Catalytic ethanolysis of Kraft lignin to small-molecular liquid products over an alumina supported molybdenum nitride catalyst

Accepted Manuscript Title: Catalytic ethanolysis of Kraft lignin to small-molecular liquid products over an alumina supported molybdenum nitride catalyst Authors: Mengmeng Chen, Wenyue Hao, Rui Ma, Xiaolei Ma, Le Yang, Fei Yan, Kai Cui, Hong Chen, Yongdan Li PII: DOI: Reference:

S0920-5861(17)30534-5 http://dx.doi.org/doi:10.1016/j.cattod.2017.08.012 CATTOD 10955

To appear in:

Catalysis Today

Received date: Revised date: Accepted date:

22-2-2017 26-7-2017 4-8-2017

Please cite this article as: Mengmeng Chen, Wenyue Hao, Rui Ma, Xiaolei Ma, Le Yang, Fei Yan, Kai Cui, Hong Chen, Yongdan Li, Catalytic ethanolysis of Kraft lignin to small-molecular liquid products over an alumina supported molybdenum nitride catalyst, Catalysis Todayhttp://dx.doi.org/10.1016/j.cattod.2017.08.012 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Catalytic ethanolysis of Kraft lignin to small-molecular liquid products over an alumina supported molybdenum nitride catalyst Mengmeng Chena, Wenyue Haoa, Rui Maa, Xiaolei Maa, Le Yanga, Fei Yana, Kai Cuia, Hong Chen b, Yongdan Lia,c* a

Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin

Key Laboratory of Applied Catalysis Science and Technology, State Key Laboratory of Chemical Engineering (Tianjin University), School of Chemical Engineering, Tianjin University, Tianjin 300072, China b

School of Environmental Science and Engineering, Tianjin University, Tianjin 300072,

China c

Department of Chemical and Metallurgical Engineering, School of Chemical Engineering,

Aalto University, Kemistintie 1 , Espoo, P.O. Box 16100, FI-00076 Aalto, Finland.



Corresponding author. Tel.: +86-22-27405613; Fax: +86-22-27405243. E-mail address:

[email protected] (Y. Li).

1

Graphical Abstract

Highlights     

An alumina supported molybdenum nitride catalyst for lignin ethanolysis The products are alcohols, esters, mono-phenols, benzyl alcohols and arenes The preparation and reaction conditions have great influences on the activity 907 mg/g lignin aliphatics and 282 mg/g lignin aromatics are obtained No significant activity loss is observed in three runs.

Abstract An alumina supported molybdenum nitride catalyst (Mo2N/Al2O3) is examined for the ethanolysis of Kraft lignin in supercritical ethanol in a batch reactor with an initial nitrogen pressure of 0 MPa (gauge). No char or tar is formed in this process. The Mo loading as well as the nitriding temperature in the catalyst preparation process has a great influence on the activity. The best performance is obtained over 30 wt% Mo2N/Al2O3 catalyst nitrided at 700 o

C. The overall yields of sulfur-free small-molecular products achieve the maximum value of

1189 mg/g lignin, with aliphatic compounds accounting for 907 mg/g lignin and aromatic compounds 282 mg/g lignin, respectively. Both the product yields and the molecular distribution are also strongly dependent on the reaction time and temperature. Furthermore, the Mo2N/Al2O3 catalyst exhibits an excellent recycle performance with no significant activity loss after at least three runs. After reaction, the bulk structure of Mo2N/Al2O3 catalyst was wellpreserved only with a slight surface oxidation. 2

Keywords: Kraft lignin; Catalytic ethanolysis; Supercritical ethanol; Small-molecular products; molybdenum nitride 1. Introduction Lignin, as one of the three essential components of lignocellulosic biomass, is a threedimensional amorphous biopolymer consisting of various methoxylated phenylpropanoid building blocks [1, 2], which are now mainly obtained from fossil-based industry. It accounts for 15-30% by weight and 40% by energy of all the lignocellulosic biomass. Kraft lignin was a bulk waste of the pulping and paper industry for a long time, and in recent three decades, it has been isolated from the black liquor due to the environmental regulations in the commercialized countries. Therefore, efficient utilization of Kraft lignin will not only address the issue of environmental impacts, but also become an attractive and promising approach for the production of alternative fuels and chemicals due to their high heating value and unique aromatic nature. However, in the present time Kraft lignin is just used as a low-grade boiler fuel or in the production of a number of low value-added materials, e.g., flocculating and dispersing reagents in construction [3-6]. A number of strategies, e.g., pyrolysis, liquid phase reforming and hydroprocessing, have been adopted in the utilization of Kraft lignin [5]. For pyrolysis, the aromatic hydrocarbon yield of 17.5% was obtained over a Mo2N/γ-Al2O3 catalyst at 700 oC [7]. For the liquid phase processes, water and ethanol were used as hydrogen-donor solvents. In supercritical water, the yield and selectivity of the products depended on the reaction conditions and capping reagents [6, 8, 9]. The yield of water-soluble products for the depolymerisation of alkali lignin reached 16.4 wt% at 390 oC for 0.1 h with p-cresol as the capping reagent. Zakzeski et al. [10] found that the solubility of Kraft lignin in a hot ethanol-water mixture was much higher than that in pure water at 225 oC, and 17.6 wt% selectivity to aromatic products was obtained over an H2SO4 promoted Pt/Al2O3 catalyst. Jongerius et al. [11] obtained 9 wt% overall yield of 3

