Novel windows for “solar commodities”: a device for CO2 reduction using plasmonic catalyst activation

No el windows o “sola commodi ies”:a de ice o CO 2 educ ion using plasmonic ca alys ac i a ion Alexande Na a e e,* a Se gio Mu˜ noz, a Luis M. Sanz-Mo al, a Jue gen J. B andne , b Pe e P ei e , b ´ Angel Ma ´ ın, a Roland Di meye b and Ma ´ ıa J. Coce o a Recei ed 12 h June 2015, Accep ed 2nd July 2015 DOI: 10.1039/c5 d00109a A no el plasmonic eac o concep is p oposed and es ed o wo k as a isible ene gy ha es ing de ice while allowing eac ions o ans o m CO 2 o be ca ied ou . Pa icula ly he e e se wa e gas shi (RWGS) eac ion has been es ed as a means o in oduce enewable ene gy in o he economy. The de elopmen o he new eac o concep in ol ed he syn hesis o a new composi e capable o plasmonic ac i a ion wi h ligh , he de elopmen o an imp egna ion me hod o c ea e a single ca alys eac o en i y, and finally he assembly o a eac ion sys em o es he eac ion. The composi e de eloped was based on a Cu/ZnO ca alys dispe sed in o anspa en ae ogels. This allows efficien ligh ansmission and a high su ace a ea o he ca alys . An effec i e ye simple imp egna ion me hod was de eloped ha allowed in oduc ion o he composi es in o glass mic ochannels. The ac i a ion o he eac ion was made using LEDs ha co e ed all he sides o he eac o allowing a high powe deli e y. The esul s o he eac ion show a s able p ocess capable o low empe a u e ans o ma ions. In oduc ion CO 2 as a enewable ene gy ec o One o he mos indus ially p omising he e ogeneous ca aly ic p ocesses is ca bon dioxide hyd ogena ion. In he p ocess o ca aly ic hyd ogena ion, hyd ogen ob ained om ca bon neu al ene gy sou ces (e.g. wind o sola ) is eac ed wi h CO 2 o ob ain p oduc s such as uels. 1 Thus, his p ocess se es a double pu pose:  s ly, as chemical s o age o he su plus ene gy gene a ed by uc ua ing enewable ene gies; and secondly, o educe emissions o CO 2 . a Uni e si y o Valladolid, Depa men o Chemical Enginee ing and En i onmen al, Technology, High P essu e P ocesses G oup, Paseo P ado de la Magdalena s/n, 47005 Valladolid, Spain. E-mail: [email protected] a.es b Ins i u e o Mic o P ocess Enginee ing, Ka ls uhe Ins i u e o Technology, Eggens ein-Leopoldsha en, Ge many This jou nal is © The Royal Socie y o Chemis y 2015 Fa aday Discuss.,2015,183,249–259 | 249 Fa aday Discussions Ci e his: Fa aday Discuss.,2015,183,249 PAPER The success ul in oduc ion o CO 2 -use echnologies equi es plausible and p o able p ocesses ha use efficien ly enewable ene gy. The mos abundan and e enly dis ibu ed o such ene gies is he one p o ided by he Sun. I s o ed in he o m o “sola commodi ies”such as me hanol o olens new oppo uni ies o he use o CO 2 could be conside ed. 2 The p oduc ion o me hanol om CO 2 and hyd ogen in ol es he ollowing eac ions: CO 2 +H 2 5CO + H 2 O; DH 0 ¼+41.19 kJ mol 1 (1) CO+2H 2 5CH 3 OH; DH 0 ¼90.70 kJ mol 1 (2) CO 2 +3H 2 5CH 3 OH + H 2 O; DH 0 ¼49.51 kJ mol 1 (3) In o de o cap u e ha ene gy in chemical bonds he e a e se e al use ul me hods. F om a chemical poin o iew, hey can be g ouped in o pho oca aly ic, he mal and elec ical ou es. In elec ical ou es, he sola ene gy is  s con- e ed in o elec ici y and hen, he esul an elec ical ene gy is used in he chemical ans o ma ion o CO 2 . 3–7 The he mal ou es concen a e sola adia- ion and con ey ha ene gy di ec ly in o he eac o . 8–10 Pho oca aly ic CO 2 con e sion in ol es ei he wa e spli ing connec ed o a CO 2 educ ion eac ion, o a p ocess combining bo h in one “single po ”. 