Catal Lett (2010) 134:184–189 DOI 10.1007/s10562-009-0208-4
Ni/NaX: A Bifunctional Efficient Catalyst for Selective Hydrogenolysis of Glycerol Jing Zhao • Weiqiang Yu • Chen Chen Hong Miao • Hong Ma • Jie Xu
•
Received: 22 May 2009 / Accepted: 30 October 2009 / Published online: 14 November 2009 Ó Springer Science+Business Media, LLC 2009
Abstract Non-noble metal Ni/NaX catalyst was prepared and used in the hydrogenolysis of aqueous glycerol. Characterization by XRD, SAED, H2 chemisorption, ICP and NH3-TPD techniques disclosed that the proper strong acid sites were responsible for the high activity and selectivity. Over Ni/NaX catalyst, conversion of glycerol reached 86.6% with 94.6% selectivity to glycols including 1,2-proplyene glycol and ethylene glycol under 6.0 MPa H2 pressure at 200 °C after 10 h reaction. Additionally, the effects of time, temperature, and H2 pressure were investigated in detail. Keywords Glycerol Hydrogenolysis Nickel NaX zeolite Acidity
1 Introduction Due to the decrease of fossil resources and the increase of atmospheric carbon dioxide caused by fuel combustion, catalytic conversion of renewable biomass-derived materials into value-added oxygenated chemicals has attracted more and more attention. Glycerol, one of the top-12 Electronic supplementary material The online version of this article (doi:10.1007/s10562-009-0208-4) contains supplementary material, which is available to authorized users. J. Zhao W. Yu C. Chen H. Miao H. Ma J. Xu (&) State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 457 Zhongshan Road, 116023 Dalian, People’s Republic of China e-mail:
[email protected] J. Zhao W. Yu Graduate University of Chinese Academy of Sciences, 100049 Beijing, People’s Republic of China
123
building block chemicals [1], is a main byproduct (10 wt%) in biodiesel production by transesterification of vegetable oils [2], which is available in large quantities accompanying the increase of biodiesel production. To enhance the economy of the whole biodiesel industry, various processes have been developed to catalytically convert glycerol to value-added chemicals [3, 4]. Among these processes, hydrogenolysis of glycerol to 1,2-propylene glycol (1,2-PG) and ethylene glycol (EG) is very fascinating. Both glycols are important chemical feedstocks, which are mostly used in the production of intermediates such as unsaturated resin and epoxy resin [5, 6]. At present, they are mainly produced from non-renewable petroleum derivatives such as propylene or ethylene. Compared with this traditional method, the process of obtaining these two glycols from renewable glycerol is more sustainable and green, and will be rapidly developed under the present situation. In literature, various catalysts, mainly based on noble metals such as Ru, Rh, and Pt [5–9], were reported for the hydrogenolysis of glycerol. Miyazawa et al. [7] found a metal–acid catalyst system, combining 5% Ru/C and Amberlyst-15, which was active in the hydrogenolysis reaction with 55% selectivity of 1,2-PG and 13% conversion. Alhanash et al. [9] have used Ru/CsPW as catalyst, the result showed 96% selectivity and 21% conversion. On the whole, with these catalysts being used, the reactivity is still not satisfactory. Furthermore, much research effort has been focused on noble metals catalysts in glycerol hydrogenolysis. Ni-based catalyst is an interesting candidate for a nonnoble metal catalyst in the hydrogenation of olefins, while there are few reports about the application of this kind of catalyst in glycerol hydrogenolysis reaction besides Perosa’s work [10], in which Raney Ni was tested and high
Hydrogenolysis of Glycerol on Ni/NaX
conversion could be obtained for a long reaction time ([40 h). Herein, inexpensive Ni-based catalysts supported on zeolites were applied to the hydrogenolysis of glycerol. The zeolite support played an important role in the enhancement of the activity, and the causes were studied. The effect of reaction parameters on the activity was also investigated.
