Reac Kinet Mech Cat https://doi.org/10.1007/s11144-018-1379-z

Hydrogenolysis of glycerol to 1,3-propanediol over Li2B4O7-modified tungsten–zirconium composite oxides supported platinum catalyst Min Zhu1 • Changlin Chen1

Received: 9 December 2017 / Accepted: 17 February 2018 Ó Akade´miai Kiado´, Budapest, Hungary 2018

Abstract A series of Pt–yLi2B4O7/WOx/ZrO2 (y = 0, 0.5, 1, 2 wt%) catalysts were prepared by varying the content of Li2B4O7 through the method of coimpregnationcalcination. The obtained catalysts were used for the selective hydrogenolysis of glycerol to 1,3-propanediol. Meanwhile, these catalysts were characterized by N2 adsorption and desorption (BET), CO chemisorption, X-ray diffraction (XRD), NH3-temperature programmed desorption (NH3-TPD), H2-temperature programmed reduction (H2-TPR), Fourier transform infrared spectroscopy (FTIR) and transmission electron microscopy (TEM). The results showed that Pt–1Li2B4O7/ WOx/ZrO2 achieved the highest activity with glycerol conversion of 90.7% at 150 °C and 4 MPa and exhibited excellent stability over 200 h. Pt/WOx/ZrO2 catalyst modified with Li2B4O7 was able to enhance catalytic activity and stability, since Li2B4O7 promoted the dispersion of Pt, increased the acid amount of the catalyst and strengthened the interaction between active components and support. Keywords Glycerol hydrogenolysis  1,3-Propanediol  Tungsten and zirconium composite oxides  Li2B4O7  Stability

Introduction With the depletion of petrochemical resources and increasing awareness of environmental issues, sustainable resources were desired urgently [1–3]. Glycerol is a by-product of the transesterification of vegetable oils and animal fats to biodiesel [4, 5]. 1,2-Propanediol (1,2-PDO) and 1,3-propanediol (1,3-PDO) produced by & Changlin Chen [email protected] 1

State Key Laboratory of Materials-Oriented Chemical Engineering, College of Chemical Engineering, Nanjing Tech University, Nanjing 210009, Jiangsu, China

123

Reac Kinet Mech Cat

catalytic hydrogenolysis of glycerol were extensively served as multifunctional chemicals. The hydrogenolysis of glycerol to 1,3-PDO has been receiving a great deal of interest in academic and society, since 1,3-PDO is increasingly demanded to synthesize polytrimethylene terephthalate (PTT) [6, 7]. PTT is an appealing polymer, which has been applied broadly in fiber and textile industries due to its good stretching, elastic recovery, lower dyeing temperature, and higher light stability than other polyesters. Especially, its biodegradability provides a huge competitive environmental and economic advantages [3, 8–11]. 1,3-PDO is traditionally produced from petroleum derivatives, such as hydroformylation of ethylene oxide or hydration of acrolein, but these two reactions must be carried out at high temperature and pressure with a low selectivity [12, 13]. Recently, Ding et al. [14] and Zheng et al. [10] perform a serious of studies on hydrogenation of dimethyl malonate to 1,3-PDO. In spite of this, the hydrogenolysis of glycerol to 1,3-PDO is still a research hotspot due to its simple process, environment-friendliness and mild reaction conditions [12, 15, 16]. This reaction has been mainly investigated through the preparation of various catalysts including Pt [15, 17–25], Ir [26–29] and Cu [30–34]. Kurosaka et al. [16] achieved a yield of 1,3-PDO up to 24% over Pt/WO3/ZrO2 in 1,3-dimethyl-2imidazolidinone (DMI) at 170 °C and 8 MPa. Zhu et al. [24, 35] investigated the effect of bifunctional catalysts containing Pt on glycerol hydrogenolysis, and the results showed that Brønsted acid played a crucial role in this conversion. Sara Garcı´a-Ferna´ndez et al. [36] obtained highly-dispersed polytungstate species by adjusting the tungsten surface density, which was able to produce Brønsted acid. Fan et al. [37] investigated the effect of crystal phases of ZrO2 in Pt–WOx/ZrO2 catalysts on hydrogenolysis of glycerol, and found that Pt–WOx/t-ZrO2 catalyst achieved a higher yield of 1,3-PDO up to 49.4% than Pt–WOx/m-ZrO2 catalyst. Priya et al. [19] studied glycerol hydrogenolysis over Pt supported on ZrO2, Sulfated ZrO2, c-Al2O3, AlPO4, activated carbon and Y-Zeolite, among which Pt/AlPO4 exhibited the best performance with 100% glycerol conversion and 1,3-PDO selectivity of 35.4% at 260 °C and 0.1 MPa. Nakagawa et al. [7] reported the combination of Ir–Re/SiO2 catalyst with sulfuric acid, and the results showed that the yield of 1,3-PDO reached 38% at 81% conversion of glycerol. Luo et al. [38] investigated the influence of impregnation sequence on Ir–Re catalyst and discovered that impregnation of Re prior to Ir possessed the best 1,3-PDO selectivity of 49.2% with 43.7% glycerol conversion. Geng et al. [32] and Niu et al. [33] found that catalysts with suitable acidity and basicity were crucial for the formation of 1,3-PDO and obtained conversion of 73.2% and 95.1%, respectively. Compared with the unmodified catalyst, the catalytic activity and stability were improved after modification. Arundhathi et al. [39] reported the highest 1,3-PDO yield up to 66% catalyzed by Pt/WO3/Boehmite and they attributed the highest yield to abundant Al–OH groups in the boehmite support. Zhu et al. [40] performed glycerol hydrogenolysis over Li exchanged H4SiW12O40 (HSiW) catalyst with 43.5% conversion and 120 h stability. Zhu et al. [41] discovered that hydrogenolysis of glycerol to 1,3-PDO over SiO2-modified catalyst achieved 54.4% conversion and 1,3-PDO selectivity of 52.0% in a fixed-bed reactor at 180 °C and 5 MPa. The selective hydrogenolysis of glycerol to 1,3-PDO was studied over Pt/SiO2 modified

