Tuning the crystallite size of monoclinic ZrO2 to reveal critical roles of surface defects on m–ZrO2 catalyst for direct synthesis of isobutene from syngas

Xuemei Wu ,Minghui Tan *,Bing Xu ,Shengying Zhao ,Qingxiang Ma ,Yingluo He ,Chunyang Zeng,Guohui Yang,*,Noritatsu Tsubaki,Yisheng Tan

1 State Key Laboratory of Coal Conversion,Institute of Coal Chemistry,Chinese Academy of Sciences,Taiyuan 030001,China

2 University of Chinese Academy of Sciences,Beijing 100049,China

3 State Key Laboratory of High-efficiency Utilization of Coal and Green Chemical Engineering,Ningxia University,Yinchuan 750021,China

4 Department of Applied Chemistry,School of Engineering,University of Toyama,Gofuku 3190,Toyama 930–8555,Japan

5 China Petroleum Chemical Industry Federation,Beijing 100723,China

Keywords:Syngas Isobutene ZrO2 catalyst Crystallite size Surface defects

ABSTRACT The effects of crystallite size on the physicochemical properties and surface defects of pure monoclinic ZrO2 catalysts for isobutene synthesis were studied.We prepared a series of monoclinic ZrO2 catalysts with different crystallite size by changing calcination temperature and evaluated their catalytic performance for isobutene synthesis from syngas.ZrO2 with small crystalline size showed higher CO conversion and isobutene selectivity,while samples with large crystalline size preferred to form dimethyl ether(DME)instead of hydrocarbons,much less to isobutene.Oxygen defects(ODefects)analyzed by X-ray photoelectron spectroscopy (XPS) provided evidence that more ODefects occupied on the surface of ZrO2 catalysts with smaller crystalline size.Electron paramagnetic resonance(EPR)and ultraviolet–visible diffuse reflectance(UV–vis DRS)confirmed the presence of high concentration of surface defects and Zr3+on m-ZrO2-5.9 sample,respectively. In situ diffuse reflectance infrared Fourier transform spectroscopy (in situ DRIFTS)analysis indicated that the adsorption strength of formed formate species on catalyst reduced as the crystalline size decreased.These results suggested that surface defects were responsible for CO activation and further influenced the adsorption strength of surface species,and thus the products distribution changed.This study provides an in-depth insight for active sites regulation of ZrO2 catalyst in CO hydrogenation reaction.

1.Introduction

Syngas conversion (CO/CO2+H2) has always been one of the most attractive routes for the production of various high-valueadded chemicals and liquid fuels.Much of the interest in syngas conversion concentrates on lower olefinsalcohols,aromatics andiso-paraffins originates from Fischer–Tropsch synthesis(FTS) [1–8] or combination of methanol-synthesis and methanol to olefins (MTO) process [9–15].Differing from the other process,the direct isosynthesis of single branched hydrocarbon with high selectivity from syngas was less attentional,which usually used metal oxides,such as thorium,zirconia and cerium oxide,as catalysts [16–18].