11–13 Selec i e use o isual ligh wi h plasmon ca alys s The su ace plasmon esonance (SPR) phenomenon is commonly ound in me allic (o ca bon) nanos uc u es and allows he ange o he sola spec um used on a gi en pho oinduced p ocess o be inc eased. 14 This effec is he esul o he esponse o he conduc ion elec ons o he oscilla ions o he elec ic eld o he ligh adia ion. Inc eased ene gy abso p ion by he elec ons is possible a selec ed wa eleng hs unde he p ope pa icle size and shape o he nano- pa icles o a gi en su ounding media (uid o ca alys ). This phenomenon p oduces a high ligh concen a ion up o he poin ha a educ ion in he amoun o semiconduc o o h ee o de s o magni ude o he same amoun o ligh has been possible. 15 On he o he hand, inc eased ligh cap u e wi h SPR o pho o he mal con e sions is leading o b eak h oughs in ene gy sys ems such as sola collec o s. 16 This wo k explo es he educ ion o CO 2 o CO as a  s s ep in a sola -based p ocess o p oduce me hanol. Thus, i is based on he e e se wa e gas shi (RWGS) eac ion as desc ibed by eqn (1). Recen app oaches ha e used gold and semiconduc o composi es o plasmonic enhancemen o he educ ion. 17,18 This p ocess is commonly ac i a ed in indus y using Cu/ZnO based ca a- lys s. 19 He e, we ha e de eloped plasmonic ca aly ic composi es in mesopo ous silica s uc u es (ae ogels). Fo his, we ha e used he plasmon- uneable Cu/ZnO ca alys epo ed by Tan e al. (2013). 20 A plasmonic mic o eac o as a ligh ha es ing de ice The efficiency o he chemical eac ions is no uled only by he ca aly ic ma e ial bu also by he eac o congu a ion and hei mass and ene gy anspo cha - ac e is ics. On many occasions, p omising ca aly ic ma e ials ha e ailed o each Fa aday Discussions Pape 250 |Fa aday Discuss.,2015,183,249–259 This jou nal is © The Royal Socie y o Chemis y 2015 indus ial success due o he disconnec ion be ween he ca aly ic s uc u e and he eac o -le el phenomena. 21,22 He e we p opose a no el concep o isual ene gy ha es ing: a plasmonic mic o eac o de ice. I in eg a es a plasmon ca alys and eac o as one en i y wi h a sole esponse o ligh (Fig. 1). Mic o eac o s allow efficien ene gy and mass anspo while a e easily scal- able (numbe ing up). Thus, he combina ion o efficien mic os uc u ed de ices and di ec plasmonic abso p ion o sola ene gy by he ca alys would ep esen a majo b eak h ough in he CO 2 -use eld. In his wo k, a plasmon- uneable composi e is in eg a ed wi h a mic ochannel based eac ion sys em unde isual LED illumina ion o he RWGS eac ion. This in ol ed he syn hesis o he new composi e; he de elopmen o an imp egna ion me hod o c ea e a ca alys - eac o en i y; and nally, he assembly o a eac ion sys em o es he eac ion. Me hods Plasmo-ca aly ic composi e syn hesis The chemicals used du ing his s age a e de ailed: zinc ace a e dihyd a e (>98%), oleylamine (70%), e ame hyl o hosilica e (98%), ammonia (28–30%) and ie- hyleneglycol (99%) we e pu chased om Sigma-Ald ich. E hyleneglycol (99.5%) (Me ck) and coppe ace a e monohyd a e (99.9%) we e pu chased om Al a Aesa . Me hanol (99.8%) was p o ided by Pan eac. All chemicals we e used wi hou u he pu ica ion. Syn hesis o he Cu/ZnO bime allic ca alys This bime allic (Cu : ZnO, 1 : 2) ca alys was syn hesised ollowing he p ocedu e p oposed by Tan e al. (2013). 