monoaromatic products in a similar process with Pt/Al2O3 and NaOH as co-catalysts at 225 oC for 2 h under 5.8 MPa argon. The same group investigated the hydrodeoxygenation performance of the lignin-derived products over CoMo/Al2O3 and Mo2C/CNF catalysts, and obtained 6 wt% and 7 wt% yields of monoaromatic products, respectively. An activated carbon supported tungsten phosphide catalyst gave 6.7 wt% yield of phenols in a hot compressed water-ethanol solvent without an additional capping reagent [12]. However, all the abovementioned works suffer from several challenges, e.g., low yields, corrosion, high temperature and pressure, and char or tar formation that causes the clogging issues. Ethanolysis strategy made an interesting advance. Ford et al. [13] converted woody biomass solids to a liquid mixture of C2-C6 aliphatic alcohols and methylated derivatives in superctitical methanol at 300-320 oC and 16-22 MPa using a copper-doped porous metal oxide. Cheng et al. [14] found that the acetone-methanol soluble products from the degradation of lignin in supercritical ethanol had a lower molecular weight than those obtained in water. Recently, we reported the catalytic ethanolysis of Kraft lignin over α-MoC1-x/AC and Mo/Al2O3 catalysts at 280 oC for 6 h, and obtained 1640 and 1390 mg/g lignin small-molecular products, respectively, on the two catalysts without any char or tar formation [15, 16]. Huang et al. [17] reported a Kraft lignin conversion result with 23 wt% yield of monomeric aromatics without formation of char or tar at 300 oC under inert atmosphere over a CuMgAlOx catalyst. Several other Mo-based catalysts, e.g., MoS2, Mo2C and NiMo/Al2O3, were reported to be active in the conversion of biomass in the hydrogen-donor solvents [18-20]. Compared with noble metals, molybdenum nitride was proved to exhibit the similar performance regarding hydrogen activation [21], and successfully applied for the hydrodeoxygenation of guaiacol, a kind of lignin model compound [22, 23]. Besides, on the premise of high yields, the synthesis of molybdenum nitride in the atmosphere of N2/H2 mixture is more accessible and cleaner than that of other Mo-based materials. 4

Herein, an alumina supported molybdenum nitride catalyst is applied to catalyze the ethanolysis reaction of Kraft lignin in supercritical ethanol. The nitriding temperature and Mo loading of the Mo2N/Al2O3 catalyst are investigated to explore the optimal preparation conditions and related catalytic activity. Besides, the effect of reaction conditions on the yield of products is also investigated. With the aid of XRD and XPS analysis, the bulk structure of Mo2N/Al2O3 catalyst and the valence state of Mo species for the catalytic ethanolysis of Kraft lignin are discussed. 2. Experimental 2.1. Materials The Kraft lignin purchased from Sigma-Aldrich (product number 471003) was used without further purification. The weight-average molecular weight (Mw) and number-average molecular weight (Mn) of the lignin are 60,000 and 10,000 g/mol, respectively. The original lignin contains 49.5 wt% C, 4.71 wt% H, 0.15 wt% N, 2.80 wt% S and 19.4 wt% ashes [15]. The Kraft lignin was dried overnight at 100 oC prior to use. The weight loss was less than 0.2%. Solvents and chemicals, including ethanol, p-cresol, ammonia water and ammonium molybdate (NH4)6Mo7O24·4H2O (AR reagent grade, Tianjin Guangfu Technology Development Co. Ltd.) were used as received. The alumina support was kindly offered by CNOOC Tianjin Chemical Research & Design Institute. The deionized water was prepared with an Ulupure ultrapure water purification machine (UPH-1-10). 2.2. Catalyst preparation The alumina supported molybdenum nitride catalyst was synthesized with an incipientwetness impregnation method. For a typical preparation, 3.00 g alumina of 40-60 mesh were impregnated at ambient temperature overnight with the 3.60 g ammonia solution containing 5

ammonium molybdate (for Mo loading of 10, 20, 30, and 46 wt%, the dosage of ammonium molybdate was 0.61, 1.38, 2.36 and 4.70 g, respectively). In the impregnation process, ammonia solution was added to improve the solubility of ammonium molybdate which was completely dissolved even at the loading of Mo up to 46%. No external materials issues were observed. Then the mixture was dried at 110 oC for 12 h. The precursor underwent nitridation in a flow of N2 and H2 (v:v = 1:5) mixture accompanied with a four-stage temperature programmed procedure, which was set from 20 to 350 oC at a heating rate of 10 oC /min, and then from 350 to 500 oC at a heating rate of 1 oC /min. Finally, the temperature was increased to the preset value at a heating rate of 2 oC /min and held at that temperature for 2 h. After cooling to ambient temperature, the samples were directly used in the reaction without exposing to air. The materials used for XRD characterization were passivated in a flow of 100 ml/min (STP) industrial nitrogen (including small amount of oxygen) at ambient temperature for 3 h. The passivation was utilized to avoid the excessive oxidation before the characterization. 2.3. Catalyst characterization X-ray diffraction patterns (XRD) of the passivated materials were measured with a powder diffractometer (Bruker AXS D8-S4), equipped with a Cu-Kα radiation source. It was operated at 40 kV and 40 mA, and the scanning was carried out in the 2θ range between 30 o and 80 o at a rate of 1.8 o/min. The texture of the samples was determined with a nitrogen adsorption-desorption experiment using a Quantachrome Autosorb-1 system. The samples were outgassed under high vacuum degree at 250 oC for 6 h before measurement.

