Alcohol promoted methanol synthesis enhanced by adsorption of water and dual catalysts

This is a e sion o a publica ion in Please ci e he publica ion as ollows: DOI: Copy igh o he o iginal publica ion: This is a pa allel published e sion o an o iginal publica ion. This e sion can di e om he o iginal published a icle. published by Alcohol p omo ed me hanol syn hesis enhanced by adso p ion o wa e and dual ca alys s Nieminen Ha i, Gi i o skiy Geo gy, Laa i A o, Koi anen Tuomas Nieminen, H., Gi i o skiy, G., Laa i, A., Koi anen, T. (2018). Alcohol p omo ed me hanol syn hesis enhanced by adso p ion o wa e and dual ca alys s. Jou nal o CO2 U iliza ion, Vol. 24, pp. 180-189. DOI: 10.1016/j.jcou.2018.01.002 Final d a Else ie Jou nal o CO2 U iliza ion 10.1016/j.jcou.2018.01.002 © 2018 Else ie L d. 1 Alcohol p omo ed me hanol syn hesis enhanced by adso p ion o wa e and dual ca alys s 1 Ha i Nieminen*, Geo gy Gi i o skiy, A o Laa i, Tuomas Koi anen 2 Lappeen an a Uni e si y o Technology, Labo a o y o P ocess and P oduc De elopmen , P.O. Box 20, FI-53851 3 Lappeen an a, Finland 4 * Co esponding au ho Tel.: +358 40 7451800, E-mail add ess: ha i.nieminen@lu . i 5 Abs ac 6 Alcohol-p omo ed me hanol syn hesis uses he e ogeneous me hanol syn hesis ca alys s in alcoholic sol en s 7 whe e he alcohols ac as a co-ca alys . In he p esence o alcohol, he eac ion p oceeds h ough alcohol o ma e 8 es e as an in e media e, allowing me hanol syn hesis a lowe empe a u es han con en ional gas-phase 9 syn hesis. In he p esen wo k, alcohol-p omo ed CO2 hyd ogena ion o me hanol was s udied expe imen ally using 10 a Cu/ZnO ca alys wi h 1-bu anol and 2-bu anol as sol en s. As wa e is known o inhibi me hanol syn hesis on 11 Cu/ZnO ca alys s, he alcohol-p omo ed p ocess was u he de eloped by in-si u adso p ion o wa e using a 3Å 12 molecula sie e. The me hanol p oduc i i y signi ican ly imp o ed as a esul o he lowe ed concen a ion o wa e . 13 The concen a ion o wa e was hus iden i ied as a key ac o a ec ing he o e all me hanol p oduc i i y. As he 14 alcohol-p omo ed me hanol syn hesis p ocess is cha ac e ized by wo sepa a e eac ion s eps, he use o sepa a e 15 ca alys s op imized o each s ep o e s an in e es ing app oach o he de elopmen o his p ocess. Such a dual- 16 ca alysis concep was es ed using a coppe ch omi e ca alys oge he wi h Cu/ZnO. P omising esul s we e 17 ob ained, as me hanol p oduc i i y inc eased wi h he addi ion o coppe ch omi e. Ca alys cha ac e iza ion was 18 ca ied ou using XRD and SEM-EDS and po en ial e ec s o obse ed changes in ca alys s uc u e du ing eac ion 19 a e discussed. 20 Keywo ds 21 CO2 hyd ogena ion, me hanol syn hesis, Cu/ZnO, liquid-phase, alcohol p omo ed, dual ca alysis, coppe ch omi e, 22 molecula sie e 23 Con lic s o in e es : none 24 25 2 1. In oduc ion 26 De elopmen o e icien and lexible ene gy s o age me hods is c i ical o a global shi om a ossil uels based 27 economy o a enewable ene gy based economy [1]. The use o su plus peak elec ici y gene a ed om luc ua ing 28 enewable ene gy sou ces, such as wind and sola ene gy, o he p oduc ion o chemical compounds would enable 29 ene gy s o age in a highly anspo able o m a high ene gy densi y. Gene a ion o hyd ogen by elec olysis o 30 wa e is he common s a ing poin in chemical ene gy s o age s a egies [2]. Howe e , due o he di icul ies and 31 haza ds associa ed wi h la ge-scale s o age and anspo a ion o gaseous hyd ogen, u he u iliza ion o hyd ogen 32 o p oduc ion o ca bon-con aining liquid uels and chemical compounds migh be p e e able. 33 Me hanol is an example o such a po en ial liquid-phase chemical ene gy ca ie [3]. Me hanol is an impo an and 34 e sa ile indus ial chemical ha can also be used as a uel in powe gene a ion and in in e nal combus ion engines 35 and uel cells [4]. Addi ionally, me hanol is a e sa ile aw ma e ial o syn hesis o a a ie y o chemical p oduc s. 36 Fo ins ance, me hanol can be ans o med in o gasoline in he me hanol- o-gasoline p ocess (MTG) [5] o in o 37 ole ins in he me hanol- o-ole ins p ocess (MTO) [6]. 38 Cu en p oduc ion o me hanol is based on ca aly ic con e sion o syn hesis gas gene a ed om ossil sou ces, 39 commonly na u al gas. The syngas is mainly composed o mix u es o hyd ogen, ca bon monoxide and ca bon 40 dioxide. In con en ional me hanol syn hesis, coppe and zinc oxide (Cu/ZnO) ca alys s a e gene ally employed a 41 eac ion empe a u es o 200-300 °C and p essu es o 50-100 ba [7]. 