In general,zirconia catalysts are the mostly used isosynthesis catalyst due to remarkable catalytic performance and safety comparing to the radioactive of thorium oxide.Therefore,extensive works had been dedicated to exploring the relationship between catalytic performance and catalyst structures as well as surface properties of ZrO2catalysts.Maruyaet al.reported that monoclinic phase on the surface of ZrO2was beneficial forisobutene formation attributed to the unsaturation of coordination sites and the strong basicity[19].By doping Y2O3and CaO on ZrO2catalyst,Jackson and Ekerdt[20]studied the required surface active sites for two isosynthesis chain growth reactions,condensation and CO insertion.They presented that Lewis acid sites and oxygen vacancies enhanced the condensation reaction.Besides,oxygen anion vacancy was one of the active sites for CO hydrogenation over ZrO2catalyst [21-24],and the selectivity of the branch hydrocarbons was influenced by a balance between the strength and quantity of acid and base sites on ZrO2catalyst [20,25].Heet al.studied the effect of various dopants,Al2O3and KOH [26,27],CeO2[18,28],Y2O3[18],Sm2O3[29,30],calcium salts [31],on isosynthesis catalytic performance of ZrO2,and proposed that acidic sites were required for the activation of CO and the formation ofn-C4hydrocarbons.The basic sites were beneficial for thei-C4hydrocarbons formation.The ratio of basic sites to acidic sites played a key role ini-C4selectivity in total C4hydrocarbons for the isosynthesis reaction [31–33].Moreover,they also presented that active lattice oxygen,increased by the incorporation of Ce4+or Y3+into the zirconia lattice,was also responsible for the condensation reaction leading to the production of C4hydrocarbons [18].Maruyaet al.[19,34–36] studied the isosynthesis reaction mechanism and propsed that isosynthesis reaction scheme involved CO insertion into a bound aldehyde or ketone as the major chain growth step.And another chain growth step involved condensation between methoxide and a surface bound enolate.

However,although the effects of acid-base,redox properties and reaction mechanism of ZrO2-based catalysts on syngas isosynthesis reaction had been extensively investigated,the in-depth knowledge of the relationship between physicochemical property and structures of ZrO2catalyst was still lacking,such as the origin of acid-base sites.Therefore,the structure–activity-selectivity relationships were still ambiguous.The previous researchers had been proposed that the unsaturation of coordination sites and oxygen vacancies should be the active sites for CO hydrogenation and isosynthesis reaction [19,25,37].During CO hydrogenation,CO molecules adsorbed on the low-coordinated surface sites on the active ZrO2catalysts (Lewis acidic sites) and interacted with surface hydroxyl groups to generate formate species [21-24].Coordinatively unsaturated surface sites also played a significant role in the activation of H2which closely associated with the catalytic CO hydrogenation reaction to lead the formation of branched hydrocarbons [38–41].Besides,the normally coordinated surface sites and some isolated metallic-Zr or Zr-cluster sites acted as the sites for molecular adsorption and hemolytic dissociative adsorption of H2,respectively [42].However,their studies remained many unsolved details questions,due to lack of systematically characterizations over the past few decades.

Generally,the particle size and shape of catalysts affected catalytic efficiency [11,43–46].The smaller nanoparticles displayed higher surface specific activities which were ascribed to a larger fraction of low-coordinated surface sites.It was demonstrated that by changing the crystallite size,the concentration of anion vacancies in ZrO2could be controlled[24,46].Witoonet al.[47,48]investigated the effect of calcination temperature of ZrO2on the selective formation of CH3OH from CO2hydrogenation at high reaction temperatures and showed that the ZrO2crystallite sizes were enlarged with increasing calcination temperature of ZrO2support from 600 to 1000 °C.Here,to study the role of surface defects of monoclinic ZrO2catalyst in isobutene synthesis from syngas,we prepared a series of monoclinic ZrO2catalysts with tailored crystallite size by altering the calcination temperature without introducing any dopant to control the amounts of unsaturated Zr sites and oxygen defects [24,47].The texture,crystal structure,crystalline size and surface defects were systematically investigated using N2physisorption,X-ray diffraction(XRD),X-ray photoelectron spectroscopy (XPS),electron paramagnetic resonance(EPR),ultraviolet–visible diffuse reflectance (UV–vis DRS) and temperature-programmed desorption of CO (CO-TPD).In situdiffuse reflectance infrared Fourier transform spectroscopy(in situDRIFTS) was also employed to investigate the surface adsorbed species.The relationship between catalytic performance and surface properties such as surface defects,unsaturated Zr sites and oxygen defects,for isobutene synthesis was discussed in this work.