2 Experimental 2.1 Catalyst Preparation Ni-based catalysts (10 wt% nominal loading) were prepared by incipient impregnation following chemical reduction with KBH4 aqueous solution, using NaMOR zeolite, NaZSM-5 zeolite, NaA zeolite, NaX zeolite, SiO2, and c-Al2O3 as supports, respectively. To study the effect of support acidity, NaX zeolite was ion-exchanged several times with 1.0 M NaNO3 solution at 80 °C for 3 h. The samples of exchanging for two times and four times were designated NaX1 and NaX2, respectively. The NaX0 zeolite was obtained by ion-exchanging with 1 M NH4NO3 solution for three times at 80 °C for 3 h. In a typical preparation, Ni-based catalyst was prepared as follows: NaX was impregnated for 24 h with equivalent volume of an aqueous nickel chlorate solution. After being dried the nickel precursor on the support was reduced by adding KBH4 (1.0 M) solution dropwise under vigorous stirring in an ice-water bath. The resulting black precipitate sample was then separated and thoroughly washed to remove the residual ions. The obtained catalyst is designated as Ni/NaX. For comparison, the unsupported Ni powder was prepared according to the above procedure without support, and Ni powder-NaX catalyst was prepared by mechanical mixing Ni powder with NaX. 2.2 Catalysts Characterization XRD data were obtained with Rigaku D/Max 3400 powder diffraction system using CuKa radiation (k = 0.1542 nm). Selected-area electron diffraction (SAED) of the catalyst was examined by transmission electron microscopy (TEM, JEOL JEM-2000EX). The nickel surface area (SNi) was determined by hydrogen adsorption with a Micromeritics ASAP 2920. The SNi of the catalysts was calculated assuming a stoichiomerty of one hydrogen molecule adsorbed per two surface nickel atoms. The amounts of Na? ions in NaX zeolites were analyzed by inductively coupled plasma (ICP, PLASAM-SPEC-II). The acidity of NaX zeolites was determined by NH3-TPD with a Micromeritics ASAP 2920 Autochem II system.
185
2.3 Catalytic Reaction and Products Analysis The reaction was investigated in a 600 mL autoclave (Parr instruments, USA). The typical procedure for hydrogenolysis was: 160 g 25 wt% glycerol aqueous solution and 2 g (7.8 mol%) catalysts (metal-based) were put into the autoclave. After replacing the air by H2 for several times, the autoclave was heated to the desired temperature and the H2 pressure was increased to the desired value. The products were analyzed by Agilent 4890 GC using capillary column HP-5 (30 m 9 0.53 mm 9 1.5 lm) equipped with a flame ionization detector. All products were confirmed by GC-MS with Agilent 6890 N GC/5973 MS detector and MS (Omnistar), and comparison of their GC retention time was made with those of the authentic samples. Liquid products including ethanol, n-propanol, acetol, EG, and 1,2-PG were detected. The gas products contain CO, CO2 and CH4. 1,2-PG and EG are the main products under these reaction conditions, and few 1,3-PG was detected. Considering the importance of 1,2-PG and EG in industrial application these two products were chosen for the further study and discussion.
3 Results and Discussion The catalytic hydrogenolysis of glycerol over nickel-based catalysts was investigated, and the results are listed in Table 1. For the tested catalysts including Ni/NaMOR, Ni/NaZSM-5, Ni/NaA, Ni/NaX, Ni/SiO2, Ni/c-Al2O3, the supports showed a strong influence on the catalytic activity and selectivity. Ni/NaMOR (entry 1) exhibited the lowest activity with only 14.0% conversion of glycerol, while Ni/NaZSM-5 (entry 2) gave the lowest selectivity of 1,2-PG (9.4%). When NaA and SiO2 (entries 3, 5) were used as support, moderate conversion was obtained. For Ni/c-Al2O3 (entry 6), it achieved remarkably high activity in terms of conversion but with 48.3% selectivity to ethanol, n-propanol, and other by-products. When using Ni/NaX, satisfactory results were obtained. The conversion was 94.5% with 72.1% selectivity for 1,2-PG, and 83.2% selectivity towards glycols (1,2-PG and EG) (entry 4). To the best of our knowledge, this is the first report using Ni/NaX for high efficient hydrogenolysis of glycerol to 1,2-PG and EG. Considering the significant difference in their catalytic performance, these Ni-based catalysts are characterized by XRD, SAED, and H2 chemisorption techniques. Figure 1a illustrates the X-ray diffraction within the range of 5–60° of all the as-prepared catalysts along with unsupported Ni powder for comparison. Besides the diffraction peaks corresponding to the supports, no characteristic diffraction peaks of metal nickel crystals were observed, but only a