42 The me hanol syn hesis p ocess can be desc ibed by he ollowing h ee equilib ium eac ions: 43 CO2+ 3 H2⇌ CH3OH + H2O Δ𝐻0= −49.8 kJ/mol (1) 44 CO + 2 H2⇌ CH3OH Δ𝐻0= −91.0 kJ/mol (2) 45 CO + H2O ⇌ CO2+ H2 Δ𝐻0=41.2 kJ/mol (3) 46 The exo he mic eac ions (1) and (2) ep esen , espec i ely, he hyd ogena ion o CO2 and CO o me hanol. 47 Reac ion (3), he wa e -gas shi (WGS) eac ion, is ele an o me hanol syn hesis as he eac ion is also ac i a ed 48 by he coppe -based me hanol syn hesis ca alys s [8]. As me hanol syn hesis is exo he mic and esul s in a 49 educ ion o mola olume, me hanol syn hesis is a o ed by low empe a u es and high p essu es. Howe e , 50 3 empe a u es abo e 200 °C a e equi ed o su icien ly high eac ion a es, and hus he he modynamic equilib ium 51 limi s he me hanol syn hesis o low con e sion le els. Hyd ogena ion o pu e CO2 o me hanol is also possible bu 52 he equilib ium con e sions a e e en lowe han o CO. Figu e 1 shows he calcula ed equilib ium con e sion o 53 s oichiome ic CO and CO2 eeds a di e en empe a u es and p essu e. The con e sions a e modelled by Soa e- 54 Redlich-Kwong equa ions o s a e, which ha e been shown o accu a ely p edic expe imen al esul s in me hanol 55 syn hesis [9]. Howe e , he hyd ogena ion o CO2 on Cu/ZnO ca alys s is highly selec i e o me hanol, wi h o he 56 he modynamically mo e a o able p oduc s such as me hane, e he s and ke ones o med only in negligible 57 amoun s [10]. 58 59 Figu e 1. E ec o empe a u e and p essu e on he equilib ium ca bon con e sion om s oichiome ic 60 CO2:H2 (1:3) and CO:H2 (1:2) mix u es. Calcula ed wi h he p edic i e Soa e-Redlich-Kwong 61 (PSRK) [11] equa ion o s a e in Aspen Plus. 62 To o e come he he modynamic limi a ions in he gas-phase me hanol p ocess, liquid-phase syn hesis p ocesses 63 ha e been p oposed as an al e na i e app oach o enable lowe eac ion empe a u es in syngas eac ions. Ea ly 64 de elopmen s u ilized highly basic ca alys sys ems such as alkali alkoxides in combina ion wi h coppe ch omi e 65 [12, 13, 14] o nickel-based ca alys s [15, 16, 17]. Me hanol syn hesis om CO/H2 a empe a u es as low as 100 66 °C and p essu es be ween 30 and 65 ba we e epo ed [18]. Howe e , he basic ca alys s a e incompa ible wi h 67 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 150 170 190 210 230 250 Con e sion Tempe a u e, °C CO₂, 40 ba CO₂, 100 ba CO, 40 ba CO, 100 ba 4 CO2 o wa e , he p esence o which, e en a ace amoun s, leads o apid ca alys deac i a ion [17]. A me hod 68 p oposed by he B ookha en Na ional Labo a o y (BNL) also u ilized a highly basic sys em o he con e sion o 69 CO o me hanol a signi ican ly low empe a u e and p essu e [19]. Fu he mo e, liquid-phase me hanol syn hesis 70 om CO2-con aining syn hesis gas in ine hyd oca bon sol en has been demons a ed in he LPMeOH p ocess 71 [20]. 72 CO2 has been iden i ied as he main ca bon sou ce in me hanol syn hesis om syngas [21]. Hence, i may be 73 expec ed ha me hanol can also be p oduced by hyd ogena ion o pu e CO2. Hyd ogena ion o CO2, cap u ed om 74 poin sou ces o e en di ec ly om he a mosphe e, would hen p o ide a sus ainable sou ce o ca bon-based uels 75 and chemicals while helping o educe he a mosphe ic concen a ion o CO2 [22]. Some pilo -scale me hanol 76 p ocesses ha can use CO2 as he s a ing ma e ial ha e been de eloped. These include he CAMERE p ocess 77 [23], which combines he e e se wa e -gas-shi eac ion and me hanol syn hesis om syngas, and he Ma sui 78 Chemicals p ocess [24], which di ec ly con e s CO2 o me hanol. Addi ionally, Ca bon Recycling In e na ional 79 es ablished comme cial me hanol p oduc ion om CO2 in 2011, and he S a sengi plan is p esen ly ope a ing a 80 a capaci y o abo e 5 million li e s pe yea [25]. The p ocess u ilizes geo he mal ene gy eadily a ailable in Iceland. 