UNIVERSITA’ DEGLI STUDI DI PARMA
DOTTORATO DI RICERCA IN
“Scienza e Tecnologia dei Ma e iali”
CICLO XXXIV
Polyaniline-based neu omo phic de ices
owa ds in e acing sensing
Coo dina o e:
P o . En ico Dalcanale
Tu o e:
D . Sal a o e Ianno a
Co-Tu o e:
D . Pasquale D’Angelo
Do o ando: Roman Sajapin
Anni academici: 2018/19 – 2021/22
No e o hanks
Be o e all, I would like o hank D . Sal a o e Ianno a o he assignmen o his challenging
and in e es ing opic, he us he placed in me o accomplish he goals ha ha e been se , and he
g ea oppo uni y o lea n and g ow beyond mysel in he p ocess. I am e y g a e ul o ou ui ul
collabo a ion and admi e his eadiness o cons uc i e discussions and he abili y o ind app oaches
o upcoming challenges. I lea ned a g ea deal abou scien i ic wo k and his guidance h oughou his
pe iod has been in aluable. I el and enjoyed a pleasan amoun o mu ual app ecia ion o he wo k
we ha e done which uly mo i a ed me all along.
A huge hank you goes o D . Pasquale D’Angelo o his willingness o co-supe ise my
hesis, he many discussions we had and he many aluable sugges ions I ecei ed om him. Some
o he mos in e es ing ideas we e bo n du ing ou con e sa ions, sha ing hough s and doub s alike.
I will always app ecia e his upbea demeanou ha was e y encou aging and a needed sou ce o
op imism o me, especially du ing he las yea o my PhD p og am.
D . Ta iana Be zina dese es my bigges g a i ude as she in oduced me o ou exci ing
esea ch ield and p ac ically showed me e e y hing he e was o know abou he ma e ials and
echniques we wo ked wi h. I will emembe he imes when we wo ked oge he as some o he mos
exci ing ones. The wo k wi h Ta iana augh me he alue o being sel -c i ical and main aining a
heal hy amoun o doub abou ob ained esul s. A he same ime, i augh me o s and by my own
con ic ions, e en i hey a e no ini ially me wi h app o al. Wi hou a doub , my hesis would no
ha e been he same wi hou he in ol emen .
I would like o hank all he o he membe s o ou esea ch g oup a IMEM and he Physics
depa men , D . Vic o E okhin, D . S e lana E okhina, D . Sil ia Ba is oni, D . Valen ina Ricci, D .
Regina Bu gano a, D . Adelia Faizullina, D . S e ania Boi, D . Da ide Vu o and D . Giuseppe
Ta abella. I was an en ichmen o wo k and exchange opinions wi h hem and I ha e lea ned lo s o
aluable lessons o e he cou se o hese h ee yea s, no only abou science, bu li e in gene al. A
wa m hank you o S e ania o ou much app ecia ed, encou aging con e sa ions. As I am su e mos
who wo ked wi h he would ag ee, he pe iod when she joined ou g oup el e y e eshing.
Special hanks go o he adminis a i e o IMEM and he Physics depa men and he
doc o al o ice o making my wo k a lo mo e ouble- ee. I also hank D . Da ide Vu o once again
o his help in ope a ing he AJP and SEM as well as D . Ma eo Cocuzza and D . Simone Ma asso
o p o iding he pho oli hog aphed elec ode samples. I hank D . Milad Takhsha o he much-
app ecia ed help wi h AFM image y. I am g a e ul o Sal a o e Van aggio om he g oup o P o .
An onella Pa isini as well as D . Ma co Villani, D . Ricca do Man edi and D . And ea Sala om
IMEM o being eady o help wi h one o he o he small hing ha I needed.
I hank D . Mi ko Bu ini and P o . Ad iana Calde a o om he Depa men o Medicine
and Su ge y o he oppo uni y o acqui e he MALDI-TOF spec a. I also hank D . Sil ia D’Au ia
o he eco ded IR spec a and o he membe s o he g oup o P o . En ico Dalcanale who ea ed me
e y kindly and made me eel welcome in he Chemis y depa men . Special hanks a e due o P o .
Dalcanale himsel o gene ously p o iding a ew aluable eagen s and, mo e impo an ly, o his
eaching and he aluable scien i ic ad ice and encou agemen I ecei ed o e hese yea s.
I would like o exp ess my deepes g a i ude o D . Tobias G ube o he mo i a ion and
suppo he ga e me be o e and du ing my doc o a e. I am g a e ul o he educa ion I ecei ed du ing
my unde g adua e s udies a he Uni e si y o F eibe g. Despi e ha ing missed ou on one o he o he
lec u e o his (which Tobias likes o emind me o some imes!), I conside mysel lucky o ha e been
his s uden . He laid he ounda ion o my iews on science and i is sa e o say ha my in e es in
esea ch was awoken du ing my bachelo hesis unde his supe ision.
A huge hanks is due o all my old iends and he new ones I made along he way. They
ha e been a e y welcome dis ac ion ha I needed a imes and hei mo al suppo has kep my spi i
up. Especially I would like o men ion Igo , who a imes seemed o be e en mo e emo ionally
in es ed in my doc o a e han mysel . I hope one day I can make up o he imes when I couldn’ be
as p esen as I wish I was.
No wo ds can desc ibe my g a i ude o my amily. I am glad you don’ expec any hing o
be said because I don’ qui e know how o exp ess i . Thank you o you unde s anding, pa ience,
and o esigh . Thank you o you uncondi ional suppo and o always being he e when I needed
you, e en a imes when we disag eed on hings. Спасибо вам за всё.
Finally, I am g a e ul o hose who mo i a ed me o ake his s ep. I was a ich sou ce o
pe sonal g ow h and p obably he bes decision I ha e e e made so a . Wi hou you, none o his
would ha e happened. I’m glad i did.

I
Abs ac
Mem is o s a e elec onic elemen s ha belong o a new gene a ion o compu a ional
sys ems wi h a g ea po en ial o con ibu e o he p omising e i al o scien i ic esea ch dedica ed
o he ha dwa e ealiza ion o A i icial Neu al Ne wo ks (ANNs) owa ds A i icial In elligence (AI)
and compu e -b ain in e aces. A dis inc i e p ope y o hese de ices is he dependence o hei
in e nal esis ance on he elec ical cha ge ha passed h ough hem. In o he wo ds, he e is a
co ela ion o he ou pu cha ac e is ics o he de ice and he his o y o i s use, in which i esembles
biological synapses and can be conside ed as hei a i icial analogue, displaying spike ime
dependen plas ici y (STDP). Fu he mo e, he abili y o con olled swi ching be ween di e en
in e nal esis i e s a es makes mem is o s some o he mos p omising candida es o he
implemen a ion in memo ies. Con a y o he con en ional Von Neumann a chi ec u e, ha dwa e-
ealized ANNs ha e he po en ial o combining he s o age and p ocessing o in o ma ion, ca ied ou
by he same kind o elemen s, mimicking biological neu ons in he b ain. Pa allel in o ma ion
p ocessing would allow o simul aneously wo k wi h a whole a ay o inpu s, a he han ca ying ou
one single ope a ion a a ime. These p ope ies pa e he way owa ds mo e ene gy- and ime-e icien
compu ing by a oiding he need o exchange in o ma ion be ween he p ocesso and a passi e
memo y. The simila i y o biological synapses sugges s an excellen biocompa ibili y and he
possibili y o emula e some unc ionali ies o biological sys ems, enabling compu e -b ain in e aces
wi h a seamless ans o ma ion and p ocessing o bioelec onic signals.
The O ganic Mem is i e De ice (OMD) is a polyme -based ep esen a i e o such elemen s.
I is a wo- e minal de ice ea u ing a conduc i e channel o polyaniline (PANI) whose esis ance is
modula ed h ough elec ochemically con olled ansi ions be ween he polyme ’s insula ing and
conduc i e s a e. These ansi ions ake place in a he e ojunc ion o PANI and a polye hylene oxide
(PEO)-based solid polyelec oly e (SPE) doped wi h a sou ce o chlo ide among o he s abili y
enhancing addi i es, p omo ing he eac ion wi h a sil e coun e elec ode. I s ad an ages wi h
espec o o he mem is o s a e i s low-cos ab ica ion and he abili y o ine- uning o he channel
esis ance, g an ed by accessing in e media y esis i e s a es.
The main goal o he p esen PhD hesis is o de elop OMDs sui able o neu omo phic
applica ions such as he in e acing sensing and signal p ocessing ha can be ealized by means o
mul ilaye pe cep on s uc u es. The combina ion o a la ge a ay o elemen s in one ne wo k as well
as wo king wi h complex biological sys ems implies he occu ence o elec ical noise, which poses
he ques ion how i in e e es wi h he unc ioning o ou de ice. P elimina y esea ch has shown ha
he cu en -s a e OMD does no possess he necessa y le el o s abili y o eliably o ca y ou such
sensi i e expe imen s. Hence, a signi ican pa o ou esea ch was de o ed o he op imiza ion o
he ma e ials employed in OMD ab ica ion. The goal is o ad ance owa ds de ices wi h highe
endu ance, i. e., ep oducible ou pu cha ac e is ics o e longe pe iods o ime, and o examine how
o design he ab ica ion echniques o imp o e biocompa ibili y.
II
A signi ican pa o ou wo k was ocused on he s abiliza ion o he labile, polye hylene
oxide-based SPE sys em by op imizing i s composi ion wi h pa icula a en ion o he dopan sal s.
This wo k ollows and builds upon an ex ensi e heo e ical s udy o he ole o he single dopan ion
species on he ope a ion mechanism o he OMD. I was demons a ed ha he o me concep which
a ibu es a signi ican ole in he swi ching p ocess o li hium ion doping is inaccu a e. Ins ead, i is
shown ha he anions a e he mos c i ical dopan species in e ms o de ice ope a ion. Fu he mo e,
h ough a combina ion o heo e ical and ex ensi e p ac ical wo k, no el ecipes o PEO-based SPEs
ha e been de eloped, p o iding unp eceden ed sho - e m and long- e m s abili y wi h a ema kable
ep oducibili y o I-V-cha ac e is ics, imp o ing he endu ance o he OMD by up o wo o de s o
magni ude. A signi ican pa o his success is due o he implemen a ion o aluminium chlo ide o
he SPE sys em, p o iding in insic acidi y and enabling o a oid doping by s ong, ola ile acids
such as HCl. This disco e y was ollowed up by in ensi e esea ch o ou es o s abilize he PEO-
based gels and impa hem wi h a ou able p ope ies o long- e m endu ance by p e en ing
agg ega ion. Du ing his esea ch, he o me ly unheeded concep o lyo opici y was in oduced o
OMDs, ecognizing i as a majo aspec o he s abili y o he polyelec oly e sys em. Fu he mo e,
silk ib oin solu ions ha e been success ully employed o he i s ime as an al e na i e o PEO-
based SPEs, u he ad ancing p io esea ch owa ds he in oduc ion o biocompa ible ma e ials
in o OMD ab ica ion.
Ano he impo an aspec o he implemen a ion o OMDs in complex sys ems o he
in e acing o sensing and signal p ocessing is he in eg abili y wi h o he elec onic sys ems such as
OECT-based senso s. This can be mos e ec i ely achie ed by shi ing he pa adigm o OMD
manu ac u ing om la gely manual owa ds high-p ecision, au oma ed ab ica ion. To his end, we
de eloped ma e ials ha can be applied by means o di ec -w i ing echniques such as Ae osol Je
P in ing (AJP). Pa icula a en ion was paid o he conduc i e polyme ic componen o he de ice.
Success ul a emp s ha e been made by de eloping AJP-compa ible inks based on p is ine PANI in
a mix u e o o ganic sol en s such as NMP, xylene and isop opanol, ha we e success ully applied
on qua z and Si/SiO2 subs a es. Fu he mo e, a wa e dispe sible composi e ma e ial o PANI and
chi osan was syn hesized, u he imp o ing he quali y o he p in ed ea u es and he co e age o
he subs a es. Bo h p in ed ma e ials ha e been implemen ed o success ully ab ica e he i s
unc ioning PANI-based OMDs ea u ing a p in ed conduc i e channel, displaying he cha ac e is ic
p ope ies o a mem is o .
The combina ion o all he expe imen al esul s and in ensi e heo e ical wo k lays he
ounda ion and pa es he way owa ds he implemen a ion o OMDs in ad anced, complex,
biocompa ible, (bio)in eg able sys ems o bio-in e acing pu poses.
III
Lis o u ilized abb e ia ions
12C4 [12]c own-4 (c own e he compound)
AJP ae osol je p in ing
ANN a i icial neu al ne wo k
APS ammonium pe oxydisul a e (o ammonium pe sul a e)
AZ ac i e zone o he OMD
BHT 3,5-di- e -bu yl-4-hyd oxy oluene (an an ioxidan )
CGF ca ie gas low (in AJP)
CPA chi osan/polyaniline composi e
CSA campho sul onic acid
DBSA p-dodecylbenzenesul onic acid
GndCl Guanidinium chlo ide (o guanidine hyd ochlo ide)
IEL elec onic cu en (in I-V cha ac e iza ion)
IG ionic cu en o ga e cu en (in I-V cha ac e iza ion)
ISD o ITOT o al cu en o sou ce-d ain cu en (in I-V cha ac e iza ion)
IJP inkje p in ing
IPA isop opanol (isop opyl alcohol)
LIB li hium-ion ba e y
LS Langmui -Schae e echnique
NEA No land Elec onic Adhesi e (p oduc name)
NMP N-me hylpy olidone (o N-me hyl-2-py olidinone)
OMD o ganic mem is i e de ice
PANI polyaniline
(-LE, EB, ES, PS) (leucoeme aldine, eme aldine base, eme aldine sal , pe nig aniline sal )
PEDOT:PSS poly(3,4-e hylenedioxy hiophene) doped wi h polys y ene sul ona e
PEO poly(e hylene oxide) (o polye hylene oxide)
TSA o TsOH osylic acid (o p- oluenesul onic acid)
SCE s anda d calomel elec ode
SDS sodium dodecyl sul a e
SF silk i oin
ShGF shea h gas low (in AJP)
SPE solid polyelec oly e (o solid-s a e polyme elec oly e)
STDP spike- ime-dependen plas ici y
IV
Table o con en s
1 In oduc ion and s a e o he a .................................................................................................... 1
1.1 Mem is o s ........................................................................................................................... 1
1.2 The O ganic Mem is i e De ice ......................................................................................... 5
1.2.1 Ma e ials employed in OMD manu ac u ing ........................................................... 6
1.2.1.1 Polyaniline as he elec ically conduc i e polyme componen .............. 6
1.2.1.2 Poly(e hylene oxide) as he solid polyelec oly e .................................. 12
1.2.2 OMD swi ching mechanism .................................................................................. 13
1.2.3 Elec ochemical model o OMD swi ching ........................................................... 16
1.2.4 Expe imen al e idence o li hium mig a ion ........................................................ 18
2 Ma e ials and me hods ................................................................................................................ 22
2.1 Chemicals .......................................................................................................................... 22
2.2 Ins umen s ........................................................................................................................ 23
2.3 Expe imen al ...................................................................................................................... 24
2.3.1 LS-deposi ion o p is ine PANI ............................................................................. 24
2.3.2 Syn hesis o he chi osan/PANI composi e ............................................................ 24
2.3.3 P epa a ion o PANI and CPA inks ....................................................................... 25
2.3.4 Ae osol Je P in ing expe imen s ........................................................................... 25
2.3.5 P epa a ion o PEO gels ......................................................................................... 26
2.3.6 OMD ab ica ion .................................................................................................... 26
2.3.7 Elec ical cha ac e iza ion ...................................................................................... 26
3 O ganic mem is i e de ices in e acing biological sys ems ....................................................... 27
3.1 Elec ical noise and OMD ope a ion ................................................................................. 28
4 PANI-based OMD wi h imp o ed s abili y and pe o mance .................................................... 34
4.1 The ole o SPE dopan addi i es in de ice unc ioning ................................................... 34
4.1.1 Li+ ee SPE con aining o he hyg oscopic sal s .................................................... 38
4.1.2 The ole o dopan ca ions ..................................................................................... 41
4.1.3 SPE con aining Li+ wi h an addi ion o [12]c own-4 ............................................. 43
4.1.4 The ole o dopan anions ...................................................................................... 44
4.2 Imp o ed SPE composi ions o inc eased de ice pe o mance ....................................... 48
4.2.1 In insically acidic AlCl3-doped SPE wi h inc eased long- e m s abili y .............. 48
4.2.2 The concep o kosmo opici y, chao opici y and he Ho meis e se ies ............. 54
4.2.3 Op imiza ion o in insically acidic SPE ............................................................... 56
6
The esis i i y o he deposi ed PANI can be uned in a na ow zone in he middle o he
channel ha is co e ed by he manually cas s ipe o a polye hylene oxide (PEO) gel ha solidi ies
and unc ions as he SPE. The sil e wi e is placed a a sho dis ance abo e he PANI channel,
subme ged in he PEO gel pe pendicula o he channel di ec ion. Because he de ice somewha
esembles a ield e ec ansis o (FET) schema ically, he co esponding deno a ion has been
adap ed o he h ee elec odes. The sil e elec ode is named he ga e (G), and he wo elec odes on
bo h ends o he PANI channel a e named he sou ce (S) and d ain (D). The ga e is sho -ci cui ed
wi h he g ounded sou ce and a ol age is applied o he d ain and dis ibu ed ac oss he channel,
causing elec ochemical eac ions a ce ain gi en po en ials. The p ocess o he esis i e swi ching
o he de ice as well as he unde lying chemical ans o ma ions o he polyaniline a e explained
below in mo e de ail.
1.2.1 Ma e ials employed in OMD manu ac u ing
1.2.1.1 Polyaniline as he elec ically conduc i e polyme componen
Polyaniline has a basic poly(p-phenyleneimineamine) s uc u e, consis ing o al e na ing
benzenoid amine and quinoid imine moie ies. As shown in Figu e 1-6 o a minimum deg ee o
polyme iza ion o 8, in ea lie wo ks, i e di e en oxida i e s a es could be isola ed and chemically
in e con e ed h ough oxida ion wi h hyd ogen pe oxide o educ ion wi h TiCl3 o phenylhyd azine
[53]–[55]. The polyme iza ion o he monome aniline (o i s hyd ochlo ide) is mos commonly
ca ied ou in acidic media wi h o oxidizing agen s such as ammonium pe oxydisul a e (APS) o
elec ochemically, yielding he polyme in i s eme aldine sal o m (PANI-ES).
The o ms o polyaniline can be dis inguished wi h spec oscopic me hods. Figu e 1-7 shows
he cha ac e is ic abso p ion spec a o he h ee polyme base o ms in he UV/Vis ange [55],[56].
Leucoeme aldine displays only one signi ican abso p ion band a 343 nm which co esponds o he
π-π* ansi ion in he a oma ic sys em. In eme aldine, his band is hypsoch omically shi ed and a
new band a 637 nm appea s which is a ibu ed o a molecula exci on associa ed wi h he quinone
diimine agmen . In pe nig aniline, he band a 530 nm is a ibu ed o a Peie ls gap ansi ion [57].
