Highly dispersed nickel boosts catalysis by Cu/SiO2 in the hydrogenation of CO2-derived ethylene carbonate to methanol and ethylene glycol

Youwei Yang ,Jingyu Zhang ,Yueqi Gao ,Busha Assaba Fayisa ,Antai Li ,Shouying Huang,2 ,Jing Lv,*,Yue Wang,2,Xinbin Ma,2

1 Key Laboratory for Green Chemical Technology of Ministry of Education,Collaborative Innovation Center of Chemical Science and Engineering,School of Chemical Engineering and Technology,Tianjin University,Tianjin 300072,China

2 Joint School of National University of Singapore and Tianjin University,International Campus of Tianjin University,Fuzhou 350207,China

Keywords:Cu-based catalyst Ethylene carbonate Methanol Hydrogenation Carbon dioxide

ABSTRACT The efficient hydrogenation of CO2-derived ethylene carbonate(EC)to yield methanol(MeOH)and ethylene glycol (EG) is a key process for indirect conversion of CO2 to MeOH.However,a high H2/EC molar ratio during the hydrogenation process (usually as 180–300) is generally required to achieve good catalytic performance,resulting in high cost and energy consumption for H2 circulation in the promising industrial application.Here,we prepared a series of Ni-modified Cu/SiO2 catalysts and explored the effects of synthesis methods and Ni contents on catalytic performance under different H2/EC molar ratios.The Cu/SiO2 catalyst with 0.2%(mass)Ni loading prepared by co-ammonia evaporation method exhibited above 99% conversion of EC,91% and 98% selectivity to MeOH and EG respectively at H2/EC ratio of 60.And no significant deactivation was observed within 140 h at a lower H2/EC of 40.It is demonstrated that a few of Ni addition could not only promote Cu dispersion and increase surface Cu+ species due to the strong interaction between Cu and Ni species,but also form uniformly-dispersed CuNi alloy species and thus enhance the adsorption and dissociation of H2.But the excess Ni species would aggregate and segregate to cover partial surface of Cu nanoparticles,leading to a significantly drop of catalytic performance in EC hydrogenation.These insights may provide guidance for further design of catalysts for the ester hydrogenation reactions.

1.Introduction

As a greenhouse gas,CO2is also a cheap,non-toxic and abundant C1 resource in nature [1,2].The chemical conversion of CO2can produce high value-added liquid fuel and chemicals,which is of great significance to alleviate global energy and environmental problems.Methanol(MeOH),as the simplest saturated C1 alcohol,is one of the most widely used chemicals and a potential clean fuel energy.The route of CO2hydrogenation to MeOH has practical significance.However,because the CO2molecule is highly stable,the C=O bond is difficult to active[3,4].And the single-pass conversion for direct hydrogenation of CO2to MeOH is limited by thermodynamic equilibrium [5,6].In recent years,researchers developed an indirect method to efficiently convert CO2to MeOH under mild reaction conditions.In this route,CO2molecule is first coupled with high-energy compound ethylene oxide to produce ethylene carbonate (EC),and then EC is hydrogenated to obtain MeOH and co-produce ethylene glycol (EG).This route has high atom economy,high CO2conversion and good MeOH selectivity [7],which has attracted great attentions from both academy and industry.EC hydrogenation is one of the key processes in this route.Developing high-efficiency catalysts for EC hydrogenation plays an important role in promoting the indirect route for CO2utilization.

Homogeneous pincer-type RuIIand IrIIIcomplex catalysts were studied in EC hydrogenation and achieved outstanding catalytic activity and product selectivity under mild conditions [8].However,due to the difficulty and high cost of separating the catalyst form product in practical operation,it will hinder their largescale industrial applications [9].Heterogeneous Cu-based catalyst has been extensively used in hydrogenation reactions such as CO2hydrogenation [10],methyl acetate hydrogenation [11] and dimethyl oxalate hydrogenation [12],and achieved good catalytic performance due to the excellent selectivity in carbon–oxygen bonds activation.