6

X-ray photoelectron spectroscopy (XPS) measurements were carried out with a PerkinElmer PHI-1600 spectrometer using monochromatic Mg Kα radiation. Binding energy was calibrated with respect to the signal for C1s at 284.8 eV. 2.4. Lignin ethanolysis reaction The reaction of Kraft lignin was carried out in a 300 mL batch reactor (Parr 4566, made of Hastelloy) equipped with a temperature controller (Parr 4848). In a typical run, 1.00 g lignin, Mo2N/Al2O3 catalyst (0.46, 0.23, 0.15 and 0.10 g corresponding to 10, 20, 30 and 46 wt% Mo loading) and 100 mL ethanol were loaded into the reactor. The reactor was purged with highpurity nitrogen for five times and sealed with an initial pressure (gauge) 0 MPa. Then the reactor was heated to the required temperature and kept for the desired reaction time under the constant stirring of 400 rpm. At the end of reaction, the reactor was quickly cooled to ambient temperature by immersing into a cold-water bath. The products in the gas phase were absorbed with alkaline hydrogen peroxide for the quantitative analysis of sulphur element. The liquid products were obtained through filtration and rotary evaporation prior to the analysis. 2.5. Product analysis The products collected from the gas phase were measured and confirmed using a HIDEN HPR20 mass spectrometer, and the sulphur content of the absorbed products in the gas phase was quantified with DX600 ion chromatography produced by Dionex. The products in the reactor were separated into ethanol soluble portion and residue (R1) after filtration. The residue was washed using diluted hydrochloric acid (pH=2) to dissolve the ashes and weighed after drying (R2). The ethanol soluble products were enriched with removing a part of the ethanol with a rotatory evaporator. Then Agilent 6890-5973 GC-MS and Agilent 6890 GC-FID were applied for qualitative and quantitative analysis of the liquid products, respectively. The compounds were confirmed by comparing the mass spectra 7

obtained from the samples to standard spectra in the system’s database (NIST 2.0). An internal standard method was adopted for the quantitative calculation of product yield. P-cresol was used as an internal standard. The working conditions and programs for both Agilent GCs were the same as reported in our previous work [15, 16]. The oven temperature program was set from an initial temperature of 45 oC to a final temperature of 250 oC at a rate of 10 oC/min, and then kept at 250 oC for 2 min. Both chromatographic column used were HP-5 MS capillary column (30m × 0.25 mm × 0.25 µm). A split ratio of 50 was used for the GC-FID and GC-MS analysis. The scanning m/z range of mass spectrometer was set from 10 to 500. 3. Results and discussion 3.1. Catalytic conversion of lignin The catalytic conversion of Kraft lignin was carried out under a pressure ranging from 9.5 to 12.5 MPa depending on the reaction temperature. Few small-molecular products were obtained when the condition was below 243 oC and 6.4 MPa (critical point of ethanol) [15]. Via one-pot ethanolysis process with Mo2N/Al2O3 catalyst in supercritical ethanol, Kraft lignin was converted into various products in different phases, i.e., gas, liquid and solid phase. The products in the gas phase detected were mainly composed of H2 together with CO, CH4, and CO2. The gas products are mainly formed from the self-transformation of ethanol. A small portion of gas products may be derived from the cleavage of lignin side-chain and the consequent reactions, e.g., the cleavage of C-O-C linkages and the elimination of methoxy groups. The ion chromatograph technique confirmed the existence of the trace amount of sulphur in the form of H2S, which was attributed to the hydrodesulfurization performance of Mo2N/Al2O3 catalyst in the ethanolysis reaction of Kraft lignin. After the reaction time of 6 h, the mass of R2 obtained through filtration, washing and drying was almost equal to that of the initial added fresh catalyst, implying that the remaining 8

unreacted Kraft lignin was at an extremely low level, and negligible amount of tar or char was formed in the process. In other words, nearly complete catalytic liquification of Kraft lignin was achieved. More than 50 kinds of molecules were identified in the liquid products, and were considered as formed from Kraft lignin conversion reactions. Among them, 29 compounds constituting more than 80% of the overall peak area of GC-FID were regarded as the main products and divided into five categories, including C6 alcohols, C8-C10 esters, benzyl alcohols, mono-phenols, and arenes. The only S-containing compound identified in the liquid products is hexyl-ethyl thioether with a yield of 29 mg that contains 6 mg sulphur, indicating that 21 wt% sulphur in the feedstock were transformed to light component and the rest was supposed to exist in the GC undetectable products. However, the yield of thioethers was not included in that of the overall products. The product yields and distribution were strongly dependent on the catalyst preparation, e.g., nitriding temperature and Mo loading, and reaction conditions, e.g., reaction time and temperature. These factors were discussed in detail as follows. 3.1.1. Effect of the factors in catalyst preparation 3.1.1.1. Effect of the nitriding temperature The overall yield obtained without catalyst at 280 oC was quite low (173 mg/g lignin), as reported in our published work [15]. The effect of the nitriding temperature on the overall and grouped product yields over the catalyst samples examined in this work is listed in Table 1. The grouped product yields present the following sequence: esters > aromatics > alcohols > thioethers. Furthermore, the share of aliphatic products in all products is higher than 75% in the investigated range of nitriding temperature. It indicates that both ethanol selftransformation and lignin depolymerisation took place in this process [24]. The self9

