DOI: h ps://doi.o g/10.1016/j.solene .2022.06.010
© 2022. This manusc ip e sion is made a ailable unde he CC-BY-NC-ND 4.0
license h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/(opens in new ab/window)
This is he accep manusc ip o he ollowing a icle ha appea ed in inal o m in
Sola Ene gy 241: 248-261 (2022),
which has been published in inal o m a
h ps://doi.o g/10.1016/j.solene .2022.06.010.
© 2022 In e na ional Sola Ene gy Socie y. Published by Else ie L d. unde CC BY-NC-ND licence
(h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/)
New insigh s in o Mn2O3 based me al oxide 1
g anula ion echnique wi h enhanced chemical 2
and mechanical s abili y o he mochemical 3
ene gy s o age in packed bed eac o s 4
Daniel Bielsaa b *, Mikel O eguib and Ped o L. A iasb
5
a Cen e o Coope a i e Resea ch on Al e na i e Ene gies (CIC ene giGUNE), Basque Resea ch and 6
Technology Alliance (BRTA), Albe Eins ein 48, 01510, Miñano (Ála a), Spain 7
b Uni e si y o he Basque Coun y (UPV/EHU), calle Alameda U quijo s/n, 48013 Bilbao, Spain 8
9
*Co esponding au ho a : dbielsa@cicene gigune.com 10
Abs ac 11
High empe a u e he mochemical ene gy s o age s ill equi es a signi ican esea ch e o . 12
Mos o he esea ch has been ca ied ou wi h ma e ials a lab-scale and p ope ma e ial 13
ab ica ion echniques need o be de eloped in o de o make easible he upscaling o he 14
echnology. Agglome a ion, ab asion o low olume ic ene gy densi y a e some o he 15
nega i e consequences obse ed when ying o pass om he powde s a e o he ma e ial 16
shape and amoun equi ed o a he mochemical eac o . In his wo k, a g anula ion 17
echnique is in es iga ed, using a Si-doped manganese oxide as ac i e ma e ial, de e mining 18
he c i ical pa ame e s ha p o ide he bes chemical and mechanical s abili y o he 19
g anules. The g anula ion p ocess uses a polyme ic binde o gi e consis ency o he g anules 20
and a e wa ds, i is emo ed o c ea e a po ous s uc u e o acili a e he oxygen di usion 21
in and ou o he g anule. We iden i ied he impo ance o dec easing he solubili y o he 22
binde o inc ease he olume ic ene gy densi y o he g anules. Fu he mo e, i was 23
obse ed ha inc easing he mechanical s abili y h ough a high empe a u e ea men 24
does no dec ease he chemical s abili y o he ma e ial. In o de o p o ide he i s insigh s 25
in o he scalabili y o he solu ion, he chemical and mechanical s abili y o he g anules has 26
been sa is ac o ily checked du ing 100 edox cycles, ou o which 50 we e ca ied ou in a 27
home-made lab-scale packed bed eac o wi h an inne diame e o 15 mm and ano he 50 28
edox cycles in a simul aneous he mal analyze . 29
30
31
Keywo ds: The mochemical ene gy s o age; Doped me al oxides; Redox eac ion; 32
Concen a ed sola powe plan ; Sin e ing inhibi ion; Packed bed eac o 33
1. In oduc ion 34
High empe a u e he mochemical ene gy s o age (TcES) is a an ea ly s age o de elopmen 35
and hus, li le expe imen al esea ch has been epo ed a eac o scale. Fo concen a ed 36
sola powe (CSP) plan s, wo exis ing possibili ies ha e been p incipally conside ed. On he 37
one hand, di ec use o sola adia ion, whe e he he mochemical ma e ial abso bs he sola 38
he mal ene gy in a ecei e , which ac s a he same ime as a eac o . On he o he hand, 39
indi ec use, whe e he he mochemical ma e ial is s o ed in a eac o and he hea exchange 40
is ca ied ou by means o an in e media e hea ans e luid (Zsembinszki e al., 2018). I has 41
o be no ed ha he selec ion o he adequa e eac o echnology canno be de ached om 42
he CSP con igu a ion, p o ided ha i should be in eg a ed in o he ope a ion o he plan . 43
The e is s ill li le wo k conce ning he con igu a ion o new CSP gene a ion a high 44
empe a u e. Ne e heless, one p elimina y conclusion is ha he TcES canno simply eplace 45
a con en ional he mal ene gy s o age sys em in a CSP such as a comme cial mol en sal 46
sys em (Schmid and Linde , 2017) (Pelay e al., 2019) (S öhle e al., 2016). O e all, aking 47
in o conside a ion he ad an ages and disad an ages o he epo ed echnologies, packed 48
bed eac o s a e he simples in cons uc ion, which makes hem an app op ia e candida e 49
o a i s echnology upscaling and in eg a ion s ep in CSP plan s. They a e easy o build and 50
ope a e bu hei main d awbacks a e he p essu e d op inside he bed, ha may induce 51
p e e ed gas channeling and he limi ed a ailable con ac su ace o he eac an s, which may 52
comp omise he he mal powe ou pu (Wokon e al., 2017b). In his ega d, he 53
he mochemical ma e ial p ope ies, namely, shape, size and s uc u al composi ion and 54
s abili y may ha e a signi ican impac on he pe o mance o a he mochemical packed bed 55
