Influence of Aging Time on the Properties of Precursors of CuO/ZnO Catalysts for Methanol Synthesis

Journal of Natural Gas Chemistry 14(2005)107- 114SCIENCE PRESSInfluence of Aging Time on the Properties of Precursorsof CuO/ZnO Catalysts for Methanol SynthesisDeren Fang-;2，Zhongmin Liul*,Shuanghe Meng'，Ligang Wang',Lei Xu',Hua Wang'1. Dalian Institute of Chermical Physics, Chinese Academy of Sciences, Dalian 116023, China;2. Institute of Applied Catalysis, Yantai University, Yantai 264005, China[Manuscript reeived May 08, 2005; revised May 16, 2005]Abstract: The aging process of pure copper precursors and copper. zinc binary precursors were studiedby XRD, TG-DTG and TPR techniques. The catalytic activity and stability of CuO/ZnO were testedusing fixed-bed flow reactor, and the physical properties of the catalysts and Cu species were charac-terized with N2 adsorption and N2O passivation method, respectively. For the Cu-Zn binary systemprepared at the precipitating condition of pH=8.0 and temperature= =80 °C, the initial phase was a mixtureof copper nitrate hydroxide Cu2(NO3)(OH)3, georgeite and bhydrozincite Zns(CO3)2(OH)6. By increas-ing the duration of its aging time, the phase of Cu2(NO3)(OH)2 first transited to georgeite, and theninterdiffused into Zns(CO3)2(OH)6 and resulted in two new phases: rosasite (Cu,Zn)2CO3(OH)2 and au-richalcite (Zn,Cu)s(CO3)2(OH)6. The former phase was much easier to be formed than the latter one,while the latter phase was more responsible for the activity of methanol synthesis than the former one. Itis found that the composition and structure of the precursors altered obviously alfter the colour transitionpoint. The experimental results showed that methanol synthesis is a structure-sensitive catalytic reaction.Key words: Cu/Zn oxide, methanol synthesis, precursor, aging time, colour transition, structure-sensitivereaction1. IntroductionThe phase and composition of the precursor caused asignificant effect on the performance of the catalyst,The copper/zinc/ aluminum oxides catalysts werewhile the preparation method and condition was thewidely used in many industrial process includ-vital factor to the precursors. Many researchers haveing the low-pressure methanol synthesis, CO low-investigated the precursors based on different viewstemperature shift reaction and dehydrogenation ofand so far achieved controversial conclusions [4- 14].alcohols, etc [1]， and moreover, it will be the ma-Okamoto et al. [4] have studied the precursorjor catalyst in the fuel cell for the production of hy-and calcined CuO-ZnO catalysts prepared by copre-drogen from oxidative steam reforming of methanolcipitation method using X-ray photoelectron spec-[2,3]. The commercial catalyst was prepared by co-troscopy and X-ray diffraction analysis, they foundprecipitation technique using the nitrate salts of cop-中国煤化工nds in the precursorsper, zinc and aluminum reacting with basic material,forfeedstock. Whilesuch as sodium carbonate to form the precursor ofHimeMYHC N M H Gi the coprecipitatedthis catalyst- mixture of CuZnAl hydroxycarbonates,hydroxycarbonate precursor of the CuO-ZnO cat-then calcining the precursor to give CuZnAl-oxides.alyst based on at. 30% copper and at. 70%* Corresponding author. Tel: +86-411-84685510; Fax: +86-411-84379289. E mail address: zml@dicp.ac.cn.108Deren Fang et al./ Journal of Natural Gas Chemistry Vol. 14 No.2 2005zine oxide and obtained a single-pbase aurichalitechange as well as the interaction between Cu and Zn(Cuo. zn0:.7)(CO3)2(OH)6. Shenet al. [8] stud-species in the precursors during the period of aging,ied the efects of the preparation condition uponwe designed a series of test.the structures of the precursors and concluded thatThe pure copper precursor was prepared by si-difference in the rate of the addition of the mixed 80-multanously dropping an aqueous solution of cop-lution of copper and zinc nitrates to the solution ofper nitrates (1.0 