2.Experimental

2.1.Catalysts preparation

The precursor of pure monoclinic ZrO2samples were obtained by a hydrothermal process as described in previous reports[24,49,50].Specifically,51.3 g ZrO(NO3)2·2H2O (Tianjin Jinke,China) and 115.2 g urea (Aladdin,China) were dissolved in 240 ml deionized water,respectively.Then the above two clear aqueous solutions were mixed up under vigorously stirring,and then the mixed solution was transferred into three Teflon–lined stainless-steel autoclaves (250 ml) equally and crystallized at 160°C for 20 h in an oven.After the three autoclaves cooled down to room temperature naturally,the precipitate was washed with distilled water by centrifugation.Then the gel was dried at 80 °C for 12 h to get Zr(OH)4precursor.The obtained Zr(OH)4powder was respectively calcined at 400,500,600,800,1000°C in a tubular furnace under Ar atmosphere.The resulting catalysts were denoted asm-ZrO2-5.9,m-ZrO2-9.2,m-ZrO2-13.9,m-ZrO2-32.6 andm-ZrO2-45.9 according to the crystallite size of ZrO2decided by XRD.

2.2.Catalysts characterization

N2adsorption–desorption isotherm was used to investigate the specific surface area and pore volume of the prepared catalysts at-196 °C using Brunauer–Emmett–Teller (BET) and the BJH model at ASAP 2020 (Micromeritics,USA).Prior to measurement,the samples were degassed under a vacuum condition at 200 °C.

The crystal phase and crystallite size of the ZrO2samples were tested in an angle range of 5°–90°by the powder X-ray diffraction(XRD)on a D/max 2500 diffractometer(Rigaku,Japan)with Cu Kα target at 40 kV,15 mA.The crystalline size was calculated using Debye–Scherrer equation according to the diffraction data.Transmission electron microscope (TEM) images of the catalysts were obtained on a Tecnai G2 F20 transmission electron microscopy(FEI,USA).

X-ray photoelectron spectroscopy (XPS) was collected using a Thermo ESCALAB 250XI instrument (Thermo Scientific,USA).Agilent Cary 5000 spectrometer (USA) equipped with an integrating sphere assembly was used to conduct ultraviolet–visible diffuse reflectance (UV–vis DRS) spectra experiments.BaSO4was used as the reference and the spectra were recorded in the range of 200–800 nm.Electron paramagnetic resonance(EPR)spectrum was collected at -173 °C on a JES-FA200 spectrometer (JEOL,Japan).

To investigate CO adsorption ability of the ZrO2samples,temperature-programmed desorption of CO (CO-TPD) was carried out on an automatic temperature-programmed adsorption instrument(BELCAT-B,Japan).The catalysts(about 100 mg,20–40 mesh)was loaded in a quartz tube and pretreated at 300 °C for 30 min under a N2flow (30 ml?min-1).After cooled to 100 °C,the N2was switched to pure CO ambient for 10 min at 100 °C to make the sample saturated with pure CO.After that,a N2flow was introduced again to remove all physical adsorbed CO molecules on the catalyst.Then the temperature-programmed desorption process was started from 100 to 700 °C at a heating rate of 10 °C·min-1,and the desorbed CO was recorded by a thermal conductivity detector (TCD).

Infrared spectra of hydroxyl groups andin situdiffuse reflectance infrared Fourier transform spectroscopy (In-situDRIFTS)analysis of CO adsorption under atmosphere were carried out at a Tensor 27 spectrometer(Bruker,Germany)(64 scans,4 cm-1resolution).The fresh catalyst powder was firstly pretreated in pure H2flow(15 ml·min-1)for 30 min at 400°C,and then swept under Ar flow (15 ml·min-1) for 1 h with the background spectrum recorded after purging.Subsequently,CO was introduced and purged for 30 min,and then being swept in Ar flow for 1 h still at 400 °C,and the corresponding IR spectra of the sample were recorded.

2.3.Catalytic performance evaluation

Fig.1.(a) XRD patterns and (b)–(f) TEM images and size distributions of ZrO2 samples calcined at different temperatures.