81 One possible way o in luence he eac ion kine ics and condi ions is o change he eac ion ou e ha leads o he 82 o ma ion o me hanol. A no el alcohol-p omo ed liquid-phase me hanol syn hesis p ocess i s p oposed by Fan e 83 al. [26] is based on he combina ion o a con en ional Cu/ZnO ca alys and alcohol as a ca aly ic sol en . The 84 alcohol p omo es me hanol syn hesis by al e ing he eac ion ou e, allowing ope a ion a lowe empe a u es. In 85 he p esence o he alcohol, he eac ion p oceeds h ough he o ma e es e o he co esponding alcohol as an 86 in e media e. As a esul , me hanol can be p oduced om syngas a empe a u es s a ing om 170 °C and 87 p essu es in he ange o 30 o 50 ba [27]. Impo an ly, he p ocess does no employ basic ca alys s sensi i e o 88 deac i a ion by CO2, allowing di ec con e sion o CO2. The ollowing eac ion s eps ha e been p oposed o his 89 p ocess [28], suppo ed by subsequen in-si u IR obse a ions [29]: 90 1. Hyd ogena ion o ca bon dioxide in o o mic acid 91 CO2+ H2⇄ HCOOH (4) 92 93 5 2. Reac ion o o mic acid wi h e hanol, o ming e hyl o ma e 94 95 HCOOH + C2H5OH ⇄ HCOOC2H5+ H2O (5) 96 97 3. Hyd ogena ion o e hyl o ma e, o ming me hanol and e hanol 98 HCOOC2H5+ 2 H2⇄ CH3OH + C2H5OH (6) 99 The ne eac ion is he hyd ogena ion o ca bon dioxide o me hanol (Eq. 1) wi h a s anda d eac ion en halpy o - 100 49.8 kJ/mol. Di e en alcohols ha e been shown o possess di e en p omo ing e ec o me hanol syn hesis. 101 Tsubaki e al. [30] ound linea alcohols o be mo e e ec i e compa ed o hei b anched coun e pa s, wi h n- 102 bu anol showing he bes esul s. Zeng e al. [31] epo ed ha he yield o bo h me hanol and he co esponding 103 es e dec eased wi h inc easing ca bon numbe o he 1-alcohols om e hanol o 1-hexanol. Fo alcohols wi h he 104 same ca bon numbe bu di e en s uc u e, 2-alcohols we e ound o ha e highe ac i i y, which was explained by 105 a combina ion o spa ial and elec onic e ec s. As a esul , 2-p opanol showed he highes p omo ional e ec . La e , 106 2-bu anol was epo ed as he mos e ec i e sol en o he con inuous me hanol syn hesis in a semiba ch eac o 107 [32]. 108 As he alcohol-p omo ed me hanol syn hesis p ocess is cha ac e ized by wo sepa a e eac ion s eps, he u iliza ion 109 o sepa a e ca alys s op imized o each eac ion could be bene icial. Such dual- o cascade ca aly ic sys ems ha e 110 been conside ed p e iously o me hanol syn hesis. Hu and San o d [33] epo ed e ec i e CO2 con e sion o 111 me hanol a 135 °C using a combina ion o homogeneous ca alys s. Chen e al. [34] used he e ogeneous ca alys s 112 in 1,4-dioxane sol en : coppe ch omi e o he hyd ogena ion o CO2 o o ma e and Cu/Mo2C o he o ma e 113 hyd ogenolysis o me hanol. This sys em was capable o me hanol p oduc ion a a es compa able o con en ional 114 gas-phase syn hesis a 135 °C and exhibi ed me hanol selec i i y abo e 75%. The me hanol syn hesis was 115 p omo ed by he addi ion o e hanol, wi h he eac ion p oceeding h ough e hyl o ma e, as epo ed in he alcohol- 116 p omo ed p ocess. On he o he hand, coppe ch omi e is known o ca alyze he hyd ogenolysis o es e s o 117 alcohols, i.e. he la e s age in he alcohol-p omo ed eac ion ou e [35]. As such, coppe ch omi e appea s an 118 in e es ing componen o a dual ca aly ic sys em o alcohol-p omo ed me hanol syn hesis. 119 6 In compa ison o CO-con aining syngas eed, CO2 hyd ogena ion o me hanol is u he complica ed by he 120 inc eased o ma ion a e o wa e . Wa e is o med as a byp oduc in me hanol syn hesis, and in he absence o 121 CO, he wa e -gas shi eac ion p oceeds in he e e se di ec ion, p oducing mo e wa e . The nega i e e ec o 122 wa e on me hanol syn hesis on Cu/ZnO-based ca alys s has been well documen ed [36]. This e ec has been 123 explained as a combina ion o kine ic inhibi ion e ec s and s uc u al ca alys deac i a ion. Wa e -de i ed hyd oxyl 124 species can block he ac i e si es on he ca alys , esul ing in kine ic inhibi ion. The p esence o wa e can also 125 accele a e he sin e ing o coppe pa icles [37], esul ing in dec eased coppe dispe sion and ca alys deac i a ion. 126 Remo al o me hanol and wa e using memb ane eac o s [38, 39] and by condensa ion a high p essu es [40] o 127 low empe a u es [41] has been p e iously desc ibed o gas-phase me hanol syn hesis. Reac i e dis illa ion [42] 128 p o ides a u he possible app oach o con inuous p oduc emo al, pa icula ly in liquid-phase p ocesses, and 129 has been p oposed in li e a u e o he me hanol syn hesis p ocess [43] and o he Fische -T opsch p ocess [44] 130 ope a ing a simila condi ions. In addi ion, selec i e emo al o wa e by adso p ion on zeoli e molecula sie es has 131 also been sugges ed in so p ion-enhanced me hanol [45] and ela ed dime hyl e he [46] syn hesis ope a ed in he 132 gas-phase. 