transformation of ethanol mainly involves the oxidation of ethanol into acetaldehyde and the dehydrogenation of ethanol into hydrogen. As reported in our published work [24], the aromatics are from lignin, and the aliphatics are derived from both the cleavage of the lignin side-chain and the self-transformation of ethanol. Unfortunately, we cannot confirm to what degree the products are from ethanol or from lignin, because there may be some interactions between the lignin-derived and ethanol-derived products based on the complicated reaction mechanism and pathways, which makes it unclear to affirm the origin of products. When the nitriding temperature was at 650 oC, the yields of sulfur-free products and thioethers were only 943 mg/g lignin and 25 mg/g lignin, respectively. Over the 700-Mo2N/Al2O3 catalyst, the overall yields of sulfur-free products and thioethers reached the maximum of 1189 and 29 mg/g lignin, respectively. The aromatic yields, being 282 mg/g lignin, accounts for 23.7 wt% of all the sulfur-free products and 28.2 wt% of the original lignin reactant loaded into the reactor. However, with the further increase of the nitriding temperature, the overall yield of sulfur-free products sharply decreased to 745 mg/g lignin and the yields of overall aromatics were only 175 mg/g lignin. It should be noted that the overall yield of of the sulfur-free compounds and the grouped products follow the same order: 700-Mo2N/Al2O3 > 650-Mo2N/Al2O3 > 750Mo2N/Al2O3. In addition, the formation of thioethers reveals that the Mo2N/Al2O3 catalyst possesses a good resistance to sulfur and presents an excellent performance of hydrodesulfurization. It is interesting to note that the sequence of the thioethers yield along with the increase of the nitriding temperature is in good accordance with that of the overall yield of the sulfur-free products, implying that in the process of lignin ethanolysis, Mo2N/Al2O3 catalyst not only cleaved the C-O-C and C-C linkages in the lignin units, but also efficiently removed the sulfur element in the form of H2S and small-molecular thioethers, resulting in sulfur-free aromatic monomers. The activity of the catalyst shows relevance to the extent of

10

nitridation of MoOx, and the nitridation was completed at 700 oC [25]. The nitriding temperature plays a key role on the nitriding degree that causes the differences in the products. Fig. 1 and Table 2 present the trend of the yields of 29 kinds of sulfur-free products versus the nitriding temperature of the catalyst in detail. We can recognize the pieces from ethanol in the product molecules. Ethanol here acts as a reactant and a capping agent that deactivates the active intermediates from lignin in case of repolymerization reaction. From Fig. 1, it can be observed that 700 oC was the optimal nitriding temperature by achieving the highest yields of alcohols and esters, and then followed by 650 oC. It is evident that the yield order of hexanol > 2-methyl-2-pentenol > 3-hexenol > 2-ethyl-butanol can be applied to all the tested catalysts. Hexanol yield was as high as 60 mg/g lignin over the 700-Mo2N/Al2O3 catalyst. 2-ethylbutanol yield was only 8 mg/g lignin when the nitriding temperature was 750 oC. The yields of 2-hexenoic and 3-hexenoic acid ethyl ester were up to 285 and 243 mg/g lignin, respectively, using the 700-Mo2N/Al2O3 catalyst. The yields of C8 unsaturated esters, i.e., hexenoic acid ethyl esters, are much higher than those of C8 saturated esters and C10 octenoic acid ethyl esters despite of different catalysts employed. The yields of both 3-methyl-valeric acid ethyl ester and octenoic acid ethyl esters are insensitive to the nitriding temperature and almost remain constant in the whole investigated range. As presented in Table 2, over the 700-Mo2N/Al2O3 catalyst, benzyl alcohols with a yield of 155 mg/g lignin accounts for 55 wt% of the aromatics, among which the yield of 2-methyl benzyl alcohol as the main product reached the highest of 71 mg/g lignin, and then followed by 4-methyl benzyl alcohol. Compared to the other aromatics, the yields of mono-phenols are the lowest for the catalysts obtained at all nitriding temperatures, ranging between 21 and 36 mg/g lignin, and fail to show that distinct tendency as benzyl alcohols do with the increase of the nitriding temperature. 1,4- and 1,2-dimethyl benzene as the main arene products, also yielded the highest of 34 and 15 mg/g lignin, respectively, over the 700-Mo2N/Al2O3 catalyst, 11

while no remarkable difference was observed for the catalysts obtained at the other two nitriding temperatures. In some cases, certain products of mono-phenols and arenes were undetected, which was ascribed to extremely low yields that were beyond the detection limit of GC-FID. In comparison with the results over -MoC1-x/AC catalyst reported in our previous work [15], the aromatic yields over Mo2N/Al2O3 catalyst are almost the same while the distribution of the aromatic products is quite different. Higher yields of benzyl alcohols and lower yields of mono-phenols are obtained over Mo2N/Al2O3 catalyst. This can be attributed to the demand of weak deoxygenation effect on benzyl alcohols and strong deoxygenation effect on mono-phenols [23]. 3.1.1.2. Effect of the Mo loading The effect of the Mo loading of the 700-Mo2N/Al2O3 catalyst on the yields of aromatic and aliphatic products is shown in Fig. 2A. The overall yields of sulfur-free products increased with the increase of the Mo loading, and reached the maximum of 1189 mg/g lignin at a 30 wt% Mo loading, and then suffered from a significant drop with the further increase of the Mo loading to 46 wt%. Aromatic products also present the same trend. The highest yield of aromatics was up to 282 mg/g lignin when the Mo loading was 30 wt%. It indicates that the increased Mo loading is conductive to the enhancement of the overall yields, especially aromatic yields. The poor activity and low yields over 46 wt% 700-Mo2N/Al2O3 catalyst may be attributed to the reduced active surface area caused by Mo2N cluster growth. Even so, it is still proved that Mo2N/Al2O3 exhibits an excellent catalytic performance in the linkage cleavages between two lignin aromatic subunits, resulting in the formation of small-molecular aromatic monomers. On the contrary, the yields of aliphatic products, including alcohols and esters, decreased monotonically with the increase of the Mo loading. The yields of aliphatic products was ranging from 992 to 541 mg/g lignin, indicating that in the investigated range, 12