123

Reac Kinet Mech Cat

with HSiW, where the yield of 1,3-PDO and glycerol conversion reached 31.4% and 81.2%, respectively [20]. Nakagawa et al. [28] achieved 33% yield of 1,3-PDO over Ir–ReOx/SiO2 with H-ZSM-5 as a co-catalyst at 120 °C and 8 MPa. Deng et al. [27] reported that the glycerol conversion reached 63.3% over Ir–Re/KIT-6-R with amberlyst-15 as an additive at 120 °C and 8 MPa. Cu loaded on HSiW was used to catalyze the hydrogenolysis of glycerol with a conversion of 83.4% at 210 °C and 0.54 MPa [30]. Catalytic hydrogenolysis of glycerol was conducted using Pt–Cu bimetallic supported on mordenite and the conversion was as high as 90% at 210 °C and 0.1 MPa [13]. Zhu et al. [42] discovered that addition of B2O3 to Cu/SiO2 could inhibited the growth of Cu particles, enhance Cu dispersion and facilitate the conversion of glycerol. Hydrothermal instability was one of the reason for the deactivation of Pt-loaded catalyst during hydrogenolysis of glycerol to 1,2-PDO [43]. Xiong et al. [44] found that the formation of graphite carbon coated oxide had more excellent hydrothermal stability. Pham et al. [45] provided a simple method of coating supported metal catalysts by controlling pyrolysis of sugars to restrain hydrolysis attack without adversely affecting the catalytic activity. Despite the high conversion and yield of conventional oxides or composite oxide-supported noble metal catalysts in glycerol aqueous solution, catalyst deactivation was still a difficult point due to the hydrothermal instability of the catalyst. Therefore, it was crucial to study the stability of the modified catalyst in aqueous solution. Li2B4O7 was widely investigated in piezo-electrics [46], gas detection [47] and thermoluminescence [48, 49] due to its important physical properties, such as high coefficient of electrochemical coupling, high mechanical strength and low electrical conductivity. Therefore, we studied Li2B4O7-modified catalysts. In this work, we prepared a series of Pt–yLi2B4O7/WOx/ZrO2 (y = 0, 0.5, 1, 2 wt%) [Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%)] catalysts and studied hydrogenolysis of glycerol to 1,3-PDO in aqueous solutions with 40 wt% glycerol. The impact of Li2B4O7 on the structure, physicochemical properties and the hydrothermal stability of the catalysts were measured by means of BET, CO chemisorption, XRD, NH3TPD, H2-TPR, FTIR and TEM. We also compared the stability of glycerol hydrogenolysis to 1,3-PDO catalyzed by Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%) within 200 h.

Experiment Catalyst preparation First, Zr(OH)4 was acquired by hydrolysis of zirconium nitrate (99%, Shanghai Chemical Reagent Co., Ltd, China) solution with the addition of ammonia (25–28%, Shanghai Ling Feng Chemical Reagent Co., Ltd, China) aqueous solution up to pH 9. The subsidence was filtered and washed with deionized water until neutral. Second, the certain amount of ammonium metatungstate (90.05%, Ganzhou Huaxing Tungsten Products Co., Ltd, China) was dissolved in deionized water and poured into Zr(OH)4. Subsequently, the paste was dried for 24 h at 85 °C and