133 In he p esen wo k, alcohol-p omo ed me hanol syn hesis was in es iga ed expe imen ally using a comme cial 134 Cu/ZnO-based me hanol syn hesis ca alys wi h 1-bu anol and 2-bu anol as he sol en s. 2-bu anol was selec ed 135 because o he p e iously epo ed high ac i i y o me hanol syn hesis, and 1-bu anol was conside ed in e es ing 136 because o he po en ially simpli ied p oduc sepa a ion due o he highe boiling poin o he alcohol. As no el 137 de elopmen s, enhancemen o he alcohol-p omo ed me hanol syn hesis by in-si u adso p ion o wa e and by he 138 use o dual ca alys s we e s udied. Wa e adso p ion was ca ied ou using a molecula sie e. Me hanol syn hesis 139 combined wi h wa e emo al has p e iously been modelled based on 4Å molecula sie es [45], and he use o 4Å 140 molecula sie es has been modelled o a ela ed dime hyl e he (DME) syn hesis [46]. Howe e , expe imen al wo k 141 o me hanol syn hesis p omo ed by wa e adso p ion has no been published ea lie o ou knowledge. A dual 142 ca alys sys em comp ising o a combina ion o Cu/ZnO and coppe ch omi e ca alys s was es ed wi h he aim o 143 imp o ing me hanol p oduc i i y by in luencing sepa a ely he o ma e o ma ion and hyd ogenolysis eac ion s eps. 144 7 2. Ma e ials and me hods 145 A Pa 4520 au ocla e eac o wi h an inne olume o 450 ml was used o he eac ion expe imen s. The eac o 146 was connec ed o a Pa 4848 con ol uni used o con ol he eac ion empe a u e and mixing speed. A mixing 147 speed o 600 pm was used in all expe imen s. Liquid samples om he eac ion mix u e we e collec ed using a 148 wa e -cooled sample collec ion essel, in which any apo s p esen in he sample we e condensed p io o collec ing 149 he sample. 150 Analysis g ade 1-bu anol and 2-bu anol, we e used as sol en s. A comme cial Cu/ZnO-based me hanol syn hesis 151 ca alys (Al a Aesa , 65.5 % CuO, 24.7% ZnO, 10.1% Al2O3, 1.3% MgO) was used. The ca alys was g ound and 152 sie ed o 150-500 µm o each expe imen . The 3Å molecula sie e (UOP, beads wi h diame e o 2 mm), was also 153 g ound and sie ed o 150-500 µm. An ini ial expe imen wi h he ung ound molecula sie e was also pe o med. 154 The molecula sie e was ac i a ed by hea ing o 250 °C o a leas 8 hou s unde ai and subsequen cooling o 155 ambien empe a u e inside a desicca o p io o use. Powde ed coppe ch omi e (Sigma-Ald ich) was used in he 156 dual ca alys expe imen s. A mixed gas con aining 75% hyd ogen and 25% ca bon dioxide was used as he eac ion 157 eed gas, and a mixed gas con aining 5% hyd ogen in ni ogen was used o ac i a ion o he ca alys s. A diag am 158 o he expe imen al se up is p esen ed in Figu e 2. 159 8 160 Figu e 2. Expe imen al se up used in he eac ion expe imen s. 161 The g ound Cu/ZnO ca alys and he coppe ch omi e ca alys we e ac i a ed in-si u in he eac o essel. Ca alys 162 ac i a ion was pe o med unde 5 ba o he 5% H2/N2 mixed gas, wi h he gas inside he eac o eplaced e e y 30 163 minu es. The empe a u e was 200 °C du ing he ac i a ion. Following ca alys ac i a ion, he eac o was cooled 164 and he ca alys s we e kep unde he ac i a ion gas un il he eac ion expe imen was execu ed. 200 ml o he 165 alcohol was quickly pou ed in o he eac o , minimizing he con ac ime o he ca alys s wi h ai . The eac o was 166 pu ged wi h ni ogen and hea ed o he eac ion empe a u e unde N2. A he eac ion empe a u e, an ini ial liquid 167 sample was collec ed and he eac o was p essu ized wi h he eed gas (CO2:H2 = 1:3) o he se eac ion p essu e, 168 which was 60 ba unless o he wise no ed. Cons an p essu e was main ained du ing he expe imen s by eplacing 169 he consumed eac ion gas wi h esh gas. The o al eac ion ime was 6 hou s and liquid samples we e collec ed 170 e e y 2 hou s. 171 An Agilen Technologies 6890N gas ch oma og aph wi h a he mal conduc i i y de ec o was used o analysis o 172 he liquid samples. A pola Zeb on ZB-WAXplus column was used o he 2-bu anol samples. An iso he mal me hod 173 wi h he column empe a u e a 70 °C and helium (1.1 ml/min) as a ca ie gas was used. Fo he 1-bu anol samples, 174 a non-pola HP-1ms column was used due o insu icien sepa a ion o bu anal and me hanol in he ZB-WAXplus 175 15 290 Figu e 7. E ec o eac ion gas pa ial p essu e on me hanol p oduc i i y wi h 10 g o Cu/ZnO ca alys in 2- 291 bu anol a 180 °C. Feed gas (CO2:H2 = 1:3), eac ion ime 6 h. 292 3.3 Wa e emo al by molecula sie e 293 Con inuous emo al o wa e om he eac ion mix u e was es ed by addi ion o a zeoli e molecula sie e. Molecula 294 sie es wi h a po e diame e o 3 Å can be used o he dehyd a ion o