eac o . Consequen ly, i may become one o he mos impo an challenges o his 56
echnology. 57
Me al oxides a e one o he mos s udied he mochemical ma e ials o high empe a u e TcES 58
in CSP plan s, since ai can be used bo h as hea ans e luid and as eac an , simpli ying he 59
sys em in eg a ion. Ne e heless, mos o he scien i ic s udies ha aimed o assess he 60
beha iou o me al oxides use only se e al mg in he o m o mic o-sized pa icle powde a 61
lab-scale(Neises e al., 2012) (Pes alozzi, 2013), whe eas he echnology upscaling up o a 62
packed bed he mochemical eac o o a eal applica ion migh equi e kg o e en ons o 63
ma e ial. This ma e ial amoun , in he shape o ine powde , may cause ex emely high 64
p essu e d op and ex a pumping powe . In addi ion, ine powde is mo e p one o 65
agglome a e, which may con ibu e o enhance channelling and hinde e-oxida ion kine ics 66
(Ca illo e al., 2014). The e o e, s udies on echnology up-scaling conside di e en ma e ial 67
p epa a ion app oaches (e. g. ma e ial pelle iza ion o g anula ion, adding suppo s), in o de 68
o p oduce pa icles in he scale o mm o cm, wi h su icien mechanical and chemical s abili y 69
capable o wi hs anding a g ea numbe o he mal cycles. 70
In he case o a he mochemical packed bed eac o , he pa icles a e subjec ed o di e en 71
s esses, namely, chemical, mechanical, and he mal s esses. The olume changes due o 72
bo h phase ansi ions and he mal expansion and sh inkage, oge he wi h he p essu e 73
induced by he weigh o he uppe pa icle laye s cause pa icle-wall o pa icle-pa icle 74
ic ion, known as a che ing. In addi ion, po e educ ion caused by pa icle sin e ing may 75
c ea e o e p essu e du ing gas elease in solid-gas he mochemical ma e ials. As a esul , 76
pa icles can c ack, o hei su ace can be e oded, leading o a dec ease on he oid ac ion 77
due o ealloca ion o pa icles. These nega i e e ec s can be signi ican on he lowes laye s, 78
which a e subjec ed o highe weigh loads. Taking in o accoun he cycling ope a ion na u e 79
o he TES sys ems and he long li e-expec ancy equi ed o CSP plan s, i can lead o a 80
signi ican inc ease o he p essu e d op o o he collapse o he con aine wall. The e o e, 81
he me al oxide pa icles o a eal applica ion, oge he wi h as kine ics, should demons a e 82
su icien mechanical s eng h o ensu e eliable s abili y du ing he li e ime o hese TES 83
sys ems, which may las se e al decades. 84
Pa icle s eng h can be inc eased by designing a p ope syn hesis ou e (e. g. sol-gel, sp ay-85
d ying), applying a p e- ea men (e. g. high empe a u e sin e ing) o by adding binde s o 86
suppo ma e ials (e. g. Al2O3, TiO2 o Z O2), p o ided ha hey do no eac wi h he ac i e 87
ma e ial leading o he mochemical deac i a ion. Ano he possibili y is s eng hening he 88
pa icle su ace by encapsula ion in a s ong po ous ma e ial o by apid hea ing and cooling 89
o he pa icles. The la e would hinde changes in he co e o he g anules ha migh damage 90
he he mochemical ma e ial p ope ies. 91
The addi ion o suppo ma e ials in manganese-based oxides has been deeply s udied bo h 92
o CO2 cap u e h ough chemical looping and TcES. The a i ion esis ance o manganese-93
i on oxide wi h he addi ion o Al2O3, MgAl2O4, CeO2, Z O2 and Y2O3−Z O2 was in es iga ed by 94
G. Azimi in a luidized bed eac o (Azimi e al., 2014)(Azimi e al., 2015). The ma e ials we e 95
p epa ed by sp ay-d ying and calcina ed up o 1200°C du ing 4 h. The esea ch showed ha 96
he c ushing s eng h does no imp o e subs an ially wi h he addi ion o hese suppo s, wi h 97
he excep ion o Z O2, whe e a sligh imp o emen was obse ed. Looking a chemical looping 98
applica ions, M. Abian obse ed ha he addi ion o TiO2 may double he c ushing s eng h o 99
manganese-i on oxides, epo ing alues o 3-5 N compa ed o he 1-2 N o he undoped 100
samples (Abián e al., 2017). 101
In ega d o TcES esea ch, N. C. P eisne s udied he e ec on he mechanical s eng h o 102
manganese-i on oxide by adding 20 w % o di e en suppo ing ma e ials, namely o Z O2, 103
CeO2 and TiO2 o a mo ing bed eac o . The ma e ial was p epa ed using a build-up 104
g anula ion echnique a 800°C du ing 10 h. Pu e manganese-i on oxide pa icles showed a 105
clea endency o agglome a e and o b eak in o ine pa icles when subjec ed o he mal 106
cycling in ai . Fu he mo e, he bed olume was inc eased by 17 % a e 30 cycles in a packed 107
bed eac o con aining 21 g o ma e ial, meaning ha coa sening happened o some ex en . 108
Ne e heless, bo h Z O2 and CeO2 con ibu ed o imp o e he a i ion s eng h, wi h a sligh 109
pa icle agglome a ion and wi hou coa sening. TiO2 eac s o o m ano he s able phase and 110
hus, i canno be conside ed o TcES (Neumann e al., 2018).