Great progress for EC hydrogenation have been achieved over Cu-based catalysts.Dai and co-workers used HMS-supported copper catalyst for EC hydrogenation in the fixed-bed reactor,and obtained good performance with EC conversion of 100%and MeOH selectivity of 74%[13].Surface properties and texture structures of supports showed great impacts on the catalytic performance of EC hydrogenation.Excessive basicity or acidity on supports surface would induce side reactions,such as decarbonylation and decarboxylation,etc.,resulting in lower product selectivity [14–16].Compared with other oxide supports,silica supported Cu-based catalysts could exhibit better catalytic performance,benefiting from the weak acidity and basicity as well as high surface areas[17–19].Moreover,the active sites of EC hydrogenation have also been investigated.It is generally accepted that both Cu0and Cu+are active sites,where the Cu0could be responsible for the dissociation of H2,while the Cu+species could act as the Lewis acid sites to activate the EC molecules,and their balanced distribution is a key to improve the catalytic performance [18,20].

Additionally,many attempts on modifing Cu/SiO2catalysts have been made to increase the activity and product selectivity for EC hydrogenation.MoOxspecies is found to play a positive role on Cu/SiO2catalysts for EC hydrogenation.Benifiting from the strong interaction between Cu and MoOxspecies,EC conversion and MeOH selectivity were greatly improved [21,22].Li and coworker found that the addition of β-cyclodextrin during the synthesis of Cu/SiO2catalysts could promote Cu dispersion and hinder the particles aggregation,thereby enhancing the long-time stability of catalysts [23].However,a high molar ratio of hydrogen to EC in feed(denoted as H2/EC,between 180–300)is usually required to obtain favorable catalytic performance for EC hydrogenation[13,18,19,21].It means that a huge hydrogen circulation flow is needed in the reaction system,which would inevitably cause the increase of H2recycling compressors cost and its driving power in potential industrial applications.A lot of effort has been devoted to reduce the molar ratio of H2/EC.A series of nanoflower-like Cu/SiO2catalysts with curved surfaces and open ends were fabricated[7].The catalyst with the highest fiber density could achieve 98%EC conversion under a low H2/EC molar ratio of 60,because its structure could enhance hydrogen adsorption and enrich hydrogen concentration among the fibers.A zeolite@copper catalysts with core/shell structure could obtain MeOH yield as high as 93% at a relatively low H2/EC molar ratio(about 110),due to the highly dispersed Cu nanoparticles and intensified mass transfer.Unfortunately,it is still difficult to fabricate these complex structured catalysts in large-scale.Developing catalysts with facile preparation method to reduce the H2/EC molar ratio is still a big challenge.For this end,using non-noble metal additives to improve the hydrogen dissociation ability of catalysts may be an effective way.

In this work,a series of Ni-modified Cu/SiO2catalysts were prepared and applied in EC hydrogenation.The effects of different synthesis methods and Ni contents on the catalytic performance were investigated.The Ni-modified Cu/SiO2catalyst prepared by coammonia evaporation method with optimized Ni content of 0.2%(-mass) exhibited above 99% conversion of EC,91% selectivity to MeOH and 98% selectivity to EG at the H2/EC ratio as low as 60.Various characterizations were conducted to investigate the distribution,structure and property of Ni in the Cu/SiO2catalysts and found that the introduction of Ni species could modulate Cu species distribution and surface element composition,thereby affecting the catalytic performance.These understandings are expected to provide further guidance for the design of industrial catalysts for EC hydrogenation.

2.Materials and Methods

2.1.Materials

Cu(NO3)2.3H2O (99%),aqueous ammonia solution (25%–28%(mass)),ammonium chloride and Ni(NO3)2.6H2O (99%) were purchased from Tianjin Kemiou Chemical Reagent Co.Ethylene carbonate (EC,99%) was obtained from Aladdin.Silica sol (30%(mass)) was purchased from Qingdao Grand Chemical Co.1,4-Dioxane (99%) was supplied by Tianjin Yuanli Chemical Co.All chemicals were handled in air and used as received.