20 B iey,  s ZnO nano ods we e p epa ed. Zinc ace a e (3 mmol) was added o 1.3 mmol o oleylamine in a wo necked ask. The oleylamine was no o high pu i y (70%), equi ing a s ep whe e he eac an s we e degassed a 80 C o 45 Fig. 1 Concep o he plasmonic mic o eac o de ice. Pape Fa aday Discussions This jou nal is © The Royal Socie y o Chemis y 2015 Fa aday Discuss.,2015,183,249–259 | 251 min unde a acuum a mosphe e. Then, he empe a u e was inc eased o 220 C unde ni ogen pu ging. Du ing he hea ing p ocess, he solu ion u ned whi e upon eaching 180 C. Ae 15 min o hea ing, he mix u e was cooled o a mo- sphe ic empe a u e, washed wi h e hanol and cen i uged in o de o isola e he p ecipi a e. I was washed wi h 6 ml o e hanol h ee imes o ensu e comple e emo al o he eac an s o byp oduc s. The ZnO nano ods hus p epa ed we e hen edispe sed in 20 ml o ie hy- leneglycol by sonica ion o wo hou s, ollowed by s i ing unde oom condi- ions o e nigh . E hyleneglycol (2 ml) was added o he ZnO dispe sion, and he mix u e was degassed a oom empe a u e o 5 min be o e hea ing o 190 C. Simul aneously, a second solu ion o coppe ace a e monohyd a e was p epa ed by dissol ing in e hyleneglycol. This mix u e equi ed sonica ion in o de o dissol e he coppe ace a e in he liquid. This mix u e was added o he ZnO mix u e in a d opwise manne o e 10 min. Ae his, 5 mo e min hea ing a 190 C was allowed be o e he composi e was washed wi h isop opanol, cen i uged o 15 minu es a 4500 pm (cen i uge Kubo a 5100, Japan) and isola ed om he mix u e. Syn hesis o he mesopo ous silica composi es Ligh ansmission o he ca aly ic s uc u es is essen ial while enough su ace a ea has o be p o ided in o de o ha e enough me al loading o cap u e ligh . T anspa en ae ogels a e mesopo ous ma e ials combining high su ace a eas and good ligh ansmission. 23 Ae ogels we e syn hesised ollowing he sol–gel ou e. The p ecu so o he silica hyd ogel selec ed was e ame hyl o hosilica e (TMOS). The mola a io o TMOS : CH 3 OH : H 2 O:NH 4 was 1 : 2.3 : 3.84 : 0.012. Fi s ly, me hanol was used o dispe se he nanopa icles o med in he ca alys syn hesis. Sonica ion (15 min) was applied o ensu e a good dispe sion o he nanopa icles in he liquid phase. Me hanol wi h he pa icles and TMOS we e mixed oge he . While his solu ion was s i ed, a second solu ion o ammonium hyd oxide and wa e was p epa ed and s i ed. Ae a ew minu es o s i ing, bo h solu ions we e mixed, and he gela ion p ocess began. In his momen , he gela ion p ocess o he silica hyd ogel has s a ed, bu i is s ill liquid o a ew minu es. This ime lapse, be o e gela ion, mus be used o imp egna e he solu ion inside he mic ochannels o he mic o eac o s. In eg a ion o he composi es and mic o eac o In o de o ha e a single in eg a ed de ice i is necessa y o in eg a e ligh ansmission and composi e ac i a ion in he same s uc u e. He e, we ha e de eloped a me hod o in eg a e anspa en ae ogels in glass mic ochannels. The me hod used o imp egna e he sol–gel ha showed he bes esul s was he suc ion o he liquid wi h a sy inge, which was p e iously adap ed o he mic o- eac o on i s op (Fig. 2). Wi h his me hod, placing he mic o eac o in a e ical posi ion, i was e y easy o ll he mic ochannel placing he ip in he liquid. Ae a sho ime, he gela ion p ocess nished and he hyd ogel o med had a good adhe ence inside he mic ochannels. I is wo h men ioning ha one o he mos impo an pa ame e s du ing his s ep is he amoun o ammonia, as i ac s as a ca alys o he gela ion p ocess. I is Fa aday Discussions Pape 252 |Fa aday Discuss.,2015,183,249–259 This jou nal is © The Royal Socie y o Chemis y 2015 necessa y o use an amoun o ammonia low enough ha allows enough ime o imp egna e he hyd ogel inside he mic o eac o , because i mus be s ill a uid. On he o he hand, he amoun o ammonia canno be e y low, because i he gela ion p ocess is oo slow, he nanopa icles s a o p ecipi a e and hey will no emain inside he silica ne . Finding an equilib ium be ween hese wo ac o s is he key o achie ing good imp