alcohols because o hei selec i e adso p ion 295 o wa e [50]. The selec i e adso p ion is based on size exclusion o molecules la ge han wa e in he inne 296 mic opo ous s uc u e o he zeoli e. 297 The limi ing e ec o wa e on he alcohol-p omo ed me hanol syn hesis p ocess was i s con i med by pe o ming 298 an expe imen wi h app oxima ely 1.4 mol/dm3 o wa e added o 2-bu anol. This concen a ion is sligh ly abo e he 299 maximum concen a ion ange o wa e ound in he expe imen s (Figu e 4). A 180 °C and 60 ba o o al p essu e, 300 he me hanol p oduc ion a e was app oxima ely 74% lowe han in he base expe imen wi h no wa e added. The 301 concen a ion o wa e did no signi ican ly inc ease du ing his expe imen bu a he emained ela i ely cons an 302 a he appa en equilib ium le el. 303 Nex , he e ec o in-si u adso p ion o wa e by he addi ion o a 3Å molecula sie e was es ed. The ela i e 304 amoun s o he ca alys and he molecula sie e we e a ied, main aining a o al solids mass o 50 g. The esul s o 305 hese expe imen s a e p esen ed in Figu e 8. A base expe imen wi h 20 g o ca alys and no molecula sie e is 306 also p esen ed o compa ison. 307 0 5 10 15 20 25 30 35 40 30 35 40 45 50 Me hanol p oduc i i y, g / kg / h CO2+ H2pa ial p essu e, ba 16 308 Figu e 8. E ec o ca alys and molecula sie e mass on me hanol and wa e o ma ion in 2-bu anol. 309 Tempe a u e 180 °C, eed gas CO2:H2 = 1:3, o al p essu e 60 ba . 310 Compa ed o he base case wi h 20 g o Cu/ZnO ca alys and no molecula sie e, he addi ion o he ung ound 311 molecula sie e inc eased he me hanol p oduc i i y om 8.2 g/kg/h o 11.2 g/kg/h. A mo e signi ican imp o emen 312 was ound wi h he molecula sie e g ound in o 150-300 µm pa icle size ange. Due o he clea e ec o he pa icle 313 size, he adso p ion o wa e appea s o be signi ican ly di usion-limi ed o he ung ound molecula sie e. Wi h 20 314 g o ca alys , he addi ion o 30 g o he g ound molecula sie e inc eases he me hanol p oduc i i y o 33.6 g/kg/h, 315 an inc ease o o e 300% o e he Cu/ZnO ca alys used wi hou a molecula sie e. Keeping he o al amoun o 316 solids (ca alys + molecula sie e) a 50 g, he me hanol p oduc i i y inc eased wi h inc easing amoun s o molecula 317 sie e. Fo ins ance, he p oduc i i y inc eased o 54.4 g/kg/h using 10 g o he ca alys and 40 g o he molecula 318 sie e. These esul s clea ly show ha he ca alys is mos e ec i ely u ilized o me hanol syn hesis when la ge 319 ela i e amoun s o he molecula sie e o he ca alys a e used. This obse a ion can be explained by he inc eased 320 wa e adso p ion capaci y o he la ge amoun o he molecula sie e, leading o dec eased concen a ions o wa e , 321 as shown in Figu e 8. 322 323 0 0.2 0.4 0.6 0.8 1 1.2 1.4 20 g Cu/ZnO 20 g Cu/ZnO, 20 g MS (ung ound) 25 g Cu/ZnO, 25g MS 20 g Cu/ZnO, 30 g MS 10 g Cu/ZnO, 40 g MS 0 10 20 30 40 50 60 70 Concen a ion, mol/dm3 Me hanol p oduc i i y, g / kg / h Me hanol p oduc i i y Me hanol concen a ion Wa e concen a ion 17 3.4 Dual ca alys s 324 To es he dual ca alysis concep o alcohol-p omo ed me hanol syn hesis, coppe ch omi e (CuC ) was used in 325 combina ion wi h he Cu/ZnO ca alys . The a ios o he wo ca alys s we e a ied: 20 g o he Cu/ZnO ca alys was 326 used wi h 10 g o CuC , and ice e sa. The expe imen s we e ca ied ou in 2-bu anol a 180 °C and 60 ba o o al 327 p essu e, co esponding o a CO2 + H2 pa ial p essu e o 50.1 ba . The esul s o hese expe imen s a e p esen ed 328 in Figu e 9. A base expe imen wi h 20 g o Cu/ZnO ca alys and no coppe ch omi e is also p esen ed o 329 compa ison. 330 331 Figu e 9. E ec o di e en amoun s o Cu/ZnO and coppe ch omi e (CuC ) ca alys s on he o ma ion o 332 me hanol and wa e in 2-bu anol. Reac ion ime 6 hou s. Tempe a u e 180 °C, eed gas CO2:H2 = 333 1:3, o al p essu e 60. An expe imen wi h 20 g o Cu/ZnO ca alys and no coppe ch omi e is 334 included o compa ison. 335 The addi ion o he coppe ch omi e ca alys clea ly inc eases he me hanol p oduc i i y. Bo h he absolu e me hanol 336 p oduc ion a e, as measu ed by he me hanol end concen a ion, and he speci ic p oduc i i y o he ca alys 337 inc ease wi h addi ion o coppe ch omi e. The inc eased p oduc i i y can be explained ei he by a syne gis ic e ec 338 be ween he wo ca alys s o by highe me hanol syn hesis ac i i y o CuC compa ed o Cu/ZnO. Howe e , a highe 339 in insic ac i i y o coppe ch omi e appea s unlikely, as he ac i i y o Cu/ZnO o me hanol syn hesis