Understanding biomass fractionation in subcritical & supercritical water

1 Unde s anding biomass ac iona ion in subc i ical & supe c i ical 1 wa e 2 Ma ía José Coce o*, Ál a o Cabeza, Ne ea Abad, Tijana Adamo ic, Luis Vaque izo, Celia M. 3 Ma ínez, Ma ía Vic o ia Pazo-Cepeda. 4 High P essu e P ocesses G oup, Depa men o Chemical Enginee ing and En i onmen al 5 Technology, Uni e si y o Valladolid (Spain). Doc o Me gelina s/n. 47011, Valladolid, Spain. 6 * Co esponding au ho . Tel: +34 983423174; ax: +34 983423013. 7 E-mail add esses: mjcoc[email p o ec ed] 8 Abs ac 9 Biomass ac iona ion in o i s indi idual building blocks poses a majo challenge 10 o he bio e ine y concep . The ecalci ance o he lignocellulose ma ix and he 11 high c ys allini y o cellulose make ypical eed s ocks di icul o sepa a e in o 12 hei componen s. Hyd o he mal p ocessing ac iona es biomass by i s 13 hyd olysis. Howe e , a deep knowledge o hyd olysis p inciples is equi ed since 14 an inapp op ia e selec ion o he ope a ing pa ame e s such as an excessi e 15 empe a u e and a long esidence imes causes d ama ic selec i i y losses. This 16 e iew is di ided in ou main sec ions which p esen he undamen als o 17 lignocellulosic biomass ac iona ion in hemicelluloses, cellulose and lignin. As 18 he biomass s uc u e plays an impo an ole, a sec ion o s udy he ex ac ion o 19 he linked phenols ha join lignin and hemicelluloses is included. 20 1. In oduc ion 21 Shi ing he chemical indus y away om pe ochemical eeds ocks owa ds 22 enewable, bio-based chemicals and ma e ials is a long- e m s a egy o he 23 Eu opean Union. This “bio e ine y” concep , despi e being p oposed as ea ly as 24 he la e 1980s, has s ill no come o ui ion because he cos and complexi y o 25 p ocessing biomass o gene a e p ac ical, usable, saleable eeds ocks makes i 26 un easible. 27 Lignocellulose is he mos abundan , cheapes and easies g own o m o 28 biomass, and i is composed o h ee main ac ions: cellulose (40-50%), 29 2 hemicellulose (25-35%) and lignin (10-30%), in addi ion o mino compounds. 30 These ac ions ep esen po en ial eeds ocks o bio-sou ced commodi y 31 chemicals, bu due do hei di e ing chemical unc ionali ies (lignin made up o 32 linked a oma ic uni s, hemicellulose o C5 suga s and cellulose o C6 suga s) 33 sepa a ion s eps a e necessa y o isola e he app op ia e ac ion and b eak i 34 in o i s indi idual building blocks (e.g. suga s o cellulose/hemicellulose and 35 a oma ic uni s o lignin). 36 37 38 39 Figu e 1: Lignocellulosic biomass s uc u e 40 This ac iona ion o biomass in o i s indi idual building blocks poses a majo 41 challenge o he bio e ine y concep , because he ecalci ance o he 42 lignocellulose ma ix and high c ys allini y o cellulose makes ypical eeds ocks 43 di icul o sepa a e in o hei componen s. Fo his eason, i ypically equi es 44 long eac ion imes ( om 30 minu es o he hyd o he mal hyd olysis o 24-70 45 hou s o enzyma ic hyd olysis) and he p esence o s ong eagen s (sodium 46 hyd oxide and sodium sul ide du ing K a pulping, o example). This leads o 47 deg ada ion o he non-cellulosic ac ions as well as la ge olumes o e luen 48 which equi es expensi e ea men o educe en i onmen al load. 49 To uly ha ness he po en ial o he bio e ine y concep , his ac iona ion s ep 50 needs o be e olu ionized. I needs o be conside ably mo e p ocess in ensi e 51 (ideally seconds pe uni olume o biomass -as opposed o minu es o hou s) o 52 enable modula uni s o deal wi h la ge olumes o biomass a decen alized 53 3 loca ions. I mus no in ol e he use o ha sh eagen s in o de o minimize 54 en i onmen al impac and cos , whils main aining quali y o he ac ions. 55 Wa e abo e i s c i ical poin (Tc 374ºC, 22 MPa), is an al e na i e sol en o 56 dissolu ion/hyd olysis o biomass. I s low iscosi y and high di usi i y acili a e 57 he pene a ion o wa e in o he complex s uc u e o he lignocellulosic ma ix, 58 whils i s low dielec ic cons an , simila o non-pola o ganic sol en s, enhances 59 solubili y o o ganic compounds. Physical p ope ies o wa e (such as densi y, 60 ionic p oduc , dielec ic cons an ) can be inely uned by a ying empe a u e and 61 p essu e. A hese condi ions, he hyd olysis o biomass ac ions is apid and 62 p esen s a mean o achie e signi ican ly mo e p ocess in ensi e ac iona ion o 63 biomass. 64 Reac ion speed – whils being an ad an age o p ocess in ensi ica ion – is also a 65 signi ican disad an age o selec i i y a longe eac ion imes, leading o 66 deg ada ion o hyd olysis p oduc s and esul ing in complex eac ion mix u es. 67 This deg ada ion and mix u e complexi y leads o ine icien eco e y o biomass 68 de i ed p oduc s and in e media es. The e is he e o e a need o unde s anding 69 he hyd o he mal ac iona ion p ocesses o imp o e p ocesses selec i i y, which 70 can ha ness he po en ial o subc i ical and supe c i ical wa e ac iona ion. 71 E en unde wa e ’s c i ical poin , ce ain ac ions o biomass ace eac ions ha 72 p oceed oo apidly o be con olled by con en ional me hods. Fo ins ance, lignin 73 unde goes apid hyd olysis and subsequen hyd olysis p oduc con e sion in less 74 han 1 second a 350 ºC. Whils poo selec i i y is common o bo h sub- and 75 supe c i ical wa e (SCW), he e a e some signi ican di e ences be ween he 76 eac ion media – mos no ably he di e ence in ionic p oduc o wa e ( o 77 example he H+/OH- concen a ion a 300ºC and 22 MPa is a ound 3·10-6 mol· L- 78 1 s 3·10-10 a 400 and 22 MPa) which means ha subc i ical wa e has a highe 79 concen a ion o ions ([H+] and [OH-]) hus a o ing ionic eac ions s he adical 80 eac ions ha a e p e alen unde SCW condi ions. 81 82 4 83 Figu e 2. Subc i ical and supe c i ical wa e p ope ies a ound he c i ical poin 84 (22 MPa). 85 This manusc ip s udies he lignocellulosic biomass undamen als ac iona ion in 86 subc i ical and supe c i ical wa e , in o de o imp o e he selec i i y o he 87 hyd o he mal biomass ac iona ion. The manusc ip p esen s ou main sec ions 88 o p esen s he ac iona ion o biomass in hemicellulose, suga s and lignin. As 89 he biomass s uc u e plays an impo an ole, a sec ion o s udy he linked 90 phenols ha join lignin and hemicelluloses is included. 91 2. Hemicellulose(s) ac iona ion undamen als 92 Hemicellulose is a biopolyme p esen in lignocellulosic ma e ials ha ac s as a 93 connec ion be ween he ibbe s (cellulose) and he 3-dimensional s uc u e 94 (lignin), cons i u ing be ween 25% and 35% o he whole biomass [1]. I is 95 cha ac e ized by hei amo phous s uc u e and by he ac ha i is ace yla ed 96 [2]. Rega ding i s composi ion, i is a biopolyme mainly composed o pen oses 97 wi h ew hexoses in be ween, wi h a maximum leng h a ound 200 o 300 98 monome ic suga s, which makes i a enewable sou ce o chemicals based on 5- 99 ca bon molecules. The maximum molecula weigh is lowe han 70 kDa in mos 100 cases [3]. Howe e , he e a e disc epancies be ween species. Fo ins ance, 101 xylose is he mos common monome in hemicelluloses o ha dwood ees, while 102 so wood ees a e p incipally composed o mannans, like mannose [2]. In 103 addi ion, he e a e wo di e en ypes o hemicellulose om he ex ac ion 104 Tempe a u e, ºC 0 100 200 300 400 500 600 700 Densi y, kG m-3 0 200 400 600 800 1000 1200 Ionic p oduc o wa e , pKw 10 12 14 16 18 20 22 24 Dielec ic cons an , e 0 20 40 60 80 100 Densi y Kw e 5 iewpoin : one hemicellulose easy o ex ac and ano he one ha is associa ed 105 wi h he ibbe s o cellulose ha can be eco e ed only when cellulose is also 106 emo ed ( empe a u es abo e 240 °C) [4–6]. Since hemicelluloses ha e some 107 po en ially acidic g oups (ace yl g oups among o he s), i can be eco e ed by 108 K a pulping. Howe e , he use o his echnique leads o a deg ada ion o 109 hemicellulose, so a di e en echnique is equi ed o ob ain i wi h a high quali y. 110 Thus, hyd o he mal ex ac ion would be one o he mos p omising op ions since 111 i only equi es wa e and mild empe a u es (160-210 °C) o ex ac i [7,8]. I he 112 ope a ional empe a u e is a ound 180 °C, 60% o he ini ial hemicellulose can be 113 eco e ed as oligome s and suga s [4,9]. Highe yields can be ob ained i 114 empe a u e inc eases bu undesi ed deg ada ion p oduc s appea [10,11]. 115 Howe e , hemicellulose can be eco e ed also a low empe a u es (90 ºC) i he 116 ope a ing ime is high enough (days) [12]. Hemicellulose ex ac ion has been 117 pe o med in bo h sys ems, ba ch and packed bed eac o s. The e o e, i should 118 be also ma ked ha wo di e en ope a ing imes can be de ined, he solid and 119 liquid ime. The o me is he ime used o ea he solid. The liquid ime has he 120 same alue as he solid ime in ba ch sys ems. Howe e , i is ixed by he 121 olume ic low when semi-ba ch o con inuous sys em a e used, being he 122 esidence ime (see appendix 1 o mo e de ails abou he di e en esidence 123 imes). Mo eo e , hemicellulose hyd o he mal ac iona ion is a complex p ocess 124 ha in ol es se e al physical phenomena [6,8,13,14], which a e p esen as in 125 ba ch as in con inuous sys ems, and a good knowledge o hem is manda o y o 126 designing a p o i able and sus ainable hemicellulose ex ac ion plan . These 127 phenomena a e: 128  Hemicellulose clea ing in o dec easing molecula weigh oligome s 129  Hemicellulose deace yla ion (au ohyd olysis) 130  Hemicelluloses dissolu ion and mass ans e be ween he solid and he 131 liquid 132  P oduc ion o suga s & suga s deg ada ion in o u u al o o he 133 subs ances 134  Po osi y changes: ex ac ion, swelling and biomass compac ion 135 6 Addi ionally, and once he phenomenology is explained, a sho summa y abou 136 he e ec o he main ope a ional a iables on hemicellulose selec i i y is 137 included. 