2.2.Catalyst preparation

The Cu/SiO2catalyst with 30%(mass)Cu loading,denoted as Cu-AE,was prepared by the typical ammonia evaporation method[11],briefly described as follows.Firstly,26.3 g of Cu(NO3)2.3H2O was dissolved in 100 ml of deionized water with 90 ml of aqueous ammonia solution (25% (mass)).Then 44.5 ml of silica sol (30%(mass))was added dropwise to the above solution.This suspension was stirred for 4 h at room temperature and then heated to 80 °C for ammonia evaporation until the pH was decreased to 6–7.The precipitate was separated by filtration,washed with deionized water,and dried at 110 °C overnight.Finally,the catalyst was calcined at 400 °C for 4 h.

The CuNi-AE catalyst was prepared by co-ammonia evaporation method,which is similar to Cu-AE,except that Ni(NO3)2.6H2O with designed content was added to Cu(NO3)2.3H2O aqueous solution at beginning.

The CuNi-IE catalyst was prepared by alkali-assisted impregnation method.Briefly,the uncalcined Cu-AE powder was ultrasonically dispersed in 100 ml of deionized water and then,a mixed solution of ammonia and ammonium chloride was added.Next,the Ni(NO3)2.6H2O aqueous solution was added dropwise to the above suspension and stirred at 70 °C for 11 h.Finally,the precipitate was separated by filtration,washed with deionized water,dried at 110 °C overnight and calcined at 400 °C for 4 h.

The CuNi-IM catalyst was prepared by impregnation method.Briefly,the uncalcined Cu-AE catalyst powder was ultrasonically dispersed in 50 ml of deionized water and then,the Ni(NO3)2.6H2-O aqueous solution was added dropwise and stirred for 4 h at room temperature.The excess water was evaporated using rotary evaporator.Finally,the dried power was calcined at 400 °C for 4 h.

2.3.Catalyst characterization

Inductively coupled plasma optical emission spectrometer(ICPOES,Varian Vista-MPX)was adopted to test the Cu and Ni contents in the samples.N2adsorption–desorption isotherms were measured at -196 °C using Micromeritics ASAP 2460 with the degassed samples.X-ray diffraction (XRD) was conducted on Rigaku Model C/max-2500 diffractometer with Cu-Kα radiation(λ=0.15406 nm) for the samples reduced at 350 °C for 4 h in H2flow.Transmission electron microscopy(TEM)image was obtained by using a TECNAI G2 F20 system electron microscope for the samples reduced at 350°C for 4 h in H2flow as well.Fourier transform infrared (FT-IR) spectra of the calcined samples were collected using a Nicolet 6700 spectrometer with a spectral resolution of 4 cm-1and 32 scans.

H2temperature-programmed reduction (H2-TPR) was performed on the Micromeritics Autochem II 2920 equipped with a thermal conductivity detector (TCD).Firstly,about 50 mg catalyst was pretreated in the quartz tube under He flow at 200 °C for 2 h and then cooled to 100 °C.Finally,the catalyst was heated to 500 °C in 10% H2/Ar with a heating rate of 10 °C.min-1.H2temperature-programmed desorption (H2-TPD) was performed on the same machine.About 50 mg catalyst was initially reduced at 350 °C for 1 h in 10% H2/Ar and temperature was decreased to 0 °C under Ar flow.Then H2was introduced for 1 h and excessive H2was purged away by Ar for 30 mins.Finally,the catalyst was heated to 800 °C with a heating rate of 10 °C.min-1in Ar flow.

X-ray photoelectron spectra (XPS) and Auger electron spectra(AES) were treated on ESCALAB Xi+equipped with a monochromatic Al Kα X-ray source (hv=1486.6 eV) to analysis surface Cu and Ni species.After reduction at 350 °C for 4 h,the sample was transferred to the holder immediately and outgassed in the chamber.The analysis was taken under a vacuum of 5 × 10-8Torr(1 Torr=133.322 Pa).The binding energy was calibrated using the C1s peak at 284.8 eV as the reference.The experimental error was within±0.2 eV.In order to measure the elemental composition in the internal-surface layers,the sample was sputtered using argon ion sputtering source.The acceleration voltage was 4 kV,and the sputtering time was 1 min.