123

Reac Kinet Mech Cat

extruded. Finally, the sample was calcined at 700 °C in the air for 3 h, and the support was denoted as WZr. Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%) catalysts were prepared by coimpregnation of the aqueous solution of H2PtCl6 (37% Pt, SCRC) and Li2B4O7 (99.99%, Aladdin), where y means the mass content of Li2B4O7. The catalyst was dried at 110 °C for 6 h then calcined at 450 °C in the air for 3 h. The Pt and W contents were fixed at 2 wt% and 10 wt% in all catalysts, respectively. yLiB/WZr (y = 0, 0.5, 1, 2 wt%) was prepared by coimpregnation of the aqueous solution of Li2B4O7 (99.99%, Aladdin). The other procedures were the same as Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%). Catalyst characterization BET surface area, pore volume, and pore size distribution were determined by N2 physisorption at 77 K by using a Belsorp-II instrument. Prior to the measurement, the catalyst was vacuumed for 2 h at 200 °C. CO chemisorption was conducted in the TP-5000-II equipped with a TCD detector. First, 0.2 g catalyst was pre-reduced by flowing H2 at 200 °C for 1 h, and then flushed with He under 450 °C for 2 h, subsequently cooled to ambient temperature. CO were pulsed periodically until the catalyst was saturated with CO. Power X-ray diffraction (XRD) patterns were recorded on a Rigaku MiniFlex 600 spectrometer with Cu Ka radiation operated at 40 kV and 15 mA. The diffraction scans were measured from 2h = 10° to 80° at the scanning rate of 20°/ min. Different crystal phases were identified by comparison with JCPDS cards. NH3-TPD was implemented on the same instrument for CO chemisorption. First, 0.5 g catalyst was pre-reduced in H2 at 200 °C for 2 h and then purged with N2 at 450 °C for 3 h, after that cooled to 50 °C in He. The NH3 adsorption was performed at 50 °C for 30 min, and then the catalyst was exposed to He for 1 h to remove the physically adsorbed NH3 at 100 °C. Finally, the catalyst was heated from 100 °C to 700 °C at a heating rate of 20 °C/min and the NH3 desorption was monitored with a TCD detector. H2-TPR was carried out in the same apparatus for NH3-TPD and CO chemisorption. Before the measurement, 0.2 g catalyst was treated in N2 at 300 °C for 30 min, and then the catalyst was cooled to 30 °C under He atmosphere. After that, 10% H2 diluted in N2 was purged into the system. The catalyst was heated to 700 °C at a heating rate of 10 °C/min and detected with a TCD detector simultaneously. FTIR spectra were measured on a Bruker TENSOR 27 infrared spectrometer. The powder catalyst was mingled with KBr and pressed into translucent flake at ambient temperature. TEM images were obtained in a JEM-2011F apparatus at 200 kV. The reduced samples were suspended in ethanol with an ultrasonic dispersion for 30 min and deposited on carbon-coated copper grids.

123

Reac Kinet Mech Cat

Catalytic reaction Hydrogenolysis of glycerol was carried out in a fixed-bed stainless steel reactor. The catalyst (20–40 mesh) was put in the center of the reaction tube and fixed with quartz sand at both ends. Before reaction, the catalyst was pre-reduced in H2 (150 ml/min) at 200 °C for 2 h and then cooled to 100 °C. Afterwards, the aqueous solution with 40 wt% glycerol was pumped into the reactor with a double plunger pump continuously. The liquid and gas were cooled and collected in the gas–liquid separator. The standard reaction conditions were listed as follows: 150 °C, 4.0 MPa H2, weight hourly space velocity (WHSV) = 0.2 h-1. The liquid products were analyzed by GC-3900 gas chromatography (Tengzhou Ruineng chromatogram analysis Co., Ltd, China) equipped with a flame ionization detector (FID) and a capillary column (Alltech ECTM-1 capillary column, 30 m 9 0.53 mm 9 1.2 lm). The quantitative analysis was based on double internal standard method using n-butanol and 1,4-butanediol as an internal standard. The conversion of glycerol, selectivity of products and yield were calculated by the following equations: conversion ¼

moles of glycerol ðinÞ  moles of glycerol ðoutÞ  100% moles of glycerol (in) selectivity ¼

moles of one product  100% moles of all products

yield ¼ conversion  selectivity Table 1 Physicochemical properties and acidities of the catalysts with different Li2B4O7 loading Catalysts

Surface area (m2/g)

Pore diameter (nm)

Pore volume (cm3/g)

Total acid amount (lmol/g)

Weak acid amount (lmol/g)

Strong acid amount (lmol/g)

Pt/WZr

91

16.1

0.371

168

62

106

Pt–0.5LiB/WZr

88

16.6

0.363

174

75

99

Pt–1LiB/WZr

86

16.7

0.355

190

86

104

Pt–2LiB/WZr

82

17.1

0.350

170

48

122

Table 2 Results of CO uptakes and Pt dispersion before and after reaction with different Li2B4O7 supported Pt/WZr catalyst Catalysts

Before reaction CO uptakes (ll/g)

After reaction Pt dispersion (%)

CO uptakes (ll/g)

Pt dispersion (%)

Pt/WZr

394.40

17.2

289.05

12.6

Pt–0.5B/WZr

472.85

20.6

376.19

16.4

Pt–1B/WZr

503.85

21.9

413.15

18.0

Pt–2B/WZr

432.55

18.8

335.88

14.6

123

Reac Kinet Mech Cat Fig. 1 TEM images and Pt particle size distribution of a, b Pt/WZr-fresh, c, d Pt/WZr-spent, e, f Pt– c 1LiB/WZr-fresh and g, h Pt–1LiB/WZr-spent