The catalytic performance was investigated in a titanium material fixed-bed continuous-flow reactor [7].2.0 g of each ZrO2catalyst(20–40 mesh)was loaded in the reactor and pretreated in 10%H2/N2(V/V,30 ml·min-1) atmosphere at 400 °C for 6 h.And then purified feed syngas (CO/H2=1,17.06 ml·min-1) was introduced into the reactor and performed at 400 °C,5 MPa.The effluent gas was analyzed by gas chromatographs.A Shimadzu 2014 gas chromatograph with a thermal conductivity detector (TCD) equipped with TDX-01 columns online was used to analyze H2,CO,CH4,and CO2.Another two gas chromatographs with a flame ionization detector (FID) were employed to separate hydrocarbons with a Al2O3/S capillary column offline and oxygenated chemicals with HP-PLOT/U column online,respectively.Online sampling and analyzing were performed after reaction of 8 h.CO conversion(ConvCO) was calculated on the carbon atom basis (Eq.(1)),where COinletand COoutletrepresented the moles of CO at the inlet and outlet,respectively.

The selectivity of individual hydrocarbon CnHm(SelCnHm)among hydrocarbons was obtained according to Eq.(2):

The formation rate of DME and hydrocarbons was calculated according to Eq.(3),whereFrepresented the inlet flow.

Fig.2.(a) N2 adsorption–desorption isotherms (b) and the pore-size distributions.

3.Results and Discussion

3.1.Characterization of XRD and TEM

Fig.3.Relationship between catalytic activity for CO hydrogenation and crystallite size of ZrO2 samples:(a)CO conversion and products distribution;(b)hydrocarbons distribution;(c) products yield versus crystallite size.

To illustrate the influence of the crystallite size and physicochemical of monoclinic ZrO2for CO hydrogenation to synthesis isobutene,we prepared a series of ZrO2catalysts with varied crystallite size by tuning the calcination temperatures.To simplify frame of research work as much as possible,the ZrO2catalysts in this work were all monoclinic phase structure.No second phase was observed.XRD patterns of the ZrO2were presented in Fig.1.The diffraction peaks at 2θ=24.4°,28.3°,31.6°,and 34.2° with a shoulder at 35.4° were assigned to monoclinic ZrO2structures(JCPDS card No.37-1484),which confirmed that all the prepared ZrO2samples were monoclinic phase structure.The increasing intensity and sharpness of these diffraction peaks also indicated that the degree of crystallinity and crystallite size of the samples both increased gradually as calcination temperature increased.The XRD pattern ofm-ZrO2-5.9 calcinedat 400 °C exhibited weak and broad peaks (Fig.1(a)).And then the peaks became more defined and shaper with increasing the calcination temperature.The estimated value of crystallite size of the ZrO2catalysts by Scherrer’s formula with X-ray diffraction line profiles was listed in Table 1.The average crystallite size varied from 5.9 to 45.9 nm.However,in Scherrer’s formula,the peak broadening was considered as a contribution of particle size only,whereas other factors that also caused broadening were not accounted[51–53].In order to accurately evalute the crystallite size and observe the detailed morphological distinctions of the samples,TEM was performed on the samples.

Table 1The size of the resulting ZrO2 samples and textural properties

Fig.1(b)–(f) showed the TEM images of these ZrO2catalysts,including corresponding particle size distribution.The average particle size of the samples estimated by TEM was exhibited in Table 1.Obviously,the particle size became larger and larger as calcination temperature increased,and the average particle size obtained from TEM ranges from 7.4 to 67.8 nm,which is apparently larger than the crystallite size calculated from XRD patterns.Moreover,the differences between particle size from TEM and crystallite size from XRD patterns were much greater at higher calcination temperature.That was probably because a particle could be made up of several different crystallites or just one crystallite.Form-ZrO2-5.9 catalyst,the crystallite size (5.9 nm) was almost consistent with the particle size (7.4 nm).This might be the case where the nanoparticles with small size were single crystalline.At higher sintering temperature,the components of the particles transformed from single crystalline to polycrystalline.Combining with the XRD characterization and TEM results,it confirmed that we successfully prepared pure monoclinic ZrO2catalysts with varied crystallite size.Based on this we evaluated the catalytic performance for isobutene synthesis from syngas to investigate the influence of crystallite size of ZrO2catalyst on CO hydrogenation reaction.