egna ion o he nanopa icles suppo ed in he silica gels inside he mic o eac o s. The mic o eac o consis ed o a 0.5 mm ID glass capilla y wi h an ex e nal diame e o 5 mm (Scho Du an, USA). In o de o ensu e good adhe ence o he ae ogel o he walls o he mic o- channels, i is necessa y o pe o m a p e ea men o he glass mic o eac o s (15.0 cm) o clean he walls o he mic ochannels. Fo his cleaning p ocess, he mos common op ion is o use a pi anha solu ion, which eac s iolen ly wi h mos o ganic ma e ials. The solu ion used was a mix u e o sul u ic acid and hyd ogen pe oxide ha can be p epa ed in diffe en p opo ions; he mos usual being 4 : 1 in concen a ed sul u ic acid. Fo 15 o 30 minu es he ma e ial is subme ged in he solu ion, hen emo ed, washed wi h plen y o Milli-Q wa e and d ied ca e ully. The emp y glass mic o eac o s we e pu inside a glass po , and he sul u ic acid was  s added. Then, he hyd ogen pe oxide was added wi h ex eme ca e because he eac ion is e y exo he mic; he empe a u e is suddenly inc eased and some apo can be o med. Ae 20 minu es, he slides we e emo ed om he pi anha solu ion, washed wi h Milli-Q wa e and d ied ca e ully. Ae he in oduc ion o he nanopa icles in he sol–gel and i s in oduc ion in o he mic o eac o s, hese we e pu in a essel wi h me hanol o aging. This essel was ca e ully closed o a oid me hanol e apo a ion, and i was hea ed o 50 C in an o en. Wi h his p ocedu e he wa e con ained in he silica ne was eplaced wi h me hanol, esul ing in alcogels. Ae 24 hou s o hea ing, he alcogels we e d ied using supe c i ical ca bon dioxide. The mic o eac o s we e pu in a high p essu e essel, and his essel was lled comple ely wi h me hanol. Ca bon dioxide was in oduced slowly in he essel o allow a good diffusion in o he me hanol. The p essu e was aised o 100 ba and he empe a u e o 40 C, abo e he c i ical poin o ca bon dioxide. Th ee cycles o 45 minu es we e pe o med, enewing he ca bon dioxide be ween each Fig. 2 Sy inge filling o he mic ochannels. Pape Fa aday Discussions This jou nal is © The Royal Socie y o Chemis y 2015 Fa aday Discuss.,2015,183,249–259 | 253 cycle o comple e he d ying p ocess. 24 Ae his, silica ae ogels we e co ec ly ob ained, keeping he adhe ence o he walls o he mic ochannels. Composi e cha ac e iza ion Scans o he bime allic ca alys s o check abso bance o isual ligh we e ca ied ou using a UV- is spec ome e (UV 2550, Shimadzu). XRD analysis was ca ied ou using a B uke Disco e D8 diff ac ome e . The po osi y measu emen s we e ca ied ou using a Su ace A ea and Po osi y Analyze (ASAP 2020, Mic ome i ics). P oo -o -concep se up Once he Cu/ZnO based plasmonic composi es we e in eg a ed in o he glass mic ochannels, a es o hese de ices was made. In o de o es he eac o concep , a eac ion sys em had been buil ha included isual LED illumina- ion and con ol o he eac ion empe a u e while a p ecise con ol o he ow and p essu e was p o ided. A scheme o he expe imen al plan is p esen ed in Fig. 3. Hyd ogen and ca bon dioxide we e in oduced in he sys em, and hei ows we e con olled wi h wo diffe en ow mass me e /con olle s (EL-Flow F-200, B onkho s ) wi h anges om 0.02 o 1 ml min 1 . Be o e he eac ion was ini ia ed, hyd ogen and ca bon dioxide we e mixed in a 3 : 1 p opo ion, and sen o he en while bo h ows we e s abilized. When he ows we e co ec ly con olled, he mix u e o he gases mo ed o he second pa o he se up. In his pa he gases we e hea ed, oge he wi h he glass mic o eac o , in a gas ch oma og aphy o en (Agilen 7890). A second en was used o ake ou he gases while he p essu e was inc eased o 20 ba . P essu e was con olled by a p