FTIR spec a a e p esen ed in Figu e 1-8 [58],[59]. The mos ele an cha ac e is ic ib a ions o he
Figu e 1-5: Schema ic ep esen a ion o he Langmui −Schae e echnique.

7
Figu e 1-7: Op ical abso bance spec a o he base o ms o polyaniline, adap ed om [56].
Figu e 1-6
: a) Gene al chemical s uc u e o PANI and b) i s oxida i e s a es o a minimum
deg ee o polyme iza ion o 8 (adap ed om [55]).
8
base polyme a e δop(C−H) on 1,4-disubs i u ed ings a 823 cm-1 and δip(C−H) a 1100−1170 cm-1,
ν(C−N) o he seconda y amine g oups wi h a band a 1299 cm−1, ν(C−N) o he imine g oups a 1378
cm−1, ν(C=C) in benzenoid ings a 1496 cm−1, ν(C=C) in quinoid ings a 1588 cm−1 [59]. The la e
wo bands and hei a io can be used o dis inguish be ween he leucoeme aldine, eme aldine and
pe nig aniline o ms [60]. The conduc i e eme aldine sal s a e can be ecognized by he appea ance
o bands a 1238 cm−1 and 1140 cm-1 ha a e assigned o he ν(C−N+•) and ν(=NH+−) ib a ions o
he pola on and bipola on s uc u es, espec i ely [59], he o igin o which is explained below.
Doping o polyaniline
The eme aldine sal is he ansien s a e wi h he highes conduc i i y and hence he mos
ele an one o OMD ope a ion, eaching conduc i i y le els in he o de o 1−5 S cm−1 [61]. All
o ms o polyaniline can be elec ochemically ans o med in one ano he and can exis in hei
espec i e base and sal o m, depending on he pH o he en i onmen , as demons a ed in
Figu e 1-9 [62]–[64]. The so-called doping o a base o m o i s espec i e sal o m occu s due o he
acid-base ac i i y o he amine and imine g oups o he polyme chains. Fo ins ance, he
comme cially a ailable nonconduc i e eme aldine base (PANI-EB) can be chemically doped wi h an
acid such as HCl o ob ain he conduc i e eme aldine sal (PANI-ES). The p ocess is ully e e sible
as ea ing he doped polyme wi h an ammonia solu ion dep o ona es i , yielding PANI-EB [61].
This can be seen in Figu e 1-8 as he spec a o PANI-ES a e la gely equal a e and p io o
Figu e 1-8: a) FT-
IR spec a o he base o ms o polyaniline: leucoeme aldine (LB), eme aldine (EB), pe nig aniline (PB), adap ed
om [58]; b) FT-IR spec a o eme aldine base (EB), eme aldine sal (ES) and eme aldine sal a e exposu e o KOH and ep o ona ion
(adap ed om [59].
9
ep o ona ion. I mus be poin ed ou ha in he pe nig aniline sal s a e, he compound becomes
uns able due o he imine g oups’ p oneness o hyd olysis in an acidic medium [60]. Howe e , he e
ha e been success ul a emp s o s abilize pe nig aniline by means o N-alkyla ion o inc ease basici y
[65] o by employing hyd ophobic ionic liquids [66],[67].
Cha ge ca ie s in polyaniline
The doping phenomenon is closely connec ed o he o ma ion and na u e o cha ge ca ie s
as well as hei ans e along he polyme chains. In li e a u e, he p ocess is mos commonly
desc ibed unde he assump ion ha he doping is ini ia ed by he p o ona ion o imine ni ogen a oms
[68]–[71]. A i s glance, his may seem con o e sial because he sp3 hyb idiza ion o amine ni ogen
a oms sugges s a highe basici y wi h espec o he s uc u ally ela ed, sp2 hyb idized imine
coun e pa s, making he o me mo e p one o p o ona ion han he la e . Howe e , i mus be
conside ed ha in a dynamic sys em, he exchange o p o ons be ween di e en basic si es can akes
place. The exclusi e p o ona ion o amine si es canno be ollowed by any edis ibu ion o cha ge
along he polyme chain, whe eas he p o ona ion o imine ni ogen a oms leads o elec onic
ea angemen s ha esul in an ene ge ically mo e a ou able s uc u e. As shown in Figu e 1-10,
he ini ial p o ona ion causes he o ma ion o (di)ca ionic quinone diiminium agmen s, named
bipola ons. They a e hen ans o med in o hei esonan adical ca ionic pola on o m which is
ene ge ically mo e a ou able due o he ex ension o he a oma ic sys em [72]. The elec ical
conduc i i y o he ma e ial is gi en by he decoupling o he c ea ed neighbou ing cha ge ca ie s
and hei delocaliza ion along he polyme chain so as o minimize Coulomb epulsion.
The p esence o he pola on as a s a e wi h unpai ed elec ons has been p o en by means o
magne ic suscep ibili y measu emen s [73],[74] and elec on pa amagne ic esonance (EPR) [75].
The coexis ence o pola ons and bipola ons was u he p o en by X- ay pho oelec on spec oscopy
Figu e 1-9: Mu ual ans o ma ions o he o ms o PANI by means o edox eac ions and chemical doping.
10
(XPS) in combina ion wi h elemen al analysis and op ical spec oscopy [76],[77]. Inoue e al. [76]
es ima ed a pola on o bipola on a io o abou 5:1 in a pe chlo a e doped PANI sample. Fu he mo e,
he XPS echnique was employed o ga he e idence o he p oposed p o ona ion mechanism, aking
ad an age o he ac ha he posi ion and shape o N(1s) line a e sensi i e o he elec onic
en i onmen o he ni ogen a oms which changes upon oxida ion and doping. The i ed line
componen s in he N(1s) co e-le el spec um o he quinoid imine, benzenoid amine and posi i ely
cha ged ni ogen a oms ha e been iden i ied by Kang e al. [78],[79], wi h band ene gies a
389.1 eV, 399.3 eV and > 400.5 eV, espec i ely. To s udy he doping p ocess, he au ho s exposed
he eme aldine base polyme o non- ola ile acids (H2SO4 o HClO4) and e alua ed he signal
in ensi ies om he N(1s) and S(2p) o Cl(2p) co e-le el spec a, which allowed o es ima e he
doping le el a di e en acid concen a ions [69]. H2SO4 and HClO4 we e chosen as doping agen s
because ola ile dopan s such as HCl a e no sui ed o measu emen s in ul a-high acuum and he
Cl-/N a io seldom exceeds 0.5 e en a high acid concen a ions [78]. Acco ding o hei da a, upon
p o ona ion, he imine signal disappea s in a ou o he eme ging cha ged ni ogen signals. The
in ensi y o he amine signal emains la gely unchanged a doping le els up o 50 % (in 1 N acid) bu
diminishes in a ou o he signals o cha ged species a highe doping le els (in concen a ed acid),
as demons a ed in Figu e 1-11 a−d). This sugges s ha amine ni ogen a oms pa icipa e in he acid-
base eac ion only a e all he imine ni ogen a oms a e p o ona ed. The au ho s also epo ha he
conduc i i y o he excessi ely doped polyme (anion/N a io > 0.5) diminishes wi h espec o a
doping le el o 50 %, which is due o he dis o ion o he π-conjuga ion and he pola on la ice by
p o ona ed amine uni s (−NH2+−). The same is epo ed by Yue e al. [80] o a sul ona ed, sel -doping
polyaniline sample: he conduc i i y is lowe when mo e han 50 % o quinoid and benzenoid ings
a e sul ona ed. In ano he wo k, Snauwae e al. [70] ha e co ela ed he pe cen age o imine ni ogen
a oms wi h he oxida ion o PANI s a ing om i s leucoeme aldine base o m, eco ding XPS spec a
Figu e 1-10: Scheme o he p ocess o doping o eme aldine base, anions a e omi ed (adap ed om [72]). a)
Eme aldine base polyme ,
b) o ma ion o he bipola on la ice upon p o ona ion, c) elec onic ea angemen o pola on s a e
, d) esul ing pola on la ice as a
consequence o cha ge delocaliza ion.
11
a e he applica ion o di e en po en ials (Figu e 1-11 ). Thei esul s coincide wi h he
abo emen ioned elec onic ans o ma ion o he polyme , al hough i mus be poin ed ou ha
acco ding o he imine pe cen age (abou 75 %), he p oduc o he second oxida ion s ep a +0.8 V
appea s o be nig aniline a he han pe nig aniline. The da a also shows ha he oxida ion occu s a
mo e o less disc e e alues a he han g adually o e he po en ial ange.
In classical semiconduc o e ms, he adical ca ionic pola ons can be conside ed as elec on
holes and hus, PANI is deno ed as a p- ype o ganic semiconduc o . Con a y o common ino ganic
semiconduc o s, PANI, like o he elec ically conduc i e polyme s, equi es high le els o doping,
eaching highes conduc i i y a 50 % p o ona ion which co esponds o a pH o 2 o lowe [81]. The
eason is he depinning o cha ge ca ie s which is bes unde s ood conside ing hei en i onmen .
Al hough delocaliza ion is possible h oughou he highly conjuga ed elec onic sys em, a low doping
le els he cha ge de ec s a e widely sepa a ed om one ano he . In o de o main ain cha ge
neu ali y, each cha ge ca ie mig a es as a pai oge he wi h i s associa ed coun e ion, ha ing he
ela i ely slow di usion o he la e as he limi ing ac o . A high doping le els, each pola on inds
i sel in he icini y o mul iple anions, no being pinned o any pa icula one, and can mig a e as e
along he polyme chain in he mean ield o he su ounding coun e ions [82]. In he bulk o he
polyme , he limi ing ac o o he conduc i i y o PANI is he elec on ans e om one chain o
ano he , which sugges s a dependence o cha ge mobili y om he mo phology o he ma e ial. In
hei wo k, Focke e al., [83] p opose a mechanism acco ding o which in e chain cha ge ans e can
be acili a ed by p o ona ion, as shown in Figu e 1-12.
Figu e 1-11: N(1s) and Cl(2p) co e-le el XPS spec a o HClO4 doped eme aldine ilms wi h a ClO4−/N a io o a,
b) abou 0.50 and
c, d) abou 0.78. e) N(1s) co e le el spec um o an eme aldine base ilm (a−e adap ed om [69]). ) XPS-
de e mined imine con en in
polyaniline as a unc ion o he applied po en ial a pH = 0 (black do s) in co ela ion wi h he CV oxida ion scan (solid li
ne), adap ed
om [70].

12
1.2.1.2 Poly(e hylene oxide) as he solid polyelec oly e
In o de o enable elec ochemical eac ions be ween he ga e elec ode and he polyaniline
channel ha would cause esis i e swi ching, he wo mus be connec ed h ough a ma e ial ha is a
dielec ic i sel bu allows o cha ge anspo om one elec ode o ano he . The ealiza ion o
mem is i e de ices in mic oelec onic appliances equi es a p e e ably solid elec oly e o be e
manageabili y. To ul il hese equi emen s, high molecula poly(e hylene oxide) (PEO) (wi h
Mw = 8 MDa) was chosen as i is a solid o ganic ma e ial ha swells in wa e o o m a iscous gel
which can be easily cas manually. The swelling occu s due o he sol a ion o he PEO chains wi h
wa e molecules. Once cas , a p is ine PEO gel g adually loses some o i s wa e con en due o
e apo a ion. Wi h dec easing wa e con en , he in e ac ions o PEO chains wi h each o he p e ail
mo e and mo e, leading o agg ega ion and c ys alliza ion which can be p e en ed by adding dopan s
(i.e., p epa e he gel in an aqueous solu ion o dopan sal s).
A li hium sal is used as he main dopan in SPEs o OMDs, which is inspi ed by he
ad ances in he ield o li hium-ion ba e ies (LIB) [84]–[86] ha ake ad an age o li hium’s small
ionic adius and i s abili y o in e cala e in ca bon o me al oxide ma ices. In he ope a ing mechanism
ha was p oposed o he OMD in ea lie wo ks, a key ole in he swi ching o he de ice is a ibu ed
o li hium ions as hey we e hough o be essen ial o he cha ge anspo be ween polyaniline and
he ga e elec ode wi hin he ac i e zone o he de ice. The li hium ions a e coo dina ed by PEOs
e he ic oxygen a oms h ough ion-dipole in e ac ions and he cha ge anspo is hough o be due o
he hopping o Li+ be ween di e en coo dina ion si es. Analogous o LIB echnology, one o he
easons why li hium sal s ha e been he dopan o choice is li hium’s small ionic adius which implies
a highe ionic mobili y. E iden ly, o enable a su icien cha ge ca ie mobili y, he PEO gel mus be
main ained in an amo phous s a e. Thus, ano he p ope y o sal s such as LiClO4 o LiBF4 is aken
ad an age o , namely hei hyg oscopici y ha main ains a ce ain necessa y amoun o wa e in he
sys em, p e en ing he agg ega ion o PEO chains. Due o he acid-base eac i i y o polyaniline i is
also necessa y o c ea e an acidic en i onmen in he PEO gel ha would o he wise dep o ona e he
Figu e 1-12: Mechanism o in e chain cha ge ans e assis ed by p o ona ion (adap ed om [83]
). Chain II is p o ona ed, enabling a
single-elec on ans e om chain I o chain II h ough esonance, a e which chain I is dep o ona ed.
13
polyaniline sal o i s nonconduc i e base o m. This equi es a second, acidic dopan such as HCl.
The aspec o doping in PEO is one o he majo opics o he p esen wo k, discussed in de ail in
sec ion 4.1.
1.2.2 OMD swi ching mechanism
Unde s anding he p ocesses ha ake place in he ac i e zone o he de ice is key o
imp o ing i in e ms o pe o mance and s abili y and making i mo e app op ia e o complex
applica ions. Fo his eason, he OMD’s swi ching mechanism has been s udied ex ensi ely in he
pas , p o iding FTIR spec oscopic [87] o X- ay luo escence (XRF) [88] da a o suppo he
cons uc ed model. Be o e p oceeding o ou line he de ails o he de ice ope a ion, i mus be poin ed
ou ha in scope o he p esen wo k, he s udy o he ole o SPE dopan ions ha is p esen ed in
sec ion 4.1 led o a numbe o indings which e eal se e al inconsis encies in he o me ly assumed
model. Howe e , in his sec ion, i appea s necessa y o p esen he de ails o he o me ly p oposed
swi ching mechanism including he espec i e suppo ing da a as o allow o a clea e discussion o
he new indings epo ed in sec ion 4.1.
As was s a ed be o e, he esis i e swi ching occu s as a consequence o elec ochemical
eac ions o he conduc i e polyme polyaniline which is ans o med om i s elec ically insula ing
leucoeme aldine s a e (PANI-LE) o i s conduc i e eme aldine sal s a e (PANI-ES) upon oxida ion
(o ice e sa upon educ ion). The p ocess can be moni o ed elec ically by plo ing he ou pu
cu en agains he bias applied o he d ain elec ode wi h espec o he g ounded sou ce and ga e
elec odes. Fo a mo e p ecise cha ac e iza ion, wo cu en s a e eco ded du ing he I-V
measu emen s – he o al cu en (ISD o ITOT) lowing h ough he de ice and sepa a ely, he ga e
elec ode cu en (o ionic cu en , IG). The ga e cu en ep esen s he amoun o cha ge ha passes
h ough he ga e elec ode du ing he elec ochemical eac ion. Calcula ing he di e ence be ween
he wo egis e ed cu en s allows o e alua e he e olu ion o he so-called elec onic cu en (IEL)
ha lows be ween he d ain and sou ce elec odes and can be exp essed as IEL = ISD − IG. As p esen ed
in Figu e 1-13, he I-V cu e (panel a) shows a hys e esis loop as a cha ac e is ic ea u e o a
mem is i e de ice, which is due o he di e ence in applied po en ials a which he elec ochemical
oxida ion and educ ion ake place. Typically, oxida ion s a s a a ound +0.3 V du ing he anodic
scan and educ ion occu s a a ound +0.1 V du ing he ca hodic scan. These po en ials co espond o
he peaks o he ionic cu en IG (panel b) which esembles a ol ammog am o polyaniline in he
gi en po en ial ange. The peaks ma k he g owing o dec easing end o he IEL cu e. The posi ion
and wid h o hese peaks is de e mined by he e en ion ime R pe ol age bias alue, i.e., he scan
a e. Fas e scanning causes a shi o highe po en ial alues o he oxida ion peak(s) and lowe
alues o he educ ion peak(s), which is due o he ac ha he sys em is no gi en enough ime o
equilib a e. On he con a y, he peaks con e ge a highe e en ion imes, as he espec i e eac ions
a e inalized soone a lowe applied ol age biases. The shape and wid h o he hys e esis loop bo h
change acco dingly.
14
Ano he cha ac e is ic ea u e o he OMD’s I-V cu e is he ec i ica ion a lowe po en ial
alues. This is gi en by he ac ha once in i s insula ing PANI-LE o m, he change o he ol age
dependen channel cu en is e y low in compa ison o he ansi ion o PANI-ES, du ing which he
conduc i i y inc eases by up o ou o de s o magni ude [89]. Ideally, only one couple o edox peaks
is expec ed as he esul o he ully e e sible ansi ion be ween hese wo o ms o polyaniline. Fo
compa ison, a cyclic ol ammog am wi h he ansi ions o polyaniline a he espec i e po en ials is
shown in Figu e 1-14. I can be seen ha when he ol age ange is ex ended beyond +0.8 V, a second
anodic peak occu s as a esul o he oxida ion o eme aldine o pe nig aniline. As men ioned abo e,
his second oxida ion s ep should be a oided because, wi h espec o PANI-ES, he ully oxidized
pe nig aniline o m con ains wice as many imine g oups ha a e p one o deg ada ion in acidic media
[59], wi h no emaining amine g oups o bu e hei p o ona ion ha , on i s u n, acili a es
hyd olysis. A smalle oxida ion peak may occu a a ound +0.6 V and is a ibu ed o c osslinking
eac ions be ween PANI chains, o ming phenazine s uc u es as demons a ed in Figu e 1-15.
Acco ding o Geniès e al. [90], a eac ion can occu be ween g owing PANI chains and aniline
ni enium ions ha a e o med as an in e media e du ing he polyme iza ion [91]. The au ho s also
assume a c osslinking eac ion be ween di e en PANI chains ha ca y ni enium ion moie ies.
Since he o ma ion o ni enium ions wi hin polyaniline chains appea s o be unlikely, due o he
likelihood o elec on ea angemen in a conjuga ed sys em, he p oposed mechanism has been
sligh ly modi ied and oxidized eme aldine chains a e assumed o eac wi h one ano he . The
phenazine s uc u es dis up he pola on la ice, educing he ma e ial’s maximum conduc i i y o e
ime as he p ocess is no ully e e sible, e en hough he p esence o a sligh B’ peak sugges s some
deg ee o e e sibili y [92]. Howe e , his p ocess akes place mo e commonly du ing he
elec ochemical syn hesis o PANI om aniline by means o cyclic ol amme y and is less likely o
be obse ed du ing OMD cha ac e iza ion.