The textural parameters of the ZrO2samples calculated according to BET and BJH method were shown in Fig.2 and Table 1.Apparently,the BET surface area and total pore volume reduced accompanied by the increase of average pore size as the crystalline size increased.Larger specific surface area and pore volumes were obtained over small crystalline size samples calcined at lower temperatures,which were beneficial for the dispersion of active sites.

Table 2XPS fitting results of ZrO2 samples

3.2.Catalytic performance

Fig.4.XPS spectra of O1s and Zr3d of the ZrO2 samples.

Fig.5.EPR and UV–vis DRS spectra of the ZrO2 samples.

Fig.3 showed catalytic performance of isobutene synthesis from CO hydrogenation reaction as the crystallite size of ZrO2catalysts changed.As seen in Fig.3(a),CO conversion decreased with increasing the crystallite size from 5.9 to 45.9 nm.The highest CO conversion was 31.9% overm-ZrO2-5.9 catalyst,which suggested that ZrO2catalyst composed of small crystallite size was rich in amounts of active sites for CO activation.Fig.3(a) and (b) showed that ZrO2catalyst with small crystallite size was favorable for hydrocarbons synthesis as well,dominated by isobutene.m-ZrO2-5.9 catalyst exhibited 32.2% hydrocarbons selectivity and 50.2%selectivity of isobutene in total hydrocarbons.As the crystallite size increased,the selectivity of hydrocarbons decreased,including the target product isobutene.But DME selectivity increased obviously simultaneously.Because CO conversion markedly decreased as crystallite size increased and the selectivity of products was relative with conversion,we further analyzed the variation of formation rate of hydrocarbons and DME to discriminate the influence of crystallite size on the product selectivity.As seen in Fig.3(c),decreasing crystallite size of ZrO2led to an increase in the hydrocarbons formation rate,but a decrease in DME.The hydrocarbons formation rate reached highest at 5.9 nm,but the formation rate of DME was highest at 45.9 nm.The reason could be that the active sites onm-ZrO2for formation of hydrocarbons and DME were different,and small crystallites of ZrO2catalyst were beneficial for CO conversion and isobutene formation.The surface properties of ZrO2catalyst varied with the crystallite size.We further took a series of characterizations to enclose the active sites for isobutene synthesis on ZrO2catalysts.

3.3.Characterization of XPS

It’s reported that the oxygen anion vacancy was an active site for CO adsorption and activation during CO hydrogenation over ZrO2catalyst [24,54].To reveal the variation of surface defects,including oxygen vacancy and unsaturated Zr sites of ZrO2catalyst as crystallite size changed,a series of characterization methods were employed.XPS measurements were carried out to study the chemical environments of O and Zr of the ZrO2samples.The chemical state of O1s and Zr3d by fitting the XPS curves of ZrO2surface was shown in Fig.4,and the relative content of surface oxygen species calculated on basis of the XPS data was listed in Table 2.As shown in Fig.4 of the Zr3d spectra,two peaks at around 181.8and 184.2 eV were assigned to the characteristic spin–orbit splitting Zr3d5/2and Zr3d3/2of Zr4+.The Zr3d peaks shifted to higher binding energy as the crystalline size increased,indicating that the stronger interaction between the surface Zr and O atoms.According to the literatures[14,55],the O1s state always contained three binding energy (BE) components,which centered nearly at 530.15,531.25 and 532.40 eV.From Fig.4 and the data in Table 2,the lower binding energy peak at 529.64 eV was assigned to the surface lattice oxygen species (OLattice),and the moderate and higher peaks at 531.47 eV and 532.95 eV were attributed to surface chemisorbed oxygen species,which were assigned to O atoms next to a defect (ODefect) and hydroxyls (OOH) [14,56,57],respectively.Apparently,ZrO2with small crystalline size showed higher ratio of chemisorbed oxygen species.Form-ZrO2-5.9 catalyst,the proportion of ODefectand OOHwere 23.1% and 7.5%,respectively.Hence,the ZrO2catalyst with smaller the crystalline size owned higher fraction of oxygen defects on the surface,which should be derived from coordinately unsaturated zirconium and/or oxygen ions [23].Generally speaking,solid defect meant the distortion of the actual crystal structure relative to the ideal lattice structure[58].ZrO2samples with larger crystalline size obtained by calcination at high temperature tended to form ideal lattice structure as a result of sintering.The better CO activation ability and higher isobutene selectivity ofm-ZrO2-5.9 catalyst should be closely related to the surface defects.Due to the insensitivity of XPS analysis to the detection of Zr3+,we further employed UV–vis DRS spectra and EPR measurement to investigate the surface defects on ZrO2samples[59].