is well-known 340 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20 g Cu/ZnO 20 g Cu/ZnO, 10 g CuC 10 g Cu/ZnO, 20 g CuC Concen a ion, mol/dm3 Me hanol p oduc i i y, g / kg / h Me hanol p oduc i i y Me hanol concen a ion Wa e concen a ion 18 and indus ially applied. Fan e al. [26] also epo ed highe me hanol yield and selec i i y o Cu/ZnO compa ed o 341 CuC in alcohol p omo ed me hanol syn hesis. Fan e al. also ound simila CO selec i i y, o RWGS ac i i y, o 342 bo h o he ca alys s. This is suppo ed by he p esen esul s, as he concen a ion o wa e was no signi ican ly 343 a ec ed by he changed a io o Cu/ZnO and CuC (Figu e 9, columns 2 and 3), suppo ing simila RWGS ac i i y 344 o he wo ca alys s. The o e all me hanol selec i i y appea s o be highe wi h he combined ca alys s, as he a io 345 o me hanol o wa e p oduced is inc eased compa ed o Cu/ZnO used alone. 346 3.5 Cha ac e iza ion o Cu/ZnO ca alys be o e and a e eac ion 347 The s uc u al ea u es o he Cu/ZnO ca alys be o e and a e eac ion we e in es iga ed by he means 348 o XRD and SEM-EDS in o de o assess he ca alys s abili y. Figu e 10 p esen s he X- ay 349 di ac og ams o he ca alys as supplied in he calcined o m, ollowing educ ion in 5% hyd ogen, and 350 ollowing use in alcohol-p omo ed me hanol syn hesis in 1-bu anol a 180 °C. I is no ed ha he same 351 ba ch o ca alys was analyzed p io o educ ion and ollowing he eac ion, while he educed ca alys 352 was p epa ed and analyzed sepa a ely. 353 354 Figu e 10. X- ay di ac og ams o he unused Cu/ZnO ca alys (A), he educed ca alys (B), and he 355 ca alys ollowing me hanol syn hesis om CO2 and H2 (1:3) in 1-bu anol a 180 °C (C). 356 19 The calcined ca alys is la gely amo phous, showing a mino pa e n co esponding o coppe (II)oxide 357 (CuO) ypical o Cu/ZnO ca alys s [51]. The pa e ns a e iden i ied based on he PDF 4+ 2018 358 c ys allog aphy da abase. The educed ca alys p esen s wi h a clea ly de ined pa e n consis en wi h 359 c ys alline, coppe (I)oxide (Cu2O), and me allic coppe . Weak c ys alline ea u es o zinc oxide a e also 360 e iden , consis en wi h p e ious s udies [52]. As he educ ion o coppe p oceeds s epwise om CuO 361 o Cu ia Cu2O [53], he p esence o Cu2O may imply incomple e educ ion, possibly due o insu icien 362 educ ion ime o empe a u e. Howe e , as he educed ca alys sample was ans e ed and analyzed 363 in con ac wi h ai , e-oxida ion o coppe c ys alli es du ing his p ocess canno be uled ou . 364 Only me allic coppe and zinc oxide is ound p esen in he used ca alys . Cu/ZnO ca alys s a e known o 365 show dynamic s uc u al changes depending on he oxida ion po en ial o he gas phase [54, 55] and 366 ongoing educ ion o he ca alys a he eac ion condi ions is possible. As he educed and used ca alys 367 analyzed he e a e no om he same ba ch o g ound and p epa ed ca alys , ba ch- o-ba ch a ia ion 368 canno be elimina ed as a cause o he obse ed s uc u al di e ences. 369 The peaks co esponding o zinc oxide a e mo e clea ly de ined compa ed o he educed ca alys , 370 po en ially indica ing con inuing c ys alliza ion o ZnO a he eac ion condi ions. Lunkenbein e al. [56] 371 iden i ied zinc oxide as he mo e dynamic phase compa ed o me allic coppe unde eac ion condi ions, 372 and ound ha c ys alliza ion o ZnO and he esul ing loss o eac i e Cu-ZnO in e aces is he main 373 mechanism o ini ial ca alys deac i a ion. The SEM-EDS elemen al maps o coppe and zinc p esen ed 374 in Figu e 11 indica e ha such a p ocess may ha e ini ia ed in he ca alys used he e. The unused 375 (calcined) ca alys shows a ela i ely homogeneous dis ibu ion o bo h coppe and zinc. Howe e , a 376 deg ee o seg ega ion o hese elemen s can be obse ed in he used ca alys , wi h he elemen al map 377 showing dis inc a eas wi h high con en o zinc (oxide) ha a e ela i ely poo in coppe . 378 20 379 Figu e 11. SEM-EDS elemen al maps o coppe and zinc in he unused Cu/ZnO ca alys (uppe ), 380 and he ca alys ollowing me hanol syn hesis om CO2 and H2 (1:3) in 1-bu anol a 180 381 °C (lowe ). Composi ion scales in weigh pe cen . 382 383 Fu he insigh is p o ided by he SEM images p esen ed in Figu e 12. Dis inc c ys als in he 384 mic ome e dimension can be obse ed, iden i ied as zinc oxide by he EDS analysis. No such ea u es 385 we e ound in he unused ca alys . I is concluded ha agglome a ion and c ys alliza ion o zinc oxide 386 du ing eac ion has occu ed, ac ing as a po en ial deac i a ion mechanism o he ca alys . Howe e , 387 as long- e m s abili y es s we e no pe o med he e, he ac ual e ec o hese s uc u al changes on he 388 ac i i y o he ca alys canno be discussed. 