Figu e 1-13: Typical I-V cha ac e is ic o an O ganic Mem is i e De ice, wi h cu es o a) he elec onic cu en IEL
ea u ing a
coun e -clockwise hys e esis and a ec i ica ion a nega i e ol ages and b) he ionic cu en IG.
15
Figu e 1-14: Cyclic ol ammog am o a HClO4 doped PANI ilm eco ded a a 200 mV s−1 sweep a e, in a 1 M aqueous HClO4
. A/A’:
peak co esponding o he ansi ion be ween PANI-LE and PANI-ES, B/B’: peak co esponding o c osslinking eac ions be w
een
PANI chains and hei pa ial e e sion, C/C’: peak co esponding o he ansi ion be ween PANI-ES and PANI-
PS (adap ed om
[92]).
Figu e 1-15: Mechanism o c osslinking a)
be ween a g owing PANI chain 1 and he ni enium ion o aniline 2 du ing he
syn hesis by means o CV and b) be ween wo PANI chains du ing oxida ion (adap ed om [90]).
22
2 Ma e ials and me hods
2.1 Chemicals
aluminium chlo ide AlCl3 Ac os O ganics
aluminium b omide AlB 3 Al a Aesa
ammonium pe sul a e (NH4)2S2O8 Ac os O ganics
ammonium hiocyana e NH4SCN Ca lo E ba
calcium chlo ide CaCl2 Fluka
ce ium chlo ide CeCl3 Ca lo E ba
caesium chlo ide CsCl VWR Chemicals
e ic chlo ide FeCl3 Ac os O ganics
e ous sul a e FeSO4 VWR Chemicals
li hium chlo ide LiCl Fluka
li hium pe chlo a e LiClO4 Sigma Ald ich
li hium e a luo obo a e LiBF4 Me ck
li hium hexa luo ophospha e LiPF6 Ald ich
li hium hiocyana e LiSCN Ca lo E ba
magnesium chlo ide MgCl2 Fluka
magnesium sul a e MgSO4 VWR Chemicals
po assium chlo ide KCl Fluka
po assium hiocyana e KSCN Sigma Ald ich
ubidium chlo ide RbCl Sigma Ald ich
sodium chlo ide NaCl Sigma Ald ich
sodium bilsul a e NaHSO4 VWR Chemicals
sodium dodecylsul a e NaSO4C12H25 Sigma Ald ich
sodium sul a e Na2SO4 Ca lo E ba
sil e (wi e, d = 50 µm and 125 µm) GoodFellow
sil e ( oil, d = 125 µm) Al a Aesa
ace ic acid (100 %) CH3COOH VWR Chemicals
hyd ob omic acid (48 %) HB Ca lo E ba
hyd ochlo ic acid (1 N) HCl Ca lo E ba
o hophospho ic acid (85 %) H3PO4 VWR Chemicals
pe chlo ic acid (60 %) HClO4 Fishe Scien i ic
sul u ic acid (98 %) H2SO4 Ca lo E ba

23
ace one Ca lo E ba
ace oni ile VWR Chemicals
aniline Me ck
chi osan (low molecula weigh ) Sigma Ald ich
ci ic acid VWR Chemicals
1,1,1,3,3,3-hexa luo oisop opanol Al a Aesa
isop opanol Ca lo E ba
NEA 121 No land P oduc s Inc.
N-me hyl-2-py olidone VWR
polyaniline (eme aldine base, Mw = 105 Da) Sigma Ald ich
poly(e hylene oxide) (Mw = 8∙106 Da, 200−500 ppm BHT) Sigma Ald ich
poly(e hylene oxide) (Mw = 106 Da) Al a Aesa
p- oluenesul onic acid Al a Aesa
oluene Ca lo E ba
xylene Ca lo E ba
The sol en pu i y was analy ical g ade o ACS eagen g ade. All chemicals we e used wi hou
u he pu i ica ion. LiSCN was d ied a 80 °C o 15 h p io o weighing.
2.2 Ins umen s
Agilen B2902A elec ical cha ac e iza ion (SMU)
B uke Sy ius MALDI-TOF mass spec ome y
Diene Elec onic Fem o low-p essu e O2/A gon plasma su ace ea men
Elma Elmasonic P ul asound ea men
He ich Ro o Silen a III cen i uga ion
Me ck Millipo e Re e ence A+ Milli-Q wa e p epa a ion
Kei hley 236 elec ical cha ac e iza ion (SMU)
Kei hley 6514 elec ical cha ac e iza ion (elec ome e )
KSV-5000 ough Langmui -Schae e hin ilm deposi ion o PANI
Leica DMS300 op ical mic oscopy
Na ional Ins umen s PXIe-1073 elec ical cha ac e iza ion (SMU)
Nikon H550L op ical mic oscopy
No ascan PSD P o Se ies UV ozone su ace ea men
Op omec Ae osol Je 200 Ae osol Je P in ing
Pe kin Elme Spec um Two FTIR spec oscopy
Veeco Dimension 3100 Nanoman A omic Fo ce Mic oscopy
Zeiss Au iga Compac Scanning Elec on Mic oscopy
24
2.3 Expe imen al
2.3.1 LS-deposi ion o p is ine PANI
Solu ions o polyaniline (1.0 mg/ml in NMP) we e p epa ed by adding PANI-EB o he
sol en po ion wise unde igo ous s i ing. E en ual p ecipi a e was allowed o se le o leas 30
min and small po ions we e aken o dilu ion o 0.1−0.2 mg/ml in NMP/ oluene (9:1 / ). S anda d
samples we e made om ec angula 15 x 7 mm (unless speci ied o he wise) qua z subs a es
co e ed by ch omium (deposi ed by CVD). Elec odes on each side o he subs a e we e made by
co e ing 2 mm wide s ipes wi h polyimide ape and e ching away he emaining ch omium wi h a
p op ie a y, ce ic ammonium ni a e-based e ching agen (Sigma Ald ich). The su ace was co e ed
by s ips o polyimide ape, lea ing a 2−3 mm ee channel. The subs a es we e washed and
ho oughly insed wi h isop opanol and wa e p io o use. The deposi ion o PANI hin ilms was
ca ied ou on a KSV-5000 ough using a modi ied Langmui -Schae e echnique. Milli-Q wa e was
used as he subphase, and he PANI solu ion in NMP/ oluene was injec ed on o i s su ace and
comp essed by Te lon ba ie s a a cons an a e o 0.5 cm/min un il a a ge su ace p essu e o 10
mN/m. The oluene in he solu ion p e en ed PANI om being d agged in o he subphase by NMP.
The o med PANI laye was sepa a ed by an ac ylic glass g id and 30−60 LS monolaye s we e
deposi ed on he subs a e, al e na ing he di ec ion o deposi ion a e e e y 5 h laye . The deposi ed
laye s we e doped o 30−40 s in 1 N HCl wice wi h an in e al o ca. 30 min. Las ly, he polyimide
ape s ips we e emo ed, lea ing a PANI-ES channel s e ching be ween he ch omium elec odes.
2.3.2 Syn hesis o he chi osan/PANI composi e
The chi osan/PANI composi e ma e ial (CPA) was syn hesized by oxida i e polyme iza ion o aniline
in he p esence o chi osan. Fi s , 0.3 g chi osan we e dissol ed in 50 ml 4 % ace ic acid unde s i ing
o 30 min. The iscous solu ion was pou ed in a 250 ml ound bo om lask. Sepa a ely, 0.9313 g
(10 mmol) aniline we e dissol ed in 50 ml 1 N hyd ochlo ic acid, he solu ion was cooled in an ice
ba h and added d opwise o he cooled lask unde s i ing. A ew d ops o a 1M FeSO4 solu ion we e
added o he mix u e. Las ly, a solu ion o 2.7382 g (12 mmol) ammonium pe sul a e (APS) in 40 ml
1 N hyd ochlo ic acid was cooled and added d opwise o he eac ion mix u e in he lask o e ca. 1
h unde s i ing, main aining he empe a u e below 5 °C. The mix u e was s i ed o 12 h and i s
colou changed om pale yellow o e blueish g een o da k g een. The excess APS was quenched by
adding 5 ml o 1 M FeSO4 solu ion. The esul ing suspension was cen i uged a 4000 RPM o 5
min. The ligh g een supe na an was disca ded, and he da k g een solid p oduc was esuspended in
1 N HCl and cen i uged. This was epea ed h ee imes. The pelle was hen esuspended once mo e
in ace one, cen i uged, and d ied a 50 °C o 24 h. Las ly, he da k g een, b i le solid p oduc was
g inded in an aga e mo a . The yield was 1.160 g (94 % ela i e o he mass sum o aniline and
chi osan educ s), al hough i is likely ha some aces o sol en s emained in he ma e ial.
25
P oduc analysis
FTIR spec a we e kindly aken by D . Sil ia D’Au ia on a Pe kin Elme Spec um Two
spec ome e in he Depa men o Chemis y, Uni e si y o Pa ma, p epa ing he sample as a KB
pelle . MALDI-TOF measu emen s we e kindly aken by D . Mi ko Bu ini on a B uke Sy ius mass
spec ome e in he Depa men o Medicine and Su ge y, Uni e si y o Pa ma. The spec a we e
acqui ed in he posi i e linea mode in he ange o 2–20 kDa wi h a 60 Hz lase equency and an
ion sou ce ol age o 20 kV, using α-Cyano-4-hyd oxycinnamic acid (HCCA) as he ma ix ma e ial.
2.3.3 P epa a ion o PANI and CPA inks
Inks o p is ine PANI we e made by p e-dissol ing PANI-EB in NMP unde s i ing. Xylene
and/o IPA we e slowly added o he s i ed solu ion in di e en p opo ions. The esul ing solu ion
was ea ed wi h ul asound (P = 120 W, = 37 kHz, as in all ollowing applica ions) o 30 min and
s o ed in a e ige a o . In some cases, he solu ion was il e ed h ough a 0.2 µm PTFE sy inge il e .
The inks we e ea ed wi h ul asound o 20 min p io o use.
CPA-inks we e p epa ed by adding CPA o wa e /IPA (8:2 / ) and ea ing wi h he sample
wi h ul asound o 10 s, esul ing in a ine suspension (a il e ing es wi h a 0.45 µm PTFE sy inge
il e yielded a clea , colou less il a e). The inks we e s o ed in a e ige a o and ea ed wi h
ul asound o 30 s p io o use. The concen a ions o PANI and CPA as well es sol en a ios a e
gi en in sec ion 6.2.
2.3.4 Ae osol Je P in ing expe imen s
15 x 7 mm qua z slides wi h ch omium elec odes o silicon slides (1 µm he mal SiO2 laye )
ea u ing T-shaped gold elec odes (by pho oli hog aphy, d = 200, 10 o 5 µm) we e used as
subs a es. P io o use, he samples we e ea ed wi h ul asound in isop opanol o 1 h, insed wi h
ace one and wa e (Milli-Q). The su ace was ea ed wi h plasma (O2/A gon, P = 3 W, = 9 min) and
UV/ozone ( = 3 min, T = 50 °C) immedia ely be o e p in ing. The deposi ion o PANI and CPA inks
was ca ied ou wi h an Op omec Ae osol Je 200 sys em wi h a ious p in ing pa ame e s, such as:
- nozzle diame e : 200 µm
- deposi ion pla e empe a u e: 90 °C
- speed: 0.5−2 mm/s
- shea h gas: 40 sccm
- ca ie gas: 30 sccm
- a omize cu en : 0.5 mA
The p in ed laye s we e doped wi h 1 N hyd ochlo ic acid o ca. 30 s.
26
2.3.5 P epa a ion o PEO gels
PEO gels we e in 3 ml sc ew-neck glass ials wi h Te lon gaske space s. Aqueous solu ions
o a ious dopan s wi h concen a ions be ween 0.1 and 1.0 mol/l we e added po ion wise and illed
up o 3 ml wi h wa e (Milli-Q, ρ = 18.2 MΩ℅cm, was used in all u he s eps). Di e en amoun s
o PEO (Mw = 8℅106 Da, 200−500 ppm BHT as inhibi o , 90−120 mg depending on inal
concen a ion) we e added po ion wise o he ial unde gen le agi a ion, ollowed by he nex po ion
o dopan solu ion o wa e . The suspension was o exed o 30−60 s and s o ed o 1−2 days o
swell un il homogenei y, being igo ously mixed wi h a s i e a ew imes by hand. The esul ing
gels we e s o ed in a e ige a o unless speci ied o he wise.
2.3.6 OMD ab ica ion
All subs a es we e ea ed wi h ul asound in isop opanol o 15-60 min and insed wi h
ace one and wa e . In he case o qua z samples, he ac i e zone o he OMD (in con ac wi h he
(poly)elec oly e was delimi ed by co e ing he PANI laye wi h s ips o polyimide ape, lea ing a
1 mm wide unco e ed s ipe in he middle. In he case o T-shaped Si/SiO2 samples, he channel was
ac i e ac oss he whole leng h. Two s eps o 2−3 laye s o na ow polyimide ape s ips we e placed
o bo h sides o he ac i e zone, pa allel o he PANI channel. Then, 1−3 laye s o PEO gel we e cas
manually on op o he ac i e zone wi h a plas ic s i e o mic opipe e ip. A sil e wi e (d = 50 µm),
wi h a piece o in was solde ed o one o i s ends as a cable connec ion poin , was s e ched and
ixa ed o e he polyimide ape s eps wi hin he ac i e zone, pe pendicula o he PANI channel. The
wi e was co e ed by ano he 1−2 laye s o cas PEO gel. The de ice was eady o be used once he
polyelec oly e solidi ied. In some cases, coa ing wi h NEA ollowed, as desc ibed in sec ion 4.3.
2.3.7 Elec ical cha ac e iza ion
The acquisi ion o elec ical da a was ca ied ou using an SMU o apply ol age biases and
egis e ou pu cu en s (bo h he o al cu en ISD and he ga e cu en IG we e egis e ed
simul aneously). Cyclic I-V cha ac e is ics we e eco ded by scanning he de ice in he po en ial
ange [−0.4 V; +0.8 V] o [−0.4 V; + 0.6 V], s a ing wi h an anodic scan om 0.0 V. The scan a e
was de e mined by he scanning s ep (0.1 V o 0.02 V) and he espec i e dwell ime, as speci ied in
he expe imen s. Kine ic measu emen s we e ca ied ou a a cons an bias (e. g. +0.5 V o oxida ion
and −0.2 V o educ ion) o 5−10 min. All measu emen s we e p eceded by an a leas 2 min long
condi ioning pe iod a −0.2 V o p epa e he sample, by comple e educ ion o PANI o he LE s a e.
Addi ional s eps we e added when pe o ming cha ac e iza ions o OMDs wi h a se ies o di e en
liquid elec oly es as desc ibed in sec ion 4.1. When changing he elec oly e, he sample and he
ough we e ca e ully insed wi h 0.1 N HCl and hen insed wice mo e wi h he nex elec oly e.
Each expe imen was ca ied ou a leas in iplica e.
27
3 O ganic mem is i e de ices in e acing biological sys ems
One o he gene al long- e m goals o ou esea ch is he de elopmen o sys ems ea u ing
O ganic Mem is i e De ices o in o ma ion p ocessing and in eg a ion wi h bioelec onic sys ems
and sensing. Such applica ions imply he use o la ge a ays o elemen s, mimicking biological neu al
ne wo ks and possibly in e acing and/o being usable as pa s o a na u al ne ous sys em. Since he
aim is o mimic some unc ionali ies o biological sys ems, as a p e equisi e o he applicabili y o
OMDs and hei compa ibili y wi h biological elemen s, a lo o p e ious wo k has been dedica ed o
he s udy o ele an analogous p ope ies. In pa icula , he simila i ies be ween an OMDs and
biological synapses ha e been demons a ed.
Acco ding o Hebb’s p inciple [2], one o he undamen al p ope ies o synap ic
ansmission be ween neu onal cells is he so-called spike- iming-dependen plas ici y (STDP),
meaning he ha he s eng h o connec ions be ween neu ons is ein o ced by he p io ac i i y o
said connec ions. This is o en simpli ied as “Wha i es oge he , wi es oge he ”, meaning ha he
connec ion be ween wo cells becomes s onge he mo e o en hese cells a e simul aneously
ac i a ed. This concep has been gene ally p o en o be applicable o mem is i e de ices, since hey
display a change in conduc ance as a esponse o a sequence o ecei ed elec ical s imuli [97]–[99].
Recen ly, Ba is oni e al. demons a ed his p ope y in O ganic Mem is i e De ices [100], showing
long- e m po en ia ion and dep ession caused by se ies o elec ical s imuli wi h a ious biologically
ele an equencies. The ope abili y o he de ice in a pulse mode has been demons a ed ea lie by
Sme ie i e al., including he abili y o gene a e an oscilla ing esponse ha is assumed o o igina e
om capaci i e e ec s [93],[101]. In he con ex o biocompa ibili y, such a beha iou can be seen
as analogous o he spike ain-like p opaga ion o ac ion po en ials in li ing cells. Las ly, in a wo k
by he g oups o Pa ma and Kazan [102], he OMD has been success ully employed o di ec ly
connec wo li ing co ical neu ons o a a . The a i icial synapse enabled he communica ion
be ween he wo s imula ed cells whe e he signal p opaga ion ully ollowed he Hebbian p inciple,
mimicking he na u al ones in ine de ails. This is ano he g ea s ep owa ds biological and
biomedical applica ions since i displays he abili y o OMDs o di ec ly emula e he unc ionali y o
biological synapse.
Fu he mo e, explo a o y esea ch has been ca ied ou o he implemen a ion o OMDs in
he ha dwa e ealiza ion o A i icial Neu al Ne wo ks (ANNs) in he o m o single and double laye
pe cep ons [103],[104], as schema ically shown in Figu e 3-1. Such sys ems allow o sol e analogue
asks such as inpu signal classi ica ion, pa ing he way he pe spec i e o ca y ou mo e complex
asks a he ha dwa e le el wi h mul ilaye s uc u es. Fu he mo e, i could be demons a ed ha
OMD-based sys ems a e capable o emula ing simple biological beha iou pa e ns, as i occu s in
he eeding o he g ea pond snail Lymnaea s agnalis [105]. Essen ially, i esembles he well-known
Pa lo ’s dog expe imen which has been la e emula ed on he basis o ino ganic mem is i e de ices
by he g oup o Ziegle e al. [106]. The equi alen elec onic ci cui mimicking he homo- and

28
he e osynap ic lea ning p ocess we e ealized wi h only one and wo OMDs, espec i ely. The esul s
a e an adequa e ep esen a ion o he inc eased esponse o a condi ioned s imulus (in his case, he
p esence o amyl ace a e) a e he condi ioning p ocedu e, when he condi ional s imulus is applied
oge he wi h he uncondi ioned one (in his case, he p esence o ood pa icles), esul ing in an
inc eased ou pu signal (co esponding o he mo emen o eeding muscles) (see Figu e 3-2).