essu e me e /con olle (EL-P ess se ies, B onkho s ). When he p essu e and ow we e s able a 20 ba , he al e o ex ac ion was closed, and he o en and LEDs we e u ned on. Fig. 3 Schema ic flow diag am o he plan . Fa aday Discussions Pape 254 |Fa aday Discuss.,2015,183,249–259 This jou nal is © The Royal Socie y o Chemis y 2015 Visual ligh s imula ion was p o ided by 36 LEDs (supe b igh , inspi ed LED) su ounding he mic o eac o as shown in Fig. 4. In o al, hey p o ided he equi alen o a nominal powe o 9780 W m 2 o whi e ligh . The eac ion began and he p oduc s o he eac ion we e measu ed in a Mic o Gas Ch oma og aph (CP-4900, Va ian) equipped wi h wo columns: a po aplo 10 m and a 5A molsie e. Be o e he mic o GC, he p essu e o he gas s eam was educed o less han 5 ba . Resul s and discussion Composi e cha ac e iza ion Ae he syn hesis UV- is scans we e ca ied ou in o de o check he abso p ion o isual ligh om bo h he ZnO nano ods and he bime allic Cu/ZnO ca alys (Fig. 5). I can be seen ha he bime allic ca alys has an abso p ion peak a 498 nm. This co esponds o abso p ion in he ange close o he g een colou . 25 The anspa en ae ogels change and acqui e colou once he composi e is o med (Fig. 6). Fig. 4 LED ligh configu a ion. Le , de ail o he inne LED dis ibu ion. Righ , mic o- eac o in he o en su ounded by he LEDs. Fig. 5 Abso p ion spec a o ZnO (le ) and Cu/ZnO ( igh ). Pape Fa aday Discussions This jou nal is © The Royal Socie y o Chemis y 2015 Fa aday Discuss.,2015,183,249–259 | 255 The XRD pa e n shows he p esence o ZnO and me allic coppe in he silica amo phous s uc u e. The ZnO planes 100 (2q¼31.7), 002 (2q¼34.4), 101 (2q¼ 36.1) and 110 (2q¼56.4) can be iden ied in Fig. 7. Coppe canno be seen due o i s lowe p opo ion in he s uc u e. BET su ace a ea and po e olume o he samples we e calcula ed om N 2 iso he ms. The adso p ion–deso p ion cu e shows a ype IV iso he m cu e ypical o mesopo ous silica ae ogels 26 (Fig. 8). The BET su ace a ea is 945.8 m 2 g 1 which indica es ha he inclusion o he bime allic ca alys does no ha e a signican inuence on he ex u al p ope ies o he ae ogel. The BJH po e olume is equal o 2.29 cm 3 g 1 ein o cing ha he s uc u e is no affec ed. Reac ion es o he concep The ull powe o he LEDs was applied and he e olu ion o he compounds we e ollowed. In o de o es he inuence o he main a iables o he p ocess, changes in ow and empe a u e we e made du ing he eac ion (Fig. 9). Fig. 6 Silica ae ogel be o e (le ) and a e ( igh ) ca alys imp egna ion. Fig. 7 XRD pa e n o he plasmonic composi e. Fa aday Discussions Pape 256 |Fa aday Discuss.,2015,183,249–259 This jou nal is © The Royal Socie y o Chemis y 2015 I could be obse ed ha he eac ion was s able a 50 C du ing mo e han 100 minu es. Then he empe a u e was inc eased a 70 C and es ed du ing he same ime span. Finally, he ow was educed o hal o he ini ial condi ions. No signican changes we e obse ed du ing almos 250 minu es a cons an ow, which indica es he sui abili y o his sys em o es se e al ca alys loads unde diffe en he modynamic and ow condi ions. I is impo an o no e ha he a e age con e sion a e a highe ow is simila o he one ob ained in o he wo ks a empe a u es a ound 200 C. 17 Thus, he in eg a ed plasmonic eac o concep p oposed he e opens new a enues o couple low empe a u e sola collec o s and chemis y as a mean o in oduce enewable ene gy in he economy, pa icula ly o he con e sion o CO 2 . Fig. 8 Adso p ion–deso p ion iso he ms o he composi es. Fig. 9 E olu ion o he eac ion. Pape Fa aday Discussions This jou nal is © The Royal Socie y o Chemis y 2015 Fa aday Discuss.,2015,183,249–259 | 257