Result and discussion Catalyst characterization Physicochemical properties of the catalysts The physicochemical properties of Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%) catalysts were summarized in Table 1. It was observed that the surface area of catalyst declined slightly with the increasing content of Li2B4O7. It may be due to the introduction of Li2B4O7 led to pore blockage or structural collapse of the composite oxides pores. Compared with Li2B4O7-free catalyst, the reduction in pore volume of the catalyst modified with Li2B4O7 could be explicated by the identical hypothesis. CO adsorption on the reduced catalyst was used to measure the dispersion of Pt, since the chemisorption of CO on Pt catalyst was not dissociated. The results were summarized in Table 2. It was found that the CO uptakes increased with the addition of Li2B4O7, reaching the maximum with Li2B4O7 loading up to 1 wt%. This was caused by the fact that Li2B4O7 had a positive influence on the dispersion of Pt. However, CO uptakes decreased with the content up to 2 wt%, which indicated that the surface of Pt would be topically overcast at higher content of Li2B4O7. It was consistent with the reports proposed by Sara Garcı´a-Ferna´ndez et al. [23]. Figs. 1a and 1e display the TEM images of fresh Pt/WZr and Pt–1LiB/WZr, respectively. Addition of Li2B4O7 to Pt/WZr increased Pt dispersion and declined Pt particle size correspondingly (Figs. 1b and 1f). It was consistent with the results of CO chemisorption. Structural characterizations of catalysts The XRD patterns of WZr and Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%) were displayed in Fig. 2. The diffraction peaks at 2h = 24.1°, 28.3°, 31.5° (JCPDS 86-1451) were attributed to monoclinic zirconia. In addition, the diffraction peaks at 2h = 30.3°, 35.2°, 50.4°, 60.2° (JCPDS 88-1007) assigned to tetragonal zirconia emerged in these catalysts. Compared the XRD patterns of Pt/WZr with WZr, the peaks at 24.1° and 31.5° decreased as tetragonal zirconia increased at 30.3°. In spite of the different content of Li2B4O7, there was no obvious change among the crystalline structures of Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%). There were no diffraction peaks related to Li2B4O7, WOx and PtOx, which implied that the crystallite sizes of Li2B4O7, WOx and PtOx were too small to be beyond the detectability of XRD and dispersed on ZrO2 surface homogeneously.

123

Reac Kinet Mech Cat

123

Reac Kinet Mech Cat

∗ tetragonal zirconia

∗ Δ Δ

Δ

(a)

∗

∗

Δ

Fig. 2 XRD patterns of (a) WZr, (b) Pt/WZr, (c) Pt– 0.5LiB/WZr, (d) Pt–1LiB/WZr, (e) Pt–2LiB/WZr

monoclinic zirconia

∗

Intensity

(b) (c) (d) (e) 10

20

30

40

50

60

70

80

2θ /(degree)

NH3-TPD The NH3-TPD technology was commonly used to detect the acid amount and strength on the catalyst surface. The acid amount could be obtained from the desorption peak area, and acid strength was evaluated from the NH3-TPD peak location. When the loading of Li2B4O7 was less than 1 wt%, the desorption peaks of ammonia could be divided into two temperature regions to differentiate the type of acid sites, including weak acid sites centered at 100–270 °C and strong acid sites in the range 270–600 °C in Fig. 3. However, when the Li2B4O7 content reached 2 wt%, the acid strength changed significantly. The NH3 desorption peak area increased gradually with an increase in Li2B4O7 and reached the maximum for Pt– 1LiB/WZr, and then decreased slightly. As shown in Table 1, the amount of weak acid increased with the increasing content of Li2B4O7. This indicated that the introduction of appropriate Li2B4O7 could transform the strong acid into weak acid, but excessive Li2B4O7 introduction would have the opposite effect. This was because the introduction of Li2B4O7 enhanced the dispersion of Pt, which promoted the dissociation of H2 to H atoms. The H atoms spilled over onto the surface of WZr Fig. 3 NH3-TPD patterns of (a) Pt/WZr, (b) Pt–0.5LiB/WZr, (c) Pt–1LiB/WZr, (d) Pt–2LiB/ WZr

(d)

TCD signal

(c) (b)

(a)

100

200

300

400

500

Temperature (oC)

123

600

700

Reac Kinet Mech Cat

forming H? and H-. In this case, a portion of reduced WOx combined with H? formed Brønsted acid W6-nOx(nH?), as reported in the literature [24, 50–52]. H2-TPR The H2-TPR profiles of WZr and Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%) are shown in Fig. 4. At the temperature range from 50 °C to 250 °C, the peaks corresponding to the reduction of PtOx were detected in all catalysts containing Pt, consistent with previous literature [53]. There was a reduction peak between 400 and 550 °C, which was probably due to the reduction of WOx. The temperature of PtOx reduction was increased after the introduction of Li2B4O7, which indicated that the interaction between the support and Pt was strengthened and that Pt was highly dispersed on the support, in line with the results of CO chemisorption and XRD. Meanwhile, the peak of WOx reduction was shifted towards lower temperature. It could be concluded that Pt and Li2B4O7 supported on WZr enhanced the capability of hydrogen spillover and promoted the reduction of WOx. FTIR spectra The FTIR spectras of yLiB/WZr (y = 0, 0.5, 1, 2 wt%) catalysts in the range of 2000–500 cm-1 were shown in Fig. 5. The band at 1630 cm-1 was assigned to the O–H bending mode due to water adsorbed on the surface, the presence of band at 746 cm-1 was ascribed to the various vibration modes of Zr-OH bonds, and the band at 960 cm-1 was attributed to W = O bond stretching vibration. As clearly shown in Fig. 5, a new band appeared at 1396 cm-1 corresponding to B-O asymmetric stretching vibration from the BO3 group and its intensity enhanced with the increasing content of Li2B4O7, in agreement with the previous literature [54].