mol/L) and an aqueous solution ofNaHCO3 exerted a marked effect upon the local con-sodium carbonate (1.0 mol/L) into a container with acentrations of the copper and the zinc cations, result-small amount of deionized water. The suspension ining in the difference in the initial states of the copper-the container was constantly stirred and kept at thecontaining precipitates and thereby in the final struc-desired pH level (8.0) and temperature (80°C) by ad-tures of the precipitates. Pollard et al. [9,10] studiedjusting the relative flow rate of the two liquids and thethe precursors prepared by the solution of CuSO4 andwater bath temperature. Taking a part of suspensionZnSO4 with Na2CO3 at 60 °C and reached the conclu-out of the container immediately after precipitation,sions that the initial product is probably georgeit withfiltered, washed with deionized water, and dried in airincorporated zinc, which conversed to malachite dur-at ambient temperature. After thoroughly dried in airing the aging of precipitated precursors in the motherthe sample was further dried at 110C for 2 h and des-liquor, the zinc-malachite is the inportant phase forignated as C-1, take some of C-1 calcined at 350°C inthe preparation of catalysts of high activity. On thethe atmosphere of air for 2 h and designated as C-1c.other hand, the results of Millar [11] revealed thatThe remaining suspension was stirred continuously atcopper hydrozincite decomposed to give zinc oxide80 °C until its colour transitted from blue to blueparticles decorated by highly dispersed, small coppergreen, then a part of the suspension was taken outoxide species. Aurichalcite appeared to result ulti-of the container and treated as the preceding sample,mately in the most intimately mixed catalyst struc-the precursor and calcined specices were designated asture whereas zincian malachite decomposed to pro-C-2 and C-2c, respectively. The remaining suspen-duce larger copper oxide and zinc oxide grains. Thesion was continuously strred at 80°C for another 30former phase contributed a great deal to the cata-minutes after the colour transition, and then treatedlyst activity. Li et al. [12] investigated the precursoras the preceding samples, the precursor and calcinedprecipitated at different pH values and temperaturesspecies were designated as C-3 and C-3c, respectively.using Cu-Zn-Al ternary nitrates co current precipita-In the same procedure as C-1, except for the fedetion method and found that the formation of copperstock was zinc nitrates, we prepared pure Zn precur-hydroxynitrate was favored at pH≤6, which led to lesssor Z-1. The copper- zinc precursors were prepared byactive catalysis; while at pH=7.0 the major phase wassimultaneously dropping a mixed aqueous solution ofthe malachite-like phase which led to the highly activecopper and zinc nitrates (total concentration was 1.0catalyst. Takexawa et al. [13,14] studied the relationmol/L, Cu/Zn atomic ratio was 2:1) and an aqueousbetween the catalytic performance and the precursors,solution of sodium carbonate (1.0 mol/L) into a con-they found that the catalysts derived from the pre-tainer with a small amount of deionized water. Thecursors containing aurichalcite exhibited high perfor-suspension in the container was constantly stirred andmance in the methanol synthesis from CO2. Exceptkept the pH at 8.0 and temperature at 80C, the otherfor the work of Pollard et al.