111
O he in e es ing app oaches in ol ed expe imen s wi h ex uded me al oxide composi es 112
(Pagkou a e al., 2015) and wi h a suppo s uc u e, i. e. honeycomb, co de i es o oams, 113
coa ed wi h he ac i e edox ma e ial, h ough which he luid can low (Ka agiannakis e al., 114
2016) (Singh e al., 2017) (Ag a io is e al., 2015b) (Ag a io is e al., 2015a) (Ag a io is e al., 115
2016). The coa ed s uc u es p esen ed a be e in eg i y o e epe i i e edox cycles, 116
al hough he olume ic ene gy s o age is limi ed by he capaci y o he ac i e ma e ial 117
loading, leading o a subs an ial dec ease in he ene gy s o age densi y. 118
Conce ning lab-scale he mochemical eac o es ing, sphe ical shape pa icles p esen lowe 119
oid ac ion and be e homogenei y wi hin he eac o bed and hus, i has been one o he 120
mos a ge ed ma e ial geome ies. Sp ay d ying is a commonly used me hod o make high-121
pe o mance, luidizable pa icles o chemical looping combus ion applica ions. Ib aheam e 122
al. s udied he e ec on he edox beha iou and sin e ing o sp ay-d ied pa icles o 123
manganese oxide wi h Al2O3, Z O2, and Fe2O3 (Ib aheam e al., 2019). All he ma e ials showed 124
e y poo mass change du ing edox cycling, since hey unde wen signi ican sin e ing du ing 125
calcina ion a 1200°C. Ne e heless, Z O2 and Al2O3 samples demons a ed enough s uc u al 126
s abili y unde he empe a u es es ed, al hough Al2O3 addi ion ends o o m a MnAl2O4 127
phase, which is s able a he es ed empe a u es. Wokon e al. s udied he kine ic 128
pe o mance o manganese-i on oxide g anules o 1-3 mm p epa ed by a build-up g anula ion 129
echnique wi hou any suppo o binde in a he mobalance (Wokon e al., 2017a). The 130
co esponding amoun s o Mn3O4 and Fe3O4 powde s we e mixed in an Ei ich mixe , based on 131
he p inciple o in ensi e mixing by an inclined a anged o a ing mixing pan, p o iding mixing 132
e ec in e ical and ho izon al di ec ions h ough he applica ion o a o a ing mic o-133
g anula o mixing ool. I was obse ed a signi ican pa icle olume inc ease a e 100 edox 134
cycles, which led o lowe densi y and mo e agile pa icles. This ac migh comp omise he 135
s abili y wi hin con inuous edox cycling a a eal scale and he e o e, hei mechanical 136
s abili y shall be enhanced. 137
Hamidi e al. syn hesized g anules o manganese-i on oxide by in ensi e mixing in an Ein ich 138
mixe , in oducing a mal odex in solu ion o ganic binde a he end o he mixing p ocess 139
(Hamidi e al., 2019). Pa icle sizes o 0.5 - 1 mm we e used o s udy he educ ion eac ion o 140
manganese-i on oxide in a small packed bed eac o con aining app oxima ely 35 g o he 141
he mochemical ma e ial. No da a ega ding he mechanical p ope ies and s abili y o he 142
pa icles unde cycling was p o ided. Ano he app oach was s udied by Gigan ino e al., who 143
modi ied a lab-scale g anula ion p ocess, known as d op echnique, o ob ain CuO pa icles o 144
1-2 mm o diame e (Gigan ino e al., 2020). The p ocess consis ed in dissol ing an o ganic 145
polyme wi h he ac i e ma e ial o c ea e a pas e ha is added d opwise o a ba h whe e he 146
immiscibili y leads o pa icle sphe ici y. Di e en combina ions o polyme s, sol en s and 147
su ac an s we e s udied, selec ing he bes combina ion which led o he highes pa icle 148
s eng h and sphe ici y, e en hough no da a ega ding he in luence on he eac ion kine ics 149
and s o age densi y was epo ed. In addi ion, in o de o educe he sin e ing e ec , he CuO 150
powde was mixed wi h a Y2O3/Z O2 s abilized powde , u ning in o less agglome a ion a e 151
100 edox cycles, o amoun s o Y2O3/Z O2 abo e 50 w %. 152
Conside ing ha adding a suppo ma e ial dec eases he olume ic ene gy s o age densi y 153
and may deac i a e he he mochemical ma e ial by undesi ed chemical eac ions, we p esen 154
a comp ehensi e s udy o he g anula ion echnique p oposed by Gigan ino e al., applied o 155
Si-doped manganese oxide. In a p e ious wo k we demons a ed he imp o ed kine ics and 156
chemical s abili y o Si-doped manganese oxide in powde s a e, ep esen ing a p omising 157
ma e ial candida e o echnology upscaling (Bielsa e al., 2020), which s o es and eleases 158
hea acco ding o he ollowing equa ion: 159
6(𝑀𝑀𝑀𝑀0.99𝑆𝑆𝑆𝑆0.01)2𝑂𝑂3(𝑠𝑠)+149 𝐽𝐽/𝑔𝑔 ↔ 4(𝑀𝑀𝑀𝑀0.99𝑆𝑆𝑆𝑆0.01)3𝑂𝑂4 (𝑠𝑠) + 𝑂𝑂2 (𝑔𝑔) Eq. 1 160
161
The in es iga ion ca ied ou in he p esen wo k aims o iden i y all he e ec s o he di e en 162
g anula ion syn hesis pa ame e s on he ma e ial beha io , paying special a en ion o he 163
chemical and mechanical s abili y and o inc ease he ac i e ma e ial con en o he g anules, 164
p o iding new insigh s in o a po en ial TcES ma e ial p epa a ion ou e o la ge scale packed 165
bed he mochemical eac o s. In he i s pa o he wo k, we e alua e he in luence o he 166
di e en syn hesis ou e pa ame e s on he ma e ial beha io , om he kine ics o he 167
chemical s abili y and mechanical beha io o he g anules. Subsequen ly, g anules exhibi ing 168
he mos p omising p ope ies we e selec ed o he upscaling s udy whe e 8 g o he ma e ial 169
was subjec ed i s o 50 h edox cycling s udy in a lab-scale packed bed eac o . Finally, a 170