Fig.6.(a) Relationship between surface-normalized concentration of CO (CCO)determined from the amount of CO desorbed in CO-TPD and the crystallite size,and(b) CO-TPD profiles of the ZrO2 samples.

3.4.Characterization of UV–vis DRS and EPR

It was reported that infrequent Ti3+was produced from the generation of oxygen vacancies [59,60].Similarly,the existence of ODefect,confirmed by XPS,might induce the reduction of Zr4+cation.UV–vis DRS spectrum was carried out to investigate the Zr valence state.Fig.5(a)showed the UV–vis DRS spectra acquired in the wavelength region of 200–800 nm for the investigated ZrO2samples.The strong absorption band at around 308 nm(4.02 eV in photon energy) ofm-ZrO2-5.9 catalyst was attributed to the interstitial Zr3+ions of the monoclinic lattice[61],which was consistent with the XPS analysis.The concentration of surface Zr3+decreased as the crystalline size increased.EPR spectroscopy is one of the most suitable techniques to investigate defects in oxides which are paramagnetic defective materials [62],especially for oxygen defects in semiconductor oxides [58].The EPR spectra of the samples were showed in Fig.5(b),and a symmetric signal,resonating atg=1.9947,was present.A possible tentative explanation for the observed signal was that electrons trapped ingrain boundaries[63–65],indicating the existence of surface defects.According to the difference of EPR signal strength,it was confirmed thatm-ZrO2-5.9 catalyst owned the maximal concentration of surface defects among these investigated ZrO2catalysts.

3.5.Characterization of CO-TPD

To further evaluate the CO adsorption and activation sites on ZrO2catalysts,CO desorption behavior of the ZrO2samples was analyzed by CO-TPD.Fig.6(a) showed the relationship between surface-normalized concentration of CO (CCO) and crystallite size,determined from the amount of CO desorbed in CO-TPD.The concentration of desorbed CO increased markedly with the decreasing crystallite size of the ZrO2samples.The results suggested that the CO adsorption and activation sites were increased as the crystallite size decreased,resulting in improved CO conversion.From Fig.6(b),there were three types of CO adsorption sites on ZrO2samples.They were weak adsorption sites atca.222 °C,moderate adsorption sites atca.331 °C and strong adsorption sites atca.467 °C.All of the peaks gradually decreased as the crystallite size increased,showing a decrease in CO activation sites.CO adsorption concentration and sites versus the crystallite size of ZrO2samples exhibited the same trend with surface O-vacancy versus the crystallite size.Thus,the active sites for CO adsorption and activation were most probably contributed to the existence of surface coordinative unsaturated Zr cations and oxygen vacancy in the lattice of ZrO2[66].The variation of CO adsorption sites should be one of the reasons for the evolution of products distribution of CO hydrogenation over ZrO2catalysts.For ZrO2with smaller crystallite size,due to higher concentrations of CO adsorption sites,hydrocarbons were more likely to be generated than DME.While for ZrO2with larger crystallite size,much more DME was produced with less CO adsorption sites.To further illustrate how the transformation of intermediates species and products distribution affected by CO adsorption,in-situ CO DRIFTS spectroscopy was employed.