123
186
J. Zhao et al.
Table 1 Results of glycerol hydrogenolysis over Ni-based catalysts supported on different materials SNi (m2/g)
Conversion (%)
Selectivity (%) 1,2-PG
EG
21.2
14.0
56.7
13.3 30.0
9.8
47.8
9.4
13.0 77.6
(a)
Others
a
(b)
Intensity (a.u.)
Entry Catalyst
(A)
1
Ni/NaMOR
2
Ni/NaZSM-5
3 4
Ni/NaA Ni/NaX
10.4 25.5
65.3 94.5
46.8 72.1
14.1 39.1 11.1 16.8
5
Ni/SiO2
27.9
56.9
44.4
8.8 46.8
(d)
6
Ni/c-Al2O3
22.2
97.1
44.2
7.5 48.3
(e) (f)
Reaction conditions: 160 g 25 wt% glycerol aqueous solution, 2.0 g catalyst (metal-based), 200 °C, 6.0 MPa H2, 10 h, stirring speed: 500 rpm a
(c)
(g) 0
10
20
30
40
50
60
2θ (degree)
Others: ethanol, n-propanol, acetol, CO, CO2 and CH4
(B) broad peak at 2h = 46°, which is attributed to the amorphous structure of the metal particles [11]. The SAED pattern (Fig. 1b) presented several diffraction cycles that further confirm the amorphous structure. The various nickel-based catalysts have different exposed nickel specific surface areas (Table 1). The catalytic performance, however, seems to be independent of these areas. Both acidity and hydrogenation capability of the bifunctional catalyst system are essential for the hydrogenolysis of glycerol to 1,2-PG [12]. H2WO4 [13] and Amberlyst [14] were identified as effective promoters to improve the activity and selectivity of the hydrogenolysis reaction. In the metal–acid bifunctional catalyst system, it is thought that the glycerol was dehydrogenated on acid sites and the generated dehydration intermediates were hydrogenated over the surface of metal sites. Similar phenomena were observed when a Cu–ZnO catalyst was employed [15]. In previous studies of our laboratory, metal oxide acidic sites had a tendency of adsorbing hydroxylcontained compounds [16]. Zeolites were believed to be employed not only as inert materials for dispersing and stabilizing metal clusters but also as necessary acidic sites suppliers. NaX might be favorable to adsorb glycerol molecules and then to increase the substrate concentration on the surface of the catalyst. The appropriate acidity of NaX was supposed to contribute to the acceleration. In order to confirm this supposition, a physical mixture of Ni powder and Ni powder-NaX was tested under identical conditions, and the results are shown in Fig. 2. When using Ni powder, the conversion was 79.5% and the selectivity of 1,2-PG was only 39.1%. This indicates that pure nickel is favorable to break C–C bonds, which leads to the formation of degradation products. The activity of the Ni powder-NaX physical mixture was comparable to that of unsupported Ni powder. The selectivity of 1,2-PG, however, was enhanced by 12.8% compared with unsupported Ni powder, and drastically enhanced by 33.0%
123
Fig. 1 a XRD patterns of Ni/NaMOR (a), Ni/NaZSM-5 (b), Ni/NaA (c), Ni/NaX (d), Ni/SiO2 (e), Ni/c-Al2O3 (f), unsupported Ni powder (g), b the SAED image of Ni/NaX catalyst
employing Ni/NaX as catalyst. Blank activity tests were carried out on NaX without nickel, which was inactive in hydrogenolysis of glycerol. The results described above show that the NaX zeolite support could facilitate the 1,2-PG production, and this might be attributed to the dehydration function of acid sites in the NaX zeolite. On the basis of the best catalytic performance presented by Ni/NaX in the hydrogenolysis of glycerol, we focused on NaX for a more detailed study. The acidity of NaX was tuned by ion-exchanging with NH4NO3 or NaNO3 solution and determined by NH3-TPD (Table 2). With the amounts of Na? ions in zeolites decreasing, the amounts of strong acid sites ([350 °C) consequentially increased in the order: NaX2 \ NaX1 \ NaX \ NaX0. Table 3 shows the catalytic performance of nickel-based catalysts supported on NaX. As expected, from Ni/NaX2 to Ni/NaX0, the conversion of glycerol gradually increased. This trend is in agreement with the results of NH3-TPD, suggesting that the catalytic activity is related to the amounts of strong acid sites in NaX zeolites. However, for the selectivity of 1,2-PG, it passes through a maximum using NaX1 as support instead of increasing monotously