sample o he es ed g anules we e subjec ed o addi ional 50 edox cycles in a STA o analyze 171
in de ail hei chemical s abili y a e a o al o 100 edox cycles. The sa is ac o y esul s 172
obse ed conce ning g anules mechanical and chemical s abili y ep esen a s ep o wa d in 173
high empe a u e he mochemical ene gy s o age upscaling, ha can be applicable no only 174
o he ma e ial unde s udy bu o o he me al oxides. 175
176
2. Expe imen al 177
178
2.1 Cha ac e iza ion echniques 179
180
The pa icles size and mo phology we e de e mined by means o a Quan a 200 FEG scanning 181
elec on mic oscope (SEM) ope a ed in high acuum mode a 20 kV and wi h a back sca e ed 182
elec on de ec o (BSED). 183
Reac ion kine ics and chemical s abili y we e s udied wi h he simul aneous he mal analyze 184
STA 449 F3 Jupi e (Ne zsch). In hese s udies, samples consis ing in 3-4 g anules accoun ing 185
o a ound 10 mg we e placed in o 85 µL open pla inum/ hodium c ucibles (Ne zsch) and 186
subjec ed o cha ging and discha ging cycles unde an ai s eam o 100 mL/min. 187
The bulk densi y and ue densi y o he di e en g anules we e measu ed using a helium 188
pycnome e AccuPyc II 1340. Fo he measu emen s a e e ence olume o 1 cm3 was 189
comple ely illed wi h he di e en composi ion g anules, esul ing in o al mass a ying om 190
0.1 o 0.3 g. 191
192
2.2 G anules syn hesis ou e 193
194
The syn hesis p ocess applied, based on he me hodology de eloped by Gigan ino e al. 195
(Gigan ino e al., 2020), consis s o he ollowing h ee s eps: i) syn hesis s ep, whe e he 196
p ecu so s a e ans o med in a ine powde o he desi ed ma e ial, ii) g anula ion s ep, 197
whe e sphe ical g anules a e p oduced, and iii) ha dening s ep, consis ing in a sin e ing 198
p ocess o enhance he mechanical s abili y o he g anules. 199
Fi s , Si-doped manganese oxide samples we e syn hesized by a sol-gel me hod, ollowing he 200
p ocedu e desc ibed in a p e ious wo k (Bielsa e al., 2021). The subsequen g anula ion 201
echnique is based on p epa ing a mix u e o his me al oxide powde and a solu ion o a 202
polyme ic binde and a sol en . Subsequen ly he mix u e is in oduced d opwise h ough a 203
sy inge in o a p ecipi a ing ba h, whe e he d ops ha den while he sol en lea es he g anule. 204
A e d ying, he g anules a e calcined o emo e he emaining o ganic ma e and o 205
enhance hei mechanical s abili y. As a consequence, he space occupied by he binde 206
becomes emp y and hen, po ous solid g anules a e ob ained. The ollowing wo condi ions 207
shall be me : he sol en mus be miscible wi h he p ecipi a ion ba h and he polyme ic 208
binde mus be soluble in he sol en and insoluble in he p ecipi a ion ba h. The p epa a ion 209
s eps a e illus a ed in Fig. 1. 210
211
Figu e 1. Me al oxide sphe ical g anules p epa a ion ou e 212
Va ious g anules o Si-doped Mn2O3 we e p epa ed a ying he p opo ions o he me al oxide 213
(MO), o ganic sol en (OS) and polyme ic binde (PB) wi h he aim o inding he bes 214
p epa a ion ou e gi ing sphe ical shape g anules wi h good mechanical and chemical 215
s abili y. The PB used was e hyl cellulose (Sigma Ald ich) and he OS 1-me hyl 2-py olidinone 216
(>99% Sigma Ald ich). The p ecipi a ion ba h was made o deionized wa e , mixed wi h a 217
su ac an (Tween 80, Sigma Ald ich) o educe i s high su ace ension and o p omo e he 218
sol en exi om he g anule. 219
I was ound ha o he amoun s o MO used (< 1 g), he a io OS:PB shall be main ained in 220
he ange 9:1. Lowe and highe OS:PB a io led o inconsis en g anules in he p ecipi a ion 221
ba h and a e he ha dening s ep, espec i ely. In p inciple, he amoun o MO in he g anules 222
is impo an because i migh a ec di ec ly o he ene gy s o age densi y, so high p opo ions 223
a e p e e ed. Ne e heless, high MO p opo ions (MO/MO+PB > 0.7) inc ease he iscosi y 224
o he mix u e esul ing in ile-like g anules (Fig. 2a), being no p ac ical o inc ease he MO 225
p opo ion abo e 0.8, since he iscosi y imposes signi ican di icul ies o he solu ion o low 226
ou o he sy inge. In o de o educe he iscosi y and main ain a high MO con en , he OS+PB 227
mix u e was hea ed a 40°C, con ibu ing o imp o e he sphe ici y o he g anules (Fig. 2b). 228
The dis ance be ween he p ecipi a ion ba h and he sy inge ip a ec ed he g anules in wo 229
ways: dis ances > 5cm esul ed in coin-shape g anules due o high mechanical shock when he 230
g anules eached he p ecipi a ion ba h su ace, and dis ances < 2 cm do no allow he 231
g anules o pene a e he ba h and emain loa ing in he su ace o some ime. The e o e, a 232
dis ance o 3 cm was selec ed o all he samples. The diame e o he sy inge ip has a s ong 233
in luence on he inal size o he g anules, which should be main ained below 1:10-20 o he 234
eac o diame e in o de o a oid gas channelling inside he he mochemical eac o 235
(Mede os e al., 2009). The e o e, a ip diame e o 2 mm was used, esul ing in g anules 236
wi hin he a e age diame e ange o 3-4 mm a e s ep 2. Remo ing he wa e du ing he 237
d ying s ep became c i ical, since i was obse ed ha he wa e con en a e he syn hesis 238