[9], there are few studiestreatment was the same as those for the copper pre-about the efct of aging time on the precursors in thecursors, but the suspension was taken out of the con-literature. In this paper we will study the change oftainer at different moments: after precipitation im-phase and composition of precursors during the agingmediately (CZ-1), under colour transition (CZ-2), 10period as well as the interactions between CuO andmin after the colour transition (CZ-3) and 30 min af-ZnO in the catalyst.ter the colour transition (CZ-4). The calcined sampleswere中国煤化工”CZ-3c and CZ-4c,2. ExperimentalrespCNMHGAn une suaupres were iipiuyed in the X-ray pow-2.1. Sample preparationder diffraction analysis, the uncalcined samples wereCatalyst precursors were prepared by co-currentused for the thermal gravimetry measurements, andprecipitation method. In order to study the phasethe calcined species were employed in the temperatureJournal of Natural Gas Chemistry Vol. 14 No. 22005109programmed reduction tests and activity tests.3. Results and discussion2.2. Sample characterization3.1. Copper precursorThe X- ray powder diffraction patterns of the solidX- ray diffraction (XRD) patterns of pure copperproducts were checked using a Shimadzu XRD6100precursors and calcined species are ilustrated in Fig-X-ray powder diffractometer (Cu K。 radiation)ure 1 and Figure 2, respectively. It is demonstratedequipped with a computer system to perform auto-from the XRD pattern that most part of the speciesmatic operation and data processing. The crystallitein C-1 was amorphous, only a few crystals of cop-sizes of CuO were estimated from the Scherer equa-per nitrate hydroxide Cu2(NO3)(OH)3 were detected.tion.In the sample of C-2, a large amount of malachiteA NETZSCH STA449 thermal analyzer was usedCu2CO3(OH)2 appeared. After 30 minutes of agingto perform differential thermal gravimetry (DTG)the phase of Cu2(NO3)(OH)3 disappeared and theand thermal gravimetry (TG) measurements simul-malachite became the unique phase.taneously in a flow of nitrogen at a heating rate of 10°C/min. The sample weight was about 10 mg.The temperature programmed reduction (TPR)●Cu2CO(OH)。Cu2(OH)NO3was performed using a TP5000 multifunction ab-sorber equipped with a computer system to performwn。)automatic operation and data recording. The catalystweight was 50 mg (in the case of pure CuO the sam-ple weight was 30 mg), and the particle size was 40 60meshes, the inside diameter of quartz reactor was 3mm with a built-in thermocouple used to detect thetemperature, a mixture of 10%H2/90%N2 was usedlile li limimnssreas the reduction gas and the rate of temperature risewas 5 °C /min. .1)|The catalytic activity of the catalysts for60methanol synthesis was evaluated in a pressurized201(0 )fixed-bed flow reactor system at3 x 10, 240 °C andFigure 1. Effect of aging time on the XRD patterna space velocity of 8000 h- I . The catalyst particleof pure copper precursorssize was 40- 60 meshes. The composition of the gas(1) C-1, (2) C-2, (3) C-3feed was V(CO)/V(CO2)/V(H2)=32/4.0/64. Beforethe gas feed was led in, the catalyst was reduced withpure hydrogen under atmospheric pressure as follows:。CuOfirst to 170 °C from room temperature with a tem-perature ramp of 5 °C /min and dwelling at this tem-3)_perature for 2 h, then rising to 250 °C and dwellingat this temperature for 1 h. After the catalytic ac-tivity test, the reactor temperature was risen to 400°C in the reaction atmosphere and dwelling at thistemperature for 2 h, then lowered to 240 °C to testthe remaining catalytic activity.2)__ ilIiieenThe BET surface area of the catalysts weremeasured by means of nitrogen adsorption with arASAP2010 type Quantachrome NOVA automated gas中国煤化Msorption system. The Cu surface area and particlesizes of the catalysts were performed by the N2O pas-HCNMHG6070201( )sivation method [15,16] using the same equipment asFigure 2. Effect of aging time on the XRD patternTPR test, while the heating rate of the reduction wasof copper oxides10 °C/min.