the lower Mo loading favours the production of aliphatic products. This may be attributed to that this range is over the optimum Mo loading based on a reasonable speculation that there exists an optimum value. 3.1.2. Effect of the reaction conditions 3.1.2.1. Effect of the reaction time The effect of the reaction time on the product yields was investigated over the 700Mo2N/Al2O3 catalyst. As illustrated in Fig. 2B, a sequential order of grouped yields as the reaction time is presented as follows: esters > aromatics > alcohols, and esters accounts for over 60 wt% of all the products from beginning to the end. Only 201 mg/g lignin aliphatics, including 15 mg/g lignin alcohols and 186 mg/g lignin esters, together with 48 mg/g lignin aromatics were obtained when the reaction time was 2 h, implying that in this condition, the depolymerisation reaction of Kraft lignin is insufficient. The yields of alcohols and esters increased to 33 and 387 mg/g lignin, respectively at a reaction time of 4 h, accompanied with 114 mg/g lignin aromatics. It is worth noting that the grouped yields obtained at 4 h almost doubles those obtained at 2 h, indicating that in the first 4 h of reaction time, the formation rate of every grouped products is at a relatively slow level and basically remains constant. When the reaction time was further prolonged to 6 h, the overall yields dramatically increased to 1189 mg/g lignin, almost four times higher than those obtained at a reaction time of 2 h. It manifests that in the last two hours, the formation rate of sulfur-free products is faster than that in the first four hours. The yields of aromatics also showed a sharp increase to 282 mg/g lignin. In addition, the mass percentage of aromatics in all sulfur-free products increases from 19.3 wt% to 23.7 wt% with prolonging the reaction time from 2 h to 6 h. In general, the yields of liquid products are strongly dependent on reaction time. Prolonging reaction time is quite conductive

13

to the enhancement of sulfur-free product yields and the improvement of aromatic products selectivity. As reported in the published literature [15, 26], Kraft lignin undergoes a two-step depolymerisation process in the supercritical ethanol. Firstly, due to the attack of radicals, Kraft lignin macromolecule is ethanolized into middle molecular-weight fragments in the m/z range of 700-1400. Then these fragments are further transformed into detectable small molecularweight compounds in the presence of catalyst and diverse radicals. In this work, it is likely that, in the first four hours of reaction, the first step represents a slow-rate reaction, accompanied with the continuous accumulation of middle molecular-weight fragments. In the last two hours, these fragments are rapidly reacted and transformed into small molecular-weight products in the presence of Mo2N/Al2O3 catalyst. 3.1.2.2. Effect of reaction temperature As depicted in Fig. 3, the yields and selectivity of aromatic products depend strongly on the reaction temperature. The yields of arenes and phenols increased with an increase of the reaction temperature, and reached the maximum of 102 and 43 mg/g lignin, respectively at 290 o

C, and then showed a slight decrease. The overall yield of aromatic products presents the same

trend. It indicates that a moderate increase of reaction temperature favours the cleavage of CO-C and C-C linkages between the lignin subunits, and enhances the yields of arene and monophenol products. However, the optimal temperature for the production of benzyl alcohols was 280 oC. Then the yields presented a drastic drop when the reaction temperature further increased to 300 oC, which might be attributed to the repolymerisation reaction of intermediates derived from Kraft lignin. As illustrated in Fig. 3B, with the increase of the reaction temperature, the selectivity of benzyl alcohols in all aromatic products shows the same variation tendency as the yields, while 14

an opposite trend occurs to arenes. This result arouses an speculation that partial mutual transformation between benzyl alcohols and arenes may take place in this conditions although no direct evidence is given. The selectivity of phenols increases monotonically with the increase of the reaction temperature, which not only is beneficial from the readily cleaved CO-C ether linkages induced by elevated tempereature to enhance the yield of phenols, but also is ascribed to the strongly inhibitory effect of elevated temperature on the yields of benzyl alcohols. 3.1.3. The recycle test The spent 30 wt% 700-Mo2N/Al2O3 catalyst was recovered via filtration and washing, and was reused for the next run under the same condition. The recyclability results are illustrated in Fig. 4. After three consecutive runs, only a slight decrease in activity is observed, which is well in the range caused by the weight loss during the catalyst collection and transfer. It also demonstrates that Mo2N/Al2O3 catalyst exhibits a high stability and activity towards the catalytic ethanolysis of Kraft lignin into small-molecular products. It should be noted that the yield of alcohols almost keeps constant in the recycle test, indicating that the weight loss has no significant impact on the production of alcohols. 3.2. Catalyst characterization The texture data of Al2O3 support and passivated 30 wt% 700-Mo2N/Al2O3 catalyst are listed in Table 3. After loading Mo2N onto the Al2O3 support, the specific surface area and total pore volume decreased from 270 m2/g and 0.759 m2/g to 169 m2/g and 0.459 cm3/g, respectively. As shown in Table 3, the average pore diameter decreases slightly from 11.3 to 10.9 nm, indicating that both samples mainly contain mesopores. The XRD pattern of passivated Mo2N presented in Fig. 5 fits well with the data in JCPDSICDD 25-1368, showing peaks at 2θ = 37.7o, 43.1o, 45.3o, 62.7o, 64.3o, 75.5o and 78.6o, which 15