Fig. 4 H2-TPR profiles of (a) WZr, (b) Pt/WZr, (c) Pt– 0.5LiB/WZr, (d) Pt–1LiB/WZr, (e) Pt–2LiB/WZr

Intensity

(e) (d) (c) (b) (a) 0

100

200

300

400

500

600

700

Temperature ( oC)

123

Reac Kinet Mech Cat Fig. 5 FTIR spectras of (a) WZr, (b) 0.5LiB/WZr, (c) 1LiB/WZr, (d) 2LiB/WZr

(d)

%Transmittance

(c) (b)

960 (a)

1396 1630 746

2000 1800 1600 1400 1200 1000

Wavenumber

800

600

(cm-1)

Catalytic performance of the catalysts Influence of Li2B4O7 on glycerol hydrogenolysis The glycerol hydrogenolysis over a series of Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%) catalysts were performed in a fixed-bed reactor at 150 °C and 4 MPa. The results of glycerol conversion, selectivity and yield of 1,3-PDO were listed in Table 3. Compared with Pt/WZr, even the slight addition of Li2B4O7 apparently increased glycerol conversion. For instance, the conversion of glycerol was up to 87.6% over Pt–0.5LiB/WZr after 100 h, while the conversion was 66.7% over Pt/WZr. Also, the conversion enhanced with the increasing Li2B4O7 loading and reached 90.7% with the 1,3-PDO yield of 45.0% over Pt–1LiB/WZr. However, further increase in Li2B4O7 loading led to its slight decline. As can be seen from Tables 1 and 2, the proper addition of Li2B4O7 was able to increase the acid amount, decrease acid strength and increase Pt dispersion in favor of the conversion of glycerol.

Table 3 Glycerol hydrogenolysis over Pt/WZr catalyst modified with different Li2B4O7 loading Catalysts

Glycerol conversion (%)

1,3-PDO selectivity (%)

Yield (%) 1,3PDO

1,2PDO

npropanol

Pt/WZr

66.7

44.8

29.9

3.6

25.3

7.9

Pt–0.5LiB/ WZr

87.6

42.7

37.3

2.6

36.9

10.6

Pt–1LiB/ WZr

90.7

49.6

45.0

1.5

33.2

10.8

Pt–2LiB/ WZr

90.0

47.4

42.7

2.0

35.5

9.9

Isopropanol

Reaction condition: 4 MPa, H2 150 ml/min, 150 °C, aqueous solution with 40 wt% glycerol, Time on stream = 100 h, WHSV = 0.2 h-1

123

Reac Kinet Mech Cat

The yield of 1,3-PDO for Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%) catalysts were displayed in Fig. 6. All Pt/WZr catalysts containing Li2B4O7 presented higher yield than the unmodified, among which the Pt–1LiB/WZr catalyst possessed the superior performance with 1,3-PDO yield of 45.0%. However, the excessive introduction of Li2B4O7 into Pt/WZr catalyst obviously declined the yield of 1,3-PDO. Consequently, the suitable addition of Li2B4O7 significantly improved the yield of 1,3PDO. Stability of glycerol hydrogenolysis All reactions of glycerol hydrogenolysis lasted for 200 h to evaluate the influence of Li2B4O7 loading on stability of catalysts and the results were displayed in Fig. 7. The conversion of glycerol over Pt/WZr significantly declined with time-evolution. The introduction of Li2B4O7 could obviously improve the stability of catalyst. For instance, the glycerol conversion decreased slightly with Li2B4O7 loading equal to 0.5 wt%, and the conversion almost remained unchanged with the content of Li2B4O7 up to 1 wt%. However, the excessive addition of Li2B4O7 showed a negative impact on the stability of catalyst. The results of CO uptakes and Pt dispersion before and after reaction with different Li2B4O7 supported Pt/WZr catalyst were summarized in Table 2. We found that CO uptakes and Pt dispersion of the catalyst after the reaction were obviously lower than those before the reaction, indicating that Pt had aggregated after the reaction. However, Li2B4O7-modified Pt/WZr catalyst inhibited the aggregation of Pt to a certain extent. From Fig. 1, we also discovered that Pt/WZrspent and Pt–1LiB/WZr-spent catalysts exhibited Pt aggregation, but the aggregation of Pt in Pt–1LiB/WZr-spent catalyst was significantly lower than that of Pt/ WZr-spent, which was consistent with CO chemisorption. The XRD results for the fresh and spent catalysts provided a clear clue as to the high stability of Pt–1LiB/WZr. As shown in Fig. 8, the diffraction peaks of Pt/WZrspent were obviously lower than the fresh. The introduction of Li2B4O7 inhibited the reduction of diffraction peaks in some extent. This indicated that Li2B4O7 could 70

Pt/WZr Pt-0.5LiB/WZr Pt-1LiB/WZr Pt-2LiB/WZr

60

Yield of 1,3-PDO (%)

Fig. 6 The yield of 1,3-PDO on glycerol hydrogenolysis with Li2B4O7-modified Pt/WZr. Reaction conditions: 150 °C, 4.0 MPa H2, the aqueous solution with 40 wt% glycerol, WHSV = 0.2 h-1