138 139 2.1. Hemicellulose clea ing in o dec easing molecula weigh 140 oligome s 141 The ollowing discussion is ocused on he beha io obse ed in a semi- 142 con inuous sys em since only globalized alues can be ob ained om a ba ch 143 eac o . Hemicellulose clea ing is one o he i s phenomena ha akes place 144 inside a chip o a pa icle o biomass. Due o he mild ope a ing empe a u e (e.g. 145 120 o 185 ºC), he bonds be ween he monome ic suga s s a b eaking 146 andomly, p oducing p og essi ely sho e oligome s. This p ocess con inues 147 un il he momen in which he oligome has a leng h low enough o be ex ac ed 148 om he solid by solubiliza ion o d agging [6,14,15]. In his momen , bo h 149 phenomena a e p esen , oligome dissolu ion and oligome clea ing, and wo 150 dis inc s ages can be di e en ia ed: (1) solid oligome clea ing, which is p esen 151 om he beginning, and (2) solid & liquid oligome clea ing wi h hemicellulose 152 dissolu ion. The ac ha hese wo phases a e p esen a he same ime explain 153 why he e is a delay in he ex ac ion p o iles (Figu e 3.a). Be o e his i s soluble 154 oligome eleasing, only aw ma e ial ee suga s and a li le numbe o clea ing 155 p oduc s (small oligome s and monome s) could be emo ed. Ne e heless, 156 he e a e cases whe e no delay is p esen due o he biomass di e si y [16]. This 157 is possible when he ini ial hemicellulose leng h is so low ha i is ini ially soluble 158 o i is so ace yla ed ha only s age 2 is p esen . The e o e, i bo h s ages a e 159 p esen , he molecula weigh e olu ion du ing he ex ac ion should ha e a 160 maximum ( he i s soluble oligome ) nea he ime ( m) when he concen a ion in 161 he liquid eaches he highes alue (Figu e 3.b). A e his molecula weigh peak, 162 i would con inuously dec ease due o he clea ing. This Beha io was al eady 163 obse ed in li e a u e [15,16]. Howe e , when only s age 2 is p esen he 164 molecula weigh would dec ease wi h ime. 165 166 7 167 168 169 170 171 172 Figu e 3: Liquid p o iles a he ou pu o a packed bed eac o du ing a 173 hyd o he mal ex ac ion p ocess: (a) TOC e olu ion, (b) molecula weigh 174 e olu ion (Mw) when bo h s ages a e p esen s and molecula weigh e olu ion 175 when only s age 2 is p esen . 176 MW TOC ,min TOC MW Mw , min Bo h s ages S age 2 a) S age 1 S age 1 & S age 2 m b) Delay 8 177 To sum up, empe a u e, he molecula weigh and de ace yl con en s plays an 178 essen ial ole in hemicellulose ex ac ion since hey di ec ly a ec hemicellulose 179 solubili y. 180 2.2. Suga p oduc ion om he clea ing p ocesses 181 As i was explained in he p e ious sec ion, he clea ing can also p oduce 182 monome ic suga s and, i empe a u e is high enough, all he hemicellulose could 183 be con e ed in o monome ic suga s. Howe e , he ope a ional condi ions 184 equi ed o achie e a comple e con e sion a e so high ha hey also imply suga 185 deg ada ions. Gallina e al. [4] s udied he op imal condi ions o he hyd o he mal 186 ac iona ion o eucalyp us in a semi-con inuous eac o . They ound ha he 187 op imum monome ic suga yield was a 185 ºC (67.41%), s a ing o dec ease a 188 highe empe a u es. Yed o e al. [9] assessed he hemicellulose ex ac ion om 189 holm oak in a ba ch sys em, ob aining ha he highes monome ic yield was a 190 170 ºC (60%) and ha deg ada ion s a ed a empe a u es as low as 150 ºC. 191 Rissanen e al. [10] analyzed he hyd o he mal deg ada ion o sp uce in he same 192 eac o as Yed o e al. [9], eaching a simila op imum. Sukhbaa a e al. [11] 193 wo ked wi h suga cane bagase also in a ba ch sys em, being hei monome ic 194 yield op imum a 180 ºC and obse ing a huge deg ada ion abo e 190 ºC. The 195 same biomass was conside ed by Vallejos e al. [17] who eached he bes 196 monome ic yield a 180 ºC oo (70%). Simila esul we e epo ed by Thomsen 197 e al.[18], dos San os Rocha e al.[19] and Makishima e al. [20] o whea s aw, 198 suga cane s aw and co ncob, espec i ely. 199 Focusing on di ec suga p oduc ion is o in e es since hey can be used o 200 p oduce uels (bioe hanol) o chemicals (like xyli ol ia hyd ogena ion). These so- 201 called “deg ada ion p oduc s” can also be he a ge [8]. Fo ins ance, u u al and 202 i s de i a i es can be used as ungicides o lub ican s [21] while lac ic acid is a 203 p ecu so o biodeg adable polyme s p oduc ion [22]. The e o e, o 204 a oid/p omo e suga deg ada ion he ope a ing empe a u e and he olume ic 205 low ( he less ime in he eac o , he lowe deg ada ion [4,7,18,20] a e he main 206 9 in ol ed a iables. I is wo h highligh ing ha when he eac o is a ba ch sys em, 207 he liquid/solid a io has he same ole as esidence ime. 208 2.3. Hemicellulose deace yla ion 209 Hemicellulose deace yla ion and clea ing ake place in pa allel, which is e lec ed 210 in a eleasing o ace ic acid du ing he hemicellulose ex ac ion, dec easing he 211 pH o he wa e . I should be ema ked ha his ace ic acid p oduc ion only 212 happens in he solid phase [6,14,23–25]. Howe e , ace ic acid can be ob ained 213 om suga deg ada ion in liquid phase oo [8]. Simila ly, u onic acid can be also 214 eleased du ing he hyd o he mal ea men [26,27]. Ne e heless, i is no 215 comple ely clea i he pH change accele a es ex ac ion o i his change is only 216 a consequence o he ex ac ion [10,12]. This phenomenon is deeply ela ed wi h 217 he hemicellulose ex ac ion p ocess selec i i y since hese acids a e a sou ce o 218 p o ons ha ca alyze he clea ing and deg ada ion eac ions in liquid phase i he 219 esidence ime is high enough [6,14]. A s a emen ha was e i ied by Song e al. 220 [28], showing ha deg ada ion is much lowe i he pH is main ained abo e 4-5. 221 Mo eo e , he eleasing o ace yl g oups also means ha he solubili y o he 222 emained pa o he hemicellulose would be lowe since he capaci y o linking 223 by hyd ogen bonds wi h wa e would be lowe . Addi ionally, he s e ic hind ance 224 also would be highe . Following his idea, i can be seen in Figu e 4. ha he 225 minimum o he pH co esponds o he maximum in he TOC p o ile. Thus, he 226 oligome s in ol ed in he s age (2) de ined in Figu e 3 will be mo e soluble since 227 hey a e smalle bu , a he same ime, hei solubili y also dec eases due o he 228 lack o ace yl g oups, explaining why he ex ac ion is mo e di icul a e he 229 maximum (dec easing slope). Mo eo e , ex ac ion would also be slowe 230 because he a ailable amoun o hemicellulose is much lowe . The eby, he 231 ace yla ion deg ee (and u onic con en ) is ano he a iable o conside . 232 16 cellulose is a wa e insoluble polyme so ha i is no possible o simply dissol e 375 and hyd olyze cellulose in wa e a ambien condi ions. As a esul , he hyd olysis 376 o cellulose in lignocellulosic biomass usually in ol es he use o s ong acids as 377 ca alys s [46], which cause a nega i e impac in he en i onmen and yields a 378 high concen a ion o deg ada ion p oduc s. Howe e , when using supe c i ical 379 wa e cellulose is mo e e ec i ely con e ed o oligome s and monome suga s 380 ins ead o yielding mainly deg ada ion p oduc s. 381 The e o e, he objec i e o his sec ion is o cla i y he mechanisms in ol ed in 382 bo h he dissolu ion and he hyd olysis o cellulose in wa e as well as o discuss 383 he in luence o he key pa ame e s which a ec bo h p ocesses. 384 4.1. Cellulose dissolu ion 385 The dissolu ion o cellulose in wa e ha e been explained [48–50] om a 386 he modynamic poin o iew. The Gibbs ee ene gy is a he modynamic 387 magni ude commonly conside ed o analyze whe he a chemical p ocess is 388 spon aneous o no . I s a ia ion is exp essed as a combina ion o he a ia ion 389 o he en halpy and he a ia ion o he en opy o he sys em: 390 ∆𝐺 = ∆𝐻 − 𝑇∆𝑆 (1) 391 “G” is he Gibbs ee ene gy, “H” he en halpy, “T” he empe a u e and “S” he 392 en opy. When he a ia ion o he Gibbs ee ene gy is nega i e, he p ocess is 393 spon aneous. In he eac ion o wo di e en compounds, he a ia ion o en halpy 394 ep esen s he hea o eac ion o he hea o mixing. In he combina ion o 395 cellulose and wa e his pa ame e is almos negligible since no addi ional hea is 396 gene a ed o consumed. The e o e, he p e ious exp ession is educed o: 397 398 ∆𝐺 = −𝑇∆𝑆 (2) 399 400 Consequen ly, he dissolu ion and hyd olysis o cellulose in wa e is ca ied ou 401 (spon aneous p ocess) when he en opy a ia ion is posi i e. F om a s uc u al 402 poin o iew, he en opy o cellulose inc eases when i s molecula con o ma ion 403 changes om a igid s uc u e o a mo e lexible one which bene i s dissolu ion. 404 Since cellulose s uc u e is cha ac e ize by i s complexi y and igidi y, a lowe 405 17 empe a u es no con o ma ional changes will be p oduced, he en opy will no 406 inc ease no he Gibbs ee ene gy will dec ease and he e o e no dissolu ion will 407 be p oduced. Only in he cases in which he empe a u e is conside ably 408 inc eased and he e o e he in e nal ene gy o he s uc u e, con o ma ional 409 changes could be p oduced. 