389 These obse a ions can be compa ed o o he indings discussed in li e a u e. P e iously, he s abili y 390 o Cu/ZnO ca alys in alcohol p omo ed me hanol syn hesis has been explo ed by Reub oycha oen e 391 al. [32] who ound he pe o mance s able du ing 40 hou s o con inuous me hanol syn hesis (a 170 392 °C), and by Jeong e al. [57] who ound no decline in ac i i y du ing 60 hou s o eac ion (150 °C). In 393 con as o ou esul s, Jeong e al. ound no changes in he XRD p o ile o he ca alys be o e and a e 394 eac ion. O he han he lowe eac ion empe a u e, he di e ing indings migh be explained by 395 di e en eed gas composi ion, as a CO- ich syngas was used in hese s udies opposed o he CO2:H2 396 21 mix u e used he e. The e o e, i is possible ha he de ec ed di e ences migh be caused by he la ge 397 amoun o wa e p esen in he eac ion sys em in he p esen s udy. 398 399 Figu e 12. SEM mic og aphs o he Cu/ZnO ca alys ollowing me hanol syn hesis om CO2 and H2 400 (1:3) in 1-bu anol a 180 °C. Zinc oxide c ys als a e highligh ed. 401 4. Conclusions 402 Me hanol syn hesis om CO2 was s udied in an alcohol-p omo ed liquid-phase p ocess using con en ional Cu/ZnO 403 and coppe ch omi e as ca alys s. 1-bu anol and 2-bu anol we e ound o ac as ca aly ic sol en s, allowing me hanol 404 syn hesis a lowe empe a u es han con en ional gas-phase p ocesses. Al hough i was no possible o de e mine 405 he exac eac ion ou e, i is expec ed ha he p omo ing e ec o he alcohols is based on a eac ion ou e 406 p oceeding h ough he in e media e o o ma e es e o he alcohol. 407 The e ec o con inuous wa e emo al using molecula sie e adso p ion was explo ed. The addi ion o a 3Å 408 molecula sie e signi ican ly enhanced me hanol p oduc i i y. G inding o he molecula sie e esul ed in imp o ed 409 esul s due o he sho e di usion pa h compa ed o he g anula ma e ial. The maximum me hanol p oduc i i y o 410 22 54.4 g/kg/h was ound when he maximum ela i e amoun o he molecula sie e (40 g) o he ca alys (10 g) was 411 used. The inal me hanol concen a ion a e 6 hou s o eac ion ime eached 0.5 mol/dm3. The ca alys was mos 412 e ec i ely used o me hanol syn hesis when he amoun o molecula sie e was maximized, which minimized he 413 concen a ion o wa e . The wa e concen a ion was ound o signi ican ly a ec he a e o me hanol syn hesis. 414 The o e all me hanol p oduc ion a e in his p ocess appea s o be limi ed by he concen a ion o wa e and i s 415 e ec s on he ca alys su ace. To p e en he nega i e e ec s o wa e , con inuous wa e emo al o de elopmen 416 o mo e wa e esis an ca alys s is i al o u he de elopmen o his p ocess. Based on he esul s, he use o a 417 3Å molecula sie e o wa e emo al appea s a p omising app oach. 418 The me hanol p oduc i i y ob ained in he cu en esea ch can be compa ed o esul s epo ed in o he s udies. 419 Yang e al. [49] ound an e en highe me hanol p oduc i i y o up o 167 g/kg/h o alcohol-p omo ed me hanol 420 syn hesis a 170 °C and 50 ba using an op imized Cu/ZnO ca alys composi ion. The di e ence o he esul s 421 p esen ed he e can be explained mainly by he di e en eed gas composi ion in hei expe imen s (CO/CO2/H2/A 422 = 32.4/5.1/59.5/3.9). Fo gas-phase CO2 hyd ogena ion o me hanol, p oduc i i y alues e en up o 1200 g/kg/h 423 ha e been achie ed [58]. Howe e , hese esul s we e ob ained a a ela i ely high empe a u e o 240 °C and a 424 high space eloci ies gi ing ela i ely low CO2 con e sions. 425 Dual ca alysis by he combina ion o Cu/ZnO wi h coppe ch omi e was also s udied in his wo k. A ema kable 426 inc ease in ca aly ic ac i i y was ound o he dual ca alys . When 20 g o coppe ch omi e and 10 g o Cu/ZnO was 427 used, he p oduc i i y inc eased by 80% compa ed o he use o 20 g o he Cu/ZnO ca alys alone. A syne gis ic 428 e ec be ween he wo ca alys s is sugges ed, which is possibly based on an inc eased o ma ion a e o he o ma e 429 es e in e media e by he coppe ch omi e ca alys . The wo ca alys s appea ed o ha e simila e e se wa e -gas 430 shi ac i i y, as he concen a ion o wa e did no change when he ela i e amoun s o Cu/ZnO and coppe ch omi e 431 we e a ied. 432 S uc u al changes in he ca alys du ing alcohol-p omo ed me hanol