p ocess may each up o 80 w % o he o al g anule mass and i s ex ac ion can des oy he 239
g anule sphe ici y, unless i is ca ied ou slowly. Fo ha eason, he g anules we e d ied a 240
oom empe a u e un il he wa e con en was almos o ally emo ed. Subsequen ly, he 241
g anules we e subjec ed o a empe a u e p og am consis ing in hea ing up o 450°C a 242
1°C/min, ollowed by an iso he mal s ep o 4 h o emo e he o ganic ma e (Fig. 2c). The 243
inal ha dening s ep consis ed in ano he hea ing s ep a 2°C/min up o he co esponding 244
calcina ion empe a u e, whe e di e en iso he mal s eps we e used o p omo e pa icle 245
sin e ing. A no iceable g anule sh inkage was obse ed du ing he wa e and o ganic ma e 246
emo al s eps, being g ea e he ewe he con en o MO, esul ing in g anules wi h a inal 247
size o a ound 1-3 mm o diame e . Table 1 shows he di e en ma e ials p epa ed in his 248
wo k. 249
a) b) c) 250
251
Figu e 2. G anules o Mn2O3 p epa ed by he d op echnique: a) p onounced ile-like g anules, b) 252
sphe ical g anules and c) g anules a e he calcina ion s ep a 450°C 253
Table 1. Desc ip ion o ma e ials p epa ed in his wo k 254
Desc ip ion
MO P opo ion
(w %)*
Diame e
(mm)
Powde
1
N/A
MO 50
50
1,83
MO 70
70
2,34
MO 80
80
2,46
*MO p opo ion (w %) = MO/(MO+PB) 255
256
2.3 Expe imen al se up 257
258
The expe imen al es ig consis s o wo main pa s: he hea ing se up and he 259
he mochemical eac o (Fig.3a). The gene a ion o a con olled gas low a e abo e 800°C 260
equi ed o ca y ou he educ ion and oxida ion o he ma e ial may become a complex issue 261
and he e o e, a u nace placed a ound he he mochemical eac o ca ies ou he 262
hea ing/cooling p ocess. The u nace se up empe a u e was a ied be ween 550°C and 800°C 263
a a hea ing/cooling a e o 20°C/min and he eac ions we e conduc ed swi ching he gas 264
low be ween N2 and O2. The he mochemical eac o consis s on an AISI 304 me al ube o 265
15 mm o inne diame e and 400 mm leng h. A ound 8 g o MO we e placed in he middle o 266
he ube leng h. Silicon ca bide pa icles we e placed below and abo e he MO, illing 267
comple ely he eac o , p e en ing any mo emen o luidiza ion o he MO. Once he gas 268
s eam c osses he eac o , i is cooled down in a wa e ba h and subsequen ly, he oxygen 269
con en is measu ed by means o a zi conia gas analyze KCD-ON320 (Senso s ecnics). The 270
eac o includes i e empe a u e measu ing poin s a di e en heigh s: T2, T3 and T4 placed 271
inside he eac o a he posi ions indica ed in Fig. 3a and T2ou and T4ou placed on he ex e io 272
su ace o he eac o a he same heigh o T2 and T4, espec i ely. The signals we e collec ed 273
wi h an acquisi ion sys em ype ABSD-MD832-81-23-HLP (Yun unyn) and egis e ed by a 274
homemade Lab iew app. The comple e se -up is illus a ed on Fig. 3b. 275
a) b) 276
277
Figu e 3. The mochemical expe imen al se up: a) componen s diag am and b) se up pho og aphy 278
279
280
281
3. Resul s and discussion 282
283
3.1 E ec s o he syn hesis pa ame e s on he g anules beha iou 284
285
A comp ehensi e s udy was made o assess he in luence o he g anule con o ma ion, he 286
MO con en and he ha dening s ep on he chemical and mechanical beha io o Si-doped 287
In o de o assess he mechanical capabili y o he di e en calcined g anules o wi hs and he 448
mechanical s esses imposed by he lab-scale packed bed eac o se -up, a laye composed o 449
3 o he di e en se o g anules was submi ed o an inc eased weigh h ough a la pla e 450
un il ailu e. The c ushing s eng h esul s a e shown in Table 2 and shall be aken only as a 451
e e ence alue o he pa icula con igu a ion es ed in he p esen wo k. To ob ain an 452
accu a e alue o he c ushing s eng h speci ic lab de ice should be used (e. g. s anda d 453
comp ession es ing machine). 454
Table 2. C ushing s eng h o he g anules calcined a di e en empe a u e p og ams 455
Sample id
Ha dening p og am
Maximum weigh
C ushing s eng h
MO 50
1050°C 4h
38 g
0,126 N
MO 50
1050°C 8h
44 g
0,146 N
MO 50
1100°C 4h
50 g
0,166 N
456
457
3.2.2 Selec ion o he g anules o he packed bed eac o expe imen s 458
459
The eac o consis ed on a s ainless-s eel ube o 13 mm o in e nal diame e and 90 mm 460
heigh , u ning in o an e ec i e olume o 11,9 cm3. In o de o de e mine he ene gy densi y 461
con ained in his olume i is necessa y o measu e he ac ual me al oxide con en in he 462
g anules and he packed bed oid ac ion. The me al oxide con en in he g anules was 463
de e mined using a e e ence olume o 1 cm3. This olume was illed wi h g anules o he 464
di e en MO:PB a ios used and he bulk densi y and ue densi y we e measu ed using a 465
helium pycnome e AccuPyc II 1340. I was obse ed ha he maximum MO in he g anules 466
eached only 16.94%, likely caused by a de icien dissolu ion o he OS in he p ecipi a ion 467
ba h. In o de o ge an inc ease on he MO con en ano he ba ch o g anules was p epa ed 468
dec easing he empe a u e o he ba h by using ice in o de o educe he solubili y o he PB 469
(MO 50-c). The esul ing g anules go a signi ican MO con en inc ease. The esul s a e 470
showed in Table 3. 471