(1) C-1c, (2) C-2c, (3) C-3c .110Deren Fang et al./ Journal of Natural Gas Chemistry Vol. 14 No. 2 2005The differential thermogravimetricanalysisloss was 25.4%, equal to the theoretic weight loss ofcurves are demonstrated in Figure 3. In the caseZns(CO3)2(OH)6.of C-1, there are two decomposition peaks, their tem-peratures were 242.7 and 281.6 °C with the per-3.2. Copper- zinc precursorscentage of the weight loss 23.89% and 8.38%, respec-tively. The lower temperature peak may belong toThe XRD pattern of Cu-Zn series precursorsthe species of copper nitrate hydroxide and georgeitewhich is an amorphous phase having a compositionand its calcined counterpart are llustrated in Fig-of Cu2CO3(OH)2; the higher peak was the malachiteure 4 and Figure 5, respectively. In the initial stagewith a better crystallinity than georgeite [9,10]. From(CZ-1)，the species was mainly composed of cop-the whole weight loss and the XRD pattern it canper nitrate bydroxide and georgeite as well as hy-e postulated that the major phase of C-1 may bedrozincite Zns(CO3)2(OH)6, although the latter ma-amorphous copper nitrate hydroxide Cu2(NO3)(OH)3terial cannot be detected by XRD techniques ow-with some amount of georgeite. After colour tran-ing to its high dispersion. In the period of coloursition (C-2) the decornposition curve was a singletransition (CZ-2) the amount of Cu2(NO3)(OH)3 de-sharp peak at 253.3 °C with the percentage of weightcreased, while no other new phase appeared on theloss 30.5% which can be attributed to the phase ofXRD pattern. According to the experimental reCu2(NO3)(OH)3 and Cu2CO3(OH)2 crystal (mala-sults of pure copper precursor in the earlier part ofchite). Owing to the coexisting and intimate contactthis paper we hypothesize that the major reactionof the two phases, the decomposition temperature ofmay be the transition of Cu2(NO3)(OH)3 to georgeiteCu2(NO3)(OH)3 shifted to the higher side and thatat this moment, of course, some substitution mayof the malachite to the lower side and then the twotake place between cupric and zinc ions, and somemerged into one. In the sample of C-3, the DTGamount of rosasite (Cu,Zn)2CO3(OH)2 or aurichal-curve showed two decomposition peaks, a major peakcite (Zn,Cu)s(CO3)2(OH)6 may be formed. Since theat 292.9 °C and a satellite peak at 317.4 °C with thephase was amorphous, it cannot be detected from thewhole weight loss 28.95%, by reference to the XRDXRD patterns. The XRD pattern of calcined speciespattern the phase can be assigned to malachite.CZ-1c and CZ-2c are also similar (see Figure 5), butthe CuO grain size of CZ-2c was smaller than that ofCZ-1c (see Table 2). This suggests that some substi-(3)tution between cupric and zinc ions indeed occurred.293.9(2ennssie_(4)tninn吕|(1)(2)●Cu(OH),NO, .242.7。Cu,CO.(OH)2▲(Cu,Zn),CO,(OH)2100200300005000 (Zn,Cu)(CO3)(OH)。Temperature (C)Figure 3. Effect of aging time on the DTG curve ofcopper precursors(1)C-1, (2)C-2, (3) C-3中国煤化工心10The XRD pattern of Zn precursor demonstratedMHCNMHGthat Z-1 is a pure hydrozincite Zns(CO3)2(OH)6.Figure 4. Effect of aging time on the XRD patternof copper-zinc precursorsThis result was also verified by the TG test. Thereis a single weight loss peak at 245 °C and the weight(1) CZ-1, (2) CZ-2, (3) CZ-3, (4) CZ-4Journal of Natural Gas Chemistry Vol. 14 No.22005111In the sample of CZ-3 the rosasite and aurichalcite●ZnOappeared in the XRD pattern, and their arnount in-。CuOcreased as the aging time extended from CZ-3 to CZ-4.(4On the other hand, the amount of Cu2(NO3)(OH)3(3decreased dramatically in CZ-3 and disappeared inCZ-4. The XRD pattern of CZ 3c and CZ-4c revealedthat the diffraction peak of ZnO overlapped with the :diffraction peak of CuO in these samples. For exam-(2)JUple, the diffraction peak of ZnO at 20= -31.60 over-lapped with the diffraction peak of CuO at 20= -32.40;while the diffraction peak of ZnO at 20= =34.3° and36.20 overlapped with the diffraction peak of CuOt 20=35.4". That is to say, a solid solution wasformed between CuO and ZnO, and their CuO grain203(406(080size decreased greatly compared to the CuO grain20/(° )size of CZ-1c and CZ-2c. Comparing the XRD pat-Figure 5. Effect of aging time on the XRD patterntern of CZ-3 and CZ-4, we find that the content ofof Cu-Zn oxides(Zn,Cu)s(CO3)2(OH)6 in CZ-4 is much more than(1) CZ-1c, (2) CZ-2c, (3) CZ-3c, (4) CZ-4cthat in CZ-3. This is very important and will be dis-cussed