410 In li e a u e, h ee main cha ac e is ics o he cellulose s uc u e a e conside ed 411 o undamen al in e es in i s dissolu ion in wa e : 412 413 1) The p esence o in a and in e molecula hyd ogen bonds [48,49,51]. 414 Cellulose is cons i u ed by glucose molecules joined oge he o ming long 415 ibbe s which a e connec ed by hyd ogen bonds. This ac esul s in a igid 416 and cohesi e s uc u e which a oids he pene a ion o wa e molecules and 417 consequen ly he dissolu ion o he s uc u e. 418 2) Cellulose is conside ed an amphiphilic molecule [48,49,52]. I s s uc u e has 419 bo h hyd ophobic and hyd ophilic zones as a consequence o he o ien a ion 420 o i s unc ional g oups. While he hyd oxyl g oups loca ed in equa o ial 421 posi ion c ea e he hyd ophilic egions, he axial glycosidic bonds p oduce 422 hyd ophobici y. This is conside ed he eason why he wa e molecules a e 423 no able o easily c ea e hyd ogen bonds wi h he cellulose which will 424 p oduce i s dissolu ion. 425 3) C ys allini y: c ys allini y has always been conside ed a key pa ame e when 426 analyzing he dissolu ion o cellulose in wa e [48,53]. The high c ys allini y 427 o he cellulose molecule is esponsible o i s igid s uc u e a oiding 428 con o ma ional changes which could acili a e i s dissolu ion. 429 Conside ing he lack o a obus model which explains he dissolu ion o cellulose 430 in wa e , se e al au ho s ha e pe o med expe imen s wi h he objec i e o 431 analyzing he in luence o he p ocess pa ame e s. 432 F om a s uc u al poin o iew, [54,55] s udied he in luence o he aw cellulose 433 used in he dissolu ion. They demons a ed ha he cellulose allomo ph di ec ly 434 a ec s he dissolu ion p ocess. Fo example, al hough cellulose I is he mos 435 abundan ype in na u e, cellulose II is mo e s able [49]. Mo eo e , no only he 436 cellulose ype in luences he dissolu ion, also he amoun o wa e has o be 437 conside ed [48,56]. Rega ding o he c ys allini y o he s uc u e, [57] analyzed 438 18 he dissolu ion o cellulose a e milling. Milling p oduces an amo phous s uc u e 439 which acili a es he ac ion o wa e . They s a ed ha he c i ical ac o is no he 440 educ ion o he pa icle diame e bu he clea age o he hyd ogen bonds and 441 he consequen gene a ion o amo phous zones. Amo phous and semi-c ys alline 442 zones a e easie o be hyd olyzed since wa e molecules can a oid he 443 hyd ophobic zones which a e p esen in he s uc u e as a consequence o he 444 amphiphilic na u e o he cellulose [58]. 445 F om an ope a ing poin o iew, he majo i y o expe imen s analyzed he 446 p ocess ocusing in he a ia ion o he p essu e, he empe a u e and he 447 eac ion ime. As i has been explained in his sec ion, due o he physical 448 s uc u e and he na u e o cellulose, i s dissolu ion is g ea ly limi ed by 449 empe a u e. Common wo king empe a u es usually ange om 200ºC o mo e 450 han 400ºC. The e o e, in o de o main ain wa e in liquid o supe c i ical s a e 451 when he wo king condi ions su passes i s c i ical poin , (Tc=374ºC, Pc=22.1 452 MPa) he p essu e shall be inc eased. The analysis o he in luence o p essu e 453 has been s udied by [53]. They p o ed ha when he p essu e is inc eased abo e 454 50MPa ( eaching p essu es up o 700MPa), e en a ela i e low empe a u es 455 he wa e molecules a e able o en e inside he cellulose s uc u e and swell he 456 polyme ic ma ix which inally collapses. When he p essu e is only conside ed in 457 o de o main ain he wa e in liquid o supe c i ical s a e, i s in luence is negligible 458 and he undamen al pa ame e s o be conside ed a e he empe a u e and he 459 eac ion ime. An inc ease o empe a u e clea ly bene i s he dissolu ion o 460 cellulose since i modi ies i s s uc u e a o ing he combina ion o cellulose and 461 wa e molecules. Howe e , i also accele a es i s hyd olysis consuming he 462 cellulose which is being dissol ed. Consequen ly, he only possibili y o dissol ing 463 cellulose and educe i s hyd olysis a e is selec ing an op imum combina ion o 464 empe a u e and eac ion ime. In li e a u e, he analysis o cellulose dissolu ion 465 a high empe a u es is gene ally combined wi h hyd olysis s udies. Hyd olysis is 466 conside ed one o he undamen al p ocesses in g een chemis y since i allows 467 ob aining high alue p oduc s om enewable esou ces such as biomass. As 468 biomass is a complex aw ma e ial and due o he lack o enough know-how in 469 his ield, he majo i y o au ho s ha e s a ed wo king wi h cellulose ins ead o 470 wi h biomass. When cellulose is mixed wi h wa e a high empe a u es, i is i s 471 19 dissol ed and subsequen ly i eac s wi h he wa e molecules p esen in he liquid 472 medium p oducing he clea age (hyd olysis) o he glycosidic bonds. As he 473 clea age o hese bonds is no comple ely simul aneous no ins an aneous, i s , 474 oligosaccha ides a e gene a ed which a e hen hyd olyzed o monosaccha ides. 475 Finally, i he hyd olysis eac ion is no s opped, he monosaccha ides a e 476 deg aded o o ganic compounds such as acids [59,60]. As i has been s a ed, i 477 is undamen al bo h in dissolu ion and in hyd olysis o ind he op imum pai o 478 empe a u e and eac ion ime alues in o de o educe he gene a ion o 479 undesi ed p oduc s. 480 The expe imen s p esen ed in li e a u e a e clea ly di ided in h ee zones: 481 subc i ical egion, icini ies o he c i ical poin and supe c i ical egion. 482 In his pape , he subc i ical egion is conside ed he one in which he empe a u e 483 emains below 320ºC. In his zone he dissolu ion and subsequen hyd olysis is 484 p oduced as a esul o he consump ion o supe icial cellulose which is able o 485 in e ac wi h wa e molecules [61]. Fu he mo e, he cellulose which can be easily 486 dissol ed is he one which was p esen in an amo phous s a e. Below 280ºC, i 487 is obse ed ha he cellulose dissolu ion a e dec eases wi h ime since wa e is 488 no able o dissol e c ys alline cellulose once he amo phous cellulose has been 489 al eady dissol ed [55]. A empe a u es be ween 280ºC and 320ºC inc easing 490 ei he he eac ion ime o he empe a u e only inc eases he deg ada ion o he 491 cellulose, mos ly amo phous, which has been al eady dissol ed [55,62]. 492 The e o e, wo king wi h low eac ion imes p oduces high DP (deg ee o 493 polyme iza ion) molecules [63]. A empe a u es below 250ºC [64] p o ed ha 494 cellulose is dissol ed bu no hyd olyzed and he e o e ha i is possible o ob ain 495 high DP molecules. Howe e , he p ocess is limi ed by he amoun o amo phous 496 cellulose a ailable. In hese cases eac ion imes in he o de o hou s a e 497 equi ed which implies he ope a ion in ba ch and semi-con inuous eac o s. 498 Finally, milling he aw cellulose c ea es amo phous zones which can be easily 499 dissol ed, e en a empe a u es below 230ºC, gene a ing high DP molecules 500 [57]. A his empe a u es, no modi ica ions a e obse ed in he solid esidue 501 when he cellulose s uc u e is c ys alline ins ead o amo phous [65]. 502 In he egion nea he c i ical poin , when he eac ion ime is inc eased, he 503 dissol ed cellulose is hyd olyzed o glucose and la ely o deg ada ion p oduc s. 504 20 I has been expe imen ally demons a ed [51] ha a empe a u es be ween 505 320ºC and 330ºC (25MPa) a ansi ion om a c ys alline o an amo phous 506 s uc u e is p oduced. This ansi ion explains he apid dissolu ion o cellulose in 507 wa e and he absence o any swelling phenomena [51]. The ac ha when 508 cellulose and wa e eac a hese o highe empe a u es du ing a sho eac ion 509 ime he inal p oduc ob ained is cellulose II when he ini ial cellulose allomo ph 510 is cellulose I is jus i ied as a consequence o he highe s abili y o cellulose II. 511 When he empe a u e is inc eased abo e 330ºC cellulose I is con e ed in o 512 amo phous cellulose. Then, when he empe a u e dec eases, he amo phous 513 cellulose is con e ed in o he mo e s able cellulose II allomo ph [59,61]. This is 514 also con i med wo king a empe a u es below 320ºC since only cellulose I is 515 ob ained [66]. 516 Finally, in he supe c i ical egion he dissolu ion and hyd olysis o cellulose when 517 wo king a low concen a ions is p oduced simul aneously, in homogeneous 518 phase and wi hou mass ans e limi a ions [53,67]. The ansi ion be ween 519 c ys alline cellulose o amo phous cellulose a 330ºC, he high empe a u es o 520 eac ion which p oduce he clea age o he hyd ogen bonds [53,68] and he 521 p ope ies o supe c i ical wa e such as high di usi i y, high densi y compa ed 522 wi h wa e in apo s a e and i s abili y o dissol e o ganic compounds, obse ing 523 he o al dissolu ion o cellulose [69,70], explain he homogenei y o he p ocess. 524 Recen ly i has been p o ed ha when he concen a ion o cellulose is inc eased 525 he dissolu ion and hyd olysis p ocesses a e no comple ely simul aneous no 526 homogeneous [71]. 