Table 3. Samples p ope ies measu emen o e a e e ence olume o 1 cm3 472
Desc ip ion
Mass g anule
g
Diame e
mm
Bulk densi y
g/cm
3
T ue densi y
g/cm
3
MO con en
%
Powde
N/A
N/A
0,3345
5,3236
N/A
MO 50
0,0023
1,835
0,735
5,116
14,38
MO 70
0,0055
2,342
0,828
5,0047
16,56
MO 80
0,0066
2,466
0,846
5,253
16,94
MO 50-c
0,002
1,396
1,402
5,653
24,81
473
Conside ing he size o he lab-scale eac o and he g anules wi h he highes bulk 474
densi y, he maximum amoun o ma e ial ha can be in oduced in he eac o wi hou 475
applying p essu e was a maximum o 8 g. The e o e, acco ding o he da a p o ided in Table 476
2, he h ee ha dening p og ams p o ide o he g anules wi h enough mechanical s abili y o 477
wi hs and he mechanical s ess imposed by he weigh o he packed bed. Consequen ly, he 478
g anules wi h he highes MO con en (MO 50-c) calcined a 1050°C o 4h, which in addi ion 479
p esen a mo e homogeneous sphe ici y, we e selec ed and syn hesized o u he cycling 480
es s in he lab-scale eac o (Fig. 9). 481
482
483
Figu e 9. Mn2O3 Si-doped g anules 50:50 (MO:PB) o eac o es ing: a) A e s ep 2 o he syn hesis 484
ou e and b) A e ha dening a 1050°C o 4 h. 485
486
3.2.3 Packed bed eac o expe imen al esul s 487
488
A summa y o he main eac o pa ame e s is p esen ed in Table 4. 489
Table 4. Lab-scale eac o pa ame e s 490
Desc ip ion
Value
Reac o adius
13 mm
Reac o leng h
90 mm
Pa icle diame e (dp)
1,39 mm
Reac o mass
7,96 g
Bulk densi y
1,402 g/cm3
Void ac ion (ɛ)
0,59
Ma e ial s o age densi y
149 J/g (Bielsa e al., 2021)
S o age capaci y
0,33 Wh
491
In o de o e alua e he chemical and mechanical s abili y o he g anules in a packed bed 492
a angemen , he eac o was subjec ed o 50 edox cycles consis ing in o : i) a hea ing s ep 493
up o 800°C a 20°C/min unde 100 mL/min o N2, ii) an iso he mal s ep a 800°C o 20 min o 494
ensu e comple e educ ion o he ma e ial, since in a p e ious s udy, he educ ion onse 495
empe a u e in N2 was iden i ied a app oxima ely 775°C (Bielsa e al., 2021) , and iii) a cooling 496
s ep down o 550°C wi h a cooling a e o 20°C/min unde a gas low o 10%/90% (N2/O2). In 497
he same wo k i was obse ed ha in such condi ions, oxida ion akes place in less han 5 498
min, so comple e ma e ial oxida ion is expec ed du ing he cooling s ep, be o e s a ing he 499
ollowing edox cycle. The empe a u e e olu ion a di e en poin s o he eac o and he 500
oxygen con en in he gas s eam lowing ou o he eac o we e eco ded by he 501
ins umen a ion du ing he 50 edox cycles and a e plo ed in Fig. 10a. Fo be e 502
unde s anding, he same pa ame e s comp ising only he 3 d and 4 h cycles a e plo ed in Fig. 503
10b. 504
505
In ega d o he educ ion eac ion, i is di icul o ob ain use ul in o ma ion om he 506
empe a u e p o iles o Fig. 10b, since i seems ha educ ion akes place when he u nace 507
eaches he empe a u e se poin and educes he powe inpu , which is ollowed by an 508
ab up empe a u e slowdown ha makes i di icul o dis inguish he e ec o he 509
endo he mic educ ion eac ion in he empe a u e p o iles o he in e nal he mocouples. 510
The eac ion occu ence can be asce ained om he oxygen a ia ion measu emen , since an 511
inc ease o he oxygen con en in he gas s eam eaching a peak o 5% was de ec ed. 512
Ne e heless, he ex en o he eac ion canno be e alua ed, since in eg a ing he a ea unde 513
he cu e, i esul s in a alue eaching only 6% o he heo e ical oxygen elease expec ed, 514
conside ing a heo e ical mass loss o 3%. In he case o he oxida ion eac ion, he in e nal 515
eco ded empe a u es (T2, T3 and T4) p esen a sligh ly lowe cooling a e han he 516
empe a u es eco ded ou side he eac o (T2ou and T4ou ) a he beginning o he cooling 517
s ep. This ac can be caused by he exo he mic na u e o he oxida ion eac ion, e en hough 518
he s a ing and inishing poin o he eac ion canno be dis inguished. Rega ding he oxygen 519
concen a ion, a sligh a ia ion du ing he oxida ion eac ion can be obse ed. The exac 520
oxygen dec ease canno be accu a ely de e mined as a consequence o he poo senso 521
esolu ion. Howe e , bo h e idences sugges s ha du ing he hea ing s ep educ ion o he 522
ma e ial is aking place, while in he cooling s ep he ma e ial is being oxidized. 523
524
The mo phological e olu ion and physical in eg i y o he g anules a e he 50 edox cycles in 525
he eac o was analyzed bo h by isual inspec ion (Fig. 10c) and SEM (Fig. 10d). None o he 526
g anules p esen ed he eddish colo ypical o he educ ion phase and hus, showing 527
incomple e oxida ion, main aining hei sphe ical shape a e emo ing hem om he 528
eac o . As can be obse ed om he SEM image, he pa icles ha e no su e ed om 529
no iceable sin e ing, since hei size emain simila o hei ini ial sa e a e he ha dening s ep 530
(Fig. 8a). The e o e, no e e sibili y loss would be expec ed. To con i m his s a emen h ee 531
g anules o he al eady cycled ma e ial in he eac o we e subjec ed o an addi ional 50 edox 532
cycles in he STA, using he same empe a u e p og am, comple ing a o al p og am o 100 533
edox cycles. The mass loss/gain du ing he STA p og am oge he wi h a SEM image o he 534