3.1 Elec ical noise and OMD ope a ion
When going owa ds la ge, complex a ays o in e connec ed de ices, i is necessa y o
conside he occu ence o some deg ee o c oss alk and in e e ences including elec ical noise. This
is especially ue when a emp ing o in eg a e OMDs wi h biological sys ems consis ing o a my iad
o in e connec ed elemen s, whe e so called noise a i ac s a e ine i able. The e m “elec ical noise”
means he luc ua ion o a signal wi hin a ce ain dis ibu ion, cha ac e ized by a spec um o
equencies and an ampli ude, wi hin which ange his luc ua ion occu s. In he case o a andom
dis ibu ion co e ing all possible equencies, he esul ing unp edic able luc ua ions a e called
“whi e noise” in analogy o whi e ligh consis ing o he whole spec um o isible wa eleng hs.
Simila ly, i is called “colou ed noise” when i s equency dis ibu ion by a single alue o e a pe iod
o ime, o “low pass il e ed noise” o a ce ain in e al o equency alues (see Figu e 3-3).
Figu e 3-1: Schema ic ep esen a ion o an elemen a y (le ) and a double laye pe cep on ( igh ) [104]. The inpu s Xi co e
spond o
di e en s imuli ha a e weighed by weigh unc ions Wi ep esen ed by OMDs, he sum o which esul s in an ou pu signal Y.
Figu e 3-2: Scheme ep esen ing a pa o he ne ous sys em* o he g ea pond snail Lymnea s aglalis
, esponsible o i s eeding
beha iou (a) wi h an equi alen elec onic ci cui ea u ing OMDs (b). c) Ci cui ou pu signal as a esponse o he condi ioned s imul
us
(CS) be o e and a e condi ioning.
*CGCs: ce eb al gian cells; SN: senso y neu ons; CBI: ce eb obuccal in e neu ons
; CPG: cen al pa e n gene a o in e neu ons. The
a ows poin a synap ic in e connec ions, emula ed by OMDs in panel (b) (adap ed om e e ence [105]).
29
Al hough i is seen mos ly as a nuisance and an in e e ence, he e a e wo ks in he ield o
neu osciences disco e ing posi i e, cons uc i e e ec s o his phenomenon [107]. In a wo k by
Hun e e al. [108] he e ec o he pa ame e s o elec ical noise on he eliabili y o he i ing pa e n
o single neu ons is s udied. The au ho s dis inguish be ween he con ibu ion o he equency ange
and he ampli ude o noise (exp essed in he o m o he so-called coe icien o a iabili y, CV). To
his end, he esponse o slowly adap ing mo oneu ons (aplysia) o bu s s o la ge ampli ude inpu s
al e na ing wi h low ampli ude inpu s is analysed. The expe imen s show a equency-independen
inc ease in spike ime eliabili y in he p esence o luc ua ions o he inpu cu en wi h su icien ly
high noise ampli ude (CV = 1). The dependence on he equency o he inpu shows when compa ing
o he wise equi alen expe imen s in which a b oadband inpu (Figu e 3-4 a) is compa ed o one
lacking he equencies a ound he neu on’s i ing a e DC (Figu e 3-4 b). I can be seen ha he i ing
eliabili y is main ained when he neu on ecei es a b oadband inpu , bu diminishes soon a e he
modula ion o he noise ampli ude o lowe alues (CV = 0.15) when he equencies a ound DC a e
no ep esen ed. The explana ion behind such a esponse is he s abiliza ion o he i ing pa e n
h ough in e e ence o he i ing a e wi h ha monics and subha monics o DC om he equency
spec um o he cu en inpu , which could be ele an when ope a ing OMDs in he oscilla o mode.
Simila esul s we e ob ained in a wo k by E men ou e al. [107], whe e popula ions o
in e connec ed li ing neu onal cells ha e been s udied by compa ing hei i ing pa e ns gene a ed
by he injec ion o cu en pulses wi h and wi hou supe imposed noise signals. The s udy shows ha
in analogy o single neu ons, despi e i s chao ic na u e, noise can c ea e o de in a sys em, causing a
ce ain deg ee o synch oniza ion o he esponse o an a ay o neu ons, as illus a ed in Figu e 3-4
c). Simila o he p e ious case, he au ho s epo ha he eason behind his phenomenon is a delay
o ad ancemen o he nex ac ion po en ial, which is caused by posi i e o nega i e in e e ence
de e mined by he phase a which he noise s imulus a i es.
Figu e 3-3: G aphical ep esen a ion o low pass il e ed and whi e noise
wi h
ampli ude alues d awn om a Gaussian dis ibu ion ( om e e ence [107]).
30
In he amewo k o bo h mimicking na u al unc ional beha iou and o in e acing ANN
wi h he biological sys ems i is qui e in e es ing o s udy he ope a ion o OMD unde he exposu e
o noise. Mo e speci ically, we in es iga ed he exposu e o noise as a sou ce o po en ial spikes ha
could d i e OMD swi ching in analogy o he expe imen s demons a ing he STDP o he de ice
[100],[105]. To his end, we s udied he kine ic esponse o he de ice a ol ages close o he
swi ching onse po en ial. As i is known, he elec ochemical ansi ion o he PANI channel in o i s
conduc i e s a e only occu s abo e his h eshold alue ( ypically a a ound +0.3 V bias), wi h as e
kine ics a highe applied po en ials (wi hin he s abili y ange o PANI-ES). Ou expec a ion was ha
by applying elec ical noise wi h a su icien ampli ude, a p ema u e swi ching o he de ice could be
achie ed a sub- h eshold po en ials due o a po en ia ion by he noise’s posi i e po en ial spikes, as
illus a ed in Figu e 3-5 a, b).
Figu e 3-5 c) shows he kine ic esponse o an OMD, ab ica ed using he classical p ocedu e
[109], a a cons an +0.3 V bias, supe imposed by a colou ed elec ical noise signal o di e en
equencies, measu ed in duplica e. A con inuous, as decay o he ou pu cu en as well as swi ching
kine ics is obse ed bo h be ween and wi hin each se o measu emen s, indica ing some sho - e m
ma e ial deg ada ion. Ou assump ion was ha his is due o he solid polyelec oly e as he mos
labile componen o he OMD, being p one o excessi e d ying as well as acid dedoping due o HCl
ola ili y. On his basis we ca ied ou an ex ensi e s udy ha is p esen ed and discussed in de ail in
chap e 4. Wi hin he amewo k o his expe imen , we a emp ed o ci cum en he decay o he
de ice’s cha ac e is ics by p o iding a s able humidi y en i onmen o he ac i e zone o he de ice
by means o capping, which is desc ibed in mo e de ail in sec ion 4.2. As i can be seen in Figu e 3-5
Figu e 3-4: Spike ime eliabili y as a unc ion o he magni ude o cu en luc ua ions
wi h a) a b oadband cu en inpu and
b) a cu en inpu lacking equencies a ound DC (adap ed om [108]). c)
Synch oniza ion e ec o an elec ical noise s imulus on he
i ing pa e n o an a ay o uncoupled model neu ons. The neu ons’ egula
and independen i ing in a diagonal pa e n is ans o med
in o a clus e ed pa e n wi h mo e synch onous i ing ac oss he popula ion (adap ed om [107]).
31
d), his app oach indeed gi es a ele an s abiliza ion o he sys em ha deli e s a much mo e s able
and ep oducible ou pu . We hence p oceeded by ca ying ou kine ics measu emen s a di e en
applied cons an ol ages close o he onse po en ial, compa ing he esponse o a bias wi h and
wi hou he applica ion o a 20 Hz noise signal wi h an ampli ude ange o 0.1 V, acqui ed in
succession. Two se s o esul s a e p esen ed in Figu e 3-6. In he i s se ies o measu emen s, no
di e ence is isible be ween he mean ou pu cu en in he case o he s imulus wi h supe imposed
noise and he esponse a he espec i e bias wi hou noise applica ion. This sugges s ha he
pos ula ed e ec o swi ching enhancemen by noise did no apply, which could be caused by a
nega ion o he po en ia ion by posi i e ol age spikes h ough he in e se e ec o nega i e ol age
spikes, be o e he swi ching can ini ia e. On he o he hand, a small di e ence can clea ly be seen in
he second se o measu emen s. Conside ing he cons an beha iou in he i s se ies, we assume
ha he sligh decay o cha ac e is ics wi hou he applica ion o noise in he second se ies migh ,
again, be caused by slow ma e ial deg ada ion.
Figu e 3-5: a) G aphical ep esen a ion o a p ema u e ini ia ion o OMD swi ching a sub-
h eshold po en ial alues, caused by a
po en ia ion by he posi i e po en ial spikes o a supe imposed noise signal (b). c) Kine ic esponse o an OMD a a cons an
bias o
+0.3 V and an applied colou ed noise o di e en equencies and an ampli ude ange o 0.1 V, exposed o ambien condi ions.
d)
Kine ic esponse o he de ice a a cons an +0.3 V bias on a capped de ice, p o iding humid condi ions a ound he ac i e zone.
38
as he ela ed eac ion kine ics a e he mos ele an aspec s o be conside ed in he con ex o he
de ice unc ioning. Hence, i becomes clea e which componen s a e essen ial o a p ope OMD
ope a ion and which ones, in u n, can be a ied in o de o imp o e he sys em in e ms o s abili y
and pe o mance. Al hough a signi ican p og ess was made by he s udy o Ci a elli e al., he e is
s ill he need o a be e unde s anding o he de ice ope a ion ha is c ucial o u he enhancing
he pe o mance o he sys em. In pa icula , one challenge is o o e come he p esen OMD
limi a ions in e ms o ep oducibili y and endu ance ha a e no ye ully adequa e o de eloping a
obus echnology. Some p oblems a ise om he ma e ial deg ada ion o he de ice’s componen s,
conce ning PEO, PANI and he sil e ga e elec ode. In all likelihood, i is he de e io a ion o he
o me ha is mos ly esponsible o he a he quick decay o he OMD’s pe o mance o e ime. In
iew o p e ious indings, in he ollowing sec ion, he e ised ope a ion mechanism is u he
in es iga ed. In pa icula , he goal is o p o ide u he p oo o he assump ions made abo e o a
be e unde s anding o he e ec s o doping addi i es in he solid polyelec oly e, and o sugges
iable app oaches o imp o e OMD pe o mance and eliabili y.
4.1.1 Li+ ee SPE con aining o he hyg oscopic sal s
I mus be poin ed ou ha , despi e no being di ec ly in ol ed in OMD swi ching, he li hium
sal plays a e y impo an ole as an SPE addi i e due o i s hyg oscopici y, as i p o ides he
necessa y wa e con en wi hin he polyelec oly e gel o lowe he ac ion o c ys allized PEO. This
is o pa amoun impo ance o he de ice’s pe o mance since, as i is well known, a PEO-based
polyelec oly e demons a es i s bes ionic conduc i i y in a ully amo phous s a e when su icien ly
hyd a ed [124]. Thus, he concen a ion o he dopan emains a signi ican pa ame e in he
ab ica ion o OMD. F om now on, when dopan concen a ions a e men ioned, hey a e always
ela ed o he p epa a ion o he gel, i.e., he s ep in which he gel is p epa ed by adding PEO o an
aqueous solu ion o he espec i e dopan sal . As he SPE is deposi ed, i solidi ies by losing he
majo i y o i s wa e con en upon e apo a ion, so ha he inal p opo ion o he dopan in he SPE
inc eases. A de e mina ion o he inal wa e con en and PEO o dopan concen a ion could be made
by p ecise weighing o he gel upon deposi ion and a e i s solidi ica ion.
To be e quali y he ole o hyg oscopic addi i es, in he p esen wo k we i s examined
how he de ice ope a es when subs i u ing LiClO4 o LiCl in PEO-based polyelec oly e wi h o he
hyg oscopic sal s. Ou expe imen al inding a e he subjec o a li e a u e wo k [125]. To his end,
we p epa ed a ious PEO gels con aining a ious sal s, namely MgCl2, CaCl2, RbCl, CsCl, CeCl3,
Mg(NO3)2, MgSO4 and Na2SO4. We also added an acid o he PEO gel o main ain he PANI
su icien ly doped upon con ac , hus p e en ing i s ansi ion o he nonconduc i e eme aldine base
s a e. Hyd ochlo ic acid was used mainly o ensu e a su icien ly low pH le el while in he case o
he gel con aining Mg(NO3)2 he HCl also p o ided he chlo ide ions which a e necessa y o bo h
hal - eac ions o occu . An excep ion was he gel con aining MgSO4 whe e H2SO4 was used ins ead
o HCl. Some o he gels, namely he ones wi h RbCl, CsCl, Mg(NO3)2, Na2SO4 and MgSO4 sal s,

39
p o ed o be insu icien ly hyd a ed in he solid s a e, leading o in ense PEO and/o sal
c ys alliza ion. The wo s case is he one based on MgSO4. In hese SPEs, swi ching was no
achie able because o he limi a ions on ionic anspo . The change om he amo phous s a e is due
o he hyg oscopici y o he men ioned sal s ha is no enough o main ain he needed le el o
hyd a ion in he gel ha , hence, unde goes a d ying p ocess when exposed o he su ounding
a mosphe e in ambien condi ion. In he con ex o swi ching kine ics, his e ec was shown by
Bu gano a e al. [126] when a ying he LiClO4 concen a ion in a s anda dly ab ica ed OMD. In
his wo k, oo low dopan concen a ions (0.025 M LiClO4) led o s onge PEO c ys alliza ion,
because he amoun o wa e ha could be e ained in he gel was no enough o main ain i s
amo phous s a e in he whole bulk o he polyelec oly e. Con e sely, he o me ly common
concen a ion o 0.1 M LilO4 also appea ed o be oo high, as in his case, he ime equi ed o all o
he ionic cu en o pass h ough he polyelec oly e was he highes . In his case, he obs uc ion o
cu en p obably was due o he o ma ion o LiClO4 sal clus e s ha , in u n, may ha e se ed as a
sou ce o nuclea ion o PEO. An in e media e concen a ion o 0.05 M LiClO4 in he gel p epa a ion
showed he as es kine ics in his wo k, making he i s s ep owa ds he op imiza ion o he de ice.
The beha iou o he gel con aining MgSO4 seems odd a i s glance, because his sal is
known o be e y hyg oscopic, so i should gua an ee a su icien wa e con en . In his case, he
c ys alliza ion is due o ano he ype o p ocess, he so-called sal -induced p ecipi a ion (“sal ing ou ”)
o PEO in he p esence o sul a e. The unde lying concep o kosmo-and chao opici y o dopan ions
is p esen ed in sec ion 4.2 o he p esen hesis wo k as i will be mo e ele an o he de elopmen
o PEO gels wi h imp o ed du abili y.
The occu ed hind ance wi h insu icien ly hyd a ed gels was p elimina ily esol ed by
c ea ing an a mosphe e wi h inc eased humidi y a ound he sample. This was achie ed by placing a
cap con aining a we ball o co on on op o e he ac i e zone. As shown in Figu e 4-5, he gels
pa ially clea ed and lique ied wi hin ew minu es, allowing o ca y ou a s anda d elec ical
cha ac e iza ion o he mem is i e de ice. Examples o cu en - ol age cu es o OMDs p epa ed
wi h he men ioned SPE a e shown in Figu e 4-6 a). All de ices ha we e ab ica ed wi h
polyelec oly es con aining chlo ide sal s displayed ypical mem is i e beha iou . I is no ewo hy
ha he OMD con aining MgCl2 shows pa icula ly sa is ac o y cha ac e is ics in e ms o he
hys e esis shape and, mo e impo an ly, ep oducibili y. This is a good indica ion o he quali y and
s abili y o he PEO gel due o a a ou able combina ion o hyg oscopici y and chao opici y o he
sal ions. Fo he de ice con aining Mg(NO3)2, no mem is i e beha iou was obse ed e en by
placing he cap wi h he we co on abo e he ac i e zone, used o main ain he PEO in i s amo phous
s a e. This may be well unde s ood by conside ing he oxidizing cha ac e o he ni a e ions, which
p e en s PANI om being educed (o main ained) o he leucoeme aldine o m when a emp ing o
swi ch o he de ice. The same e ec , al hough weake , likely occu s in he de ices wi h PEO
con aining MgSO4 and Na2SO4. When applying he cap o lique y he SPE and ca ying ou s anda d
cu en - ol age measu emen s, hese de ices display cha ac e is ics ha a e somewha simila o
40
hose o a ypical OMD. Figu e 4-6 b) shows an I-V cu e o a de ice ab ica ed wi h PEO con aining
MgSO4, eco ded wi h a dwell ime o 20 s, which exhibi s some swi ching, cha ac e ized by he
ec i ica ion o he cu e a nega i e applied biases and a na ow hys e esis. This indica es ha he
de ice shows as e o e all kine ics as compa ed o hose in panel a) a simila measu ing condi ions.
When he dwell ime is dec eased o 5 s pe ol age s ep, a signi ican b oadening o he hys e esis is
obse ed, ye he de ice is s ill as enough o swi ch. On he con a y, inc easing he dwell ime o
60 s causes an e en s onge na owing o he hys e esis o a poin whe e i is ba ely p esen . This is
because du ing he backwa d sweep, he kine ics o PANI educ ion is as e han he dwell ime. I
is no ewo hy ha he elec onic cu en in he “o ” s a e a nega i e ol age biases emains in he
Figu e 4-5: a) Pho og aph o a commonly ab ica ed OMD, b) co e ed wi h a plas ic cap wi h a we co on ball unde i s sealing; Top-
iew mic opho og aphs o he OMD c) be o e and d) a e he exposu e o he humid a mosphe e h ough capping (SPE made wi h 3
0
mg/ml PEO concen a ion and 0.1 mol/l RbCl).
Figu e 4-6: a) Hys e esis cu es o a se ies o OMDs wi h SPE made om PEO solu ions (as colo -
coded and labelled in he igu e).
The samples wi h RbCl and CsCl we e mea
su ed unde humid a mosphe e by applying a cap wi h a we ball o co on on op o he
ac i e zone , dwell ime o each acquisi ion s ep o 0.1 V = 30 s.
b) I-V cha ac e iza ion o an OMD wi h SPE made om a PEO solu ion con aining 0.05 M MgSO4, measu e
d unde humid a mosphe e
wi h di e en dwell imes pe ol age s ep as labelled.
41
o de o µA (compa ed o nA o a ypical OMD) in all h ee cases. Compa able esul s we e ob ained
o he de ices made wi h PEO con aining Na2SO4. They indica e ha he swi ching-o ook place
only pa ially and an equilib ium s a e was es ablished in which some o he ma e ial emains in i s
oxidized, conduc i e o m. In sec ion 4.1.3, he ole o he sul a e ion as an oxidizing agen will be
discussed in u he de ail.
4.1.2 The ole o dopan ca ions
Based on he abo e expe imen al esul s, we obse ed ha li hium ions, in speci ic, a e no
necessa y o OMD ope a ion, since de ices ab ica ed wi h SPE con aining o he hyg oscopic
chlo ide sal s also showed swi ching. Howe e , his does no exclude he possibili y ha in hese
cases, he ini ially sugges ed ole o Li+ [93],[94] is pe o med by ano he ca ion (Mg2+, Ca2+, Cs+
e c.), as i has been implied in he undamen al wo k dedica ed o explo ing he de ice ia XRF [88].