527 4.2. Cellulose hyd olysis 528 I is no ed ha in his eac ion zone he hyd olysis o biomass has gained a lo o 529 a en ion [67,68]. In ac , special a en ion has been paid o he hyd olysis o 530 cellulose, since i is he majo componen o lignocellulosic biomass and he e o e 531 is he key o be e unde s and he eac ion mechanisms, kine ics and 532 pe o mance o supe c i ical wa e hyd olysis o eal biomass [71,72]. 533 4.2.1. P oduc ion o suga s om cellulose hyd olysis in supe c i ical 534 wa e 535 21 The con e sion o cellulose o suga s in supe c i ical wa e has been ex ensi ely 536 s udied using di e en kinds o eac o s. The hyd olysis in ba ch- ype eac o s is 537 usually ca ied ou wi h long eac ion imes, a o ing he decomposi ion o glucose 538 o deg ada ion p oduc s [73,74]. Howe e , he low- ype sys em makes i possible 539 o educe he eac ion ime and he e o e inc easing he yields o suga s ins ead 540 o deg ada ion p oduc s [61,70]. Recen ly ou esea ch g oup de eloped an 541 expe imen al se up o pe o m he hyd olysis o cellulose suspensions in 542 supe c i ical wa e by using a con inuous mic o- eac o , gi ing as a esul a o al 543 con e sion o cellulose in milliseconds and yielding a suga s p oduc ion o 96 % 544 w/w [67]. This con inuous mic o- eac o is shown in Figu e S2, whe e i can be 545 seen ha he eac ion sec ion consis ed o a ee junc ion (M) whe e he cellulose 546 (o biomass) was ins an aneously hea ed up by mixing i wi h a supe c i ical wa e 547 s eam. In o de o e ec i ely s op he hyd olysis eac ion, a sudden 548 dep essu iza ion h ough a needle al e was ca ied ou , so ha he e luen was 549 immedia ely cooled down om 400ºC o a ound 100ºC and he e o e eac ion 550 was o e . Then, depending on he dimensions o he pipe be ween he junc ion 551 and he dep essu iza ion al e, he eac ion ime was calcula ed as a unc ion o 552 eac o olume and low o he eac o , so ha jus by changing he dimensions 553 o he eac o o he pumped low, di e en eac ion imes would be p o ided. In 554 e ms o suga s yield om cellulose hyd olysis in hyd o he mal medium, se e al 555 condi ions we e es ed by changing empe a u e, p essu e and eac ion ime in 556 he mic o- eac o men ioned abo e. As a esul , i was ound ha he op imal 557 condi ions o ob ain soluble suga s (up o six uni s o glucose) we e achie ed a 558 400 ºC wi h ex emely sho eac ion imes (a ound 0.01 s). I he eac ion ime 559 was inc eased, he suga s we e hyd olyzed and he yield dec eased, as i can be 560 seen in Figu e 8. The combina ion o supe c i ical wa e medium and he e ec i e 561 me hod o he eac ion ime con ol allowed such a high suga s yield om 562 cellulose hyd olysis. This ac can be explained aking in o accoun han unde 563 hose condi ions, he cellulose hyd olysis kine ics a e imp o ed and he glucose 564 hyd olysis kine ics a e slow enough so ha using he sudden expansion mic o- 565 eac o is possible o s op he eac ions a e comple e cellulose hyd olysis bu 566 be o e glucose deg ada ion [67]. I was also p o en ha cellulose hyd olysis 567 eac ions we e highly in luenced by empe a u e, meanwhile p essu e did no 568 a ec ed cellulose hyd olysis a e in he s udied ange [75,76]. 569 22 570 Figu e 8. Suga s yield om cellulose hyd olysis in hyd o he mal medium along 571 eac ion ime. Expe imen empe a u e: ed = 400ºC; yellow = 350 ºC; blue = 572 300ºC. Expe imen p essu e ( ♦ ) 27 MPa; ( ■ ) 25 / 23 MPa and ( ▲ ) 23 / 18 MPa 573 [76]. 574 575 4.2.2. Cellulose hyd olysis kine ics in supe c i ical wa e 576 Cellulose was hyd olyzed ollowing he main hyd olysis eac ion pa hway in 577 supe c i ical wa e which is shown in Figu e 9 [71], whe e i can be seen ha 578 cellulose is i s ly hyd olyzed in o oligosaccha ides and hen in o glucose. Once 579 he glucose has been p oduced, i can be isome ized o uc ose and hen 580 con e ed in o dehyd a ed (5-HMF) o e o-aldol condensa ion p oduc s 581 (glycolaldehyde, py u aldehyde and/o glyce aldehyde). As men ioned abo e, 582 wo king a 400 ºC and e y sho eac ion imes, he eac ion would be s opped 583 a glucose. Howe e , i he eac ion ime is inc eased, e o-aldol condensa ion 584 p oduc s would be p oduced, yielding aldehydes as glycolaldehyde, 585 py u aldehyde and/o glyce aldehyde. The e o e, he con ol o eac ion ime was 586 he key ac o o selec i ely hyd olyze cellulose in supe c i ical wa e . 587 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0.01 0.1 1 10 100 Suga s Yield, % w·w-1 , s 400-27 400-25 400-23 350-27 350-23 350-18 300-27 300-23 300-18 23 Glucose O H H H H OH OH H OH OH OH OOH OH H H OH H OH H H OH F uc ose O H H H H OH OH H OH O OH O H H H HOH H OH OH OH Cellobiose O H H H H OH OH H OH O OH OH H H HOH H OH O OH O H H H H OH H OH O OH O H H H HOH H OH OH OH Oligosaccha ides O H H H H OH H OH O OH O H H H HOH H OH OH OH Cellulose O OH O 5-HMF O OH OH Glyce aldehyde OH O Glycolaldehyde CH3 O O Py u aldehyde 588 Figu e 9. Reac ion pa hway o cellulose hyd olysis in supe c i ical wa e based 589 on [71]. 590 The p ope ies o wa e may a y conside ably when changing he condi ions 591 o m subc i ical o supe c i ical, a ec ing o he p oduc s yielded om cellulose 592 hyd olysis [72]. Jus by changing p essu e and empe a u e, di e en eac ion 593 mechanisms a e a o ed. Wa e a 25MPa and empe a u es below 300 ºC has 594 a densi y a ound 800 kg/m3 and an ionic p oduc (as pK) be ween 11 and 14. 595 Unde hose condi ions, wa e is highly dissocia ed and H+/OH- ions a e highly 596 a ailable in he eac ion medium and he e o e ionic eac ions a e a o ed [77,78]. 597 Howe e , when empe a u e is inc eased up o 400 ºC a cons an p essu e, he 598 densi y conside ably dec eases (being a ound 150 kg/m3) and he ionic p oduc 599 inc eases up o 21 [79]. This change in he ionic p oduc a ec s he kine ics o 600 glucose and uc ose deg ada ion, a oiding he ionic deg ada ion eac ions 601 (which a e he go e ning chemis y when using acid ca alys s) and a o ing he 602 adical eac ions [72]. In ac , i was ound ha he concen a ion o H+/OH- due 603 o wa e dissocia ion was a de e mining ac o in he selec i i y o cellulose 604 hyd olysis in supe c i ical wa e [76]. So a , kine ic models o cellulose 605 hyd olysis only conside ed he concen a ion o cellulose and i s de i ed p oduc s 606 24 in o he equa ions, so ha i s o de kine ics we e selec ed o p edic cellulose 607 hyd olysis in supe c i ical wa e . Following hose adi ional kine ic models 608 Can e o e al. [76] ound an incong ui y o he kine ic cons an s o uc ose 609 dehyd a ion o 5-HMF when ca ying ou he hyd olysis o cellulose in supe c i ical 610 wa e a empe a u es be ween 300 – 400 ºC and 25 MPa. In Figu e 10 i can be 611 seen he i ed kine ic cons an s o 5-HMF o ma ion (khm ) e sus he ecip ocal 612 empe a u e, acco ding o A henius law. A b eak poin can be clea ly obse ed 613 in Figu e 10a, co esponding o he su oundings o he c i ical poin o wa e , 614 which ep esen s a de ia ion om A henius law. So ha , he adi ional models 615 whe e only cellulose concen a ion was aken in o accoun in a i s o de kine ics 616 equa ion we e only capable o p edic he kine ic cons an s o uc ose 617 dehyd a ion o 5-HMF a subc i ical condi ions. Tha sugges ed ha ano he 618 ac o was no aken in o accoun in o he kine ic equa ion. To sol e he p oblem, 619 he concen a ion o p o ons and hyd oxide ions we e added o he kine ic model, 620 u ning i in o a second o de kine ic equa ion. As a consequence o ha 621 ans o ma ion, he kine ic cons an s ollowed he A henius law o he ull 622 empe a u e spec a, meaning ha he dehyd a ion o 5-HMF unde bo h 623 subc i ical and supe c i ical condi ions was lineally i ed as i can be obse ed in 624 Figu e 10b. Tha would sugges ha he selec i i y o he p ocess was s ongly 625 a ec ed by he p o ons and hyd oxide ions concen a ion in he eac ion medium, 626 so ha imp o ing he unde s anding o he eac ion mechanisms o he hyd olysis 627 o cellulose in supe c i ical wa e . In ha way, e oaldol condensa ion eac ions 628 om glucose and uc ose ( o p oduce aldehydes) a e no e y demanding o ions 629 and he e o e hey a e a o ed when wa e is highly associa ed (as i occu s a 630 supe c i ical s a e). On he o he hand, isome iza ion glucose- uc ose and 631 dehyd a ion eac ions a e no a o ed since hese eac ions ake place o ming 632 ansi ion s a es wi h OH- and H+ and hus hey a e diminished when wa e is 633 highly dissocia ed [76]. 