pa icles a e he 100 edox cycles a e shown in Fig. 11a and Fig. 11b, con i ming no loss o 535
he chemical s abili y o he g anules wi h no no iceable change on he pa icle mo phology. 536
537
a) b) 538
539
c) d) 540
541
542
543
Figu e 10. Si-doped Mn2O3 g anules edox cycling in he he mochemical eac o : a) he mocouples 544
signals eco d du ing he whole 50 edox p og am, b) magni ica ion o he he mocouple signals and 545
oxygen senso du ing he 3 d and 4 h edox cycle, c) Si-doped g anules oge he wi h SiC pa icles 546
appea ance a e cycling and emo ing om he eac o , and d) SEM image showing he 547
mo phological e olu ion a e he 50 h edox cycle 548
549
a) b) 550
551
Figu e 11. Si-doped Mn2O3 g anules subjec ed o 50 addi ional edox cycles in he STA: a) mass 552
loss/gain eco ded in he STA, and b) SEM pic u e a e he addi ional 50 edox cycles 553
554
4. Conclusions 555
In his wo k, a g anule p epa a ion ou e was s udied o he mochemical ene gy 556
s o age upscaling using a no el Si-doped manganese oxide o concen a ed sola powe 557
plan s. The esea ch wo k comp ises ma e ial p epa a ion and he mochemical pe o mance 558
e alua ion in a lab-scale packed bed eac o . The p ocess succeeded in ob aining sphe ical 559
po ous g anules o 1-2 mm wi h di e en ac i e ma e ial con en . The esul s iden i y he 560
c i ical pa ame e s o he syn hesis p ocess which p o ide he bes mechanical and chemical 561
s abili ies o he g anules in o de o be used in a he mochemical packed bed eac o . I was 562
obse ed ha he ha dening s ep aimed o inc ease he mechanical s abili y o he g anules 563
does no a ec signi ican ly hei chemical s abili y. This ac con i ms ha he se e i y o he 564
ha dening p ocess could be u he inc eased, leading o mo e mechanically s able g anules, 565
equi ed o la ge scale eac o s. Fu he mo e, we obse ed ha dec easing he solubili y o 566
- 10 µm ---
- 10 µm ---
he polyme ic binde in he syn hesis ba h almos double he ac i e ma e ial con en in he 567
g anules and consequen ly, enhances he olume ic ene gy s o age capaci y. We achie ed 568
an ac i e ma e ial con en o 24.81%, and his esul sugges s ha he e is a way o keep 569
inc easing he ene gy s o age densi y o he g anules, which need o be u he de eloped. In 570
addi ion, 8 g o Si-doped g anules we e subjec ed o 50 edox cycles in a lab-scale packed bed 571
eac o , showing sa is ac o y mechanical and chemical s abili y, which was con i med o e 572
addi ional 50 edox cycles in a he mobalance, wi h comple e e-oxida ion o e he whole 573
p og am. In summa y, bo h he Si-doped manganese oxide and he g anule p epa a ion ou e 574
conduc ed ha e demons a ed enough eliabili y o be used on a la ge scale and hus, 575
con ibu e o push he high empe a u e he mochemical s o age echnology o a nex 576
gene a ion o concen a ed sola powe plan s. 577
578
579
Acknowledgemen s 580
This wo k has been suppo ed by he Depa men o Economic De elopmen and 581
In as uc u es o he Basque go e nmen , h ough he unding o he ELKARTEK CIC 582
Ene gigune-2017 esea ch p og am. The au ho s exp ess hei since e hanks o C is ina 583
Luengo, Mikel In xau ie a and Yagmu Pola o hei echnical suppo . 584
585
Bibliog aphy 586
Abián, M., Abad Secades, A., Izquie do Pan oja, M.T., Gayán Sanz, P., Diego Poza, L.F., Ga cía 587
Labiano, F., Adánez Elo za, J., 2017. Ti anium subs i u ed manganese- e i e as an 588
oxygen ca ie wi h pe manen magne ic p ope ies o chemical looping combus ion 589
o solid uels. Fuel 195, 38–48. 590
Ag a io is, C., Becke , A., Roeb, M., Sa le , C., 2016. Exploi a ion o he mochemical cycles 591
based on solid oxide edox sys ems o he mochemical s o age o sola hea . Pa 5: 592
Tes ing o po ous ce amic honeycomb and oam cascades based on cobal and 593
manganese oxides o hyb id sensible/ he mochemical hea s o age. Sola Ene gy 139, 594
676–694. 595
Ag a io is, C., Roeb, M., Sa le , C., 2015a. Hyb id sensible/ he mochemical sola ene gy 596
s o age concep s based on po ous ce amic s uc u es and edox pai oxides chemis y. 597
Ene gy P ocedia 69, 706–715. 598
Ag a io is, C., Roeb, M., Schmuecke , M., Sa le , C., 2015b. Exploi a ion o he mochemical 599
cycles based on solid oxide edox sys ems o he mochemical s o age o sola hea . 600
Pa 2: Redox oxide-coa ed po ous ce amic s uc u es as in eg a ed he mochemical 601
eac o s/hea exchange s. Sola Ene gy 114, 440–458. 602
Azimi, G., Leion, H., Ma isson, T., Rydén, M., Snijke s, F.M.M., Lyng el , A., 2014. Mn–Fe 603
Oxides wi h Suppo o MgAl2O4, CeO2, Z O2 and Y2O3–Z O2 o Chemical-Looping 604
Combus ion and Chemical-Looping wi h Oxygen Uncoupling. Ind. Eng. Chem. Res 53, 605
10358–10365. 606
Azimi, G., Ma isson, T., Leion, H., Rydén, M., Lyng el , F.M.M., 2015. Comp ehensi e s udy o 607
Mn–Fe–Al oxygen-ca ie s o chemical-looping wi h oxygen uncoupling (CLOU). 608
In e na ional Jou nal o G eenhouse Gas Con ol 34, 12–24. 609
Bielsa, D., Zaki, A., A ias, P.L., Faik, A., 2021. De elopmen o a kine ic eac ion model o 610
educ ion and oxida ion o Si doped Mn2O3 o he mochemical ene gy s o age in 611
concen a ed sola powe plan s. Jou nal o Ene gy S o age 43, 103271. 612
