(NO)(NO3)3/TiO2 > RuCl3/C ≈ RuCl3/ TiO2. Characterization with XRD, XANES, EXAFS, and TEM revealed that small ruthenium metal particles were formed during the lignin gasification in supercritical water. The sizes of reduced metal particles obtained from TEM images of the RuCl3/C and RuCl3/TiO2 catalysts after the gasification were 11.9 and 11.4 nm, respectively. On the other hand, those of the Ru(NO)(NO3)3/C and Ru(NO)(NO3)3/TiO2 catalysts were 2.3 and 1.7 nm, respectively. Activities of RuCl3/C and RuCl3/TiO2 catalysts in the presence of hydrochloric acid were low because (i) large ruthenium metal particles were formed having fewer active sites and (ii) chloride ions adsorbed on the ruthenium metal particles poisoned the gasification. Acknowledgment. EXAFS measurements were done by the approval of the PAC committee (proposal no. 2006G324). EF8001263