634 25 635 636 Figu e 10. Kine ic cons an s A henius i ing o uc ose dehyd a ion o 5-HMF 637 a 25 MPa and empe a u es be ween 300 and 400ºC [76]. a) Kine ic e alua ion 638 jus conside ing cellulose and de i ed p oduc s concen a ion. b) Kine ic 639 e alua ion also conside ing p o ons and hyd oxide ions concen a ions as 640 eagen s. 641 4.2.3. Cellulose concen a ions as a mass ans e limi a ion 642 Ano he ac o ecen ly e ised conce ning cellulose hyd olysis and dissolu ion in 643 supe c i ical wa e was he e ec o cellulose concen a ion i sel [71]. So a , 644 exis ing models desc ibing he con e sion a e o cellulose assumed ha he 645 hyd olysis o cellulose pa icles akes place a hei su ace and he e o e he 646 pa icle size was conside ed he key pa ame e o he con e sion a e. Tha 647 sh inking-co e model implied he use o a noncon en ional kine ic equa ion 648 [66,72]. On he o he hand, o ake in o accoun he eagen concen a ion, a i s 649 o de kine ic was assumed o desc ibe he con e sion a e o cellulose in 650 supe c i ical wa e . As i can be seen in Figu e 11, expe imen al esul s o 651 cellulose hyd olysis in supe c i ical wa e a 400 ºC and 25 MPa and di e en 652 concen a ions we e i ed o he i s o de kine ic equa ion by plo ing he 653 loga i hm agains he eac ion ime. In all cases, a linea dependence was ound, 654 whe e he slope ep esen ed he kine ic cons an , k. In Figu e 11 i can be 655 obse ed ha when inc easing he cellulose inle concen a ion he eac ion a e 656 is slowe , sugges ing ha mass ans e esis ances mus ha e an impo an 657 e ec o e cellulose hyd olysis kine ics. Also, combining hose da a wi h he ones 658 om a p e ious wo k [67] i was possible o calcula e he so called mass ans e 659 limi o cellulose hyd olysis in hyd o he mal media. Those calcula ions a e 660 a b subc i ical supe c i ical 32 hou he lowes yield was gi en when jus wa e was used as sol en . The bes 836 pe o ming sol en was a wa e -phenol mix u e, which achie ed nea ly o al 837 supp ession o cha o ma ion wi h 99% TS molecules [96]. 838 Fang a al. ollowed decomposi ion o o ganosol lignin in wa e /phenol solu ion 839 in mic o- eac o coupled wi h op ical mic oscopies a empe a u es up o 600˚C 840 and wa e densi ies up o 1165 kg/m3. The mic o eac o , diamond an il cell (DAC) 841 allows o in-si u obse a ions o samples in he ully- isible chambe ia op ical 842 mic oscopy. The DAC consis ed o a hole and sealed by comp ession o wo 843 opposing an ils made o diamond. The chambe was apidly hea ed by wo 844 elec ic mic ohea e s by cu ing powe , which is con enien o he s udy o phase 845 beha io and chemical eac ions. Expe imen s ha e been done a di e en wa e 846 densi ies, hea ing a e, maximum empe a u es and lignin concen a ion. Th ee 847 di e en ypes o p oduc s we e ob ained: a non-dissol ed black esidue, a 848 p ecipi a ed esidue and eddish oil. A homogenous phase was o med o he 849 phenol + lignin sys em whe e phenolic cha p ecipi a ed as he main p oduc . 850 Adding wa e o his sys em de-polyme iza ion o lignin was p omo ed by 851 hyd olysis in a homogeneous phase and i s e-polyme iza ion was inhibi ed by 852 phenol. The homogenous phase was no ound in he case o lignin + wa e 853 sys em. A e ini ial dissolu ion a abo e 377 °C lignin unde wen hyd olysis and 854 py olysis o phenolic, which a e u he changed o oil in he aqueous phase. A 855 highe empe a u es, solid pa icles p ecipi a ed om he aqueous ia 856 homogeneous e-polyme iza ion o he phenolics and wa e soluble compounds 857 o o m a phenolic cha . A hese same condi ions, non-dissol ed lignin unde wen 858 he e ogeneous py olysis and o med polya oma ic cha . Highe wa e densi y 859 dec ease lignin dissolu ion. The e o e, polya oma ic cha , wi h a ligh e molecula 860 weigh was he main p oduc along wi h a smalle ac ion o phenolic cha . I can 861 be conclude ha o wa e and phenol mix u es, lignin can be comple ely 862 solubilized and unde goes homogeneous hyd olysis and py olysis ha p e en s 863 u he e-polyme iza ion [94]. 864 5.2.2. Wa e wi hou co-sol en 865 Sasaki and Go o p esen ed a wo k in which he chemical con e sion o alkali 866 lignin in nea and supe c i ical wa e a 350 ºC and 400 ºC and a p essu e o 25- 867 33 40 MPa using a ba ch eac o wi hou ca alys , ha ing 5-240 minu es esidence 868 ime was s udied. The p oduc s we e sepa a ed in o wo ac ions, me hanol 869 soluble (MS) and me hanol insoluble (MI). The main p oduc s obse ed in he MS 870 ac ion we e ca echol, phenol, and o, m, p- c esols, while MI p oduc was de ined 871 as a esidual solid. I was p oposed ha ca echol is o med ia hyd olysis o 872 guaiacol which is he main compound in s uc u e o lignin. In u he hyd olysis, 873 phenol, m, p and o-c esol we e ob ained. Dependence o eac ion ime showed 874 ha he yield o ca echol apidly inc eased wi h eac ion ime ( ill 30 min) and hen 875 dec eased, especially a 400 ˚C, while he yields o phenol, m, p and o-c esol 876 inc eased wi h eac ion ime. A e 90 min he yields o m, p and o-c esols we e 877 almos cons an while he yield o phenol sligh ly inc eased. A 400 ˚C a e 878 ca echol was consumed, he majo i y o he eac ion mos likely e mina ed. The 879 dec easing o ca echol was no ollowed by he inc easing o phenol, m, p and o- 880 c esol signi ican ly. Wa e densi y in luence yields o p oduc s whe e he yield o 881 ca echol was g adually dec eased wi h inc easing he wa e densi y a 350 ˚C 882 and d ama ically dec eased a 400 ˚C. The yields o phenol, m, p and o-c esol 883 inc eased g adually wi h inc easing he wa e densi y a 350 ˚C and 400 ˚C. I 884 was sugges ed ha an inc ease in wa e densi y enhanced he hyd olysis a e o 885 e he and ca bon-ca bon bonds o alkylphenol in lignin. Acco ding o his esul s 886 i was p oposed eac ion mechanism showed in scheme below (Figu e 13) whe e 887 lignin was deg aded in o i s de i a e compounds by dealkyla ion and hyd olysis 888 eac ion. Unde SCW condi ions hyd olysis akes place a e he and es e bonds 889 in lignin. Hyd olysis is accele a ed by a high ion p oduc o wa e . Dealkyla ion o 890 lignin gi es ca echol, which is hen hyd olyzed in o phenol. This eac ion pa hway 891 sugges s ha some use ul chemical in e media es (MS ac ion) migh be 892 eco e ed in a apid and selec i e manne by changing he empe a u e, eac ion 893 ime unde nea and supe c i ical wa e condi ion. A he same ime, e- 894 polyme iza ion o low molecula weigh compounds occu s as seen by he 895 o ma ion o cha h ough condensa ion eac ion [97]. 896 897 34 898 Figu e 13. P oposed scheme o deg ada ion o lignin unde nea and 899 supe c i ical condi ion 900 Lignin con e sion was also in es iga ed in he con inuous sys em o sho 901 esidence ime 0.5-10 s, p essu e o 25 MPa and di e en empe a u es unde 902 supe c i ical condi ions a 390˚C and 450˚C and subc i ical condi ion a 300 ˚C 903 35 and 370 ˚C [98] [99]. Tempe a u e plays an impo an ac o in deciding he 904 dominan pa hway because o he exis ence o he pa allel ionic and adical based 905 pa hways. Lignin p oduc s we e di ided in cha , gas, TOC, phenolic and a oma ic 906 hyd oca bons. Unde hyd o he mal condi ion lignin was apidly con e ed in o 907 lowe molecula weigh p oduc s o all empe a u es which was ollowed wi h 908 high yields o TOC, phenolic compounds and a oma ic hyd oca bons, while 909 decomposi ion was accele a ed unde supe c i ical condi ion [99]. Inc easing 910 decomposi ion a e wi h empe a u e ollows A henius beha io o lignin 911 deg ada ion wha was al eady ob ained by Zhang and Ramaswamy [100]. The 912 apid depolyme iza ion is cause by clea age o e he bonds om abundan β-a yl 913 e he (β-O-4) linkages in so wood lignin [101] [102]. The low dissocia ion 914 en halpies o he e he bond in he β-O-4 linkage ini ia ed he eac ion o o m a 915 phenoxy adical and a seconda y alkyl a oma ic adical [103]. The A henius 916 beha io shown by lignin decomposi ion unde hyd o he mal condi ions e en in 917 subc i ical egion u he suppo ed he conclusion ha he ini ial decomposi ion 918 was a adical eac ion. TOC yield dec eased wi h empe a u e and he yield was 919 much highe unde subc i ical condi ion. The TOC yield in subc i ical wa e 920 inc eased wi hin sho esidence ime and emained s able o dec ease slowly 921 despi e longe esidence ime. The inc easing in he polyme iza ion du ing he 922 inc ease o empe a u e should be e lec ed in he TOC yield. Low TOC yield 923 unde supe c i ical condi ion sugges ed he occu ence o seconda y eac ion. 924 This could be due o he c oss-linking be ween eac i e deg ada ion agmen s 925 ob ained om he lignin depolyme iza ion o p oduce agmen s wi h highe 926 molecula weigh s. Inc ease in he lowe molecula weigh compounds du ing he 927 ime esul ed wi h he simul aneous o ma ion o he highe molecula weigh 928 compounds because o he epolyme iza ion. The minimal dec ease in TOC yield 929 o subc i ical empe a u es implied ha he c osslinking eac ions be ween hese 930 lowe -molecula weigh compounds did no ake place ac i ely unde hese 931 condi ions. This sugges ed he signi icance o adical’s in ol emen in enhancing 932 he eac ion [98][99]. 