later.In order to confirmn this hypothesis further, we in-The DTG curve of sample CZ-3 and CZ-4 ilus-vestigated the decomposition process by differentialtrated that with the increase of aging time, the de-thermogravimetric technique and the results are il-composition temperature shifted to the higher side.lustrated in Figure 6. There is a single decompositionAccording to the results by other authors [11] thispeak at 249 °C with a weight loss of 30.97% in thephenomenon may be explained by the formation ofDTG curve of CZ-1, it may be ascribed to copper ni-rosasite and aurichalcite.trate hydroxide and hydrozincite. While in the case ofFigure 7 demonstrated the temperature-CZ-2, there are two decomposition peaks at 237 andprogrammed reduction (TPR) profile of the calcined245.5 °C with weight loss of 16.83% and 6.82%, re-samples, curve C was for pure CuO made by the samespectively. Although we cannot exactly ascribe themethod as others except that the aging time was 60decomposition peak of CZ-2 to some substance atmin. There are three reduction peaks in CZ-1c, thepresent, but the phase transition had indeed occurred.lower temperature peak (170-180°C) can be attri-_3340.8()(2317.4后(1249.0中国煤化工10020030040000MYHCNMHG250Temperature (C)Temperature(C)Figure 6. Effect of aging time on the DTG curve ofFigure 7. Effect of aging time on the TPR profilecopper-zinc precursors(1) CZ-1. (2) CZ-2, (3) CZ-3, (4) CZ-4(1) CZ-1c, (2) CZ-2c, (3) CZ-3c, (4) CZ 4c, (5) C.112Deren Fang et al./ Journal of Natural Gas Chemnistry Vol. 14 No.2 2005buted to the amorphous surface CuO, the second re-XRD pattern of our test we can observe that the phaseduction peak around 220 °C may be ascribed to CuO-of (Cu,Zn)2CO3(OH)2 is easier to form than that ofZnO solid- solution formed from the rosasite, while the(Zn,Cu)s(CO3)2(OH)6. .last peak (260 °C ) can be ascribed to the CuO crystalin bulk. For sample CZ-2c, the TPR profile reveals3.3. Catalytic activitythat the state of CuO has changed a great deal withrespect to CZ-1c, the amount of bulk CuO has de-The catalytic activity of CuO/ZnO was testedcreased dramatically, at the same time a new peakwith fixed-bed flow reactor, the results are given inat about 190- 200 °C appeared which can be ascribedTable 1. The results demonstrated that the initialto the CuO-ZnO solution formed from the decomposi-activity and stability of the catalyst was increasedtion of aurichalcite. Comparing the four TPR profilesmonotonously as the aging time became longer. It isin Figure 7, we find that the peak intensity of theclear that the specific activity (activity per unit areatwo kinds of CuO -ZnO solid-solution increased withof Cu) is not just simply a direct proportion to the Cuthe aging time, especially the CuO-ZnO solid solutionsurface area. The specific activity is in proportion toof the lower temperature peak formed from the de-the Cu surface area in which the Cu particle has inti-composition of aurichalcite. This result is in agree-mate contact with the ZnO, that is, dependent on thement with that of XRD. As for the diference of thestructure of the catalyst. The stronger the synergytwo kind of CuO in reduction temperature, it may bebetween Cu and ZnO becomes, the higher the cat-explained as follows: in the case of CuO originatedalytic activity is. The other useful result is that thefrom the (Zn,Cu)s(CO3)2(OH)6, the content of CuOstability of the catalyst also dramatically increasedis smaller compared to ZnO, so the grain size of CuOas the aging time became longer. This is very im-produced in this way was fine and it was surroundedportant, because the commercial catalyst must haveby a plenty of ZnO grains which offer active H bya sustained activity for a long-term methanol syn-spillover to CuO to enhance the reduction of CuO.thesis operation. Comparing the composition of theWhile in the case of CuO-ZnO solid solution decom-Cu-Zn precursors, we may find that the CuO-ZnOposed from (Cu,Zn)2CO3(OH)2, the