As discussed by Ci a elli e al. [114], he p esence o Li+ ions in he polyelec oly e is no
necessa y o he unc ioning o he OMD, whe eas he ole o he sil e ga e elec ode and he
p esence o chlo ide ions is c ucial. We aim a demons a ing wi h addi ional e idence he no ion ha
dopan ca ions do no play a decisi e ole in he elec ochemical eac ion de ining he OMD’s wo king
p inciple shown in Equa ion 7 and Equa ion 8. Ou app oach is o show he independence o he
OMD’s swi ching kine ics om he ype o ca ion p esen in he elec oly e. F om he XRF
expe imen s o e e ence [88] i is known ha he mig a ion eloci y o he bulkie Rb+ in PEO is
lowe wi h espec o Li+ ions. Hence, i he ca ions pa icipa e in he OMD’s swi ching mechanism,
we expec o obse e a di e ence in he cha ac e is ics o he co esponding de ices. A di ec
compa ison o de ices wi h a ious composi ions o he PEO-based polyelec oly e (as discussed in
sec ion 4.1.1) is di icul because o he di e en hyg oscopici y o he dopan sal s de ining he SPE’s
inal phase composi ion. E en wi h equal ion concen a ions and p o iding a humid a mosphe e using
he cap, he condi ions o he polyelec oly es could s ill be signi ican ly di e en om each o he .
Fo his eason, we decided o make he compa ison using equally concen a ed liquid elec oly es
whe e he sal hyg oscopici y has no e ec a all, lea ing whiche e di e ences may occu o be
a ibu able solely o he dopan ions wi h di e en ionic adii and mobili ies. This goal equi ed a
modi ied expe imen al se up simila o he one p oposed in e e ence [88], which is shown Figu e
4-7.
Figu e 4-7: a)
Pho og aph and b) schema ic ep esen a ion o he se up o kine ics cha ac e iza ions o de ices wi h liquid elec oly es.
The well dep h and wid h is 3 mm.
42
A se ies o solu ions o mono alen me al chlo ides (LiCl, NaCl, KCl, RbCl, CsCl) in
aqueous HCl we e chosen o bo h simpli y and uni y he sys em, he la e meaning ha he only
di e ence in SPEs would be he added ca ion. Rela i ely low concen a ions o bo h HCl and he
dopan (0.01 N and 0.001 M, espec i ely) we e chosen because highe concen a ions could possibly
e en ou g adien s and mask e en ual kine ics di e ences caused by di e en dopan ions. Because
o he need o an acidic en i onmen , we could no educe u he he HCl concen a ion, since i
would lead o he educ ion o PANI-ES o PANI-EB and, hence, o a loss o conduc i i y. To p o ide
e en be e compa abili y, all measu emen s we e pe o med on he same PANI ilm, ho oughly
insing i when changing he elec oly e and ca ying ou each measu emen epea edly in o de o
moni o he p esence o a possible aging o he sample. We ca ied ou h ee se ies o expe imen s
using he same PANI ilm, epea ing kine ic measu emen s when a ying polyelec oly e solu ions
one by one in each unique se ies. The ob ained kine ics o all he se ies show he same end.
The ypical kine ic esponses o he OMD o di e en me al chlo ide elec oly es a e shown
in Figu e 4-8 a). They appea o be qui e simila wi h di e ences in he no malized cu es ha a e
wi hin he ma gin o e o , as con i med by he i pa ame e s in Table 2. We could no obse e any
end in co ela ion wi h he ionic adii and he es ablished ionic mobili ies [127]. Howe e , when
conside ing he cu es wi hou no maliza ion, one pa icula i y mus be poin ed ou . The sa u a ion
cu en esul s o be sligh ly highe o he elec oly es con aining caesium and ubidium chlo ide in
mos cases. I is especially peculia ha no end wi h espec o ionic adii is obse ed h oughou
he es o he se ies. This obse a ion could no be adequa ely explained in he amewo k o he
showed s udy. As a pe spec i e, i would be in e es ing o assess whe he his e ec could be
associa ed wi h he s eng h o he in e ac ions o hese ca ions wi h chlo ide ions. A s onge
in e ac ion wi h chlo ide ions compensa ing he posi i e cha ges on oxidized PANI chains could
cause a s onge depinning o pola ons om hei coun e ions, inc easing cha ge mobili y and
conduc i i y.
Figu e 4-8: a) No malized and b) no no malized kine ic esponse o OMD wi
h liquid elec oly es con aining di e en alkali me al
chlo ide sal s a an applied ol age bias o +0.5 V. Reduc ion pe iods a −0.2 V a e omi ed.
43
Table 2: Double exponen ial i pa ame e s ((τ ep esen a e cons an s) o cu es esul ing om kine ics measu emen s as in Figu e
4-8 a).
Fi equa ion: 







A leas nine single measu emen s ha e been ca ied ou o each elec oly e wi h a o al numbe o 52 measu emen s used o he
s a is ics. In he las ow, a mean alue o he calcula ed a e age τ pa ame e s can be seen wi h he co esponding s anda d e o .
4.1.3 SPE con aining Li+ wi h an addi ion o [12]c own-4
The esul s shown up o now indica e ha he e is no signi ican e ec o he ca ion adius
on he kine ic esponse o he de ice. Howe e , he obse ed inconsis encies in he sa u a ion cu en s
o de ices ab ica ed wi h SPE con aining di e en ca ions could gi e ise o some doub s. We hence
ca ied ou ano he expe imen , simila in design o he p e ious one, wi h he idea o obse e he
kine ic esponse o a de ice when adding a subs ance ha in e e es signi ican ly wi h he ca ionic
mobili y. In de ail, we compa ed he esponse o de ices wi h wo equal elec oly es con aining LiCl
as he solu e, one o which con ained [12]c own-4 (12C4) o equimola concen a ion. C own e he s
a e ep esen a i es o he class o co onands ha a e known o o m s able complexes wi h alkali
me al ions (among o he s) depending on he size o hei ca i y. The 12C4 is one such co onand,
ha ing a ca i y adius o 0.72−0.81 Å [128], ma ching he ionic adius o Li+ (0.74−0.86 Å), hence
sui able o bind li hium ions, as shown in Figu e 4-9.
The e a e nume ous epo s in li e a u e showing ha such a complexa ion a ec s he
mobili y o ionic species [129]–[132]. Following he o me assump ion ha li hium ions play a ole
in he ope a ion mechanism o he OMD, a signi ican di e ence should be expec ed in he kine ic
esponse o he de ices when compa ing he wo abo e men ioned elec oly es. We hen ca ied ou
he same measu emen s shown in Figu e 4-9 on he same subs a e, using he app oach discussed in
he p e ious sec ion. We he e o e chose he se up wi h he liquid elec oly es o he same easons
discussed in sec ion 4.1.2 (Figu e 4-7). The no malized cu es co espond a he well wi h one
ano he , keeping he same shape and ends, sugges ing ha he addi ion o [12]c own-4 o he
elec oly e causes no signi ican change in he kine ic esponse. The co esponding double
exponen ial i pa ame e s a e epo ed in Table 3. Some insigni ican di e ences a e obse ed in he
I/Imax sa u a ion alues be ween he se ies cha ac e izing he elec oly e wi h and wi hou he c own
e he , whe e he la e a e sligh ly lowe han hei coun e pa s. This is likely due o changes in he
ma e ial i sel because all he measu emen s we e ca ied ou in succession and a consis en , small
decay can also be obse ed wi hin bo h sepa a e se ies.
dopan a g. τ
1
σ(τ
1
) a g. τ
2

LiCl 7.37 0.78 108.02 7.15
NaCl 7.64 1.03 124.92 6.48
KCl 7.53 0.76 123.18 4.82
RbCl 8.03 1.02 113.65 6.52
CsCl 6.76 0.81 106.03 5.47
mean 7.38 0.19 115.16 3.85

44
Table 3: Double exponen ial i pa ame e s (τ ep esen a e cons an s) o cu es esul ing om kine ics measu emen s as in Figu e
4-9.
Fi equa ion: 







4.1.4 The ole o dopan anions
In he abo e-desc ibed expe imen s, he choice o a liquid elec oly e may lead o hink ha
ha e en ual di e ences in he de ice cha ac e is ics could be masked o e en neu alized by he as
(ca ionic) di usion p ocesses occu ing in he liquid s a e, e en a ela i ely low concen a ions o
dopan s. In o de o p o ide a u he insigh on he ole o ionic species, we also in es iga ed whe he
changing he anion would a ec he mem is i e cha ac e is ics o he OMD. Fo a be e compa ison
o he esul s, we decided o con on anions wi h he same cha ge bu o di e en size. Fo ou
sys em, he choice o monop o ic acids o be compa ed wi h he s anda dly used HCl is somewha
limi ed. Common o ganic acids such as ace ic acid a e oo weak o p o ide a pH ha would dope
PANI a app op ia e concen a ions, while he hyd ogen halides HF and HI a e oxic, co osi e o
could display some edox chemis y o hei own, which could in e e e wi h he measu emen . O he
possible candida es as dopan s o PANI a e s ong o ganic (sul onic) acids. Al hough he e a e
epo s [132],[133] o polyanilines doped wi h CSA, TSA o DBSA, hese a e ypically syn hesized
compounds, made by polyme iza ion o aniline in he espec i e acid solu ion, yielding he sel -doped
elec oly e cu e τ
1
τ
2
LiCl
1 15.64 164.00
2 12.24 117.51
3 13.03 107.76
LiCl
+ 12C4
1 11.45 106.08
2 12.37 101.91
3 12.76 98.61
Figu e 4-9
: Kine ic esponse o OMD wi h liquid elec oly es con aining LiCl wi h ( ed) and wi hou (blue) he equimola addi ion o
[12]c own-4 a an applied ol age bias o +0.5 V. Reduc ion pe iods a −0.2 V a e omi ed. The molecula s uc u e o he li hium-
12C4 complex is shown in he inse . Sandwich complexes o one li hium ion wi h wo 12C4 molecules a e also common.
45
polyme o m. Ou de ices a e ab ica ed wi h p is ine PANI wi h highe mean molecula weigh
(Mw = 105 Da) in i s eme aldine base o m, he doping o which was no success ul wi h any o hese
h ee sul onic acids in ou expe imen s. We assume ha his is due o he high densi y o he deposi ed
PANI ilm and he low mobili y o he espec i e anions. Fo hese easons, he ange o elec oly es
ha we used was limi ed o HCl, HB , HClO4 and H3PO4.
In a i s se ies o expe imen s, a compa ison o he de ice’s kine ics in liquid elec oly es
ea u ing di e en anions (i.e., di e en dilu ed acids) was ca ied ou ollowing he measu ing
ou ine desc ibed in Chap e 2. To p o ide equal condi ions in each expe imen , we p epa ed
elec oly es a he same pH and concen a ion o 0.1 mol/l, assuming he ull dissocia ion o he
s ong acids HCl, HB , and HClO4. In he case o he ela i ely weak H3PO4 (pKa1 = 2.124 [134]), he
concen a ion o choice was 1.43 mol/l in o de o p o ide equal H3O+ and anion (H2PO4-)
concen a ions o 0.1 mol/l. While he second and hi d dissocia ion s eps in oduce HPO42- and PO43-
o he sys em, hei concen a ions a e negligibly low and hence a e no conside ed. As i is shown in
Figu e 4-10 a), he de ice swi ches on and o in each case wi h a clea di e ence in he kine ics
depending on he elec oly e. This becomes e en mo e e iden upon i ings he acqui ed cu es by a
double exponen ial unc ion, gi ing he a e cons an s τ1 and τ2 lis ed in Table 4. We obse ed ha in
he case o HCl and HB , he maximum cu en was decaying du ing he se ies. We ound ha his is
p obably due o he ola ili y o hese compounds because he same was no obse ed o he non-
ola ile HClO4 and H3PO4. I appea s ha he esponse is qui e simila in HCl and HB , which is in
good co ela ion wi h he qui e simila ela i e ionic mobili ies o chlo ide and b omide, epo ed in
Table 4 [135]. The small di e ence obse ed in he onse ime can be conside ed as negligible. I
occu s a ew seconds ea lie in he case o HB , likely due o he di e en eac ion kine ics o chlo ide
o b omide on ei he one o he elec odes (PANI, Ag o bo h). An ins an swi ching onse is obse ed
o HClO4, which we assume o be due o he oxidizing cha ac e o pe chlo ic acid, which p omo es
he ansi ion o PANI o i s conduc i e s a e e en a low applied ol ages. This is u he con i med
by he ela i ely high cu en s du ing he educ ion pe iod a −0.2 V, sugges ing ha he ansi ion
om he conduc i e eme aldine sal s a e o he nonconduc i e leucoeme aldine s a e is incomple e
and some conduc i i y is main ained, analogous o solid s a e de ices p oduced wi h sul a e sal s,
men ioned in sec ion 4.1.1. The swi ching kine ics, howe e , appea s o be slowe han in HCl o HB
(see slopes in he inse o Figu e 4-10 a), which co esponds o he lowe ela i e ionic mobili y o
pe chlo a e wi h espec o chlo ide o b omide [135]. A clea e indica ion o he co ela ion o
kine ics wi h he ionic mobili y is obse ed in H3PO4 which does no ea u e any addi ional edox
ac i i y o i s own. The onse o he swi ching occu s la e han ha obse ed o HCl and HB , again,
possibly because o slowe eac ion kine ics on Ag and/o PANI. Once he swi ching is ini ia ed, he
cu en inc eases signi ican ly slowe in compa ison o he o he acids. This is in good co ela ion
wi h he signi ican ly lowe ionic mobili y o dihyd ogen phospha e (mo e han a ac o 2).
46
Table 4: Double exponen ial i pa ame e s (τ1 and τ2 ep esen he a e cons an s o he i ings) o cu es esul ing om kine ics
measu emen s in Figu e 4-10 a). The ionic mobili y alues uX/uK+ ( ela i e o ha o K+) a e aken om e e ence [135].
To u he suppo ou indings, in a second se ies o expe imen s, we compa ed he kine ics
o he de ices in a mono- and a dip o ic acid (HCl and H2SO4). Al hough he expe imen s wi h MgSO4
discussed in sec ion 4.1.1 show ha H2SO4, like HClO4, also has an oxidizing cha ac e , PANI doped
wi h H2SO4 shows e y simila conduc i i y o HCl doping, sugges ing ha he second oxida ion s ep
owa ds he nonconduc i e pe nig aniline s a e does no ake place wi hou any applied bias ol age.
Keeping in mind he di alen cha ac e o he sul a e ion, he espec i e po assium and magnesium
sal s (KCl and MgCl2 o HCl, K2SO4 and MgSO4 o H2SO4) we e added o he elec oly e o a mo e
meaning ul compa ison. To keep he pH equal in each expe imen , we chose an acid concen a ion o
0.01 N, assuming comple e dissocia ion o H2SO4. The e o e, he amoun o sal addi i e was chosen
o ob ain ha he o al concen a ion o he chlo ide and sul a e ions, espec i ely, we e also equal in
each case (0.02 M). As a consequence, due o he di e en alencies o he ca ions, he espec i e
ca ion concen a ions also di e ed (e.g., when measu ing wi h HCl, 0.005 M MgCl2 o 0.01 M KCl
we e added o achie e an equal o al chlo ide concen a ion o 0.02 mol/L). This s ep allowed us o
u he assess he ole o he concen a ion, size and elec ic cha ge o he ca ions. As shown in Figu e
4-10 b), no signi ican di e ence could be obse ed when compa ing elec oly es con aining he same
acid bu di e en ca ions, which u he p o es he conclusions o sec ion 4.1.2. A he same ime,
elec oly e τ
1
τ
2
u
X
/u/
K
+
0.1 M HCl 1.89 49.86 1.04
0.1 M HB 1.91 53.24 1.06
0.1 M HClO
4
3.35 80.66 0.92
1.43 M H
3
PO
4
7.59 79.20 0.45
Figu e 4-10: a Kine ic esponse o OMD in HCl, HB , HClO4 and H3PO4 a +0.5 V (oxida ion pe iod, 5 min) and −
0.2 V ( educ ion
pe iod, 5 min). b) Kine ic esponse o OMD wi h HCl (blue, cyan) and H2SO4 ( ed, magen a)
based elec oly es a +0.5 V (oxida ion
pe iod, 10 min) and −0.2 V ( educ ion pe iod, 5 min).
Fo each elec oly e, om a se ies o h ee consecu i e measu emen s, only he
i s is shown in he igu e.
47
he di e ence be ween he pe o mance o he OMD wi h elec oly es con aining di e en anions is
e y p onounced – a signi ican ly as e esponse is obse ed in H2SO4. This could be in a i s
ins ance a ibu ed o he bi alen na u e o he sul a e ion, sugges ing ha a la ge cha ge pe uni o
ime is ca ied om one elec ode o ano he compa ed o he chlo ide ion. Howe e , his is only one
pa o he pic u e since he sul a e ion is also signi ican ly less mobile in aqueous solu ions [135].
The inal balancing is unclea and hence i is di icul o come o a conclusi e unde s anding. As
shown in Figu e 4-10 b), he e is a clea decay o he cu en a e he maximum achie ed a e he
swi ching-on. This sugges s ha a po ion o he PANI-ES is u he oxidized o he nonconduc i e
pe nig aniline s a e a he applied ol age bias (+0.5 V). Besides, when applying a nega i e d ain
ol age, he measu ed nega i e cu en is compa a i ely high, indica ing an e ec simila o ha
obse ed o he de ice ea u ing HClO4 (see Figu e 4-10 a). The oxidizing cha ac e o dilu ed
sul u ic acid mus be conside ed, since i p esumably hinde s he educ ion o PANI-ES o PANI-LE
when swi ching o he de ice, gi ing ise o he la ge cu en a nega i e d ain ol age. On his basis,
i is deba able whe he he ema kable accele a ion o he kine ics is due o he na u e o he anion
alone o assis ed by his ex aneous elec ochemis y. Fu he expe imen s wi h sul u ic acid go
beyond he scope o his wo k since we es ablished ha sul a e sal s a e no sui ed o be used in SPE
o mulas based on PEO.
To summa ize, we p o ided a deepe insigh and u he de ailed da a ha a e consolida ing
and alida ing he mechanism p oposed by Ci a elli e al. [114]. The o dina y unc ioning o he
de ice was success ully shown in a se ies o expe imen s wi h PEO gels con aining a se ies o
di e en dopan sal s. Ou indings [125] s ongly indica e ha LiClO4, commonly used in OMD
de ices, has he sole unc ion o p o iding a ou able condi ions o PEO-based polyelec oly e o
main ain i s amo phous s a e, and hence i can be eplaced by ano he su icien ly hyg oscopic sal .