N2O titration was operated on Micromeritics Autochem II 2920 to determine metallic surface area.In brief,50 mg of the sample was firstly reduced at 350 °C for 1 h and cooled down to 70 °C in Ar flow,then exposed to N2O for 1 h,ensuring the surface metallic Cu and Ni was entirely oxidized to Cu2O and NiO.After purged for 30 min with Ar flow,the sample was reduced again at 350 °C by pulse injection of 10% H2/Ar.Based on the surface atomic ratio of Cu and Ni measured by XPS and the total hydrogen consumption measured by N2O titration,the surface area of metallic Cu and metallic Ni of the catalyst could be calculated respectively.

Magnetic measurement system was conducted on Squid-Vsm apparatus with high magnetic field and extremely low-test temperature.The magnetic signal for reduced samples was recorded by radio frequency SQUID.The magnetic sensitivity is 10-9emu.

2.4.Catalytic performance test

The catalytic performance for EC hydrogenation was evaluated in a stainless-steel fixed-bed reactor using a thermocouple inserted into the catalyst bed to monitor the reaction temperature.The catalyst pellets(0.5 g,300–450 μm)were loaded into the reactor with an inner diameter of 8 mm and reduced in H2flow at 350°C for 4 h.Then,the catalyst was cooled down to the reaction temperature of 170 °C.EC (10%(mass) in the 1,4-dioxane solution) was continuously pumped to the reactor at the weight liquid hourly space velocity (WHSV) of EC of 0.6 h-1and different H2flows were fed into the reactor to change the molar ratio of H2/EC.The liquid products were condensed and analyzed off-line with an Simadzu GC2010 PRO instrument with an WondaCap FFAP capillary column(0.53 mm×30 m×1.00 μm).To ensure the accuracy and repeatability of the results,2–3 samples were measured under each reaction condition.Before the catalytic performance tests,the internal and external diffusion limitation and heat transfer effect were first excluded.We used the Carberry number,Weisz-Preter criterion[12,22,24] and Mears’ criterion [22,25–27] to evaluate the influence,respectively(details are listed in the Table S1 of Supplementary Material).The space time yield of MeOH (STYMeOH) is calculated as gram of MeOH per gram of catalyst per hour(g.g-1.h-1)

STYMeOHis calculated as follows:

Where WHSVECis the weight liquid hourly space velocity of EC(g.(g cat)-1.h-1),

MMeOHis the molecular weight of MeOH (g.mol-1),

MECis the molecular weight of EC (g.mol-1).

3.Results and Discussion

3.1.Catalytic performance of Ni-modified Cu/SiO2 catalysts for EC hydrogenation

The preparation method usually has a great influence on the physical and chemical properties of the catalyst,resulting in different catalytic performance.Here,we used Cu-AE prepared by typical ammonia evaporation method as contrast,and applied three methods including co-ammonia evaporation,alkali-assisted impregnation and impregnation to conduct the Ni modification.The Cu and Ni contents of different catalysts measured by ICP-OES were basically the same,which were 30%(mass) and 0.2%(mass) respectively (Table S2).

Before testing the catalytic performance,the limitations of mass transfer and heat transfer were first ruled out(see Table S1 in Supplementary Material).The Carberry number is less than 0.05 and the Weisz–Prater criterion is less than 0.1,indicating the catalytic performance of catalyst is not limited by mass transfer [12,22,24]Moreover,the Mears’criterion is less than 0.15,meaning the effect of heat transfer limitation can be neglected [22,25–27].As displayed in Fig.1,with the H2/EC molar ratio decrease,all Nimodified catalysts showed better EC conversion than the contrast Cu-AE catalysts.Especially when the H2/EC molar ratio was less than 100,EC conversion of the Cu-AE catalyst dropped rapidly,while the CuNi-AE and CuNi-IM catalysts could enable EC to be almost completely consumed at low H2/EC molar ratio of 60.EG selectivity of different catalysts were almost the same of about 98%but the MeOH selectivity of different catalysts showed obvious changes.The best catalytic performance was obtained on CuNi-AE catalyst with 99% EC conversion and 91% MeOH selectivity at the H2/EC molar ratio as low as 60.It is suggested that the introduction of Ni species on Cu/SiO2could effectively reduce the H2/EC molar ratio and improve the catalytic performance in EC hydrogenation.

Fig.1.Catalytic performance of Ni-modified Cu/SiO2 catalysts prepared by different methods:conversion of EC(a),selectivity to MeOH(b)and selectivity to EG(c)(Reaction conditions:170 °C,3 MPa,WHSV=0.6 h-1).