Hydrogenolysis of Glycerol on Ni/NaX Conversion Selectivity of 1,2-PG Selectivity of EG
80
60
40
20
Ni powdera
Ni powder-NaXb
Ni/NaX
Fig. 2 Hydrogenolysis of glycerol over Ni-based catalysts prepared by different methods. Reaction conditions: 160 g 25 wt% glycerol aqueous solution, 2.0 g catalyst (metal-based), 200 °C, 6.0 MPa H2, 10 h, stirring speed: 500 rpm, a prepared the same way as Ni/NaX catalyst but without NaX support, b prepared by mechanical mixing of Ni powder and NaX support Table 2 The contents of Na? ions and acid amounts of NaX zeolites Sample Na? content (wt%)a Acid amounts (lmol g-1)b 150–250 °C 250–350 °C [350 °C NaX2
9.49
30
51
21
NaX1
9.32
40
40
41
NaX
8.78
17
35
60
NaX0
3.21
32
102
398
a
Determined by ICP
b
Determined by NH3-TPD
Table 3 Hydrogenolysis of glycerol over Ni based catalysts supported on NaX zeolites containing different amounts of Na? ions Catalyst
Conversion (mol %)
Selectivity (mol %) 1,2-PG
EG
Othersa
Ni/NaX2
79.8
72.1
10.5
17.4
Ni/NaX1
86.6
80.4
14.2
5.4
Ni/NaX Ni/NaX0
94.5 96.8
72.1 64.9
11.1 11.2
16.8 23.9
Reaction conditions: 160 g 25 wt% glycerol aqueous solution, 2.0 g catalyst (metal-based), 200 °C, 6.0 MPa H2, 10 h, stirring speed: 500 rpm a
100 1,2-PG 80 80 Glycerol
60
60 40
Selectivity (%)
0
achieved at 140 lmol H? and the additive of Amberlyst 70 was almost saturated under these conditions. Considering the tradeoff of conversion and selectivity, the appropriate amount of strong acidity in NaX was found to be essential for glycerol hydrogenolysis. Temperature has a significant effect on the glycerol reactivity and products selectivity over Ni/NaX catalyst as shown in Fig. 3. The increase in reaction temperature from 140 to 220 °C resulted in a monotonous increase in glycerol conversion from 22.9 to 95.6%. The selectivity toward 1,2-PG slightly decreased from 87.9 to 72.1% from 140 to 200 °C and then drastically declined to 39.7% at 220 °C. This is probably due to excessive hydrogenolysis converting 1,2-PG to lower alcohols such as methanol, ethanol at high reaction temperatures [12]. EG selectivity slightly increased from 8.6 to 11.1% as the temperature increased from 140 to 200 °C, and then declined to 6.8% at 220 °C. It was reported that more EG is obtained at lower temperatures than at higher temperatures, possibly attributing to rapid depleting of EG to other products at higher temperatures [18]. The influence of the H2 pressure on products distributions in the hydrogenolysis of glycerol is shown in Fig. 4. The conversion of glycerol drastically increased with increasing H2 pressure up to 6.0 MPa, and slowly increased when the pressure was further enhanced to 7.0 MPa, indicating that H2 was saturated the active sites at 6.0 MPa or higher. The selectivity towards 1,2-PG passed a maximum at 6.0 MPa H2 pressure. An excess H2 pressure resulted in a decrease of the 1,2-PG selectivity from the maximum of 72.1%. In view of the EG selectivity, there is no considerable change in the pressure range 3.0-7.0 MPa. However, Huang et al. [19] have reported a decrease of the EG selectivity at higher pressure.