The e o e, i does no play a speci ic ole in he mem is i e mechanism. Addi ional e idence is also
gained om kine ics s udies p o ing ha he na u e o he ca ion (i.e., i s size and alence) ha e no
signi ican e ec s on he mem is i e esponse o he OMD. A u he , possibly mo e con incing
con i ma ion o his conclusion is gi en by he ac ha no signi ican di e ence could be egis e ed
when compa ing OMD wi h elec oly es con aining Li+ and i s 12C4 complex. Based on ou indings,
one could conclude ha in he OMD swi ching mechanism, he ele an cha ge ans e is p o ided
by he chlo ide ions, con i ming wha was p oposed by Ci a elli e al. Addi ional e idence o suppo
his mechanism is gi en by he expe imen s discussed he e whe e he choice o he anion has a
p onounced e ec on he OMD’s kine ics. The e ec is e iden when chlo ide ions a e eplaced by
sul a e ions wi h a di e en ionic cha ge and/o mobili y in he elec oly e, indica ing he ele ance
o he anion o he OMD’s ope a ion. O e all, his wo k con ibu es o a mo e de ailed unde s anding
o he chemical p ocesses unde lying he swi ching o he de ice ha is absolu ely necessa y o bo h
imp o e he sys em’s pe o mance and o e come i s weak poin s in e ms o pe o mance and
endu ance.
54
O e all, he a emp s o eplace aluminium chlo ide as he acidic componen o he SPE we e
unsuccess ul as his dopan appea s o ha e a supe io combina ion o p ope ies such as acidi y and
hyg oscopici y. One las acidi y-enhancing addi i e, namely p- oluenesul onic acid (TSA), has also
been employed, and he ela ed de ice pe o mance will be discussed below, in sec ion 4.2.3.3.
4.2.2 The concep o kosmo opici y, chao opici y and he Ho meis e se ies
As i has been es ablished abo e, he hyg oscopici y o he dopan is a c ucial pa ame e o
p o ide a necessa y con en o wa e o he SPE sys em. Howe e , expe imen s ca ied ou wi h gels
con aining e y hyg oscopic sul a e sal s, such as Na2SO4 and MgSO4, esul ed in SPEs ha we e
ne e heless e y p one o c ys alliza ion upon d ying. This impo an piece o in o ma ion shows
ha he e mus be ano he pa ame e o be conside ed when de eloping a new polyelec oly e mix u e.
Concep ually, he issue wi h sul a e ions lies in he same a ea as hyg oscopici y, being he in e ac ions
o sul a e ions wi h wa e molecules. As i has been men ioned abo e, he kosmo- o chao opici y o
he dopan ions has been es ablished as he second c ucial concep o he selec ion o app op ia e
addi i es in a PEO gel composi ion.
The concep s o kosmo opici y ( om “kosmos” – o de ) and chao opici y ( om “chaos” –
diso de ) was concei ed in he ield o p o ein pu i ica ion [142]–[144]. Acco ding o hese concep s,
he addi ion o ce ain sal s o an aqueous solu ion o mac omolecules, such as p o eins, can in luence
he s abili y o hei na i e s a e. A dis inc ion is made be ween he so-called “sal ing ou ” caused by
kosmo opic agen s, meaning ha he added solu e causes he p ecipi a ion o he p o ein, and “sal ing
in” caused by chao opic agen s and implying a s abiliza ion in he dissol ed s a e. Depending on
hei kosmo opic o chao opic ac i i y, ionic solu es as well as o ganic amphiphilic compounds,
hyd o opes, su ac an s and e en o ganic sol en s can be ca ego ized and o de ed in a so-called
lyo opic se ies. The lyo opic se ies o ca ions and anions is commonly e e ed o as he Ho meis e
se ies, acco ding o which di e en ions a e o de ed acco ding o hei capaci y o p ecipi a e p o eins.
O e all, anions appea o ha e a s onge e ec han ca ions. In ea lie s udies, he kosmo opic and
chao opic e ec s we e said o s em om he o de ing o diso de ing o he s uc u e o wa e and
inc easing o dec easing i s su ace ension, espec i ely [145]. Mo e ecen s udies indica e ha ions
do no a ec he bulk wa e p ope ies [146]–[149], indica ing ha di ec in e ac ions be ween he
ions and mac omolecules mus be conside ed o unde s and lyo opic phenomena [150]. The
indi idual e ec s o he ions depend on he mac omolecule. Fo ins ance, in p o eins, many di e en
amino acid moie ies a e a ached o he pep ide backbone and may be hyd ophobic, pola , o cha ged.
The sum o he in e ac ions o said side chains dic a es he na i e olding and he e ia y/qua e na y
s uc u e o he mac omolecule. Chao opic ions in e ac wi h he backbone o side chains o he
p o ein in such a way ha he na u al amewo k o wa e -media ed in e ac ions is dis up ed. The
hyd ophobic e ec is weakened and he e ia y s uc u e un olds, p omo ing he solubiliza ion and
causing a dena u a ion o he p o ein. Con e sely, kosmo opes, such as he commonly used

55
ammonium sul a e, s eng hen he hyd ophobic e ec , causing he p ecipi a ion o he p o ein, ideally
p ese ing i s na i e s uc u e.
In he case o he SPE con aining MgSO4, co espondingly, he p ecipi a ion o PEO is
caused by sul a e ions as an ea ly membe o he Ho meis e se ies. These ions in e ac mo e eadily
and equen ly wi h wa e molecules han he PEO chains. As wa e becomes less p one o in e ac
wi h he pola −O− uni s o PEO molecules, he hyd ophobic effec is s eng hened, dec easing he
chain mobili y and hyd a ion, and hence causing s onge PEO agg ega ion [151],[152]. Con e sely,
he o me s anda d gels con aining LiClO4 displayed a e y a ou able mo phology as he chao opic
ClO4− ion dec eases he hyd ophobic e ec and p e en s PEO chains om agg ega ing. The e ec o
kosmo opic and chao opic ions on he mo phology o PEO is shown in Figu e 4-15. The chlo ide
ion is usually lis ed as a neu al anion in e ms o lyo opici y, meaning ha he endency o uncapped
AlCl3-based gels o c ys allize, as desc ibed in sec ion 4.2.1, is likely due o he Al3+ ion, despi e he
high hyg oscopici y o he sal . In e e ence [153], Al3+ is ea u ed as a kosmo ope, gi en i s small
size, high ionic cha ge and low pola izabili y ha sugges s ong in e ac ions wi h wa e . Howe e ,
his s udy ocuses on he in e ac ion o ions wi h wa e and he p esen ed lyo opic se ies o ca ions
se e ely con adic s he end gi en in o he wo ks [154]. Fo compa ison, wo di e en se s o
anionic and ca ionic Ho meis e se ies a e shown in Figu e 4-16. The di e en esul s ob ained in
a ious wo ks indica e ha he p e alen e ec s o ions canno always be gene alized and depend on
he sys em a hand. Howe e , he e appea s o be a mo e solid consensus conce ning he e ec o
anions.
Sha ma e al. [155] p esen ed a s udy o he e ec s o ino ganic ions and o he compounds
(such as su ac an s) on he cloud poin o a solu ion con aining a PEO-PPO-PEO-block-copolyme .
Thei esul s a e consis en wi h he anionic Ho meis e se ies. This con i ms ha he concep o
kosmo- and chao opici y can be applied o he in e ac ions o addi i es in PEO-based SPEs.
Howe e , cau ion mus be aken when e alua ing he applicabili y o some o he esul s p esen ed in
ha e e ence. Fo ins ance, he au ho s epo an inc ease o he cloud poin when sodium dodecyl
sul a e (SDS) is added, whe eas in ou case, he addi ion o SDS o he PEO gel caused apid
Figu e 4-15
: Simpli ied schema ic ep esen a ion o he in e ac ions o chao opic and kosmo opic ions wi h wa e and PEO chains.
Chao opes
end o in e ac mo e wi h PEO chains and less wi h wa e molecules, PEO chains a e main ained hyd a ed and amo phous.
Kosmo opes end o in e ac mo e equen ly wi h wa e , making i less a ailable o in e ac ions wi h PEO, leading o a s
eng hening
o he hyd ophobic e ec and agg ega ion o PEO chains.
56
p ecipi a ion a e i s deposi ion. The wo sys ems mus be di e en ia ed on he basis o he di e en
molecula weigh o he (co)polyme s and i s wa e con en . Because wa e is signi ican ly less
abundan in ou sys em, he hyg oscopici y o he solu es s a s o play an impo an ole, as i was
es ablished abo e. The lack o hyg oscopici y may cause ha he solu e claims a signi ican po ion
o wa e o i s own hyd a ion wi hou con ibu ing o i s o e all con en , hus cancelling possible
posi i e lyo opic e ec s. I is also wo h no ing ha speci ic p e e en ial in e ac ions such as he
o ma ion o ion-speci ic coo dina ion complexes may cause de ia ions om he classical Ho meis e
se ies [142],[152]. This ci cums ance makes compa isons o he kosmo- and chao opic cha ac e o
ions di icul , especially in he case o ca ions. Fo ins ance, he Mg2+ ion is p esen ed as a kosmo ope
in some s udies [151], while con e sely, i is said o ha e a he chao opic p ope ies in o he wo ks
[144].
The ollowing sec ion shows he use ulness o he Ho meis e se ies as a concep ha
complemen s he abo emen ioned impo ance o he hyg oscopici y o he dopan sal . I pa es he
way o he pe spec i e o u he explo ing a ou able SPE composi ions con aining solu es ha will
elimina e he nega i e e ec s o sal o polyme agg ega ion and p ecipi a ion caused by d ying o
sal ing ou p ocesses. These de elopmen s a e undamen al o ab ica ing de ices sui able o
ad anced applica ions in neu omo phic echnologies, as ep oducibili y and endu ance emain some
o he mos ele an challenges in cu en s a e OMDs.
4.2.3 Op imiza ion o in insically acidic SPE
The implemen a ion o aluminium chlo ide as an addi i e in he i s gene a ion o no el
PEO gels is an impo an s ep in he de elopmen o high pe o mance SPEs. While ul illing he
hyg oscopici y equi emen , he dopan also endows he gels wi h in insic acidi y, allowing o a oid
he addi ion o a ola ile acid o he p epa a ion. SPEs con aining AlCl3 enable as e swi ching
kine ics and show inc eased du abili y, leading o a mo e s able ope a ion, as he polyaniline emains
doped a all imes. Howe e , p oblems eme ge as he hyg oscopici y o he sal compe es wi h he
Figu e 4-16: Two se s o anionic and chao opic se ies om e e ences [154] (some anions om he o iginal publica io
n a e omi ed)
and [153]. No e ha he ca ionic se ies s ongly di e s be ween he wo e e ences.
57
kosmo opic cha ac e o he aluminium ion, which can cause he PEO o agg ega e when he
su ounding a mosphe e is insu icien ly humid. The de elopmen o s able and pe o ming SPEs
equi es he ul ilmen o some undamen al c i e ia o he selec ion o dopan candida es:
 Solubili y: he addi i e mus be wa e -soluble o be employed in wa e -based PEO gels,
especially conside ing he compa a i ely low le el o hyd a ion a e deposi ion;
 app op ia e anion: he anion mus pa icipa e in bo h hal - eac ions in he ac i e zone
(i.e. i bo h se es as a coun e ion o PANI-ES and eac s wi h he Ag ga e elec ode);
 in insic acidi y: a oiding he addi ion o ola ile acid compounds ensu es a s able pH;
 hyg oscopici y: upon solidi ica ion o a deposi ed gel, he dopan mus main ain enough
wa e in he sys em o keep PEO in i s amo phous s a e unde ambien condi ions;
 lyo opici y: he ions cons i u ing he dopan mus ha e a ne chao opic cha ac e , in
o de o dec ease he hyd ophobic e ec o PEO chains and o p e en agg ega ion.
4.2.3.1 AlCl3-based SPE wi h seconda y doping
Among all possible dopan candida es, aluminium chlo ide has been selec ed as he main
dopan because i mee s all he abo e equi emen s besides he lyo opici y, which may be
compensa ed by he in oduc ion o a seconda y addi i e. Fo his, we used he Ho meis e se ies as
an aid o choose sui able sal s showing lyo opici y. While i is no necessa y ha he seconda y solu e
pa icipa es in PANI swi ching o p o ides acidi y, i mus be aken in o accoun ha i also mus
emain hyd a ed a e he d ying p ocess. As i claims a pa o he wa e con en o i sel , a
combina ion o su icien hyg oscopici y and chao opici y is desi able also o he seconda y sal .
Fo example, in PEO gels wi h a mix u e o AlCl3 and MgCl2, an in ense agg ega ion s ill occu ed,
as i can be seen in Figu e 4-17. In his pa icula se ies o gels, i appea s ha a highe ac ion o
MgCl2 imp o es he quali y o he gel. Howe e , e en hough he magnesium sal is bo h hyg oscopic
and has a chao opic e ec [144],[152],[154], ei he one o bo h p ope ies appea o be insu icien ly
p onounced o ully compensa e he d awback o aluminium. I has been epo ed ha he a ini y o
Mg2+ o p o eins may be in luenced by he pH because he ion a ini y depends on he cha ge o he
Figu e 4-17: Mic og aphs o solidi ied PEO gels (20 mg/ml) made om aqueous solu ions o a) 0.09 M AlCl3
,
0.01 M MgCl2, b) 0.75 M AlCl3, 0.25 M MgCl2, c) 0.05 M AlCl3, 0.05 M MgCl2
. The quali y o he gel inc eases wi h a dec easing
Al:Mg a io. The AlCl3 concen a ion was no lowe ed u he o main ain an accep able pH.
58
p o ein [156]. A lowe ed pH c ea ed by aluminium chlo ide migh cause a simila e ec , al e ing he
in e ac ions be ween Mg2+ and PEO.
In he ollowing expe imen s, he gels’ mo phology was s udied unde he mic oscope and
used as a measu e o he assessmen o hei quali y. The gene al co ela ion be ween mo phology
and pe o mance is qui e eliable, since excessi ely d ying gels wi h agg ega es always pe o m
wo se in e ms o swi ching kine ics and achie able peak cu en s. Conside ing he good expe ience
made wi h he li hium pe chlo a e-based gels, widely used so a , we decided o use mix u es o AlCl3
and LiClO4 in he second gene a ion o high pe o mance SPEs. In ac , li hium pe chlo a e is an
excellen dopan by i sel i we dis ega d he necessi y o add he ola ile HCl o p o ide bo h acidi y
and chlo ide ions in ea lie p epa a ions. The sal has a a ou able combina ion o bo h hyg oscopici y
and lyo opici y since he pe chlo a e ion is a la e membe o he Ho meis e se ies (see Figu e 4-16).
We ha e obse ed ha he gels p epa ed wi h mix u es o sal s we e no ably mo e luid han hei
single-doped coun e pa s, which is p obably due o he ex a added wa e con en . Inc eased luidi y
poses he isk o poo ly en eloping he ga e elec ode, hus inhibi ing he p ope swi ching o he
de ice. Fo his eason, he PEO concen a ion was inc eased om 20 o 30 mg/ml wi h he aim o
inc easing he gels’ iscosi ies and acili a ing hei deposi ion.
Th ee gels ha e been p epa ed; The i s gel was made wi h a PEO concen a ion o 30
mg/ml in a 0.05 M AlCl3 and 0.05 M LiClO4 solu ion; we named i as P30-Alli-0111, (P30 s ands o
he concen a ion o PEO; Alli s ands o Aluminium and Li hium; he i s wo digi s, 01, deno e he
o e all sal concen a ion (0.1M) and he las wo digi s (11) indica e hei a io. The second and hi d
gels a e made o 0.075 M AlCl3 and 0.025 M LiClO4 (P30-Alli-0131) and 0.025 M AlCl3 and 0.075
LiClO4 (P30-Alli-0113).
Figu e 4-18 demons a es a compa ison o he mo phology o hese h ee gels. I is wo h
no ing ha he quali y o he gel inc eased wi h inc easing LiClO4 con en which is in acco dance
wi h i s chao opici y. P30-Alli-0113 and P30-Alli-0111 emained amo phous and anspa en a e
Figu e 4-18: a) Gel P30-Alli-0111 30 min a e deposi ion; b),c) p og essing d ying o P30-Alli-0111. d) P30-Alli-
0113 30 min a e
deposi ion. e) P30-Alli-0131 30 min a e deposi ion; ) p og essing d ying o P30-Alli-0131.
59
deposi ion while P30-Alli-0131 c ys allized a e a sho pe iod o ime. De ices made wi h he i s
wo gels bo h showed p omising esul s in e ms o endu ance wi h espec o AlCl3-only SPEs. Figu e
4-19 and Figu e 4-20 show he I-V cha ac e is ics o wo se ies o h ee di e en de ices made wi h
gels P30-Alli-0113 and P30-Alli-0111, espec i ely, demons a ing a good le el o cycle- o-cycle
ep oducibili y o e 300−500 cycles. One de ice was cycled o e a o al o 1500 imes ( wo se ies o
500 and 1000 consecu i e cycles), which co esponds o o e 80 hou s o con inuous ope a ion. I is
no ewo hy ha despi e an e en ual decline o peak cu en , any 20 consecu i e cycles we e almos
iden ical o one ano he . In gene al, he shape o he I-V cu e emained unchanged o e he whole
se ies o measu emen s. A sho e se ies o 20 measu ing cycles was ca ied ou on some de ices 2−3
days a e he endu ance se ies o show ha hei unc ionali y is main ained e en when he de ice is
le in ambien condi ions in an idle s a e. The e en ual luc ua ions o ou pu cha ac e is ics a e
caused by sligh changes o he measu ing condi ions h oughou he se ies ( empe a u e, humidi y)
since he expe imen s we e in en ionally ca ied ou in a non-con olled en i onmen . Such an
endu ance le el is p ecluded o s anda d de ices wi h SPEs con aining LiClO4 and HCl. In ou
expe ience, such de ices usually las ed o only abou 50 equi alen cycles be o e no swi ching could
Figu e 4-19: Endu ance se ies o I-V cha ac e iza ions o di e en OMDs ab ica ed wi h gel P30-Alli-
0111 as he SPE, measu ed
wi h a dwell ime o 2 s/0.02 V. a,b) 500 consecu i e cycles, ollowed by a sho se ies o 20 cycles acqui ed a e 2 and 3 da
ys,
espec i ely. c) Se ies o 500 consecu i e cycles ollowed by ano he 1000 cycles acqui ed 2 days a e he i s se ies. The
inse
demons a es cycles 501-1500. d) De elopmen o he peak ou pu cu en Iel o he se ies in panel c). The discon inui y
a 500 cycles
is caused by he b eak be ween he wo se ies o measu emen s, du ing which pa ial d ying o he sample has occu ed.

60
be obse ed, due o he PANI dedoping caused by HCl e apo a ion. The endu ance o ou modi ied
SPE is unp eceden ed o solid s a e OMDs, su passing he eco d se by Lapkin e al. [157] who ha e
measu ed a PANI- and PEO-based mic ode ice o e 11000 sho on/o cycles (10.5 s each,
amoun ing o ca. 32 h in o al). A guably, he compa ison is no en i ely co ec as we con on kine ic
on/o cycling wi h hys e esis measu emen s. Howe e , he ep oducibili y o he shape o he
hys e esis and he ela ed s abili y o he ou pu cu en o e ime leads o hink ha ou de ice may
wo k p ope ly on e y long imescales. The inc ease o he ou pu cu en in he beginning o he
se ies (bes seen in Figu e 4-20 c) and Figu e 4-19 c)) is due o a condi ioning p ocess du ing which
he SPE mo phology adjus s o he en i onmen al condi ions and he de ice pa ame e s s abilize.