According to literature reports,different loadings of additives could lead to different interaction between metal-additives or metal-support,thereby affecting the catalytic activity [28,29].Thus,we further investigated the effect of different Ni loadings(0.1%,0.2%,0.4% and 1.6%(mass),measured by ICP-OES and listed in Table S3) in CuNi-AE catalysts for EC hydrogenation.As shown in Fig.2,at a high H2/EC molar ratio of 140,all catalysts exhibited above 99% EC conversion.As the H2/EC molar ratio decreased,the EC conversion of Cu-AE and 1.6NiCu-AE catalysts dropped rapidly,while CuNi-AE catalysts with Ni loading less than 0.4%(mass)could maintain 99%EC conversion at low H2/EC molar ratio of 60.MeOH selectivity was influenced greatly by the H2/EC molar ratio,where the highest value of 93% was achieved on 0.1NiCu-AE at H2/EC molar ratio of 80.Notably,similar MeOH selectivity of 91% was obtained on 0.2NiCu-AE at a lower H2/EC molar ratio of 60.We selected the optimal 0.2NiCu-AE catalyst to further evaluate the long-term stability under the condition of incomplete EC conversion at a low H2/EC molar ratio of 40.The result in Fig.2(d)showed that the 0.2%(mass) Ni modified Cu/SiO2catalyst possessed excellent stability with no obvious deactivation during 140 h reaction.

Fig.2.Catalytic performance of xNiCu-AE catalysts with different Ni loadings:conversion of EC(a),selectivity to MeOH(b)and selectivity to EG(c)(Reaction conditions:170°C,3 MPa,WHSV=0.6 h-1),(d) stability test of the 0.2NiCu-AE catalyst at H2/EC molar ratio of 40.

3.2.Structural properties of the Ni-modified Cu/SiO2 catalysts

As shown in Fig.S1 and Table S2,all the Ni-modified Cu/SiO2samples synthesized by different methods showed a similar N2adsorption–desorption isotherms and pore size distributions,indicating that the three different methods didn’t obviously affect the pore structure of Cu-AE catalyst.FT-IR spectra of these samples in Fig.3 showed an adsorption peak at 670 cm-1,assigned to the δOHbands in copper phyllosilicates [30],denonstrating that all the samples were composed of copper phyllosilicate.Notably,the CuNi-AE catalyst prepared by the co-ammonia evaporation method had an absorption peak at 665 cm-1,assigned to Si-O-Ni bond of nickel phyllosilicate [31],while this peak did not appear in the other two Ni-modified catalysts.According to the XRD and TEM results shown in Fig.S2 and Fig.S3,all reduced samples showed an average Cu particle size about 3.2 nm,indicating Cu species was highly dispersed (Table S2),These results implyed the introduction of Ni species by the three methods would not cause the growth and aggregation of Cu species,which may be due to the low Ni content.

Fig.3.FTIR spectra of the calcined Ni-modified catalysts prepared by different methods.

In order to have a deeper understanding of the structureperformance relationship of Ni-modified Cu/SiO2catalyst,a series of xNiCu-AE catalysts with different Ni contents prepared by the co-ammonia evaporation method were characterized.The pore structure of xNiCu-AE catalysts was detected by N2adsorption–desorption method and the results were listed in Fig.S4 and Table S3.Compared to Cu-AE,the changes of texture parameters for samples at low Ni loadings were negligible,whereas for the sample with 1.6%(mass)Ni loading,the average pore size and pore volume were decreased slightly.

There was one peak at about 180°C in H2-TPR profiles of all catalysts(Fig.S5),which could be attributed to the overlapped reduction peaks of the well-dispersed CuO to Cu0species,copper phyllosilicate to Cu+species as well as oxidized Ni species to metallic Ni.With the increase of Ni loading,the reduction peak firstly shifted to lower temperature and then to higher temperature and no additional reduction peak of Ni species was observed.It is suggested the addition of Ni species may promote the dispersion and reduction of Cu species,while excessive Ni would cover the Cu surface and cause the reduction temperature to increase.