Conversion (%)
Conversion and Selectivity (%)
100
187
40
Others: ethanol, n-propanol, acetol, CO, CO2 and CH4
20
EG 20
with the acidity. As discussed above, NaX plays a role of glycerol dehydration. The highest selectivity towards 1,2PG obtained with the Ni/NaX1 catalyst can be explained by assuming a sufficient amount of strong acid sites to be provided by the support. Miyazawa et al. [17] observed a similar trend in that the highest yield of 1,2-PG was
140
160
180
200
220
o
Temperature ( C)
Fig. 3 Influence of temperature on the catalytic performance of Ni/NaX in glycerol hydrogenolysis. Reaction conditions: 160 g 25 wt% glycerol aqueous solution, 2.0 g catalyst (Ni metal-based), 6.0 MPa H2, 10 h, stirring speed: 500 rpm
123
188
J. Zhao et al.
100
- H2O
80
90
NaX
Glycerol
80
60
1,2-PG
70 40 60 EG
Selectivity (%)
Conversion (%)
HO
20
50 40
3
4
5
6
7
0
H2Pressure (MPa)
Fig. 4 Influence of H2 pressure on glycerol hydrogenolysis over Ni/NaX. Reaction conditions: 160 g 25 wt% glycerol aqueous solution, 2.0 g catalyst (metal-based), 200 °C, 10 h, stirring speed: 500 rpm
The effect of reaction time on the hydrogenolysis of glycerol was investigated. As illustrated in Fig. 5, the conversion of glycerol reached 25% with 83.4% selectivity for 1,2-PG and 12.7% selectivity for EG after 2 h. The glycerol conversion gradually increased with time. When the reaction time was up to 10 h, a high level of glycerol conversion (94.5%) with slightly decreased selectivity of 1,2-PG was achieved, suggesting that the prolonged reaction time resulted in the decomposition of 1,2-PG. The selectivity towards EG was almost invariant during the 10 h reaction time, indicating that EG is mainly produced by glycerol degradation [9]. The present study confirmed the mechanism proposed by Dasari et al. [12] including dehydration of glycerol to form an acetol intermediate on acid sites and consecutive hydrogenation to 1,2-PG over metal sites. In our work, 100 100
80
1,2-PG
60 Glycerol
60
40 40
EG
Selectivity (%)
Conversion (%)
80
0 4
6
8
O
+ H2 Ni
OH (Process A) OH
OH OH
OH (Process B) OH
Scheme 1 Proposed glycerol hydrogenolysis route
trace amounts of acetol intermediate were detected by GCMS. The identification of acetol in the reaction products confirmed the mechanism that the glycerol is dehydrated to an acetol intermediate. This reaction scheme differs from that proposed by Montassier [20], involving dehydrogenation of glycerol to glyceraldehyde, followed by dehydration of the latter to 2-hydroxyacrolein, and subsequent hydrogenation of 2-hydroxyacrolein to 1,2-PG. The reasonable reaction pathway for the hydrogenolysis of glycerol over Ni/NaX catalyst is proposed in Scheme 1, which encompasses a two-stage catalytic reaction. 1,2-PG was formed in consecutive reactions from the dehydration of glycerol to intermediates on acid sites of NaX and subsequent hydrogenation of the intermediates on the nickel surface. When the acidic zeolite support HNaX was introduced, process A was accelerated in favor of glycerol dehydration, and then a high yield of 1, 2-PG was obtained. In addition, as mentioned above, the formation of EG was thought to be via direct cleavage of C–C bonds in glycerol. This is supported by the work of Miyazawa et al. [7, 17] on the hydrogenolysis of 1,2-PG and 1,3-PG, in which EG was not formed. Chiu et al. [18] also reported the absence of EG when acetol was used as the reagent.