grain size of CuOsolid- solution produced from (Cu,Zn)2CO3(OH)2 andwas larger and the H atom offered by ZnO through(Zn,Cu)5(CO3)2(OH)6 contributed more activity tospillover was lttle, so the reduction of CuO was more the catalyst than that formed from Cu2(NO3)(OH)3difficult than the former case.and georgeite; furthermore, the CuO-ZnO solid-From the results of TPR profile and XRD patternsolution formed from (Zn,Cu)s(CO3)2(OH)6 is morewe can reach the conclusion: during the period of ag-stable than that formed from (Cu,Zn)2CO3(OH)2.ing the phase of Cu2(NO3)(OH)3 first transited toThis is the reason why CZ-4c has a much more stabil-georgeite, then some substitution between the phaseity than CZ- 3c, although they have almost equal to-of Cu2CO3(OH)2 and the phase of Zns(CO3)2(OH)6tal amount of CuO ZnO solid solution. These resultstook place and resulted in the formation of twostrongly suggest the methanol synthesis is a structure-new eutectic compounds: (Cu,Zn)2CO3(OH)2 andsensitive catalytic reaction [19]. Besides this, the cat-(Zn,Cu)s(CO3)2(OH)6. This result differs fromalytic activity has a sharp increase after the colourthe conclusion of Pollard and Spencer [9,10] whotransition, this result must be associated with a struc-consider zinc malachite (Cu,Zn)2CO3(OH)2 as thetural change. For the reason of this it will be discussedunique phase. Besides this, from the TPR profile andlater.Table 1. Catalytic activity and heat stabilityInitial activityRemaining activityReservedSamplemmolcHgOH/(gcat .h)mmolcH3OH/(m2cu.b)(mmol/(geat:h))fractionCZ-1c0.07CZ-2c4.00.10中国煤化工CZ-3c12.90.28MYHCNMHG0.2CZ-4c21.20.320.50Journal of Natural Gas Chermistry Vol. 14 No. 2 2005113The BET surface area as well as Cu surface areapoint of crystallinity. Since the crystallinity of theand particle size are given in Table 2.precursor of CZ-3 is lower than that of CZ-4, so the .The data of Table 2 showed that the BET surfacedecomposition products may thus have a high poros-area was increased as the aging time became longer,ity and surface area. On the other hand, the averageexcept for CZ-4c, with its BET surface area lower thanpore diameter was increased monotonously with thethat of CZ-3c. This may be explained from the view-increase of the aging time.Table 2. Physical properties of catalystsBET methodN2O passivation methodXRD methodSample Surface area Pore volumeAverage poreCu surface area Cu particle Cu dispersion CuO particlem2/g)(ml/g)diameter (nm)(m2/g)size (nm)(%)CZ-1c .9.00.0210.531.021.822.2CZ-2c12.60.0411.640.916.46.19.7CZ-3c64.60.2616.046.014.76.87.0CZ-4c5.5.0.2417.010.05.466.8The results of N2O passivation method showedthe curves have a break in the vicinity of the colourthat the Cu surface area and Cu dispersion increasedtransition point. By making a comprehensive con-as the aging time became longer, and the Cu par-sideration of XRD, DTG, TPR and catalytic activ-ticle size decreased with the aging time.The reity experimental results, this phenomenon may be ex-sults showed that with the increase of the agingplained as follows: at the moment of colour transition,time, the interaction between CuO-ZnO would be-a large amount of Cu2(NO3)(OH)3 was convertedcome stronger, so the aggregation of Cu particles dur-to amorphous (Cu,Zn)2CO3(OH)2, i.e. georgeite ining the period of reduction may be inbibited by ZnO.a short time. The structural adjustment of copperFrom the results of BET and N2O passivation asspecies accompanying the process was significant andwell as activity data, we may conclude that the syn-provided many chance for the substitution of Cu2+ergy between Cu and ZnO gradually increases withand Zn2+ in the precursors. This may be the causethe aging time.why the catalytic activity increased sharply after theWe traced the pH value of mother liquor in the pe-point of colour transition, that is, a large amount ofriod of aging with an acidimeter, the result was ilus-(Cu,Zn)2CO3(OH)2 and (Zn,Cu)s(CO3)2(OH)6 