50 40 30 20 10 0

50

100

150

200

Time on stream (h)

123

Reac Kinet Mech Cat 120

Pt/WZr Pt-0.5LiB/WZr Pt-1LiB/WZr Pt-2LiB/WZr

100

Conversion (%)

Fig. 7 The catalytic stability on glycerol hydrogenolysis with Li2B4O7-modified Pt/WZr. Reaction conditions: 150 °C, 4.0 MPa H2, the aqueous solution with 40 wt% glycerol, WHSV = 0.2 h-1

80

60

40 0

50

100

150

200

Time on stream (h)

strengthen the interaction between active components and support, in agreement with H2-TPR result. No diffraction peaks associated with Pt, WOx and Li2B4O7 were detected, which implied that Pt, WOx and Li2B4O7 were homogeneously dispersed on the support surface and did not apparently agglomerated during the reaction. Effect of water on the support Hydrothermal instability was one of the reason for the deactivation of catalyst. Consequently, we investigated the influence of hydrothermal stability on catalysts. Catalysts in deionized water were heated for 300 h at 200 °C to evaluate the hydrothermal stability. The sample after hydrothermal treatment was defined as yLiB/WZr-HT (y = 0, 1 wt%). The effect of hydrothermal treatment on surface area of WZr and 1LiB/WZr was studied and the results were listed in Table 4. The surface area of 1LiB/WZr did not change significantly before and after hydrothermal treatment, which were 86 and 89 m2/g. However, the surface area of WZr decreased from 92 to 83 m2/g after Fig. 8 XRD patterns of (a) Pt/ WZr-fresh, (b) Pt/WZr-spent, (c) Pt–1LiB/WZr-fresh, (d) Pt– 1LiB/WZr-spent

Intensity

(d) (c) (b) (a) 10

20

30

40

50

2θ /(degree)

123

60

70

80

Reac Kinet Mech Cat Table 4 The results of surface area of WZr and 1LiB/WZr before and after hydrothermal treatment Supports

Fresh Surface area (m2/g)

After 300 h hydrothermal treatment Surface area (m2/g)

WZr

92

83

1LiB/WZr

86

89

hydrothermal treatment. The XRD patterns of support before and after hydrothermal treatment were displayed in Fig. 9. The diffraction peak at 2h = 30.3° of WZr-HT was much lower. However, the introduction of Li2B4O7 inhibited the influence of hydrothermal treatment, and it was well consistent with the results of XRD in Fig. 7. It was reasonable to speculate that the introduction of Li2B4O7 improved the hydrothermal stability of support. Effect of WHSV Effect of WHSV on the catalytic stability of glycerol hydrogenolysis over Pt–1LiB/ WZr catalyst was shown in Fig. 10. The conversion of glycerol was relatively high and stable when WHSV was equal to 0.1 h-1 and 0.2 h-1. However, WHSV was increased to 0.3 h-1 further, the conversion of glycerol was significantly declined with time-evolution. The yield of 1,3-PDO for Pt–1LiB/WZr catalyst with different WHSV were displayed in Fig. 11. It was find that, compared to WHSV of 0.1 h-1 and 0.3 h-1, the yield of 1,3-PDO was the highest when WHSV was equal to 0.2 h-1. Our results therefore suggested that WHSV should be equal to 0.2 h-1 accompanied by the highest 1,3-PDO yield and glycerol conversion. Mechanism of glycerol hydrogenolysis to 1,3-PDO According to the literature [24, 36, 40, 55, 56], the ideal Pt-based catalyst for catalytic hydrogenolysis of glycerol to 1,3-PDO possessed the characteristics of Fig. 9 XRD patterns of (a) WZr, (b) WZr-HT, (c) 1LiB/ WZr, (d) 1LiB/WZr-HT

(d)

Intensity

(c) (b) (a) 10

20

30

40

50

60

70

80

2θ /(degree)

123

Reac Kinet Mech Cat 120

WHSV = 0.1 h-1 WHSV = 0.2 h-1 WHSV = 0.3 h-1

100

Conversion (%)

Fig. 10 Effect of WHSV on the catalytic stability of glycerol hydrogenolysis over Pt–1LiB/ WZr catalyst. Reaction conditions: 150 °C, 4.0 MPa H2, the aqueous solution with 40 wt% glycerol

80

60 0

50

100

150

200

Time on stream (h)

70

Yield of 1,3-PDO (%)

Fig. 11 The yield of 1,3-PDO over Pt–1LiB/WZr catalyst with different WHSV. Reaction conditions: 150 °C, 4.0 MPa H2, the aqueous solution with 40 wt% glycerol