E en ually, each se ies showed a dec easing end o he ou pu cu en which is caused by
he slow d ying o he AlCl3/LiClO4 gel. We a ibu e his o he lowe ed a io o dopan s and PEO,
since he concen a ion o PEO was inc eased om 20 o 30 mg/ml. I appea s ha he esul ing
dec eased o e all sal :PEO a io o abou 3.3∙10-3 mmol/mg is insu icien o he hyd a ion o he
Figu e 4-20: Endu ance se ies o I-V cha ac e iza ions o di e en OMDs ab ica ed wi h gel P30-Alli-
0113 as he SPE, measu ed
wi h a dwell ime o 10 s/0.1 V. a) 200 cycles in he bias in e al [−0.6 V; +0.8 V] ollowed by 200 cycles in he bias in e al [
−0.4 V;
+0.6 V]. The s ong dec
ease o peak cu en a e he i s 200 cycles is due o he lowe applied ol age a he end o he anodic scan.
The swi ching onse emains una ec ed by he changed ol age ange. b,c) 300 and 500 consecu i e cycles, espec i ely.
d) De elopmen o he peak ou pu cu en Iel o he se ies in panel c).
61
SPE unde changing su ounding condi ions. Fo his eason, he concen a ions o he added sal s
ha e been adjus ed acco dingly o a summed dopan concen a ion o 0.15 mol/l, co esponding o
he ini ial sal :PEO a io o 5∙10-3 mmol/mg. Figu e 4-21 shows wo o such gels wi h di e en sal
a ios. The gel P30-Alli-01521 emains anspa en wi h minimal c ys alliza ion on he s ipe p o ile,
while P30-Alli-01512 ends o c ys allize in he whole bulk o he laye . This sugges s ha he bene i
o added mois u e in he gel wi h a 2:1 AlCl3:LiClO4 a io is highe han ha o he excess o
chao opic agen in he gel wi h a e e sed a io o 1:2. Wi h a u he inc eased LiClO4 concen a ion,
he gel solidi ies e y slowly and o ms inhomogeneous agg ega es.
Gi en he imp o ed s abili y o e ime, we ab ica ed a de ice wi h he P30-Alli-01521 gel
as he solid polyelec oly e and an endu ance measu emen se ies was ca ied ou o de e mine he
long- e m s abili y/ ep oducibili y o he de ice esponse. We es ablished ha he humidi y o he
su ounding a mosphe e has a s ong in luence on he quali y o he SPE, which is why i is desi able
o keep he measu ing condi ions s able. In addi ion, he conduc i i y o PANI is also dependen on
humidi y, meaning ha a gel wi h a highe wa e con en should enhance he pe o mance. This
p ope y has been used o ab ica ed PANI-based humidi y senso s as wa e is said o acili a e he
p o ona ion o PANI’s imine ni ogen a oms h ough hyd ogen bonding [158],[159]. To show he
po en ial o ou modi ied PEO gels unde s able condi ions, we ca ied ou ano he endu ance
expe imen in which a de ice ab ica ed wi h he gel P30-Alli-01521 was subjec ed o an e en longe
se ies o measu emen s. In o de o p o ide s able and a ou able humidi y le els h oughou he
se ies, he de ice was placed in a home-made closable chambe oge he wi h a pe i dish illed wi h
wa e . The expe imen al se up is shown in Figu e 4-22. A e ins alling he de ice and closing he
chambe , he sys em was le o equilib a e o an hou . This was done in o de o a oid la ge
luc ua ions o he elec ical esponse, as i was he case du ing he endu ance se ies in Figu e 4-19
c) and Figu e 4-20 c). The esul s o he con inuous, 3000 cycles long endu ance se ies co esponding
o almos 7 ull days (abou 167 h) o con inuous ope a ion a e p esen ed in Figu e 4-23. As i can be
seen in panel a), some equilib a ion s ill ook place o e he i s 500 cycles, du ing which he onse
ol age shi ed owa ds lowe alues.
Figu e 4-21: Gels a) P30-Alli-01512 , b) P30-Alli-01521 and c) P30-Alli-0211 six hou s a e deposi ion.
62
Figu e 4-22: Expe imen al se up o
he 3000 cycles endu ance se ies. Holes we e d illed in he pe ime e o he chambe o inse
cables. A wa e - illed pe i dish was used as a sou ce o humidi y o p o ide mo e s able measu ing condi ions.
Figu e 4-23: The esul s o he endu ance es o a de ice ab ica ed wi h he gel P30-Alli-
01521 as he SPE which was cycled 3000
imes wi h a dwell ime o 2 s pe s ep o 0.02 V in he ol age ange [
−0.4 V; +0.6 V]. The measu emen s ook place in a pa ially
ai -condi ioned labo a o y ( u ned on du ing weekdays and o du ing he nigh ). Shown a e he I-
V cha ac e is ics o he de ice
du ing cycles a) 1-3000, b) 501-2500 and c) 1201-1400. d) I- p o iles o he maximum and minimum cu en s IEL,max and IEL,min
h oughou he se ies.
63
The middle 2000 cycles (501-2500) display a ela i ely s able esponse in e ms o he shape
o he hys e esis and onse ol age. Al hough a slow decay o peak ou pu cu en o e ime is e iden ,
he di e ences be ween any 200 successi e cycles wi hin his span a e negligible, displaying a
ema kable ep oducibili y o I-V cha ac e is ics. The de elopmen o he maximum and minimum
egis e ed ou pu cu en s du ing he anodic and ca hodic scans in panel d) e eals an in e es ing de ail
abou he pe o mance o he de ice. Fi e pla eau-like spikes (one in he i s and ou in he second
hal o he se ies) can be seen in he p o ile o he maximum ou pu cu en IEL,max. These spikes a e
caused by empe a u e di e ences be ween day and nigh since he ai condi ioning in he labo a o y
was no con inuous h oughou he day. I is known ha he conduc i i y o polyaniline as an o ganic
semiconduc o is empe a u e-dependen , inc easing a highe empe a u es ( alid o alues a ound
oom empe a u e) [160]. The collec ed da a e lec his p ope y, as he conduc i i y ises e e y o he
day excep o days 2 and 3 when he ai condi ioning emained o and he empe a u e emained
low. On he o he hand, IEL,min also displays spikes owa ds less nega i e alues (lowe absolu e
alues) a he same imes as IEL,max spikes. Highe (absolu e) cu en s du ing he ca hodic scan a e an
indica ion o an incomple e educ ion o he polyme , causing some esidual conduc i i y. Since he
eac ion kine ics depend on he di usion o ions h ough he SPE and PANI laye , i can be a gued
ha he iscosi y o PEO, which also depends on empe a u e, is ano he ac o o such beha iou .
This seems pa icula ly plausible because he de ice is measu ed close o he scan a e limi wi h a
dwell ime o 2 s/0.02 V. A high scan a e means ha he educ ion and oxida ion o he PANI laye
will s a lagging behind he scan a subop imal condi ions. This is u he con i med by he ac ha
IEL,min is g adually es o ed o lowe absolu e alues once he empe a u e is inc eased. Con e sely,
dis ega ding he empe a u e spikes, IEL,max displays a s eady decay h oughou he whole se ies,
indica ing a slow deg ada ion p ocess, he cause o which is no ully unde s ood. On he one hand, a
g adual d ying o he SPE could be a eason o a p og essing wo sening o he de ice’s pe o mance.
Howe e , i is unlikely in his case, as his endu ance es was ca ied ou in a closed chambe o
ela i ely low olume, wi h a con olled humidi y. This indica es ha some o m o ma e ial
deg ada ion is he mo e likely cause.
The decomposi ion o PANI appea s o be mo e p obable due o he hyd olysis and clea age
o he polyme chains in i s eme aldine sal s a e, as i has been discussed abo e. Wi h a lowe a e age
chain leng h, he cha ge ans e be ween di e en chains becomes mo e p edominan as he limi ing
ac o o conduc i i y. The decomposi ion o PANI seems especially plausible when conside ing he
inc eased wa e con en o he PEO gel gi en by he humid en i onmen . O he undesi ed eac ions
canno be excluded, such as c osslinking [90],[91], and co alen binding o chlo ine o benzenoid o
quinoid ings in PANI [161], bo h o which dis u b he egula elec onic s uc u e o he polyme .
Acco ding o e e ence [161], he la e eac ion can ake place a su icien ly high acid
concen a ions, e en hough he chlo ide ion is a poo nucleophile.
Ano he sou ce o deg ada ion could be he sil e wi e as i is also subjec ed o con inuous edox
cycles du ing which he po osi y o he AgCl laye may g adually change. Though, gi en he ac ha
70
hexa luo ophospha e ha e p o en hemsel es as ye ano he class o dopan s unsui able o OMDs.
The applica ion o LiBF4 and LiPF6 in LIB echnology is possible because, in his case, nonaqueous
sol en s ha do no eac wi h he solu es a such meaning ul a es, such as e he s o o ganic
ca bona es, a e employed [169].
4.2.3.3 SPE wi h b omide as he p ima y o seconda y dopan anion
Fo he ou h gene a ion o ou SPEs wi h inc eased s abili y, we decided o swi ch ou
a en ion back owa ds al e na i es o aluminium chlo ide as he p ima y dopan . Wi h he expe ience
gained om ou p e ious esea ch, adhe ing o he c i e ia ha quali y an addi i e as an app op ia e
dopan , ou idea was o eplace AlCl3 by a sal wi h simila p ope ies. We we e sea ching o a Lewis-
acidic, hyg oscopic sal wi h highe chao opici y, whose anion would pa icipa e in he edox
eac ions in ol ing bo h he PANI channel and he sil e elec ode. To minimize he numbe o
sys ema ic changes b ough by he eplacemen o he dopan , we used a homologue o aluminium
chlo ide. A he same ime, he sal should ha e no edox chemis y o i s own in he po en ial ange
be ween −0.4 V and +0.6 V, i.e., whe e he OMD is ope a ed. Fo his eason, we e ained om
using aluminium iodide because o he low s anda d elec ode po en ial o he couple 3 I−/ I3−
(E° = +0.53 V [170]). On he o he hand, a simila side eac ion is no expec ed o b omide because
i emains s able unde ou expe imen al condi ions. Acco ding o he elec ochemical se ies o
s anda d elec ode po en ials, he oxida ion o B − o B 2 occu s a po en ials abo e +1.0 V (+1.087 V
in aqueous solu ions [171]. Aluminium b omide is compa ably hyg oscopic, wi h he [Al(H2O)6]3+
ion p o iding acidi y while ha ing a la ge , mo e pola izable and hence mo e chao opic anion han
chlo ide. OMD ope a ion media ed by b omide ions has al eady been shown in sec ion 4.1.4 o his
wo k, whe e measu emen s in aqueous HB ha e been ca ied ou success ully. The expe imen s ha e
also shown an o e all simila beha iou in compa ison wi h an equimola HCl elec oly e, gi en ha
chlo ide and b omide ha e compa able ionic mobili ies [135]. Ne e heless, hose esul s (ob ained
using liquid elec oly es) mus be handled wi h cau ion, since he obs uc ion o he bulkie b omide
(ionic adii: (B -) = 1.95 Å, (Cl-) = 1.81 Å [172]) could be s onge in a solid polyelec oly e. Ano he
sou ce o conce n is he b omide’s highe chemical ac i i y, since i is a much be e nucleophile han
chlo ide, possibly causing subs i u ion eac ions o he clea age o PEO chains, which will be
assessed below.
Figu e 4-30: Mic opho og aphs o gel P30-AlliBF-
0211 a) 15 min a e solidi ica ion, b) 30 min a e solidi ica ion;
c) gel P30-AlliPF-01521 30 min a e solidi ica ion.

71
We s a ed by p epa ing a 30 mg/ml PEO gel con aining 0.1 M AlB 3 as he sole dopan
(P30-AB-01). Su p isingly, al hough chlo ide and b omide a e neighbou ing each o he in he
lyo opic se ies, he deposi ed gel was conside ably mo e s able han i s equimola AlCl3 analogue
om sec ion 4.2.1, emaining anspa en a e solidi ica ion. The gels displayed sa is ac o y long-
e m s abili y, showing some mode a e d ying on he edges o he deposi ed laye a e 11 days o
s o age in ambien condi ions. Mo eo e , a no ewo hy p ope y o he AlB 3-based gel was a supe io
esis ance o hea , as i emained s able upon hea ing he sample o 60 °C on a ho pla e, as i is shown
in Figu e 4-31 c). Howe e , OMDs assembled wi h P30-AB-01 as he SPE displayed no swi ching.
We assume ha he acidi y o he gel was insu icien o main ain polyaniline in i s conduc i e PANI-
ES o m, because AlB 3 solu ions had highe pH le els han he equimola solu ions o AlCl3. In an
a emp o lowe he gel’s pH, we inc eased he dopan concen a ion, p epa ing gels wi h 0.2 M (P30-
AB-02), 0.5 M (P30-AB-05) and 1.0 M AlB 3 (P30-AB-10). Figu e 4-31 d- ) show ha he esul ing
SPE became mo e homogeneous and anspa en , while also becoming mo e and mo e so , o he
poin whe e P30-AB-05 and P30-AB-10 emained in he o m o a iscous liquid. This also mean
ha i became mo e di icul o ensu e ha he ga e elec ode was in good con ac wi h he SPE when
assembling de ices, as i would low down om i a he han en eloping i , as shown in Figu e 4-32.
Thus, such gels we e no used u he .
The issue wi h he assumed insu icien acidi y o he gel could no be sol ed by using
mix u es o AlB 3 and AlCl3. Low concen a ions o AlCl3 had no conside able e ec on he acidi y
o he gel while highe concen a ions caused excessi e hyg oscopici y, so ha he gels would no
solidi y su icien ly. On he con a y, low concen a ions o AlB 3 did no p o ide enough
chao opici y, leading o poo s abili y and c ys alliza ion o e ime. Fo his eason, i was decided
o use ano he addi i e as a sou ce o acidi y. As we ha e es ablished abo e, he addi i e mus be
Figu e 4-31: Mic opho og aphs o AlB 3-doped SPE wi h 30 mg/ml PEO.
Gel P30-AB-01 (0.1 M AlB 3) a) 30 min a e deposi ion, b) 11 days a e deposi ion, c) a e 15 min a 60 °C.
d) P30-AB-02 (0.2 M AlB 3), e) P30-AB-05 (0.5 M AlB 3), ) P30-AB-10 (1.0 M AlB 3) 1 h a e deposi ion.
72
non- ola ile and ha e no sal ing-ou e ec on PEO. When discussing he concep o lyo opici y in
sec ion 4.2.2, hyd o opes ha e been men ioned among o he subs ances (besides chao opic sal s)
ha a e capable o sal ing-in o mac omolecules. One such hyd o ope is he o ganic, non- ola ile p-
oluenesul onic acid (TSA o p-TsOH, see Figu e 4-33), whose abili y o sal -in s ems om i s
amphiphilic s uc u e, allowing i o in e ac wi h bo h wa e and unipola moie ies. TSA is known o
be somewha soluble in wa e and o ms sols in alcohols and e he s as well as saline solu ions. I s pKa
is es ima ed by di e en echniques o be be ween −1.4 and −6.2 [173], which is in any case
su icien ly low o main ain PANI in i s doped s a e in he ac i e zone.
The AlB 3- and TSA-based PEO gels equi ed some op imiza ion in e ms o he con en o
bo h sal s. The mic og aphs o some examples a e shown in Figu e 4-34 a-d). Howe e , he bes SPEs
we e ob ained using concen a ion a ios simila o he ones amilia om ou p e ious esea ch, i.e.,
he gels P30-ABT-01521 (0.1 M AlB 3, 0.05 M TSA) and P30-ABT-0211 (0.1 M AlB 3 and 0.1 M
TSA). While we can empi ically de e mine he dopan :PEO a io o be op imal a ound
5-7∙10-3 mmol/mg, a ho ough examina ion o op imal le els o hyg oscopici y and lyo opici y
equi e he s udy o a ious pa ame e s o a wide a ay o sal s, which goes beyond he scope o his
wo k o hesis. The wo men ioned gels ha e shown ou s anding long- e m and he mal s abili y and
wi hs ood he con inuous exposu e o empe a u es as high as 110 °C o o e 16 hou s, as i is shown
in Figu e 4-34 e, ). Such a p ope y is o pa icula in e es o he s andpoin o au oma ed
manu ac u ing, since one o he s eps ( he p in ing o polyaniline) equi es he subs a e o be hea ed
du ing he ilm deposi ion, as will be shown in Chap e 4.
The ope a ion o OMDs assembled wi h AlB 3- and TSA-based SPEs was unsuccess ul. We
assumed ha he eason was one o he conce ns exp essed when we i s employed aluminium
b omide, namely ha i is bulkie han chlo ide which migh hinde i om mig a ing h ough PEO
o enable he swi ching eac ion. Fo his eason, gi en he excep ional he mal s abili y o
polyelec oly es con aining TSA, we decided o e u n o AlCl3 as he p ima y dopan wi h an addi ion
Figu e 4-32: Schema ic ep esen a ion o he d
eposi ion o a PEO gel in he ac i e zone o an OMD a), ollowed by an o dina y d ying
p ocess wi h he ga e elec ode s aying eme ged in he SPE (b). c) Supposed d ying p ocess o gels wi
h excessi e dopan
hyg oscopici y and/o chao opici y (such as P30-AB-05, P30-AB-
10), leading o insu icien ly con ac o he ga e elec ode wi h he
elec oly e.
Figu e 4-33
: Chemical s uc u e o
p
- oluenesul onic acid (TSA).
73
o he sul onic acid. As expec ed, due o he lowe o e all chao opici y o he dopan s, such gels
we e gene ally less s able han hei b omide-based analogues, which can be seen in he s uc u e o
he solidi ied P30-ACT-01521 gel (0.1 M AlCl3, 0.05 M TSA) in Figu e 4-35. Since he P30-ACT-
0211 gel emained s able, we can assume ha he TSA i sel exe s a sal ing-in e ec , e en wi hou
he p esence o ano he chao opic sal . Mo eo e , he anspa en appea ance o he gel P30-ACT-
022101 e eals ha his e ec mus be e y p onounced, since a ela i ely low concen a ion o TSA
(0.02 mol/l) is enough o coun e ac he e ec o he 0.2 M AlCl3. As i was shown ea lie in Figu e
4-28, such high concen a ions o AlCl3 in gels wi hou a seconda y dopan caused he c ys alliza ion
o he SPE a e deposi ion. Howe e , i s he mal s abili y was no as high as ha o AlB 3-/ TSA-
based ones. Gel P30-ACT-0211 on he o he hand showed ema kable s abili y e en a e 15 h a
Figu e 4-34: Mic opho og aphs o AlB 3- and TSA-doped SPE wi h 30 mg/ml PEO. a) Gels P30-ABT-
01251 30 min a e deposi ion;
b) P30-ABT-01521 30 min a e deposi ion; c) P30-ABT-0211 30 min a e deposi ion; d) P30-ABT-
011101 30 min a e deposi ion;
e) P30-ABT-01521 a e 16 h a 110 °C; ) P30-ABT-0211 a e 16 h a 110 °C.