watrated in Figure 8. The pH value of the mother liquorformed at this moment and resulted in the formationincreased monotonously with the aging time, whileof CuO ZnO solid-solution in the catalyst which wasthere is a minimum value in the vicinity of colour tran-the active phase for the methanol systhesis [17,18]. .sition point (about 9 min after precipitation). Com-As for the break on the pH curve around the point ofparing the curves of catalytic activity, physical prop-colour transition it can be explained by the followingerties of the catalyst as well as the pH value with thereaction model [9,12].aging time, we find an interesting phenomenon: allCu(NO3)2 + 3Cu(OH)2(s) - →3Cu(OH)2(s) + Cu2+ + 2NO3(1)Cu2+ + 20H- - ，Cu(OH)2()(22Cu2+ + CO3- + 20H-一 →Cu2CO3(OH)2(s)(3)Cu(OH)2() + Zn2+ + CO3-一 →(Cu,Zn)2(OH)2CO3(4)Zns(CO3)2(OH)6 + Cu2+一 →(Zn,Cu)s(CO3)2(OH)6 + Zn2+(5H+ +CO3- →HC((6)中国煤化工HCO3+H+一→H2CC(7TYHCNMHGReactions (1)-(5) occurred simultaneously in theresulting in the relative excess of H+ ion in the solu-vicinity of the colour transition point, a large amounttion so the pH was decreased. While the accumulationofCO3 - and OH- was consumed by various reactions,of H+ ion will induce the equilibrium of reaction (6),114Deren Fang et al./ Journal of Natural Gas Chemistry Vol. 14 No. 2 2005(7) shifting to right to form H2CO3, which decom-solution originated from (Zn,Cu)s(CO3)2(OH)6 con-posed to carbon dioxide and water, so the pH valuetributes more to the activity of the catalyst than theof the solution increased sharply at the later periodones from (Cu,Zn)2CO3(OH)2.of colour transition.(3) The methanol synthesis is a structure-sensitivecatalytic reaction. There is synergy between Cu and8.4ZnO and its strength gradually increases with aging2time.8.0References7.8年7.66[] Chinchen G C, Denny P J, JenningsJ R et al. ApplCatal, 1988, 36: 17.42] Velu s, Suzuki K, Okazaki M et al. J Catal, 2000,194: 3733] Mascaros S M, Navarro R M, SainneroL G et al. JCatal, 2001, 198: 3380152030[4] Okamoto Y, Fukino K, Imanaka T et al. J PhysAging time (min)Chem, 1983, 87: 3740Figure 8. Plots of aging time us pH of mother liquor[5] Himelfarb P B, Simmons G W, Klier K et al. J Catal,1985, 93: 442[6] Idem R O, Bakhshi N N. Ind Eng Chem Res, 1994,4. Conclusions33: 20477] Trifiro F, Vaccari A. J Catal, 1984, 85: 2608] Shen G Ch, Fujita S, Takezawa N. J Catal, 1992, 138:(1) The initial phases of pure copper precur-754sor were in the form of copper nitrate hydrox-[9] Pollar A M d, Spencer M S, Thomas R G et al. Applide Cu2(NO3)(OH)3 and georgeite. On aging, theCatal A, 1992, 85:1Cu2(NO3)(OH)3 first transited to georgeite then[10] Spencer M S. Catal Lett, 2000, 66: 255changed to malachite Cu2CO3(OH)2.[11] Millar G J, Holm 1H, UwinsP J R et al. J Chem Soc(2) In the case of copper -zinc precursors, theFaraday Tran, 1998, 94: 593initial phases of precipitate were copper nitrate hy-12] LiJ L, Inui T. Appl Catal A, 1996, 137: 105droxide Cu2(NO3)(OH)3, georgeite and hydrozincite [13] Fujita S I, Kanamori Y, Satriyo A M, Takezawa NZns(CO3)2(OH)6. After colour transition, owing toCatal Today, 1998, 45: 241the mutual interdiffusion of copper precursors and[14] Fujita S I, Moribe s, Kanamori Y et al. Appl CatalA, 2001, 207: 121zinc precursors, resulted in the formation of twomew phases: rosasite (Cu,Zn)2CO3(OH)2 and au-[15] Bond G C, Namijo S N. J Catal, 1989, 118: 507richalcite (Zn,Cu)s(CO3)2(OH)6, both can be con-[16] Velu S, Suzuki K, Osaki T. Catal Lett, 1999, 62: 159[17] Choi Y, Futagami K, Fujitani T et al. J Appl Catalverted to the CuO-ZnO solid-solution under calcina-A, 2001, 208: 163tion. It is the reason why the catalyst activity in-[18] Fujitami T, Nakamura J. Appl Catal A, 2000, 191:creased sharply after the colour-transition. The phase111of (Cu,Zn)2CO3(OH)2 is formed more easily than that[19] Chinchen G C, Spencer M S. Catal Today, 1991, 10:of (Zn,Cu)s(CO3)2(OH)6, while the CuO-ZnO solid-293中国煤化工MYHCNMHG