WHSV = 0.1h-1 -1

WHSV = 0.2h

60

WHSV = 0.3h-1 50 40 30 20

0

50

100

150

200

Time on stream (h)

high Pt dispersion, Brønsted acid sites and strong interaction between active components and support. Nimlos et al. [57] reported that the proton affinity for the internal hydroxyl group in glycerol was 195.4 kcal/mol, while that for the terminal hydroxyl group was 194.8 kcal/mol, which means that the internal hydroxyl group in glycerol was easier to activate. Based on the literature [15, 36, 40] and our experimental results, a plausible reaction mechanism of glycerol hydrogenolysis to 1,3-PDO was presented in Fig. 12. In this mechanism, H2 molecules were adsorbed on Pt to form dissociative H atoms, which underwent spillover onto the surface of WZr [21, 58]. One spillover H atom reached a Lewis acid site and donated an electron forming H?. A portion of reduced WOx was combined with H? formed Brønsted acid W6-nOx(nH?). Another spillover hydrogen atom reacted with the electron trapped by the Lewis acid site forming H-, which could stabilize on the Lewis acid site. At the same time, glycerol was adsorbed on tungsten active sites

123

Reac Kinet Mech Cat

Fig. 12 Reaction mechanism of glycerol hydrogenolysis to 1,3-PDO

forming an alkoxide. Subsequently, a proton coming from Brønsted acid W6-nOx(nH?) protonated the internal hydroxyl group of the alkoxide. After dehydration, a secondary carbocation was formed. The carbocation was attacked by H- and further hydrolyzed to 1,3-PDO.

Conclusion A series of Pt–yLiB/WZr (y = 0, 0.5, 1, 2 wt%) catalysts were prepared by varying the content of Li2B4O7 through the method of coimpregnation-calcination. The catalysts were used for selective hydrogenolysis of glycerol to 1,3-PDO. Hydrogen spillover induced by Pt could promote the reduction of WOx to Brønsted acid W6-nOx(nH?). The appropriate introduction of Li2B4O7 was conducive to improving Pt dispersion and increasing the Brønsted acid amount, both of which were in favor of glycerol conversion. Compared to Pt/WZr, the activity and stability over Li2B4O7-modified catalysts increased remarkably. 1,3-PDO yield of 45.0% and glycerol conversion of 90.7% were achieved for the aqueous solution with 40 wt% glycerol over Pt–1LiB/WZr catalyst at 150 °C and 4 MPa when WHSV was equal to 0.2 h-1. Pt–1LiB/WZr showed long-term stability of 200 h due to the introduction of Li2B4O7. This was because that Li2B4O7 could restrain the reduction of tetragonal zirconium after hydrothermal treatment and reaction by strengthening the interaction between active components and support.

123

Reac Kinet Mech Cat Acknowledgements Financial support by the Research and Development of Prospective Research Project of Jiangsu Province, China (BY2015005-08) is gratefully acknowledged.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.

Wang YL, Zhou JX, Guo XW (2015) RSC Adv 5:74611–74628 Zhu SH, Zhu YL, Hao SL, Zheng HY, Mo T, Li YW (2012) Green Chem 14:2607–2616 Ruppert AM, Weinberg K, Palkovits R (2012) Angew Chem 51:2564–2601 Boffito DC, Neagoe C, Edake M, Pastor-Ramirez B, Patience GS (2014) Catal Today 237:13–17 Nakagawa Y, Tomishige K (2011) Catal Sci Technol 1:179–190 Liu LJ, Zhang YH, Wang AQ, Zhang T (2012) Chin J Catal 33:1257–1261 Nakagawa Y, Shinmi Y, Koso S, Tomishige K (2010) J Catal 272:191–194 Lee CS, Aroua MK, Daud WMAW, Cognet P, Pe´re`s-Lucchese Y, Fabre PL, Reynes O, Latapie L (2015) Renew Sust Energy Rev 42:963–972 Zhang YH, Zhao X-C, Wang Y, Zhou LK, Zhang JY, Wang J, Wang AQ, Zhang T (2013) J Mater Chem A 1:3724–3732 Zheng SN, Zhu KN, Li W, Ji Y (2017) New J Chem 41:5752–5763 Kurian JV (2005) J Polym Environ 13:159–167 Mizugaki T, Yamakawa T, Arundhathi R, Mitsudome T, Jitsukawa K, Kaneda K (2012) Chem Lett 41:1720–1722 Priya SS, Bhanuchander P, Kumar VP, Bhargava SK, Chary KVR (2016) Ind Eng Chem Res 55:4461–4472 Ding TM, Tian HS, Zhao BQ (2016) React Kinet Mech Cat 118:497–508 Garcı´a-Ferna´ndez S, Gandarias I, Requies J, Soulimani F, Arias PL, Weckhuysen BM (2017) Appl Catal B Environ 204:260–272 Kurosaka T, Maruyama H, Naribayashi I, Sasaki Y (2008) Catal Commun 9:1360–1363 Fan YQ, Cheng SJ, Wang H, Ye DH, Xie SH, Pei Y, Hu HR, Hua WM, Li ZH, Qiao MH, Zong BN (2017) Green Chem 19:2174–2183 Wang J, Lei N, Yang CJ, Su Y, Zhao XC, Wang AQ (2016) Chin J Catal 37:1513–1519 Priya SS, Kumar VP, Kantam ML, Bhargava SK, Chary KVR (2014) Catal Lett 144:2129–2143 Zhu SH, Zhu YL, Hao SL, Chen LG, Zhang B, Li YW (2012) Catal Lett 142:267–274 Qin L-Z, Song M-J, Chen C-L (2010) Green Chem 12:1466–1472 Edake M, Dalil M, Darabi Mahboub MJ, Dubois J-L, Patience GS (2017) RSC Adv 7:3853–3860 Garcı´a-Ferna´ndez S, Gandarias I, Tejido-Nu´n˜ez Y, Requies J, Arias PL (2017) ChemCatChem 24:4508–4519 Zhu SH, Gao XQ, Zhu YL, Li YW (2015) J Mol Catal A Chem 398:391–398 Oh J, Dash S, Lee H (2011) Green Chem 13:2004–2007 Pamphile-Adria´n AJ, Florez-Rodriguez PP, Pires MHM, Perez G, Passos FB (2017) Catal Today 289:302–308 Deng CH, Duan XZ, Zhou JH, Zhou XG, Yuan WK, Scott SL (2015) Catal Sci Technol 5:1540–1547 Nakagawa Y, Ning X, Amada Y, Tomishige K (2012) Appl Catal A Gen 433–434:128–134 Luo WT, Lyu Y, Gong LF, Du H, Wang T, Ding YJ (2016) RSC Adv 6:13600–13608 Huang L, Zhu YL, Zheng HY, Ding GQ, Li YW (2009) Catal Lett 131:312–320 Feng YH, Yin HB, Wang AL, Shen LQ, Yu LB, Jiang TS (2011) Chem Eng J 168:403–412 Geng GL, Wei RP, Liang T, Zhou MH, Xiao GM (2015) React Kinet Mech Cat 117:239–251 Niu L, Wei RP, Li C, Gao LJ, Zhou MH, Jiang F, Xiao GM (2015) React Kinet Mech Cat 115:377–388 Zhou JX, Guo LY, Guo XW, Mao JB, Zhang SG (2010) Green Chem 12:1835–1843 Zhu SH, Qiu YA, Zhu YL, Hao SL, Zheng HY, Li YW (2013) Catal Today 212:120–126 Garcı´a-Ferna´ndez S, Gandarias I, Requies J, Gu¨emez MB, Bennici S, Auroux A, Arias PL (2015) J Catal 323:65–75 Fan YQ, Cheng SJ, Wang H, Tian J, Xie SH, Pei Y, Qiao MH, Zong BN (2017) Appl Catal B Environ 217:331–341 Luo WT, Lyu Y, Gong LF, Du H, Jiang M, Ding YJ (2016) React Kinet Mech Cat 118:481–496 Arundhathi R, Mizugaki T, Mitsudome T, Jitsukawa K, Kaneda K (2013) ChemSusChem 6:1345–1347