are associated with the (112), (200), (004), (220), (204), (312) and (116) planes, respectively. It should be noted that the oxidation during the passivation only took place on the surface as no sign of bulk oxidation, e.g., MoO3 or MoO2 peaks, emerged in the XRD pattern, confirming that the precursor was completely converted to molybdenum nitride [27]. The diffraction peaks of Mo2N phase are observed for the high loading samples, and no diffraction peaks of any other oxides are observed, fairly clear as depicted in Fig. 5, indicating that Mo2N phase is well formed. The crystallite size of 30 wt% 700-Mo2N/Al2O3 estimated from XRD is about 10 nm. However, for the samples with lower Mo loading, only the diffraction peaks of Al 2O3 support are observed in the XRD pattern, which is due to the high dispersion of Mo2N species over the support [28]. It can also be seen that the bulk structure of Mo2N remains the same after reaction. However, a slight change occurs to the diffraction pattern of the residue. The newly recognized peak centred at 35o cannot be related to the Mo containing phases, and thus may be assigned to the crystals of the ashes from Kraft lignin. The surface Mo 3d XPS spectra of binding energy (BE) region are presented in Fig. 6, which depicts a partial superposition of doublet (Mo 3d5/2 and Mo 3d3/2) with an intensity ratio of 3:2 with a Mo 3d5/2 BE of 232.7 and 229.1 eV. The exact values depend on the BE reference chosen for sample correction, and are susceptible to the support. The difference of BE between Mo6+ in MoO3 and Mo4+ in the nitrided sample, 3.2±0.4 eV, is in close proximity to that reported in literature [27, 29]. The occurrence of a detectable signal at 236.5 eV is in good agreement with a partial oxidation to Mo6+ as a result of passivation. The Mo4+ peak at BE = 229.1 eV is appreciably weak after reaction while the Mo6+ peak at BE = 232.7 eV is well defined, demonstrating the nearly complete oxidation of Mo4+ occurred on the surface of Mo2N/Al2O3, and the oxidation process may result in the slight decrease of catalytic activity in the recyclability test of Mo2N/Al2O3. However, it should be noted that after reaction, the peak intensity of Mo6+ located at both 232.7 and 236.5 eV is much weaker than that obtained before 16

reaction because of the lower Mo element content in the solid sample, which can be attributed to the dilution effect of ashes from Kraft lignin. Román-Leshkov’s group manifested that the unstable Mo4+ species are inactive in the hydrodeoxygenation reaction of phenolic compounds, and Mo5+ species are responsible for the hydrodeoxygenation reaction of biomass-derived oxygenates over a MoO3 catalyst [30]. Our group proposed a common mechanism that Mo5+ is the actual active site in the catalytic ethanolysis process of Kraft lignin over different Mo-based catalysts [24]. In the ethanolysis reaction, lignin biomacromolecule is incapable of diffusion into the pores due to sterichinerance effect, and traditional theory of heterogeneous catalysis cannot apply to this reaction system. The exterior catalyst surface provides primary active sites at the beginning of reaction. Due to steric hindrance, the linkage cleavage of Kraft lignin into radical fragments initially took place on the exterior catalyst surface before solid Mo-containing species can dissolve into the solvent. Therefore, a novel feature of Mo-based catalysts is postulated that solid Mocontaining species on the surface of catalyst are dissolved into the ethanol solvent, and then act on the lignin linkages in the form of dissociative molybdenum (V) ethoxide. In other words, the contributions of presumable active sites are expected to originate from Mo5+ species, which are derived from the partial surface redox during the reaction. However, the direct evidence, e.g., ICP characterization, cannot be given to prove this speculation due to the deposition of dissociative molybdenum (V) ethoxide back to solid state at ambient temperature after reaction. Blind tests without catalyst are unable to shed valuable data for further interpretation and analysis [15]. Similarly, blind tests with catalyst and ethanol only (without lignin) yielded acetaldehyde, ethyl acetate, butanol, and 1,1-diethoxyethane as main products. However, no quantified products were formed [24]. The presence of Mo4+ in Fig. 6 indicates that the Mo2N/Al2O3 catalyst undergoes a gradual surface oxidation process from Mo4+ to Mo5+ and Mo6+. As for the existence of Mo6+, it can be ascribed to the excessive oxidation of Mo2N 17

during the catalyst recovery and drying. In this work, the Mo5+ species are not detected by the XRD and XPS methods, which may be attributed to the unstable state of Mo5+ species, excessive oxidation and ex-situ characterization. Detailed reaction pathway of Kraft lignin ethanolysis reaction over the Mo2N/Al2O3 catalyst is under rationalization in our lab [24]. In addition, the chemical pathway may involve some other reactions, e.g., saturation of aromatics and ring-opening. 3.3. General discussion Compared to the traditional catalytic strategies that feature the mono-phenols compounds as the main or sole products [31-34], the ethanolysis route produces alcohols and esters, as well as aromatic compounds, i.e., benzyl alcohols, arenes and mono-phenols [15, 16, 24]. This specialty endows ethanolysis strategy with unique advantages and powerful attraction in the biorefining process of Kraft lignin. The inert atmosphere and alcoholic solvents, e.g., methanol and ethanol, are common for the ethanolysis reaction of lignin, and no additional gaseous hydrogen molecule is introduced. Hydrogen was proved to be obtained from methanol reforming and water gas shift reactions over copper-doped porous metal oxide catalyst [35] or ethanol hydrogen-donor reactions over CuMgAlOx catalyst [17]. Furthermore, the formed hydrogen plays an important role on the hydrodeoxygenation process of lignin. Ford et al. [35] demonstrated that in-situ hydrogen from hydrogen-donor solvents exhibits a better activity than additional gaseous hydrogen, which also provides a reasonable explanation for the fact that inert atmosphere is more conductive to the ethanolysis reaction than hydrogen atmosphere due to the inhibitory effect of additional hydrogen on the in-situ hydrogen. Nevertheless, in the action of various Mo-based catalysts, our group reported that ethanol shows a better behaviour than that of methanol [15, 24], indicating that ethanolysis performance fails to follow the same rule as hydrogen-donor ability. 18