Fig.4(a) showed the XRD patterns of the reduced catalysts.All catalysts had two peaks at 2θ of 36.4° and 43.3°,assigned to Cu2O species and metalic Cu species respectively [18].The TEM images (Fig.5) showed that average Cu particle sizes of xNiCu-AE catalysts slightly changed between 3.1 to 3.6 nm,in good agreement with the results calculated from XRD patterns (Table 1).No characteristic diffraction peaks of metallic Ni or oxidized Ni were found,which may be due to low Ni loading or high dispersion.To identify whether the Ni species in the reduced samples were metal or oxides,we conducted the magnetic characterization.As shown in Fig.4(b) the saturated magnetic susceptibility gradually increased with the increased Ni loading,indicating that most Ni species in the reduced xNiCu-AE catalysts was reduced to metallic Ni,consistent with the H2-TPR results.

Fig.4.(a) XRD patterns and (b) magnetization curves of the reduced CuNi-AE catalysts with different Ni loadings (1 emu=10-3 A.m2).

Fig.5.TEM images of the reduced CuNi-AE catalysts with different Ni loadings.

The chemical states and distribution of surface active Cu species is one of the most important factors affecting the catalytic performance.As shown in Fig.6(a),two peaks of binding energy centered at 932.6 and 952.6 eV were observed for all catalysts,which could be attributed to Cu2p1/2and Cu2p3/2peaks,respectively [21].No peak at 942–944 eV existed,indicating that Cu2+species were completely reduced to Cu+and Cu0species under the reduction condition.It was consistent with XRD result.Notably,after the introduction of Ni species,the binding energy of Cu2p was gradully shifted to higher value,implying strong interaction and charge transfer between Cu and adjacent Ni atoms.Since the binding energies of Cu0and Cu+species in Cu2p XPS spectra are quite close,the Cu LMM AES spectra were used to analyze the changes of Cu+and Cu0distribution induced by Ni addition.The asymmetric peaks of samples were deconvoluted to two peaks centered at 573.1 and 570.1 eV,which could be attributed to Cu+and Cu0species,respectively(Fig.6(b)).The molar ratio of surface Cu+speciescalculated from the CuLMM AES firstly increased and then decreased with the increase of Ni loading,and reached the highest value of 46.8% at 0.2%(mass) Ni loading (Table 1),suggesting the strong interaction and charge transfer between Cu and Ni species [32].

Fig.6.(a) Cu 2p XPS and (b) CuLMM AES spectra of the reduced CuNi-AE catalysts with different Ni loadings.

The surface metallic Cu areawas measured by N2O titration.Since the surface metallic Ni species could be oxidized to NiO by N2O and then reduced to metallic Ni in the following with H2consumption,it would result in a larger value of calculatedthan the actual value.Thus,we here combined the surface atomic ratio of Cu/Ni of the reduced catalyst measured by XPS and the amount of H2consumption in the second reduction in N2O titration to obtain reliable surface areas of metallic Cu and metallic Ni species.As listed in Table 1,compared to Cu-AE catalyst,of all Nimodified samples was increased.When the Ni loading was less than 0.4%(mass),was gradually increased and reached the largest value of 71.2 m2g-1at 0.2% (mass) Ni loading.But when the Ni loading was 0.4%(mass)or higher,was gradually decreased.It is demonstrated that excess Ni species might cover Cu particle surface and lead to the decrease of.Furthermore,the molar ratio of Cu/Ni on the catalyst surface was decreased with the increased Ni loading and the surface metallic Ni areawas gradually increased.

Table 1 Structures and catalytic performance in the reduced CuNi-AE catalysts

In order to further identify the changes of elemental composition between the inner and outer surface layers,all the reduced samples were sputtered for 1.0 min to remove surface cover and measure the XPS spectra again.As shown in Fig.7,before and after sputtering,the molar ratio of Cu/Ni on 0.1NiCu-AE and 0.2NiCu-AE catalysts changed little,whereas that of 0.4NiCu-AE catalyst was increased obviously compared to the non-sputtered sample,and this change on 1.6NiCu-AE was the greatest among these samples.It is indicated that the Ni species was uniformly dispersed with Cu and formed CuNi alloy species on the two catalysts with 0.1 and 0.2% (mass) Ni loadings,but with a larger amount of Ni addition,the excessive Ni species would segregate and aggregate on the catalyst surface and cover partial Cu surface.Consequently,at high Ni loading,thewas gradually decreased andwas increased slowly,which is agreed with the calculated results.