933 Cha has signi ican ly highe yields in he supe c i ical egion and o ma ion was 934 enhanced a ele a ed empe a u es. Fo ma ion o cha om lignin ollows 935 A henius beha io and i is no a ec ed wi h change in wa e p ope ies unde 936 36 subc i ical condi ion wha poin adical eac ion. In o de o examine he 937 sugges ed hypo hesis o o ma ion o low molecula -weigh agmen s and 938 o ma ion o highe molecula weigh agmen s by c oss linking o he smalle 939 agmen i was de e mine he yields o he phenolic compounds and a oma ic 940 hyd oca bons. The main phenolic compound om lignin decomposi ion is 941 guaiacol, which is ollowed by mino composi ion o o he phenolic compounds 942 such as o, m, p-c esol, ca echol and phenol. Fo ma ion o guaiacol was highe in 943 supe c i ical empe a u e, bu apidly dec eased a longe esidence ime [98] 944 [99]. Guaiacol is an in e media e deg ada ion p oduc and highly eac i e, since 945 he me hyl C−O bond is he weakes in he guaiacol uni and is suscep ible o 946 unde go clea ing. The alipha ic C−O bond o he me hoxyl g oup is mo e likely o 947 eac because he bond ene gy o he alipha ic C−O bond (245 kJ/mol) is smalle 948 han ha o he a oma ic C−O bond (256 kJ/mol). This was concluded in he s udy 949 o Wahyudiono e al. whe e also was ound ha guaiacol showed a as 950 decomposi ion a e and he o ma ion o high-molecula -weigh subs ances 951 e o med o cha was impo an o he guaiacol decomposi ion o each 952 equilib ium [104]. Howe e , high yield o guaiacol was also ob ained unde 953 subc i ical condi ions. The high yield o guaiacol unde wo sepa a e egions o 954 empe a u e (subc i ical and supe c i ical) wi h di e en wa e p ope ies 955 indica ed guaiacol o ma ion ia wo di e en pa hways. The o ma ion o guaiacol 956 om lignin p obably p oceeded h ough hyd olysis unde subc i ical condi ions 957 because o he high ionic p oduc and dielec ic cons an o wa e . On he 958 con a y, unde supe c i ical condi ion and high empe a u e ee adical eac ion 959 should be enhanced ha lead o he o ma ion o guaiacol om lignin [98]. In bo h 960 s udies i was showed ha he decomposi ion o lignin occu ed apidly wi h 961 esidence ime below one second, which indica e ha kine ic s udy should be 962 done o esidence ime below 1s. 963 5.2.3. Wa e and base ca alys 964 In o de o enhance he ob aining o monome ic phenols, basic compounds such 965 as hyd oxides a e used as ca alys [105][106] [107]. S udies on lignin model 966 compound dihyd o-diisoeugenol, showed ha he basic agen caused e he and 967 C–C bond clea age which yielded ola ile phenols [108]. Fu he mo e, he 968 37 analysis o p oduc s om model compound eac ions e ealed ha phenyl e he 969 linkages we e e ec i ely b oken in he base ca alyzed hyd olysis eac ion while 970 C-C linkages we e less a ec ed [109]. In ano he s udy, i was concluded ha in 971 alkaline depolyme isa ion o lignin, e he bonds a e hyd olyzed a andom, mos 972 likely om he ou side o he oligome and no in he sequence o hei bond 973 s eng hs, o ming i s la ge uni s and hen smalle subuni s [106]. In addi ion, i 974 was s a ed ha he o ma ion o monome s is di ec ly p opo ional o he 975 concen a ion o sodium hyd oxide in he aqueous medium. Fu he mo e, a 976 mechanism o he NaOH ca alyzed b eakdown o he e he bonds o lignin is 977 p oposed explaining he p e e en ial o ma ion o sy ingol de i a i es, based on 978 he s abilizing e ec ha he me hoxyl g oups p o ides o he ansi ion s a es o 979 he ca benium ions. I was also concluded ha he p oduc ion o monome s is 980 limi ed by he oligome iza ion and polyme iza ion eac ions o he p oduc s 981 o med. 982 Mille e al. showed ha in he alkali depolyme iza ion o lignin using wa e as 983 sol en he mos impo an ac o in lignin depolyme iza ion was base 984 concen a ion. Mo eo e , i was obse ed ha concen a ion excess o a s ong 985 base ga e be e esul s on lignin depolyme iza ion. In addi ion, a li le amoun o 986 a s ong base (NaOH) oge he wi h a la ge amoun o less expensi e base 987 (Ca(OH)2) p oduced posi i e esul s [105]. 988 Sil a e al. s udied he ca aly ic depolyme iza ion o o ganosol lignin wi h bo h 989 NaOH ca alys and wi h bo ic acid as a capping agen , aiming o p oduce oils o 990 monome ic and dime ic p oduc s. In he case o eac ions wi h NaOH and no 991 capping agen , he highes oil yield was ob ained a 300 ºC wi h a esidence ime 992 o 4 minu es. This ga e a yield o 23% oil and no cha o ma ion. Lignin 993 con e sion inc eased s eadily wi h inc easing empe a u e bu cha was o med 994 as well as oil. In o de o inc ease oil yields, bo ic acid was used as a capping 995 agen . Wi hou base, he bo ic acid inc eased he yield o oils o a maximum o 996 36% a e 40 minu es a 300 ºC, bu a longe esidence imes o highe 997 empe a u es he yield dec eased again. The esul s showed ha he molecula 998 weigh s o he oils om he bo ic acid ca alyzed eac ions we e a ound 500 Da, 999 compa ed o 300 Da o he base ca alyzed depolyme iza ion [88]. 1000 38 In con as o a basic en i onmen , leading o dep o ona ion o phenolic hyd oxyl 1001 g oups and dec eased hyd ogen bonding, he acidic en i onmen enhances he 1002 deg ee o in e nal hyd ogen bonding. As esul , he p obabili y o acid-ca alyzed 1003 clea age o e he bonds is educed compa ed o base-ca alyzed clea age. Thus, 1004 in acid-ca alyzed hyd olysis he p ima y p oduc s p oduced a e la ge (dime s o 1005 e ame s) han in he base-ca alyzed ou e. Fo bo h cases, he p ima y p oduc s 1006 unde go easy addi ion and condensa ion eac ions leading o highe molecula 1007 weigh p oduc s [110]. 1008 Unde supe c i ical and subc i ical condi ion lignin is hyd olyzed and di e en 1009 phenolic and o he a oma ic compounds could be ob ained. These hyd olysis 1010 eac ions occu in he sho es ime han he esidence ime ha has al eady been 1011 used in li e a u e (mo e han one second). I is e y impo an o ha e be e 1012 unde s anding o eac ion pa hways, in e media e eac ion p oduc s and eac ion 1013 p oduc s o i s milliseconds o eac ion ime, hus he speci ic weaknesses and 1014 s eng hs o he polyme and i s in e media es – i.e. he subs uc u es which a e 1015 he mos suscep ible o chemical a ack. Kine ic models ha ha e been ob ained 1016 un il oday a e jus i ied wi h he inal eac ion p oduc s, wi hou in o ma ion abou 1017 in e media e p oduced. 1018 Conside able e o is s ill equi ed o add ess he sepa a ion challenges 1019 associa ed wi h lignin depolyme iza ion. The supe c i ical wa e ul a as 1020 hyd olysis could open a new way o imp o e he unde s anding o lignin 1021 depolyme iza ion, as has been done in he cellulose hyd olysis. 1022 1023 Acknowledgemen s 1024 The au ho s hank MINECO and FEDER p og am o he inancial suppo 1025 P ojec s CTQ2013-44143-R and CTQ2016-79777-R. 1026 1027 1028 1029 1030 1031 1032 39 1033 Re e ences: 1034 [1] P. Bajpai, S uc u e o Lignocellulosic Biomass, in: P e ea . Lignocellul. 1035 Biomass Bio uel P od., Sp inge Singapo e, Singapo e, 2016: pp. 7–12. 1036 doi:10.1007/978-981-10-0687-6_2. 1037 [2] O. Boble e , Hyd o he mal deg ada ion o polyme s de i ed om plan s, 1038 P og. Polym. Sci. 19 (1994) 797–841. 1039 h p://ca .inis . /?aModele=a icheN%7B&%7Dcpsid =4247470. 1040 [3] J. V Rissanen, H. G énman, C. Xu, S. Will ö , D.Y. Mu zin, T. Salmi, 1041 Ob aining Sp uce Hemicelluloses o Desi ed Mola Mass by using 1042 P essu ized Ho Wa e Ex ac ion, ChemSusChem. 7 (2014) 2947–2953. 1043 doi:10.1002/cssc.201402282. 1044 [4] G. Gallina, Á. Cabeza, P. Biasi, J. Ga cía-Se na, Op imal condi ions o 1045 hemicelluloses ex ac ion om Eucalyp us globulus wood: hyd o he mal 1046 ea men in a semi-con inuous eac o , Fuel P ocess. Technol. 148 1047 (2016) 350–360. doi:h p://doi.o g/10.1016/j. up oc.2016.03.018. 1048 [5] C. Wyman, S. Decke , M. Himmel, J. B ady, C. Skopec, L. Viika i, 1049 Hyd olysis o Cellulose and Hemicellulose, in: Polysaccha ides, CRC 1050 P ess, 2004. doi:doi:10.1201/9781420030822.ch43. 1051 [6] A. Cabeza, C.M. Pique as, F. Sob ón, J. Ga cía-Se na, Modeling o 1052 biomass ac iona ion in a lab-scale bio e ine y: Solubiliza ion o 1053 hemicellulose and cellulose om holm oak wood using subc i ical wa e , 1054 Bio esou . Technol. 200 (2016) 90–102. 1055 doi:h p://doi.o g/10.1016/j.bio ech.2015.09.063. 1056 [7] F. Ca alhei o, L.C. Dua e, F. Gí io, P. Moniz, Chap e 14 - 1057 Hyd o he mal/Liquid Ho Wa e P e ea men (Au ohyd olysis): 1058 A Mul ipu pose P ocess o Biomass Upg ading A2 - Mussa o, Solange I, 1059 in: Biomass F ac iona ion Technol. a Lignocellul. Feed. Based Bio e ine y, 1060 Else ie , Ams e dam, 2016: pp. 315–347. doi:h p://doi.o g/10.1016/B978- 1061 0-12-802323-5.00014-1. 1062 40 [8] C.M. Pique as, Á. Cabeza, G. Gallina, D.A. Can e o, J. Ga cía-Se na, 1063 M.J. Coce o, Online in eg a ed ac iona ion-hyd olysis o lignocellulosic 1064 biomass using sub- and supe c i ical wa e , Chem. Eng. J. 308 (2017) 1065 110–125. doi:h p://doi.o g/10.1016/j.cej.2016.09.007. 1066 [9] F.M. Yed o, H. G énman, J. V Rissanen, T. Salmi, J. Ga cía-Se na, M.J. 1067 Coce o, Chemical composi ion and ex ac ion kine ics o Holm oak 1068 (Que cus ilex) hemicelluloses using subc i ical wa e , J. Supe c i . Fluids. 1069 (n.d.). doi:h p://doi.o g/10.1016/j.sup lu.2017.01.016. 1070 [10] J. V Rissanen, H. G énman, S. Will ö , D.Y. Mu zin, T. Salmi, Sp uce 1071 Hemicellulose o Chemicals Using Aqueous Ex ac ion: Kine ics, Mass 1072 T ans e , and Modeling, Ind. Eng. Chem. Res. 53 (2014) 6341–6350. 1073 doi:10.1021/ie500234 . 1074 [11] B. Sukhbaa a , E.B. Hassan, M. Kim, P. S eele, L. Ing am, Op imiza ion o 1075 ho -comp essed wa e p e ea men o bagasse and cha ac e iza ion o 1076 ex ac ed hemicelluloses, Ca bohyd . Polym. 101 (2014) 196–202. 1077 doi:h p://doi.o g/10.1016/j.ca bpol.2013.09.027. 