This may be ascribed to the formation of dissociative molybdenum (V) ethoxide and the transition from heterogeneous to homogeneous catalysis on account of higher solubility of Kraft lignin at elevated temperature [36]. The solvent substitution of both pure water and isopropyl alcohol for ethanol brings about remarkably low yields. This also testifies the importance of dissociative molybdenum (V) ethoxide on the other side although no direct supporting evidence is provided. Therefore, the ethanolysis activity of Kraft lignin is strongly dependent on the atmosphere, solvent and catalyst. The selection and confirmation of these conditions will be the research topics of the ethanolysis route. In addition, it is still underway to find a common mechanism that can be applied to various catalytic systems. 4. Conclusion A Mo2N/Al2O3 catalyst has been employed in the ethanolysis of Kraft lignin to smallmolecular products. The 30 wt% 700-Mo2N/Al2O3 sample displays the best performance in the product yields. The highest overall yields of sulfur-free products, being 1189 mg/g lignin, were obtained at 280 oC for 6 h, including 907 mg/g lignin aliphatic products and 282 mg/g lignin aromatic monomers. The yield of thioethers also reached the maximum of 29 mg/g lignin under the same condition. The Mo2N/Al2O3 catalyst exhibits a good activity of hydrodesulfurization and hydrodeoxygenation. The spent catalyst can be reused at least for three runs without distinct activity loss. Evidence for the surface oxidation of the Mo2N/Al2O3 catalyst after reaction is presented while the bulk structure remained the same. Acknowledgements The financial support from the Ministry of Science and Technology of China (2011DFA41000) and the National Natural Science Foundation of China (21336008) is gratefully acknowledged. References 19

[1] K.E. Achyuthan, A.M. Achyuthan, P.D. Adams, S.M. Dirk, J.C. Harper, B.A. Simmons, A.K. Singh, Molecules 15 (2010) 8641-8688. [2] L. Yang, K. Seshan, Y. Li, Catal. Today http://dx.doi.org/10.1016/j.cattod.2016.11.030. [3] W.O.S. Doherty, P. Mousavioun, C.M. Fellows, Ind. Crops Prod. 33 (2011) 259-276. [4] L.H. Hu, H. Pan, Y.H. Zhou, M. Zhang, BioResources 6 (2011) 3515-3525. [5] C. Li, X. Zhao, A. Wang, G.W. Huber, T. Zhang, Chem. Rev. 115 (2015) 11559-11624. [6] H. Pińkowska, P. Wolak, A. Złocińska, Chem. Eng. J. 187 (2012) 410-414. [7] Y. Zheng, D. Chen, X. Zhu, J. Anal. Appl. Pyrolysis. 104 (2013) 514-520. [8] Wahyudiono, M. Sasaki, M. Goto, Chem. Eng. Process: Process Intensification 47 (2008) 1609-1619. [9] J. Barbier, N. Charon, N. Dupassieux, A. Loppinet-Serani, L. Mahé, J. Ponthus, M. Courtiade, A. Ducrozet, A.A. Quoineaud, F. Cansell, Biomass and Bioenergy 46 (2012) 479491. [10] J. Zakzeski, A.L. Jongerius, P.C. Bruijnincx, B.M. Weckhuysen, Chemsuschem 5 (2012) 1602-1609. [11] A.L. Jongerius, P.C.A. Bruijnincx, B.M. Weckhuysen, Green Chem. 15 (2013) 30493056. [12] X. Ma, Y. Tian, W. Hao, R. Ma, Y. Li, Appl. Catal. A 481 (2014) 64-70. [13] T.D. Matson, K. Barta, A.V. Iretskii, P.C. Ford, J. Am. Chem. Soc. 133 (2011) 1409014097. [14] S. Cheng, C. Wilks, Z. Yuan, M. Leitch, C. Xu, Polym. Degrad. Stab. 97 (2012) 839-848. [15] R. Ma, W. Hao, X. Ma, Y. Tian, Y. Li, Angew. Chem., Int. Ed. 53 (2014) 7310-7315. [16] X. Ma, K. Cui, W. Hao, R. Ma, Y. Tian, Y. Li, Bioresour. Technol. 192 (2015) 17-22. [17] X. Huang, T.I. Koranyi, M.D. Boot, E.J. Hensen, Chemsuschem 7 (2014) 2276-2288. [18] G. Veryasov, M. Grilc, B. Likozar, A. Jesih, Catal. Commun. 46 (2014) 183-186. 20

[19] M. Grilc, B. Likozar, J. Levec, Appl. Catal., B 150-151 (2014) 275-287. [20] M. Grilc, G. Veryasov, B. Likozar, A. Jesih, J. Levec, Appl. Catal., B 163 (2015) 467477. [21] J.B. Claridge, A.P.E. York, A.J. Brungs, M.L.H. Green, Chem. Mater. 12 (2000) 132-142. [22] I. Tyrone Ghampson, C. Sepúlveda, R. Garcia, J.L. García Fierro, N. Escalona, W.J. DeSisto, Appl. Catal. A 435-436 (2012) 51-60. [23] C. Sepúlveda, K. Leiva, R. García, L.R. Radovic, I.T. Ghampson, W.J. DeSisto, J.L.G. Fierro, N. Escalona, Catal. Today 172 (2011) 232-239. [24] X. Ma, R. Ma, W. Hao, M. Chen, F. Yan, K. Cui, Y. Tian, Y. Li, ACS Catal. 5 (2015) 4803-4813. [25] R.N. Panda, S. Kaskel, J. Mater. Sci. 41 (2006) 2465-2470. [26] Q. Song, F. Wang, J. Cai, Y. Wang, J. Zhang, W. Yu, J. Xu, Energy Environ. Sci. 6 (2013) 994. [27] F. Cárdenas-Lizana, D. Lamey, S. Gómez-Quero, N. Perret, L. Kiwi-Minsker, M.A. Keane, Catal. Today 173 (2011) 53-61. [28] S. Gong, H. Chen, W. Li, B. Li, T. Hu, J. Mol. Catal. A: Chem. 225 (2005) 213-216. [29] D. McKay, J.S.J. Hargreaves, R.F. Howe, Catal. Lett. 112 (2006) 109-113. [30] T. Prasomsri, M. Shetty, K. Murugappan, Y. Román-Leshkov, Energy Environ. Sci. 7 (2014) 2660-2669. [31] J.E. Miller, L. Evans, A. Littlewolf, D.E. Trudell, Fuel 78 (1999) 1363-1366. [32] Q. Lu, Y. Zhang, Z. Tang, W. Z. Li, X. F. Zhu, Fuel 89 (2010) 2096-2103. [33] Y. Ye, J. Fan, J. Chang, Anal. Appl. Pyrolysis 94 (2012) 190-195. [34] A. Toledano, L. Serrano, J. Labidi, J. Chem. Technol. Biotechnol. 87 (2012) 1593-1599. [35] K. Barta, T.D. Matson, M.L. Fettig, S.L. Scott, A.V. Iretskii, P.C. Ford, Green Chem. 12 (2010) 1640-1647. 21

