Catalytic Conversion of Lignin in an Isopropanol/formic Acid Medium With Nimo Catalyst Promoted by W Species

23 Background : Large amounts of enzymatic hydrolysis lignin (EHL) are generated with 24 the production of cellulosic bioethanol. Efficient degradation and upgrading of EHL is 25 significant for the sustainable and stable development of energy supply. 26 Results : In this study, hydrodeoxygenation (HDO) of EHL to biofuels was carried out 27 promoted by the in situ hydrogen donor produced from the decomposition of formic 28 acid over NiMo catalysts. Results showed that active sites (derived from the support 29 SiO 2 , W, and NiMo species) had remarkable effect on lignin conversion, and the highest 30 oil yield (57.2 wt%) was gained over NiMo/W-SiO 2 catalyst. 31 Conclusions : The product evolution demonstrated that active metal sites (derived from 32 NiMo species) favored hydrogenolysis and deoxygenation via leading in situ hydrogen 33 to attack C-O-C bonds, while acid sites (derived from the support) adsorbed and 34 activated chemical bonds in lignin, resulting in the linkage cleavage caused by the 35 heating program. The obtained bio-oil was rich in alkyl guaiacols (6.7 wt%), containing 36 stable chemical properties and high quality. 37 42 Conceptualization, X.L.; methodology, H.M.R.; software, H.G.; validation, D.W., P.X. 459 and J.C.; formal analysis, C.X.; investigation, X.G.; resources, X.G.; data curation, X.L.; 460 writing-original draft preparation, X.L.; writing-review and editing, X.G.; visualization, 461 H.M.R.; supervision, X.G.; project administration, C.X.; funding acquisition, X.G.

SiO2-NC is slow without obvious hysteresis loop, which is due to pore blocking cause 112 by the presence of surfactant (CTMAB), therefore, causing the uneven distribution of 113 pore size (Fig. 1f). Calcination at 550 o C is an effective method to remove CTMAB, 114 leading to an increasing N2 adsorption, an enlarged surface area ( Table 1), and a 115 uniform distribution of pore size (⁓ 30 Å) (Figs. 1g). Furthermore, variations in pore 116 volume and pore size displayed in Table 1 also indicate the change in textural properties 117 after calcination. With the incorporation of Ti/W species, the surface area decreases 118 from 654.2 to ⁓ 400 m 2 g -1 , indicating the pore blocking, thus leading to a decrease in 119 pore volume (from 0.5 to 0.3 cm 3 g -1 ) and a decline in N2 adsorption (Figs. 1h, i). 120 Moreover, with the addition of NiMo species, a further decrease in pore volume and 121 surface area can be observed, which indicates the formation of metal particles blocking 122 internal pores. However, Fig. 1e) shows that N2 adsorption of NiMo/W-SiO2 increases 123 significantly as compared to that of W-SiO2, which is due to the formation of different 124 porous structures derived from metal particles dispersed on the support surface [18].
125 Table 1 Physical properties and acid strength of synthesized catalysts.   Meanwhile, FTIR spectra of supports are presented in Fig. 4. Obviously, all 151 samples show a broad peak at 1089 cm -1 , a sharp peak at 461 cm -1 , and a weak one at 152 806 cm -1 , which are assigned to the internal asymmetric stretching, bending vibration, 153 and symmetric stretching of Si-O-Si linkage, respectively [22][23][24]. In addition, the peak 154 located at 1632 cm -1 is attributed to Si-OH group [ for Ti-SiO2, which is lower than those of W-SiO2. It is suggested that with the 172 incorporation of metal cations into silica, acidity (including L and B acidity) in these 173 formed mixed oxides will be induced, which depends on the coordination of 174 incorporated heteroatom and its amount of addition (Table 1, 6.82 wt% for W, 2.28 wt% effect for Ti-SiO2. Therefore, more acidity in W-SiO2 (0.485 mmol g -1 ) has been 177 detected than that in Ti-SiO2 (0.339 mmol g -1 ). Catalyst support with higher acidity is 178 more beneficial for supported metal dispersion, lignin-derived oxygen-containing 179 groups adsorption, and C-O linkages breaking [30,31], which make W-SiO2 potential 180 for utilization as a support of NiMo catalysts during HDO process.    The results of H2-TPR of NiMo/W-SiO2 are shown in Fig. 6d. As depicted in Fig.   215 6d, a sharp peak at around 300-500 o C is recorded, which is the characteristic adsorption 216 of the reduction of Mo 6+ to Mo 4+ species, while one small peak at around 600-700 o C 217 is assigned to the characteristic peak of NiO species reduction, which is bound to   For testing the recyclability of catalyst, lignin depolymerizaiton was performed at the 246 same condition over NiMo/W-SiO2 for five times and each experiment was conducted 247 for three times, and catalytic evaluation is presented in Table 2. With the increase of 248 catalyst recyclability, surface area decreases dramatically with a decrease in pore 249 volume, that is due to the blocking of porous structure caused by the coke deposition. catalyst is worse than that of fresh one due to active phases sintering and leaching.
258 Table 2. Physical properties of spent catalyst and reactivated one.   and hemicellulose would cause a negative effect on lignin depolymerization and the 307 processing steps were presented in our previous research [46]. After that, the extracted lignin was dried at 105 o C for 12 h, and lignin used in the following experiment was 309 assigned to extracted lignin. The proximate, ultimate, and component analyses of raw 310 and extracted lignin are presented in Table 3.    353 The physical properties (i.e., Brunauer-Emmett Teller (BET) surface area, pore volume, 354 and pore size) and adsorption-desorption isotherms of synthesized samples were 355 measured by a Tristar II series, micrometric analyzer at N2 atmosphere (77K, -196 o C).

356
Surface area was determined using BET method and the distribution of pore size was  Each experiment was repeated for three times ensure the reproducibility.

392
The obtained products were in components of gas, liquid, and solid (including 393 residual lignin, char, and catalyst). As the content of gaseous products was less than 1 394 wt% (based on initial weight of lignin), thus they were ignored in the following analysis.

395
In a typical separation step (Fig. 10), gaseous products were released into the air after 396 opening the reactor. Liquid and solid phases were separated by washing the reactor with 397 30 mL dichloromethane (DCM, 99.5 %) for three times followed by filtration.   SBET -the surface area measured by BET method.

659
Spore -the size of pore.          Catalytic evaluation of lignin depolymerization over different synthesized catalysts.

Figure 9
Possible reaction mechanism of lignin depolymerization over supported metal catalysts with assistance of formic acid used as an internal hydrogen donor.

Figure 10
Separation steps of depolymerized phases.