1078 [12] J. V Rissanen, D.Y. Mu zin, T. Salmi, H. G énman, Aqueous ex ac ion o 1079 hemicelluloses om sp uce – F om ho o wa m, Bio esou . Technol. 199 1080 (2016) 279–282. doi:h p://doi.o g/10.1016/j.bio ech.2015.08.116. 1081 [13] W. Reynolds, H. Singe , S. Schug, I. Smi no a, Hyd o he mal low- 1082 h ough ea men o whea -s aw: De ailed cha ac e iza ion o ixed-bed 1083 p ope ies and axial dispe sion, Chem. Eng. J. 281 (2015) 696–703. 1084 doi:h p://doi.o g/10.1016/j.cej.2015.06.117. 1085 [14] A. Cabeza, F. Sob ón, F.M. Yed o, J. Ga cía-Se na, Two-phase 1086 modelling and simula ion o he hyd o he mal ac iona ion o holm oak in 1087 a packed bed eac o wi h ho p essu ized wa e , Chem. Eng. Sci. 138 1088 (2015) 59–70. doi:h p://doi.o g/10.1016/j.ces.2015.07.024. 1089 [15] X. Chen, M. Lawoko, A. an Heiningen, Kine ics and mechanism o 1090 au ohyd olysis o ha dwoods, Bio esou . Technol. 101 (2010) 7812–7819. 1091 doi:h p://doi.o g/10.1016/j.bio ech.2010.05.006. 1092 [16] X.J. Ma, X.F. Yang, X. Zheng, L. Lin, L.H. Chen, L.L. Huang, S.L. Cao, 1093 Deg ada ion and dissolu ion o hemicelluloses du ing bamboo 1094 41 hyd o he mal p e ea men , Bio esou . Technol. 161 (2014) 215–220. 1095 doi:h p://doi.o g/10.1016/j.bio ech.2014.03.044. 1096 [17] M.E. Vallejos, F.E. Felissia, J. K uyeniski, M.C. A ea, Kine ic s udy o he 1097 ex ac ion o hemicellulosic ca bohyd a es om suga cane bagasse by 1098 ho wa e ea men , Ind. C ops P od. 67 (2015) 1–6. 1099 doi:h p://doi.o g/10.1016/j.indc op.2014.12.058. 1100 [18] M.H. Thomsen, A. Thygesen, A.B. Thomsen, Hyd o he mal ea men o 1101 whea s aw a pilo plan scale using a h ee-s ep eac o sys em aiming 1102 a high hemicellulose eco e y, high cellulose diges ibili y and low lignin 1103 hyd olysis, Bio esou . Technol. 99 (2008) 4221–4228. 1104 h p://linkinghub.else ie .com/ e ie e/pii/S0960852407007158. 1105 [19] M.S.R. dos San os Rocha, B. P a o, R. de Sousa Júnio , R.M.R.G. 1106 Almeida, A.J.G. da C uz, A kine ic model o hyd o he mal p e ea men o 1107 suga cane s aw, Bio esou . Technol. 228 (2017) 176–185. 1108 doi:h p://doi.o g/10.1016/j.bio ech.2016.12.087. 1109 [20] S. Makishima, M. Mizuno, N. Sa o, K. Shinji, M. Suzuki, K. Nozaki, F. 1110 Takahashi, T. Kanda, Y. Amano, De elopmen o con inuous low ype 1111 hyd o he mal eac o o hemicellulose ac ion eco e y om co ncob, 1112 Bio esou . Technol. 100 (2009) 2842–2848. 1113 h p://linkinghub.else ie .com/ e ie e/pii/S0960852408010791. 1114 [21] A. Eseyin E., P. S eele H., An o e iew o he applica ions o u u al and 1115 i s de i a i es, 2015. 3 (2015) 6. doi:10.14419/ijac. 3i2.5048. 1116 [22] F.A. Cas illo Ma inez, E.M. Balciunas, J.M. Salgado, J.M. Domínguez 1117 González, A. Con e i, R.P. de S. Oli ei a, Lac ic acid p ope ies, 1118 applica ions and p oduc ion: A e iew, T ends Food Sci. Technol. 30 1119 (2013) 70–83. doi:h ps://doi.o g/10.1016/j. i s.2012.11.007. 1120 [23] P. Gao, G. Li, F. Yang, X.-N. L , H. Fan, L. Meng, X.-Q. Yu, P epa a ion 1121 o lac ic acid, o mic acid and ace ic acid om co on cellulose by he 1122 alkaline p e- ea men and hyd o he mal deg ada ion, Ind. C ops P od. 48 1123 (2013) 61–67. doi:h p://doi.o g/10.1016/j.indc op.2013.04.002. 1124 [24] J.C. Pa ajó, G. Ga o e, J.M. C uz, H. Dominguez, P oduc ion o 1125 xylooligosaccha ides by au ohyd olysis o lignocellulosic ma e ials, 1126 48 Cellulose Hyd olysis in Supe c i ical Wa e , ChemSusChem. 8 (2015) 1316 1026–1033. doi:10.1002/cssc.201403385. 1317 [77] N. Akiya, P.E. Sa age, Roles o Wa e o Chemical Reac ions in High- 1318 Tempe a u e Wa e , Chem. Re . 102 (2002) 2725–2750. 1319 [78] A. K use, A. Gawlik, Biomass Con e sion in Wa e a 330−410 °C and 1320 30−50 MPa. Iden i ica ion o Key Compounds o Indica ing Di e en 1321 Chemical Reac ion Pa hways, Ind. {&} Eng. Chem. Res. 42 (2003) 267– 1322 279. doi:10.1021/ie0202773. 1323 [79] C. P omdej, Y. Ma sumu a, Tempe a u e E ec on Hyd o he mal 1324 Decomposi ion o Glucose in Sub- And Supe c i ical Wa e , Ind. {&} Eng. 1325 Chem. Res. 50 (2011) 8492–8497. 1326 h p://pubs.acs.o g/doi/abs/10.1021/ie200298c. 1327 [80] D.A. Can e o, C. Ma ínez, M.D. Be mejo, M.J. Coce o, Simul aneous and 1328 selec i e eco e y o cellulose and hemicellulose ac ions om whea 1329 b an by supe c i ical wa e hyd olysis, G een Chem. 17 (2015) 610–618. 1330 doi:10.1039/c4gc01359j. 1331 [81] A. Rome o, D.A. Can e o, A. Nie o-Má quez, C. Ma ínez, E. Alonso, M.J. 1332 Coce o, Supe c i ical wa e hyd olysis o cellulosic biomass as e ec i e 1333 p e ea men o ca aly ic p oduc ion o hexi ols and e hylene glycol o e 1334 Ru/MCM-48, G een Chem. 18 (2016) 4051–4062. 1335 doi:10.1039/C6GC00374E. 1336 [82] R.J.A. Gosselink, E. De Jong, B. Gu an, A. Abäche li, Co-o dina ion 1337 ne wo k o lignin - S anda disa ion, p oduc ion and applica ions adap ed 1338 o ma ke equi emen s (EUROLIGNIN), Ind. C ops P od. 20 (2004) 121– 1339 129. doi:10.1016/j.indc op.2004.04.015. 1340 [83] M.P. Pandey, C.S. Kim, Lignin Depolyme iza ion and Con e sion: A 1341 Re iew o The mochemical Me hods, Chem. Eng. Technol. 34 (2011) 29– 1342 41. doi:10.1002/cea .201000270. 1343 [84] J.H. Lo a, W.G. Glasse , Recen indus ial applica ions o lignin: A 1344 sus ainable al e na i e o non enewable ma e ials, J. Polym. En i on. 10 1345 (2002) 39–48. doi:10.1023/A:1021070006895. 1346 49 [85] J. Ralph, J. Peng, F. Lu, Isoch oman s uc u es in lignin: A new β-1 1347 pa hway, Te ahed on Le . 39 (1998) 4963–4964. doi:10.1016/S0040- 1348 4039(98)00968-X. 1349 [86] M.M. Campbell, R.R. Sede o , Va ia ion in Lignin Con en and 1350 Composi ion (Mechanisms o Con ol and Implica ions o he Gene ic 1351 Imp o emen o Plan s)., Plan Physiol. 110 (1996) 3–13. 1352 doi:10.1104/pp.110.1.3. 1353 [87] W.-J. Liu, H. Jiang, H.-Q. Yu, The mochemical con e sion o lignin o 1354 unc ional ma e ials: a e iew and u u e di ec ions, G een Chem. 17 1355 (2015) 4888–4907. doi:10.1039/C5GC01054C. 1356 [88] E.A.B. da Sil a, M. Zabko a, J.D. A aújo, C.A. Ca e o, M.F. Ba ei o, M.N. 1357 Belgacem, A.E. Rod igues, An in eg a ed p ocess o p oduce anillin and 1358 lignin-based polyu e hanes om K a lignin, Chem. Eng. Res. Des. 87 1359 (2009) 1276–1292. doi:10.1016/j.che d.2009.05.008. 1360 [89] G. Gonzalez, J. Sal ado, D. Mon ane, Reac ions o anillic acid in sub- 1361 and supe c i ical wa e , J. Supe c i . Fluids. 31 (2004) 57–66. 1362 doi:10.1016/j.sup lu.2003.09.015. 1363 [90] G.L. Huppe , B.C. Wu, S.H. Townsend, M.T. Klein, S.C. Paspek, 1364 Hyd olysis in supe c i ical wa e : iden i ica ion and implica ions o a pola 1365 ansi ion s a e, Ind. Eng. Chem. Res. 28 (1989) 161–165. 1366 doi:10.1021/ie00086a006. 1367 [91] Wahyudiono, T. Kane ake, M. Sasaki, M. Go o, Decomposi ion o a Lignin 1368 Model Compound unde Hyd o he mal Condi ions, Chem. Eng. Technol. 1369 30 (2007) 1113–1122. doi:10.1002/cea .200700066. 1370 [92] L. Pan, Z. Shen, L. Wu, Y. Zhang, X. Zhou, F. Jin, Hyd o he mal 1371 p oduc ion o o mic and ace ic acids om sy ingol, J. Zhejiang Uni . Sci. 1372 A. 11 (2010) 613–618. doi:10.1631/jzus.A1000043. 1373 [93] K. Yoshida, J. Kusaki, K. Eha a, S. Saka, Cha ac e iza ion o low 1374 molecula weigh o ganic acids om beech wood ea ed in supe c i ical 1375 wa e , Appl. Biochem. Bio echnol. 121–124 (2005) 795–806. 1376 [94] Z. Fang, T. Sa o, R.L. Smi h, H. Inoma a, K. A ai, J.A. Kozinski, Reac ion 1377 50 chemis y and phase beha io o lignin in high- empe a u e and 1378 supe c i ical wa e , Bio esou . Technol. 99 (2008) 3424–3430. 1379 doi:10.1016/j.bio ech.2007.08.008. 1380 [95] M. Saisu, T. Sa o, M. Wa anabe, T. Adschi i, K. A ai, Con e sion o Lignin 1381 wi h Supe c i ical Wa e - Phenol Mix u es, Ene gy & Fuels. (2003) 922– 1382 928. 1383 [96] K. Okuda, M. Ume su, S. Takami, T. Adschi i, Disassembly o lignin and 1384 chemical eco e y - Rapid depolyme iza ion o lignin wi hou cha 1385 o ma ion in wa e -phenol mix u es, Fuel P ocess. Technol. 85 (2004) 1386 803–813. doi:10.1016/j. up oc.2003.11.027. 1387 [97] Wahyudiono, M. Sasaki, M. Go o, Reco e y o phenolic compounds 1388 h ough he decomposi ion o lignin in nea and supe c i ical wa e , Chem. 1389 Eng. P ocess. P ocess In ensi . 47 (2008) 1609–1619. 1390 doi:10.1016/j.cep.2007.09.001. 1391 [98] T.L.K. Yong, M. Yukihiko, Kine ic Analysis o Lignin Hyd o e mal 1392 Con e sion in Sub- and Supe c i ical Wa e , Ind. Eng. Chem. Res. 52 1393 (2013) 9048–9059. 1394 [99] T.L.K. Yong, Y. Ma sumu a, Reac ion Kine ics o he Lignin Con e sion in 1395 Supe c i ical Wa e , (2012) 0–8. 1396 [100] B. Zhang, H.J. Huang, S. Ramaswamy, Reac ion kine ics o he 1397 hyd o he mal ea men o lignin, Appl. Biochem. Bio echnol. 147 (2008) 1398 119–131. doi:10.1007/s12010-007-8070-6. 1399 [101] J. Li, G. Hen iksson, G. Gelle s ed , Lignin 1400 depolyme iza ion/ epolyme iza ion and i s c i ical ole o deligni ica ion o 1401 aspen wood by s eam explosion, Bio esou . Technol. 98 (2007) 3061– 1402 3068. doi:10.1016/j.bio ech.2006.10.018. 1403 [102] S. Kang, X. Li, J. Fan, J. Chang, Classi ied sepa a ion o lignin 1404 hyd o he mal lique ied p oduc s, Ind. Eng. Chem. Res. 50 (2011) 11288– 1405 11296. doi:10.1021/ie2011356. 