Bielsa, D., Zaki, A., A ias, P.L., Faik, A., 2020. Imp o ing he edox pe o mance o 613
Mn2O3/Mn3O4 pai by Si doping o be used as he mochemical ene gy s o age o 614
concen a ed sola powe plan s. Sola Ene gy 204, 144–154. 615
Bielsa, D., Zaki, A., Faik, A., A ias, P.L., 2019. E iciency imp o emen o Mn2O3/Mn3O4 616
edox eac ion by means o di e en ope a ion s a egies. 617
h ps://doi.o g/10.1063/1.5117750 618
Ca illo, A.J., Se ano, D., Piza o, P., Co onado, J.M., 2014. The mochemical hea s o age 619
based on he Mn2O3/Mn3O4 edox couple: in luence o he ini ial pa icle size on he 620
mo phological e olu ion and cyclabili y. J. Ma e . Chem. A 2, 19435–19443. 621
Ca illo, A.J., Se ano, D.P., Piza o, P., Co onado, J.M., 2015. Imp o ing he The mochemical 622
Ene gy S o age Pe o mance o he Mn2O3/Mn3O4 Redox Couple by he Inco po a ion 623
o I on. ChemSusChem 8, 1947–1954. 624
Gigan ino, M., Sas B unse , S., S ein eld, A., 2020. High-Tempe a u e The mochemical Hea 625
S o age ia he CuO/Cu2O Redox Cycle: F om Ma e ial Syn hesis o Packed-Bed Reac o 626
Enginee ing and Cyclic Ope a ion. Ene gy & Fuels 34, 16772–16782. 627
Hamidi, M., Bayon, A., Wheele , V.M., K eide , P., Wallace, M.A., Tsuzuki, T., Ca chpole, K., 628
Weime , A.W., 2019. Reduc ion kine ics o la ge sphe ical 2:1 i on–manganese oxide 629
edox ma e ials o he mochemical ene gy s o age. Chemical Enginee ing Science 201, 630
74–81. 631
Ib aheam, A.S., Eh ha , B.D., Wa d, B.J., Bayon, A., Wallace, M.A., Bade , R., K eide , P., 632
Weime , A.W., 2019. Pa icle design and oxida ion kine ics o i on-manganese oxide 633
edox ma e ials o he mochemical ene gy s o age. Sola Ene gy 183, 17–29. 634
Ka agiannakis, G., Pagkou a, C., Hale as, E., Bal zopoulou, P., Kons andopoulos, A.G., 2016. 635
Cobal /cobal ous oxide based honeycombs o he mochemical hea s o age in u u e 636
concen a ed sola powe ins alla ions: Mul i-cyclic assessmen and semi-quan i a i e 637
hea e ec s es ima ions. Sola Ene gy 133, 394–407. 638
Mede os, F.S., Anchey a, J., Chen, J., 2009. Re iew on c i e ia o ensu e ideal beha io s in 639
ickle-bed eac o s. Applied Ca alysis A: Gene al 355, 1–19. 640
Neises, M., Tesca i, S., de Oli ei a, L., Roeb, M., Sa le , C., Wong, B.Y., 2012. Sola -hea ed 641
o a y kiln o he mochemical ene gy s o age. Sola Ene gy 86, 3040–3048. 642
Neumann, N.C., Block, T., Linde , M., Leion, H., 2018. S abilizing pa icles o manganese-i on 643
oxide wi h addi i es o he mochemical ene gy s o age. Ene gy Technology 6. 644
Pagkou a, C., Ka agiannakis, G., Zygogianni, A., Lo en zou, S., Kons andopoulos, A.G., 2015. 645
Cobal oxide based honeycombs as eac o s/hea exchange s o edox 646
The mochemical Hea S o age in Fu u e CSP Plan s. Ene gy P ocedia 69, 978–987. 647
Pelay, U., Luo, L., Fan, Y., S i ou, D., Cas elain, C., 2019. In eg a ion o a he mochemical 648
ene gy s o age sys em in a Rankine cycle d i en by concen a ing sola powe : Ene gy 649
and exe gy analyses. Ene gy 167, 498–510. 650
Pes alozzi, F., 2013. Expe imen al In es iga ion o Mn-Based Redox Reac ion Kine ics o 651
The mochemical Ene gy S o age. 652
Schmid , M., Linde , M., 2017. Powe gene a ion based on he Ca(OH)2/ CaO 653
he mochemical s o age sys em – Expe imen al in es iga ion o discha ge ope a ion 654
modes in lab scale and co esponding concep ual p ocess design. Applied Ene gy 203, 655
594–607. 656
Schoen elde , C.W., Swishe , J.H., 1973. Kine ics o The mal Decomposi ion o TiH2. Jou nal 657
o Vacuum Science and Technology 10. h ps://doi.o g/10.1116/1.1318443 658
Singh, A., Tesca i, S., Lan in, G., Ag a io is, C., Roeb, M., Sa le , C., 2017. Sola 659
he mochemical hea s o age ia he Co3O4/CoO looping cycle: S o age eac o 660
modelling and expe imen al alida ion. Sola Ene gy 144, 453–465. 661
S öhle, S., Haselbache , A., Jo ano ic, Z.R., S ein eld, A., 2016. The E ec o he Gas-Solid 662
Con ac ing Pa e n in a High-Tempe a u e The mochemical Ene gy S o age on he 663
Pe o mance o a Concen a ed Sola Powe Plan . Ene gy & En i onmen al Science 9, 664
1375–1389. 665
Va sano, F., Al ani, C., la Ba be a, A., Masi, A., Padella, F., 2016. Li hium manganese oxides 666
as high- empe a u e he mal ene gy s o age sys em. The mochimica Ac a 640. 667
h ps://doi.o g/10.1016/j. ca.2016.07.018 668
Wokon, M., Block, T., Nicolai, S., Linde , M., Schmücke , M., 2017a. The modynamic and 669
kine ic in es iga ion o a echnical g ade manganese-i on bina y oxide o 670
he mochemical ene gy s o age. Sola Ene gy 153, 471–485. 671
Wokon, M., Kohze , A., Linde , M., 2017b. In es iga ions on he mochemical ene gy s o age 672
based on echnicalg ade manganese-i on oxide in a lab-scale packed bed eac o . Sola 673
Ene gy 153, 200–214. 674
Wong, B., 2011. THERMOCHEMICAL HEAT STORAGE FOR CONCENTRATED SOLAR POWER. 675
Golden, CO (Uni ed S a es). h ps://doi.o g/10.2172/1039304 676
Zsembinszki, G., Sole, A., Ba eneche, C., P ie o, C., Fe nandez, A.I., Cabeza, L.F., 2018. 677
Re iew o Reac o s wi h Po en ial Use in The mochemical Ene gy S o age in 678
Concen a ed Sola Powe Plan s. Ene gies 11, 2358. 679
680