[36] J. Zakzeski, B.M. Weckhuysen, Chemsuschem 4 (2011) 369-378. Figure Caption

22

Table and Figure captions Fig. 1. Detailed yields of alcohols (A) and esters (B) obtained from the valorization of Kraft lignin over different catalysts at 280 oC for 6 h. Fig. 2. Yields of the products obtained from the valorization of Kraft lignin over 700Mo2N/Al2O3 catalyst with different Mo loadings (A) and for different reaction time (B). Fig. 3. The effect of reaction temperature on the yields (A) and selectivity (B) of aromatic products over 30 wt% 700-Mo2N/Al2O3 for 6 h. Fig. 4. Recyclability of the spent 30 wt% 700-Mo2N/Al2O3 catalyst for the valorization of Kraft lignin at 280 oC for 6 h. Fig. 5. XRD patterns of the Al2O3 support and the Mo2N/Al2O3 catalyst samples. Fig. 6. XPS spectra of the Mo 3d energy region of the 30 wt% 700-Mo2N/Al2O3 samples before (A) and after (B) reaction at 280 oC for 6 h.

23

.

(A)

(B)

Fig. 1. Detailed yields of alcohols (A) and esters (B) obtained from the reaction of Kraft lignin over different catalysts at 280 oC for 6 h.

24

(B)

Fig. 2. Yields of the products obtained from the valorization of Kraft lignin over 700Mo2N/Al2O3 catalyst with different Mo loadings (A) and for different reaction time (B).

25

Fig. 3. The

effect of reaction temperature on the yields (A) and selectivity (B) of aromatic products over 30 wt% 700-Mo2N/Al2O3 for 6 h.

26

Fig. 4. The recycle test of the spent 30 wt% 700-Mo2N/Al2O3 catalyst for the valorization of Kraft lignin at 280 oC for 6 h.

27

Fig. 5. XRD patterns of the Al2O3 support and the Mo2N/Al2O3 catalyst samples.

28

Fig. 6. XPS spectra of the Mo 3d energy region of the 30 wt% 700-Mo2N/Al2O3 samples before (A) and after (B) reaction at 280 oC for 6 h.

29

Table 1, Yields of the grouped products obtained from the reaction of Kraft lignin over different catalysts at 280 oC for 6 h. The non-sulfur product yields (mg/g lignin) Entry Catalyst a Thioethers Overall Alcohols Esters Aromatics 1 650-Mo2N/Al2O3 943 128 604 211 25 2 700-Mo2N/Al2O3 1189 148 759 282 29 3 750-Mo2N/Al2O3 745 73 497 175 18 a o X-Mo2N/Al2O3: ‘X’( C) means nitriding temperature.

Table 2, Detailed yields of the aromatics obtained from the reaction of Kraft lignin over different catalysts at 280 oC for 6 h. Total yields are in boldface. Yield(mg/g lignin) Component 650-Mo2N/Al2O3 700-Mo2N/Al2O3 750-Mo2N/Al2O3 Benzyl alcohols 127 155 83 Benzyl alcohol 5 24 4 Benzyl alcohol, 2-methyl65 71 39 30

Benzyl alcohol, 4-methylBenzyl alcohol, 4-ethylBenzyl alcohol, 2,4,5-trimethylMono-phenols Phenol, 2-methoxyPhenol, 2-methoxy-4-methylPhenol, 2-methoxy-4-ethylPhenol, 2-methoxy-4-propylPhenol, 2-methoxy-3,4,5trimethylPhenol, 2,3,4,6-tetramethylArenes Benzene, 1,2-dimethylBenzene, 1,4-dimethylBenzene, 1-ethyl-3-methylBenzene, 1-propenylBenzene, 1-ethyl-2,4-dimethylBenzene, 2-ethyl-1,4-dimethylBenzene, 1,2-dimethyl-4-vinylu.d. means undetected.

26 9 22 27 3 17 u.d. 7 u.d.

32 7 21 36 3 u.d. 6 7 11

17 10 13 21 2 u.d. u.d. 9 10

u.d. 57 11 31 u.d. u.d. 4 11 u.d.

9 91 15 34 2 6 5 12 17

u.d. 71 11 33 2 u.d. 4 10 11

Table 3, The texture data of Al2O3 and 30 wt% 700-Mo2N/Al2O3 determined by nitrogen adsorption-desorption. Average pore Specific surface Total pore Sample diameter area (m2/g) volume (cm3/g) (nm) Al2O3 270 0.759 11.3 30 wt% 700-Mo2N/Al2O3 169 0.459 10.9

31

32