1406 [103] T. Fa a elli, A. F assolda i, G. Miglia acca, E. Ranzi, De ailed kine ic 1407 modeling o he he mal deg ada ion o lignins, Biomass and Bioene gy. 1408 51 34 (2010) 290–301. doi:10.1016/j.biombioe.2009.10.018. 1409 [104] Wahyudiono, M. Sasaki, M. Go o, The mal decomposi ion o guaiacol in 1410 sub- and supe c i ical wa e and i s kine ic analysis, J. Ma e . Cycles 1411 Was e Manag. 13 (2011) 68–79. doi:10.1007/s10163-010-0309-6. 1412 [105] J. Mille , L. E ans, J.E. Mudd, K.A. B own, Ba ch mic o eac o s udies o 1413 lignin depolyme iza ion by bases. 2. Aqueous Sol en s, Sandia Na l. Rep. 1414 (2002). doi:10.2172/800964. 1415 [106] V.M. Robe s, V. S ein, T. Reine , A. Lemonidou, X. Li, J.A. Le che , 1416 Towa ds quan i a i e ca aly ic lignin depolyme iza ion, Chem. - A Eu . J. 1417 17 (2011) 5939–5948. doi:10.1002/chem.201002438. 1418 [107] A. Toledano, L. Se ano, J. Labidi, O ganosol lignin depolyme iza ion 1419 wi h di e en base ca alys s, J. Chem. Technol. Bio echnol. 87 (2012) 1420 1593–1599. doi:10.1002/jc b.3799. 1421 [108] B.F. Wa d, Tall Oil- Chemicals om a na u al, enewable sou ce, P oc. 1422 Eigh Cellul. Con . Wood Chem. Chall. 378 (1975) 332–334. 1423 [109] J.E. Mille , L. E ans, A. Li lewol , D.E. T udell, Ba ch mic o eac o s udies 1424 o lignin and lignin model compound depolyme iza ion by bases in alcohol 1425 sol en s, Fuel. 78 (1999) 1363–1366. doi:10.1016/S0016-2361(99)00072- 1426 1. 1427 [110] X.E. I ia e, S udy o lignin as high added alue chemical compounds 1428 sou ce, Thesis 2014 UPV San Sebas ian Spain. 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 52 1440 Figu es cap ions 1441 Figu e 1: Lignocellulosic biomass s uc u e 1442 Figu e 2. Subc i ical and supe c i ical wa e p ope ies a ound he c i ical poin . 1443 Figu e 3: Liquid p o iles a he ou pu o a packed bed eac o du ing a 1444 hyd o he mal ex ac ion p ocess: (a) TOC e olu ion, (b) molecula weigh 1445 e olu ion (Mw) when bo h s ages a e p esen s and molecula weigh e olu ion 1446 when only s age 2 is p esen . 1447 Figu e 4: Rela ion be ween he pH and he ex ac ed biomass. 1448 Figu e 5: Hyd o he mal ex ac ion o eucalyp us in a semi-con inuous eac o 1449 (solid ime o 90 min): e olu ion o he hemicellulose ex ac ion yield (Yield o ), 1450 he yield o hexoses (C6), pen oses (C5) and deg ada ion p oduc s o eucalyp us 1451 wi h empe a u e (a) and esidence ime a 185 °C (b) [4] 1452 Figu e 6. E ec o subc i ical wa e empe a u e on he ex ac ion o di e en 1453 phenolic compounds om de a ed ice b an ( esidence ime = 10 min). Ob ained 1454 om Pou ali e al [40] 1455 Figu e 7. Cellulose o mula 1456 Figu e 8. Suga s yield om cellulose hyd olysis in hyd o he mal medium along 1457 eac ion ime. Expe imen empe a u e: ed = 400ºC; yellow = 350 ºC; blue = 1458 300ºC. Expe imen p essu e ( ♦ ) 27 MPa; ( ■ ) 25 / 23 MPa and ( ▲ ) 23 / 18 MPa 1459 [76]. 1460 Figu e 9. Reac ion pa hway o cellulose hyd olysis in supe c i ical wa e based 1461 on [71]. 1462 Figu e 10. Kine ic cons an s A henius i ing o uc ose dehyd a ion o 5-HMF 1463 a 25 MPa and empe a u es be ween 300 and 400ºC [76]. a) Kine ic e alua ion 1464 jus conside ing cellulose and de i ed p oduc s concen a ion. b) Kine ic 1465 e alua ion also conside ing p o ons and hyd oxide ions concen a ions as 1466 eagen s. 1467 53 Figu e 11. Kine ic analysis o cellulose concen a ions o 5, 15 and 20 % w/w 1468 (co esponding o 1.5, 4.5 and 6 % w/w a he eac o inle ). The eg ession 1469 coe icien s we e: 0.90, 0.81 and 0.96, espec i ely [71]. 1470 Figu e 12. Typical s uc u e o lignin de i ed om ha dwood (le ) and so wood 1471 ( igh ) [87] 1472 Figu e 13. P oposed scheme o deg ada ion o lignin unde nea and 1473 supe c i ical condi ion 1474 1475 54 Appendix 1. Solid and liquid esidence ime 1476 1477 Du ing an ex ac ion o eac ion p ocess whe e a packed bed eac o is in ol ed (Figu e S1) wo 1478 di e en esidence imes can be de ined, one o he solid and ano he one o he liquid. The 1479 solid esidence ime co esponds o he amoun o ime spen du ing he ope a ion since i is 1480 ixed inside he eac o . Fo ins ance, he solid esidence ime in he wo k o Cabeza e al. [1] 1481 was 94 min because hey ea ed 5 g o holm oak wi h ho p essu ized wa e du ing 94 min. In 1482 con as , he liquid is con inuously lowing h ough he eac o . The e o e, he liquid esidence 1483 depends on he eac o olume (V), he eac o po osi y () and he olume ic low (Q) ed, 1484 being his ime de ined as V·/Q. Fo his eason, i was be ween 2 and 15 min in he wo k o 1485 Cabeza e al. [1] since each expe imen was done wi h a di e en olume ic low. 1486 1487 To sum up, he solid esidence ime e e s o he ime ha he solid is being ea ed wi h he 1488 liquid. And he liquid esidence ime is he ime ha he liquid is inside he eac o . 1489 1490 1491 1492 1493 1494 Figu e S1: packed bed eac o scheme 1495 1496 [1] A. Cabeza, F. Sob ón, F.M. Yed o, J. Ga cía-Se na, Two-phase modelling and simula ion 1497 o he hyd o he mal ac iona ion o holm oak in a packed bed eac o wi h ho 1498 p essu ized wa e , Chem. Eng. Sci. 138 (2015) 59–70. 1499 doi:h p://doi.o g/10.1016/j.ces.2015.07.024. 1500 1501 1502 55 Appendix 2. Sudden expansion mic o- eac o 1503 1504 P-1 P-2 Biomass Tank Wa e Tank CV-1 CV-2 PI-1 PI-2 V-1 M PT-1 HE-1 Flash HE-2 HE-3 V-2 V-3 Hea e SV-1 SV-2 Vapo Ou le Liquid Ou le TT TTTT TT TT TT TT TT PI-3/PT-2 1505 Figu e S2. Expe imen al se -up whe e a mic o- eac o was used o hyd olyze cellulose and biomass a 1506 supe c i ical wa e condi ions [1]. 1507 1508 The esul s o cellulose and biomass hyd olysis in supe c i ical wa e discussed in he main 1509 manusc ip [1–7] we e pe o med in he con inuous plan o he FASTSUGARS p ocess, able o 1510 hyd olyze biomass in SCW a empe a u es up o 400 ºC and p essu es up o 30 MPa. A scheme 1511 o he expe imen al se -up designed by he High P essu e P ocesses G oup is shown in Figu e 1512 S2. 1513 B ie ly, wa e and a biomass suspension we e con inuously pumped o he eac o a he ope a ing 1514 p essu e (25 MPa). A he inle o he eac o (as a ee junc ion-M-) he biomass was 1515 ins an aneously hea ed up by mixing i wi h a SCW s eam, eaching in ha way he ope a ing 1516 empe a u e (400 ºC). A e he desi ed eac ion ime was achie ed, he eac o e luen was 1517 suddenly dep essu ized h ough a high empe a u e al e (V-1) ob aining an ins an aneous 1518 cooling and he e o e, s opping he eac ions. The cooling me hod was an impo an pa o he 1519 FASTSUGARS p ocess, because i was he mechanism used o e ec i ely s op he eac ions, 1520 a oiding uncon olled eac ions and he dilu ion o he p oduc s, which would occu i hey we e 1521 cooled down by quenching. 1522 An elec ic hea e was used o con ol he empe a u e o he wa e s eam wi h an adjus able 1523 powe up o 10 kW. Also, a hea exchange (HE-1) was used o bo h p ehea he wa e s eam 1524 and cool down he p oduc , in oducing in ha way a hea in eg a ion sys em. SCW was supplied 1525 up o a maximum low a e o 5 kg/h by pump P-2 and biomass suspension was ed o a maximum 1526 low a e o 3 kg/h by pump P-1. 1527 Finally, a lash chambe sepa a o was ins alled a e he eac o , allowing he sepa a ion o he 1528 p oduc s in o wo phases: a apo phase mainly composed o wa e and a liquid phase wi h he 1529 concen a ed p oduc . A e his s age, wo hea exchange s we e used o cool down he sample 1530 o oom empe a u e (HE-2 and HE-3). 1531 1532 [1] C.M. Ma ínez, D.A. Can e o, M.D. Be mejo, M.J. Coce o, Hyd olysis o cellulose in 1533 supe c i ical wa e : eagen concen a ion as a selec i i y ac o , Cellulose. 22 (2015) 1534 2231–2243. doi:10.1007/s10570-015-0674-3. 1535 [2] D.A. Can e o, M.D. Be mejo, M.J. Coce o, Go e ning Chemis y o Cellulose Hyd olysis 1536 in Supe c i ical Wa e , ChemSusChem. 8 (2015) 1026–1033. 1537 doi:10.1002/cssc.201403385. 1538 Reac ion Sec ion 56 [3] D.A. Can e o, M.D. Be mejo, M.J. Coce o, Kine ic analysis o cellulose depolyme iza ion 1539 eac ions in nea c i ical wa e , J. Supe c i . Fluids. 75 (2013) 48–57. 1540 h p://www.sciencedi ec .com/science/a icle/pii/S0896844612003841. 1541 [4] D.A. Can e o, Á. Sánchez Tapia, M.D. Be mejo, M.J. Coce o, P essu e and empe a u e 1542 e ec on cellulose hyd olysis in p essu ized wa e , Chem. Eng. J. 276 (2015) 145–154. 1543 doi:10.1016/j.cej.2015.04.076. 1544 [5] D.A. Can e o, M. Dolo es Be mejo, M. José Coce o, High glucose selec i i y in 1545 p essu ized wa e hyd olysis o cellulose using ul a- as eac o s, Bio esou . Technol. 1546 135 (2013) 697–703. doi:10.1016/j.bio ech.2012.09.035. 1547 [6] D.A. Can e o, C. Ma ínez, M.D. Be mejo, M.J. Coce o, Simul aneous and selec i e 1548 eco e y o cellulose and hemicellulose ac ions om whea b an by supe c i ical wa e 1549 hyd olysis, G een Chem. 17 (2015) 610–618. doi:10.1039/c4gc01359j. 1550 [7] A. Rome o, D.A. Can e o, A. Nie o-Má quez, C. Ma ínez, E. Alonso, M.J. Coce o, 1551 Supe c i ical wa e hyd olysis o cellulosic biomass as e ec i e p e ea men o ca aly ic 1552 p oduc ion o hexi ols and e hylene glycol o e Ru/MCM-48, G een Chem. 18 (2016) 1553 4051–4062. doi:10.1039/C6GC00374E. 1554 1555