3.6.Characterization of in-situ CO DRIFTS

Fig.7(a) showed the hydroxyl region (4000–3500 cm-1) of the infrared spectra obtained after the samples being pretreatedin situunder Ar flow to remove surface-adsorbed impurities.The intensity variation of the hydroxyl groups peaks was concordant with the XPS results.In-situCO DRIFTS spectroscopy was carried out to investigate the variation of surface adsorbed species over the ZrO2samples.The negative bands at 3721 and 3655 cm-1were associated with terminal and bridged hydroxyl groups,respectively.On the basis of references,adsorbed CO reacted with surface hydroxyl group and then formate species were formed.As can been seen from Fig.7(b),typical bridged formate characteristic peaks at 2867 cm-1for the C–H stretching (ν(CH)) mode and the peaks at 1590,1381 and1351 cm-1represented the COO-asymmetric(νas(OCO)),the C-H in plane deformation (δ(CH)) and the COOsymmetric stretching (νs(OCO)) vibrations confirmed the formation of formate species.Besides,combination modes of formate species,namely,ν1(combi)(νas(COO+δ(CH))at 2965 cm-1and ν2-(combi)(νs(COO)+δ(CH))at 2745 cm-1were observed as well.As crystalline size increased,the intensity of bridged formate vibration peaks decreased.Moreover,the vibration of νas(OCO) shifted to lower wavenumbers region as crystalline size increased,which indicated formate species more weakly bounded on the ZrO2samples with small crystalline size.In contrast to the results of CO-TPD and catalytic performance,we could assume that weaker adsorption of formate species caused by higher concentrations CO adsorption on the defect-rich ZrO2catalyst was favorable for hydrocarbons formation.ZrO2catalyst with fewer surface defects showed stronger formate species adsorption and tended to generate DME.Based on these,we proposed that the active sites on ZrO2catalyst were different for hydrocarbons and DME formation.Surface defects of ZrO2catalysts were required for hydrocarbons formation,but not for DME Furthermore,isobutene was dominated in hydrocarbons,in that weaker adsorption strength of formate species on defect-rich ZrO2sample should also be beneficial for isobutene generation.

Fig.7.(a)Infrared spectra of hydroxyl groups and(b)in situ CO DRIFTS analysis of CO adsorption under atmosphere at 400 °C for 30 min on ZrO2 samples.

4.Conclusions

In summary,this work investigated the effect of surface defects,including unsaturated Zr sites and oxygen defects,by tuning crystallite size of ZrO2catalyst on CO activation and products distribution in isobutene synthesis from CO hydrogenation reaction.It was demonstrated that ZrO2catalyst with small crystalline size exhibited better CO conversion and isobutene selectivity.Its better catalytic performance was contributed to the existence of higher concentration of unsaturated Zr sites and oxygen defects on the catalyst surface,confirmed by a series of characterization methods.The results of CO-TPD indicated that the decreased concentration of surface defects not only reduced the amounts of CO activation sites,but also weaken its adsorption strength.The analysis ofin situCO DRIFTS confirmed that the amounts of formed formate species decreased and adsorption strength was enhanced as the concentration of surface defects decreased caused by increasing crystalline size.Combined with the catalytic evaluation results,it could be assumed that weaker adsorption of formed formate species on defect-rich ZrO2catalyst was beneficial for hydrocarbons formation,especially for isobutene.Our work presented the new insights for tailoring the structure of ZrO2catalyst for CO hydrogenation to specific products.

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

This work was financially supported by the Natural Science Foundation of China (21978312,21908235 and 21802155),the Key Research Program of Frontier Sciences,CAS (QYZDB–SSW–JS C043),as well as Foundation of State Key Laboratory of Highefficiency Utilization of Coal and Green Chemical Engineering(2019-KF-05 and 2018-K22).Research Project Supported by Shanxi Scholarship Council of China and Fund Program for the Scientific Activities of Selected Returned Overseas Professionals in Shanxi Province are also greatly appreciated.