4 Conclusion In summary, Ni/NaX catalyst was found to be effective in the catalytic hydrogenolysis of glycerol to glycols. The acidity of the zeolite support played an important role on the performance of the catalyst. Under optimized conditions, 86.6% conversion with 94.6% selectivity for glycols was obtained at 200 °C under 6.0 MPa H2 after reaction for 10 h. Acknowledgments We thank the Knowledge Innovation Program of the Chinese Academy of Sciences (K2006D1) support for our research.
20
20 2
OH
10
References
Reaction time (h)
Fig. 5 Influence of reaction time on glycerol hydrogenolysis over Ni/NaX. Reaction conditions: 160 g 25 wt% glycerol aqueous solutions, 2.0 g catalyst (metal-based), 200 °C, 6.0 MPa H2, stirring speed: 500 rpm
123
1. 2. 3. 4.
Ritter SK (2004) Chem Eng News 82:31 Huber GW, Iborra S, Corma A (2006) Chem Rev 106:4044 Corma A, Iborra S, Velty A (2007) Chem Rev 107:2411 Iriondo A, Barrio VL, Cambra JF, Arias PL, Guemez MB (2009) Catal Commun 10:1275
Hydrogenolysis of Glycerol on Ni/NaX 5. Maris EP, Davis RJ (2007) J Catal 249:328 6. Maris EP, Ketchie WC, Murayama M, Davis RJ (2007) J Catal 251:281 7. Miyazawa T, Kusunoki Y, Kunimori K, Tomishige K (2006) J Catal 240:213 8. Schlaf M, Ghosh P, Fagan PJ, Hauptman E, Bullock RM (2001) Angew Chem Int Ed 40:3887 9. Alhanash A, Kozhevnikova EF, Kozhevnikov IV (2008) Catal Lett 120:307 10. Perosa A, Tundo P (2005) Ind Eng Chem Res 44:8535 11. Liaw BJ, Chiang SJ, Chen SW, Chen YZ (2008) Appl Catal A 346:179 12. Dasari MA, Kiatsimkul PP, Sutterlin WR, Suppes GJ (2005) Appl Catal A 281:255
189 13. Chaminand J, Djakovitch L, Gallezot P, Marion P, Pinel C, Rosier C (2004) Green Chem 6:359 14. Kusunoki Y, Miyazawa T, Kunimori K, Tomishige K (2006) Catal Commun 6:64 15. Wang S, Liu HC (2007) Catal Lett 117:62 16. Sun ZQ, Xu J, Du ZT, Zhang W (2006) Chin J Catal 27:299 17. Miyazawa T, Koso S, Kunimori K, Tomishige K (2007) Appl Catal A 329:30 18. Chiu CW, Tekeei A, Ronco JM, Banks ML, Suppes GJ (2008) Ind Eng Chem Res 47:6878 19. Huang L, Zhu YL, Zheng HY, Li YW, Zeng ZY (2008) J Chem Technol Biotechnol 83:1670 20. Montassier, C, Giraud D, Barbier, J (1988) Heter Catal Fine Chem 165
123