123

Reac Kinet Mech Cat 40. Zhu SH, Gao XQ, Zhu YL, Zhu YF, Xiang XM, Hu CX, Li YW (2013) Appl Catal B Environ 140–141:60–67 41. Zhu SH, Gao XQ, Zhu YL, Cui JL, Zheng HY, Li YW (2014) Appl Catal B Environ 158–159:391–399 42. Zhu SH, Gao XQ, Zhu YL, Zhu YF, Zheng HY, Li YW (2013) J Catal 303:70–79 43. Checa M, Marinas A, Marinas JM, Urbano FJ (2015) Appl Catal A Gen 507:34–43 44. Xiong HF, Schwartz TJ, Andersen NI, Dumesic JA, Datye AK (2015) Angew Chem 54:7939–7943 45. Pham HN, Anderson AE, Johnson RL, Schwartz TJ, O’Neill BJ, Duan P, Schmidt-Rohr K, Dumesic JA, Datye AK (2015) ACS Catal 5:4546–4555 46. Robertson DS, Young IM (1982) J Mater Sci 17:1729–1738 47. Mohandoss R, Dhanuskodi S, Renganathan B, Sastikumar D (2013) Curr Appl Phys 13:957–963 48. Ozdemir A, Altunal V, Kurt K, Depci T, Yu Y, Lawrence Y, Nur N, Guckan V, Yegingil Z (2017) Radiat Phys Chem 141:352–362 49. Celik MG, Yilmaz A, Yazici AN (2017) Radiat Meas 102:16–26 50. Barton DG, Soled SL, Iglesia E (1998) Top Catal 6:87–99 51. Hadjiivanov K, Lukinskas P, Kno¨zinger H (2002) Catal Lett 82:73–77 52. Xi YJ, Zhang QF, Cheng HS (2013) J Phys Chem C 118:494–501 53. Ebitani K, Hattori H (1991) B Chem Soc Jpn 64:2422–2427 54. Zhigadlo ND, Zhang M, Salje EKH (2001) J Phys Condens Mat 13:6551–6561 55. Priya SS, Bhanuchander P, Kumar VP, Dumbre DK, Periasamy SR, Bhargava SK, Lakshmi Kantam M, Chary KVR (2016) ACS Sustain Chem Eng 4:1212–1222 56. Gong LF, Lu Y, Ding YJ, Lin RH, Li JW, Dong WD, Wang T, Chen WM (2010) Appl Catal A Gen 390:119–126 57. Nimlos MR, Blanksby SJ, Qian XH, Himmel ME, Johnson DK (2006) J Phys Chem A 110:6145–6156 58. Triwahyono S, Yamada T, Hattori H (2003) Appl Catal A Gen 242:101–109

123