syn hesis we e ound by he means o XRD 433 and SEM-EDS in es iga ions. EDS elemen al analysis showed ha seg ega ion o coppe and zinc oxide had aken 434 place, and bo h XRD analysis and SEM imaging p o ided e idence ha c ys alliza ion o zinc oxide occu ed. Such 435 phenomena has p e iously been iden i ied as cause o ca alys deac i a ion due o he loss o eac i e Cu-ZnO 436 in e aces [56]. Howe e , comp ehensi e ca alys s abili y es s we e no pe o med in he cu en s udy, and hus 437 23 he e ec o he obse ed changes on ca aly ic ac i i y canno be de e mined conclusi ely. I is clea ha s abili y 438 es s a di e en eac ion empe a u es and, impo an ly, a di e en eed gas composi ions a e necessa y o u he 439 cha ac e ize he alcohol-p omo ed me hanol syn hesis p ocess. 440 Acknowledgemen s 441 The Au ho s a e g a e ul o Finnish Academy o Science o “Mic o- and millis uc u ed eac o s o ca aly ic 442 oxida ion eac ions” MICATOX p ojec unding, numbe : 269896. Funding p o ided by he Lappeen an a 443 Uni e si y o Technology Doc o al School is also g a e ully acknowledged. 444 24 Re e ences 445 [1] R. Schlögl, "The sola e ine y," in Chemical Ene gy S o age, Be lin/Bos on, Wal e de G uy e GmbH, 2013, pp. 1-34. [2] F. Schü h, "Ene gy s o age s a egies," in Chemical Ene y S o age, Be lin/Bos on, Wal e de G uy e GmbH, 2013, pp. 35-48. [3] G. Olah, "Beyond oil and gas: The me hanol economy," Angew. Chem. In . Ed., ol. 44, pp. 2636-2639, 2005. doi:10.1002/anie.200462121 [4] H. O e manns, L. Plass and M. Be au, "F om aw ma e ials o me hanol, chemicals, and uels" in Me hanol: The Basic Chemical and Ene gy Feeds ock o he Fu u e, Be lin Heidelbe g, Sp inge -Ve lag, 2014, pp. 1-7. doi:10.1007/978-3-642-39709-7 [5] L. Reichel and F. Schmid , "Me hanol- o-gasoline p ocess," in Me hanol: The Basic Chemical and Ene gy Feeds ock o he Fu u e, Be lin Heidelbe g, Sp inge -Ve lag, 2015, pp. 440-453. doi:10.1007/978-3-642-39709-7 [6] F. Schmid and C. Pä zold, "Me hanol- o-ole ins p ocesses," in Me hanol: The Basic Chemical and Ene gy Feeds ock o he Fu u e, Be lin Heidelbe g, Sp inge -Ve lag, 2015, pp. 454-472. doi:10.1007/978-3-642-39709-7 [7] J. O , V. G onemann, F. Pon zen, E. Fiedle , G. G ossmann, Ke sebohm, D.B., G. Weiss and C. Wi e, "Me hanol," in Ullmann's Encyclopedia o Indus ial Chemis y, Weinheim, Wiley-VCH Ve lag GmbH & Co. KGaA, 2012, pp. 1-27. doi:10.1002/14356007 [8] S. Lee, "Me hanol syn hesis om syngas," in Handbook o Al e na i e Fuel Technology, Taylo & F ancis G oup, LLC, 2007, pp. 297-321. [9] J. Sk zypek, M. Lachowska, M. G zesik, J. Sloczynski and P. Nowak, "The modynamics and kine ics o low p essu e me hanol syn hesis," Chem. Eng. J., ol. 58, pp. 101-108, 1995. doi:10.1016/0923-0467(94)02955-5 [10] E. Kunkes and M. Beh ens, "In oduc ion o me hanol syn hesis and s eam e o ming," in Chemical Ene gy S o age, Be lin, Wal e de G uy e GmbH, 2013, pp. 415-417. [11] T. Holde baum and J. Gmehling, "PSRK: A g oup con ibu ion equa ion o s a e based on UNIFAC," Fluid Phase Equilib., ol. 70, pp. 251-265, 1991. doi:10.1016/0378-3812(91)85038-V [12] V. Paleka , "Alkali compounds and coppe ch omi e as low- empe a u e slu y phase me hanol ca alys s," Appl. Ca al., A, ol. 103, no. 1, pp. 105-122, 1993. doi:10.1016/0926- 860X(93)85177-Q [13] Y. Zhao, L. Bai, Y. Hu, B. Zhong and S. Peng, "Ca aly ic pe o mance o CuC /CH3ONa used o low empe a u e me hanol syn hesis in slu y phase," J. Na . Gas Chem., ol. 8, no. 3, pp. 181-187, 1999. [14] W. Chu, T. Zhang, C. He and Y. Wu, "Low- empe a u e me hanol syn hesis (LTMS) in liquid phase on no el coppe -based ca alys s," Ca al. Le ., ol. 79, pp. 129-132, 2002. doi:10.1023/A:1015384015528 [15] M. Ma chionna, L. Basini, A. L. M. A agno and F. Ancillo i, "Mechanis ic s udies on he homogeneous nickel-ca alyzed low empe a u e me hanol syn hesis," J. Mol. Ca al., ol. 75, no. 2, pp. 147-151, 1992. doi:10.1016/0304-5102(92)80116-X [16] E. Lee and K. Aika, "Low- empe a u e me hanol syn hesis in liquid-phase wi h a Raney Nickel– alkoxide sys em: E ec o Raney Nickel p e ea men and eac ion condi ions," J. Mol. Ca al. A: Chem., ol. 141, pp. 241-248, 1999. doi:10.1016/S1381-1169(98)00267-2 [17] S. Ohyama, "Low- empe a u e me hanol syn hesis in ca aly ic sys ems composed o nickel compounds and alkali alkoxides in liquid phases," Appl. Ca al. A, ol. 180, pp. 217-225, 1999. doi:10.1016/S0926-860X(98)00338-X