3.3.Structure-performance relationship of the Ni-modified Cu/SiO2 catalyst

Based on the above catalytic performance and characterizations analysis,the addition of appropriate Ni content could greatly enhance EC conversion and MeOH selectivity,especially under low molar ratio of H2/EC.The comparison of the catalytic performance of various Cu-based catalysts in the fixed bed reactor for EC hydrogenation is listed in Table S4.The results show that most of the copper-based catalysts in the literature require a high H2/EC molar ratio (>100).The 0.2NiCu-AE catalyst could achieve high EC conversion and MeOH selectivity at H2/EC molar ratio as low as 60.To obtain such an excellent catalytic performance,the ability of H2adsorption and activation plays an important role.H2-TPD was conducted and the results is shown in Fig.8 to investigate the influence of Ni loading on H2adsorption.There was a broad desorption peak at 350–600 °C in all catalysts,ascribed to desorption of chemisorbed H2.Compared with the unmodified Cu/SiO2catalyst,the H2desorption peak areas of the Ni-modified ones were increased.Specially,the H2desorption peaks on 0.1NiCu-AE and 0.2NiCu-AE catalysts were gradually shifted to higher temperature,indicating the enhanced capability of H2adsorption and activation.This may be due to the formation of uniformly dispersed CuNi alloy species and enhanced metal dispersion at low Ni loading,consistent with XPS and N2O titration results.Interestingly,when the Ni loading was further increased,the broad peak gradually became two separate peaks,implying that excessive Ni species might form other sites,such as metallic Ni nanoclusters or nanoparticles,for H2activation.And the H2desorption peak in 0.4NiCu-AE and 1.6NiCu-AE samples were gradually shifted to lower temperatures and the peak areas were decreased,which might be due to the aggregation of excess Ni species and segregation on the surface Cu nanoparticles.

Fig.7.Cu/Ni atomic ratio obtained from XPS over the reduced CuNi-AE catalysts before and after surface ion sputtering.

It is notable that a volcano-like trend of MeOH selectivity with the decreasing H2/EC molar ratio was observed in Fig.2 over the Ni-modified catalysts,espcially over the ones with 0.1% and 0.2%(mass) Ni loadings.It is because the promoting effect of Ni on H2dissociation.At high H2/EC ratios,the dissociation of H2could be greatly accelerated,which may facilitate the further hydrogenation of MeOH to methane,resulting in a decrease of MeOH selectivity.It is also reported in other works that methane was detected in EC hydrogenation [15].In the hydrogenation of propylene carbonate,the strong dissociation ability of H2also caused the hydrogenation of MeOH to generate a large amount of methane[35].Thus,a high H2/EC molar ratio resulted in enhanced further hydrogenation of MeOH to methane,while at the low H2/EC molar ratio,the insufficient hydrogenation capacity led to a decrease in MeOH selectivity as well,demonstrating a volcano-like trend in this work.

Fig.8.H2-TPD profiles of the reduced CuNi-AE catalysts with different Ni loadings.

Fig.9.TOF and STY of MeOH against the surface Cu+/(Cu++Cu0) molar ratio of CuNi-AE catalysts.

Detailed characterizations revealed that the introduction of Ni could not only affected the texture properties of the Cu/SiO2catalyst,but also changed the chemical states distribution of active Cu species due to the strong interaction between Cu and Ni species.In order to gain more insight into the structure-performance relationship,the turnover frequency (TOF) and space time yield of MeOH(STYMeOH) as a function of surface Cu+/(Cu++Cu0) molar ratio was investigated.As shown in Fig.9,along with the Cu+/(Cu++Cu0) increase,both TOF and STYMeOHexhibited similar volcanolike curves,which was in good agreement with the literature reports [18,22].The 0.1NiCu-AE catalyst with Cu+/(Cu++Cu0) of 41.4% displayed the highest TOF of 12.0 h-1and STYMeOHof 0.75 g MeOH.(g cat)-1.h-1,which was among the best catalysts reported in literatures [7,18,20,22,28].It is suggested that both Cu+and Cu0species are the active sites in EC hydrogenation where Cu0species could facilitate the dissociation of H2,and the Cu+species could adsorb and activate the C=O/C-O bonds.Balanced proportion of Cu+and Cu0species plays an important role in enhancing the catalytic performance in EC hydrogenation.Specifically,although the SC0uof 0.2NiCu-AE catalyst was similar to 0.4NiCu-AE and higher than 0.1NiCu-AE catalyst,its TOF was the lowest among the three catalysts,which could be ascribed to the too high Cu+proportion.On the contrary,the 1.6NiCu-AE with a larger SC0uthan Cu-AE behaved a lower TOF,which may be attributed to too low Cu+proportion.

From Fig.2,we could found that EC could be completely consumed over the CuNi-AE catalysts with low Ni addition when the H2/EC molar ratio was as low as 60.However,over the CuNi-AE catalyst with high Ni loading (1.6%(mass)),when the H2/EC molar ratio was less than 100,the EC conversion dropped rapidly,even lower than the unmodified Cu-AE catalyst.According to the above various characterizations,at low Ni loading,Ni species could not only promote the dispersion of Cu species,but also highly disperse and form uniform CuNi alloy species.The close contact of Ni and Cu species might promote the H2dissociation and spill over to the surrounding Cu sites.The enriched activated H on Cu sites could quickly react with the EC or intermediates adsorbed on the adjacent Cu+sites to improve the catalytic performance in EC hydrogenation.When the Ni loading was further increased,excessive Ni species could aggregate on the catalyst surface and segregate from CuNi alloy to cover partial Cu nanoparticles.Consequently,the amount of adsorbed H2was decreased and the Cu and Ni species dissociated H2separately,as shown by two H2desorption peaks in Fig.8,resulting in weakened H2activation and spillover.Furthermore,it was difficult for the dissociated H to react with EC or intermediates activated on Cu+sites,because the Cu+species was greatly reduced shown by the XPS result,resulting in the rapidly decreased catalytic performance over 1.6NiCu-AE catalyst with the low molar ratio of H2/EC in EC hydrogenation.

4.Conclusion

In summary,we prepared a series of Ni-modified Cu/SiO2catalysts and investigated the effects of synthesis methods and Ni contents on catalytic performance for hydrogenation of CO2-derived EC to produce MeOH and EG.Ni-modified Cu/SiO2catalysts prepared by co-ammonia evaporation method exhibited the best catalytic performance.Cu/SiO2catalyst modified with 0.2%(mass) Ni displayed outstanding EC conversion of 99%,selectivity to MeOH of 91% and selectivity to EG of 98% at H2/EC ratio as low as 60.And no significant deactivation was observed within 140 h at a lower H2/EC molar ratio of 40.It is demonstrated that with a few of Ni species introduced into Cu/SiO2catalyst,Ni species could be highly dispersed and form uniform CuNi alloy species to enhance ability of H2adsorption and dissociation.However,excess Ni addition would aggregate and segregate from CuNi alloy to cover partial surface of Cu nanoparticles,resulting in a significantly drop of catalytic activity.Moreover,due to strong interaction between Cu and Ni species,the distribution of Cu0and Cu+species could be altered as well,which played a crucial role in EC hydrogenation.These insights may provide guidance for further design of catalysts for ester hydrogenation reactions.

CRediT Authorship Contribution Statement

Youwei Yang:Writing– original draft,Data curation,Methodology,Writing–review&editing.Jingyu Zhang:Writing–original draft,Data curation,Visualization.Yueqi Gao:Visualization,Investigation.Busha Assaba Fayisa:Formal analysis.Antai Li:Formal analysis.Shouying Huang:Writing– review &editing.Jing Lv:Writing– review &editing,Supervision.Yue Wang:Writing–review &editing,Supervision,Resources.Xinbin Ma:Supervision,Resources.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

We are grateful for the supports from the National Natural Science Foundation of China (22022811,U21B2096 and 21938008)and the National Key Research&Development Program of China (2018YFB0605803).

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cjche.2022.01.017.