Figu e 4-35: Mic opho og aphs o a ious AlCl3- and TSA-based PEO gels. a) P30-ACT-
01521 30 min a e deposi ion; b) a e 30
min a 90 °C. c) P30-ACT-022101 30 min a e deposi ion; d) a e 2 h a 70 °C. e) P30-ACT-
0211 30 min a e deposi ion; ) a e 2
h a 70 °C; g) a e 15 h a 110 °C; h) a e 30 min a 10 °C.
74
110 °C on a ho pla e. The sample also emained s able a e 30 min a 10 °C. I is no ewo hy ha
du ing he long hea ing a 110 °C, he subs a e became co e ed by a pale deposi o each side o he
cas SPE laye . The same could be seen in Figu e 4-34 e, ), al hough less p onounced. This is likely
due o he sublima ion o aluminium chlo ide in he o m o Al2Cl6 (al hough, acco ding o li e a u e,
his no mally occu s in ha she condi ions [174].
Despi e he excellen s abili y o he SPE, de ices assembled wi h P30-ACT-0211 we e no
unc ional, as i was he case wi h analogous b omide-based SPEs. Gi en he ac ha b omide-based
liquid elec oly es allowed swi ching in sec ion 4.1.4, i is hence mo e p obable ha he swi ching is
p e en ed by TSA in b omide- and chlo ide-based gels. In a s udy by S ö le and Lu [175], a ious
ou es o chemical oxida ion o TSA wi h hyd ogen pe oxide a e shown. Conside ing he elec on-
wi hd awing cha ac e o he −SO3− g oup, a simila elec ochemical p ocess (possibly assis ed by
AlCl3 o AlB 3) seems o be plausible in ou case, p e en ing he oxida ion o PANI-LE o PANI-ES.
Al hough he PEO gels discussed in his sec ion could no be used o ab ica e OMDs, he imp essi e
s abili y o gels con aining b omide and/o TSA is encou aging o u he esea ch in his di ec ion.
Fo ins ance, TSA could be eplaced by o he , mo e s able sul onic acids such as campho sul onic
acid, which would p o ide he necessa y le el o humidi y, chao opici y and, possibly, he mal
s abili y.
4.2.3.4 AlCl3-based SPE wi h a ca ionic chao ope as a seconda y dopan
Ou p e ious a emp s o eplace LiClO4 as he seconda y dopan led o he de elopmen o
new SPEs based on di e en chao opic anions. Despi e he ema kable s abili y o some o hese
SPE, hey could no be employed in OMDs because he swi ching did no occu o a ious easons.
Thiocyana e and TSA ha e been shown o display some edox chemis y o hei own which could
ha e p e en ed polyaniline om swi ching; b omide, al hough shown o be unc ional in liquid
elec oly es, p obably did no p o ide he necessa y le el o acidi y in o m o AlB 3 o main ain
PANI p o ona ed; e a luo obo a e and hexa luo ophospha e mos likely decomposed due o
hyd olysis, eleasing kosmo opic luo ide ions ha led o p ecipi a ion. Pe chlo a e i sel has shown
he disad an age o being a s ong oxidan , which likely leads o he deg ada ion o PEO o e ime.
As shown by a ew examples o gels such as P30-ACT-022101, a ela i ely low
concen a ion o he (anionic) chao opic agen can be enough o s abilize a gel wi h a high
concen a ion o AlCl3. This leads us o he conclusion ha al hough ca ions a e gene ally known o
ha e a weake sal ing in-/ou e ec on mac omolecules han anions, a s ong ca ionic chao ope migh
be able o p o ide he necessa y condi ions as well. Hence, in he i h gene a ion o ou modi ied
PEO-based SPEs, we decided o ocus on in oducing he chlo ide sal o a chao opic ca ion o he
sys em in o de o s abilize PEO’s s uc u e. Acco ding o li e a u e (mos ly dedica ed o p o ein
esea ch), guanidinium chlo ide (GndCl) is a s ong chao ope [176]–[179], capable o sal ing-in o
p o eins, dec easing he hyd ophobic e ec and causing he un olding o hei e ia y s uc u e, hence
dena u alizing hem. While his e ec is undesi ed in p o ein pu i ica ion and isola ion, i is he goal
75
in ou case as i keeps PEO om p ecipi a ing and c ys allizing. The cause o guanidinium’s
chao opici y is i s in e ac ions wi h he mac omolecules and wa e . The ca ion is highly pola izable,
since i s cha ge is e enly dis ibu ed o e he whole molecule, as shown in Figu e 4-36.
Unlike some o he dopan s we used be o e (TSA, AlB 3), guanidinium chlo ide has a high pKa o
13.71 [180], which is why he acidi y in he SPE had o be p o ided en i ely by he p ima y dopan
AlCl3. Fu he mo e, based on p e ious esul s, adjus men s o he dopan concen a ions had o be
made o make up o guanidine hyd ochlo ide’s lack o hyg oscopici y. We ound ou empi ically
ha concen a ions o AlCl3 abo e 0.2 mol/l o GndCl abo e 0.1 mol/l caused c ys alliza ion due o
an o e sa u a ion o he mix u e. We p epa ed a se ies o 30 mg/ml PEO gels ha a e p esen ed in
Table 5. Once cas , mos o hese gels d ied ou and c ys allized a e a ew hou s. Some o he mo e
s able SPEs a e shown in Figu e 4-37. The gel P30-ACG-0231 eme ged as he one wi h he mos
a ou able composi ion as i emained anspa en 2 days a e deposi ion, while all he o he ones
(including hose in Figu e 4-37 a-d) e en ually began o d y ou du ing he same pe iod.
Table 5: Composi ions o 30 mg/ml PEO gels based on AlCl3 and GndCl, and hei isual appea ances 2 hou s and 2 days a e cas ing
on qua z subs a es in ambien condi ions.
gel symbol
[AlCl
3
]
(mol/l)
[GndCl]
(mol/l)
appea ance
a e 2 h a e 2 d
P30-ACG-0131 0.025 0.075 d y edges −
P30-ACG-0111 0.05 0.05 d y edges −
P30-ACG-01512 0.05 0.1 e y d y edges −
P30-ACG-01521 0.1 0.05 d y ends; sligh ly d y edges −
P30-ACG-0213 0.05 0.15 o e all u bid; d y edges −
P30-ACG-0211 0.1 0.1 anspa en ; sligh ly d y ends d y, c ys allized
P30-ACG-0231 0.15 0.05 anspa en ansp.; one end sligh ly d y
P30-ACG-02532 0.15 0.1 anspa en * d y, c ys allized
P30-ACG-02541 0.2 0.05 anspa en d y pa ches
P30-ACG-0321 0.2 0.1 anspa en small, d y pa ches
*cas ing o his gel was pa icula ly di icul due o low iscosi y/densi y
Figu e 4-36: Chemical s uc u e o he guanidinium ca ion.

76
The he mal s abili y o he gel P30-ACG-0231 (0.15 M AlCl3, 0.05 M GndCl) was es ed
by placing he cas sample on a hea ing pla e a 60 °C o 5 min. Much unlike he gels con aining
aluminium b omide and/o TSA, his gel c ys allized wi hin ens o seconds, as shown in Figu e 4-37
). Howe e , he sample egene a ed almos comple ely a e 10-15 min in a humid chambe (wi h a
wa e - illed pe i dish, as used in he endu ance se ies in sec ion 4.2.3.1). Looking ahead, such a
beha iou is sa is ac o y when planning he implemen a ion o such gels in au oma ed OMD
manu ac u ing. The gel P30-ACG-0231 was hen chosen o assess he ope abili y o OMDs ea u ing
AlCl3- and GndCl-doped PEO. I was done using he gel as he elec oly e in i s semiliquid o m in
he se up shown in Figu e 4-7. The de ices ope a ed o dina ily in he bias ange o [−0.4 V; +0.6 V]
wi h a egula hys e esis shape and kine ics simila o ea lie examples. This means ha he pH o he
SPE was su icien and no undesi ed side eac ions inhibi ing he swi ching ook place. A PEO-
concen a ion o 30 mg/ml has p o en o be oo low in he case o AlCl3- and GndCl-based gels
because hei iscosi y and densi y was gene ally a he low, making he deposi ion by manual cas ing
mo e di icul . Thus, we inc eased he concen a ion o PEO o 35 o 40 mg/ml while main aining he
o e all sal :PEO a io (6.67∙10-3 mmol/mg) and p opo ion o dopan s. The ad an age o dense gels
besides he be e applicabili y is ha hey solidi y as e han he mo e liquid 30 mg/ml analogues.
The gel P35-023331 (0.175 M AlCl3, 0.058 M GndCl) was used o assemble an OMD ha was
subjec ed o an endu ance es o e 1000 cycles. The de ice was placed in a p o ec i e chambe , bu
wi hou humidi ying he a mosphe e inside wi h a wa e - illed pe i dish like i was he case wi h he
de ice ea u ing gel P30-Alli-01521 in sec ion 4.2.3.1.
The se ies s a ed o wi h ela i ely slow kine ics and a low ou pu cu en which no malized
wi hin he i s 10 cycles h ough condi ioning o he sys em ( o ins ance, e eshing Ag/AgCl
su ace). A slow bu s eady decay o peak cu en h oughou he se ies o 1000 con inuous I-V-cycles
can be obse ed, becoming almos linea a e ca. 27 h. I is no ewo hy ha no pla eau-like peaks
we e obse ed like in he 3000-cycle se ies in Figu e 4-23. The cu en luc ua ions in ha I- p o ile
Figu e 4-37: Mic opho og aphs o a ious AlCl3- and GndCl-based PEO gels 2 h a e deposi ion: a) P30-ACG-
0211;
b) P30-ACG-02532; c) P30-ACG-02541; d) P30-ACG-0321; e) P30-ACG-0231. ) P30-ACG-0
231 5 min a e 60 °C on he hea ing
pla e; g) P30-ACG-0231 egene a ed o 15 min in humid condi ions.
77
we e clea ly dependen on he ime o day and he ai condi ioning in he labo a o y. We a gued ha
he cause o such beha iou was ei he he empe a u e o humidi y and emphasized he o me ,
easoning ha he wa e emaining in he pe i dish wi hin he chambe should ha e p o ided an equal
le el o humidi y h oughou he se ies. Howe e , he mo e ecen esul s o he de ice shown in
Figu e 4-38 indica e ha he humidi y has a much la ge e ec on he pe o mance o he OMD han
he empe a u e. This en i e endu ance 1000-cycles se ies was eco ded wi hou any ai condi ioning,
howe e , some empe a u e luc ua ions be ween day and nigh s ill occu ed. The lack o an
addi ional sou ce o humidi y in he chambe combined wi h he linea decay o he cu en sugges s
ha he empe a u e luc ua ions alone had no isible e ec . On he con a y, in he p esence o a
sou ce o humidi y, an inc ease o empe a u e has a isible e ec in ha he ela i e humidi y
inc eases as well, causing a s onge hyd a ion o PEO and acili a ing di usion p ocesses. Rega dless
o his de ail, he OMD ea u ing he SPE P35-ACG-023331 displays qui e good endu ance,
conside ing ha he measu emen s ook place in non-op imized, ambien condi ions. The g adual
decay o ou pu cha ac e is ics is mos likely caused by pa ial d ying and nuclea ion o agg ega es
om which he c ys alliza ion could p opaga e, as demons a ed in Figu e 4-38 d). The shape o he
Figu e 4-38: Endu ance se ies o I-V cha ac e iza ions o an OMDs ab ica ed wi h gel P35-ACG-
0231 as he SPE, measu ed wi h a
dwell ime o 2 s/0.02 V. a) 1000 consecu i e cycles wi h he bias in e al [
−0.4 V; +0.6 V] ollowed b) Cycles 400−600. c)
De elopmen o he peak ou pu cu en IEL,max
h oughou he se ies. d) Mic opho og aph o he ac i e zone o he de ice wi h pa ial
nuclea ion/ c ys alliza ion by he end o he endu ance se ies.
78
hys e esis showed good ep oducibili y h oughou he whole se ies, ecognizable by he swi ching
onse a he exac same bias in any 200 consecu i e cycles (excep he i s ew).
Summa izing he esul s o sec ion 4.2 i can be s a ed ha he in oduc ion o in insically
acidic PEO dopan s was e y success ul a elimina ing one o he majo sou ces o ins abili y. The
bes achie ed esul s a e compiled in Table 6. The ollowing a emp s o imp o e he SPE
composi ion o be e pe o mance and endu ance led o a deepe unde s anding o he e ec s ha
a ec he s abili y o he gel, as well as ac o s ha can cause ma e ial deg ada ion. The mos no able
esul s we e ob ained wi h gels con aining aluminium b omide and/o p- oluenesul onic acid,
displaying ou s anding he mal s abili ies. Al hough hese SPEs did no allow OMD swi ching in one
way o ano he , hei in e es ing p ope ies mo i a e o wo k wi h o he , simila subs ances o c ea e
op imized, unc ioning gels wi h supe io endu ance. Finally, he in oduc ion o guanidinium
chlo ide o AlCl3-doped SPEs allowed o ab ica e de ices wi h sa is ac o y s abili y in ambien
condi ions, enabling mo e sophis ica ed expe imen s in he u u e.
Table 6: Summa y o he mos ep esen a i e esul s o ou esea ch o no el, mul iply doped, PEO-based SPE.
gel symbol
gel composi ion (30 mg PEO/ml) s abili y applicabili y
in OMD
p im. dopan sec. dopan sho - e m (2 h) long- e m (2 d+)
Alli-0113 0.025 M AlCl3 0.075 M LiClO4 good poo yes
Alli-01521 0.1 M AlCl3 0.05 M LiClO4 e y good good yes
AlAS-0111 0.05 M AlCl3 0.05 M NH4SCN bad bad n. a.
AlKS-0113 0.025 M AlCl3 0.075 M KSCN poo bad n. a.
AlliS-0211 0.1 M AlCl3 0.1 M LiSCN e y good e y good no
AlliBF-0211 0.1 M AlCl3 0.1 M LiBF4 e y bad bad n. a.
AlliPF-01521 0.1 M AlCl3 0.05 M LiPF6 bad bad n. a.
ABT-0211 0.1 M AlB 3 0.1 M p-TsOH excellen excellen ** no
ACT-0211 0.1 M AlCl3 0.1 M p-TsOH excellen e y good** no
ACG-0231 0.15 M AlCl3 0.05 M GndCl* e y good good yes
* guanidinium chlo ide, ** e y good he mal s abili y a 90 °C
4.3 Inc eased de ice s abili y h ough coa ing wi h NEA
So a , we ha e shown ha some o he gels such as AlCl3-/ LiClO4-based ones display a
ema kable ep oducibili y o de ice cha ac e is ics o e long pe iods o ime when p o ided wi h a
a ou able le el o humidi y su ounding he sample. In his espec , he deg ee o swelling is a
p ope y o PEO-based SPEs ha poses a ce ain cause o ins abili y, as i makes he sys em dependen
on he a ailable con en o wa e . As i was shown in some o ou expe imen s, he ela i e humidi y
o he a mosphe e su ounding he sample can signi ican ly a ec he SPE’s s uc u e and he
pe o mance o he ela ed OMD, speci ically he ou pu cu en and he swi ching kine ics. This
79
sou ce o luc ua ions in he de ice esponse is in insic o PEO-based SPEs and is ine i able,
independen om he dopan s. Recen ly, mic ode ices ea u ing liquid elec oly es (aqueous HCl
solu ions) ha e been p esen ed by Ba is oni e al. [181]. While such an app oach bea s ad an ages
like inc eased swi ching kine ics and o e all ep oducibili y, i is a he mo e sui ed o model de ices
o heo e ical, p oo -o -concep expe imen s because o he imp ac icali y o liquid elec oly es in
elec onic de ices and sys ems. Besides, he issues o dopan e apo a ion (in his case HCl), which
was sol ed in he p esen wo k, would pe sis in such de ices, albei p obably less p onounced han
in HCl-doped, PEO-based SPE.
One way o ci cum en ing he issue o he en i onmen al in luence could be he sealing o
he de ices a e he assembly, main aining he SPE in an op imal s a e and lea ing i una ec ed by
ambien humidi y. Sealing would also hinde he di usion o ai oxygen in o he sys em, p e en ing
oxida i e deg ada ion.
To apply he concep o de ice sealing o ou OMDs, we ha e chosen he comme cially
a ailable NEA 121 (No land Elec onic Adhesi e). Acco ding o he manu ac u e (No land P oduc s
Inc.), he p oduc is a u e hane- ela ed esin-based o mula ion and consis s o a mix u e o
benzophenone and a numbe o me cap o es e s ( he exac composi ion is no disclosed). A oom
empe a u e, i appea s as a iscous, colou less liquid wi h a dis inc smell. I con ains a ca alys ha
allows i o be cu ed by i adia ion wi h UV ligh o a ele a ed empe a u es (e.g., 10 min a 125 °C
in a con ec ion o en o 3 h a 80 °C). Cu ing h ough hea ing has he ad an age o a mo e
homogeneous p ocess, i.e., he polyme iza ion p og esses e enly h oughou he whole hickness o
he deposi ed laye . Cu ing by UV ligh depends on he hickness o he laye and ac s s onge on he
laye s ha a e close o he ligh sou ce. Ne e heless, due o he sensi i i y o mos o ou SPEs o
high empe a u es, we eso ed o UV-i adia ion. The polyme iza ion o monome s occu s in he
wa eleng h ange om 320 o 380 nm wi h peak sensi i i y a ound 365 nm, esul ing in a anspa en ,
solid polyme coa ing. The cu ing is said o be a e y exo he mic p ocess. The manu ac u e claims
ha NEA 121 displays e y good adhesion on glass, me als, p in ed ci cui boa ds and many plas ics,
making i an app op ia e candida e o he ype o manu ac u ing echniques en isioned o ou
de ices. Some o he p ope ies o he coa ing ma e ial a e p esen ed in Table 7.
Table 7: P ope ies o NEA 121 (sou ce: No land P oduc s Inc. da a shee ).
p ope y
iscosi y a 25 °C 300 cps
elonga ion a ailu e 30 %
modulus o elas ici y 160000 psi
ensile s eng h 3500 psi
dielec ic cons an (1 MHz) 4.04
86
5 Kine ic and dynamic aspec s in OMD ope a ion
5.1 The ole o he de ice’s geome y
The kine ic beha iou o OMDs has been a subjec o heo e ical esea ch in he pas . The
elec ochemical model p oposed by Sme ie i e al. [93] and u he de eloped by Demin e al. [94]
(see sec ion 0) adequa ely desc ibes he OMD’s swi ching p ocess. Al hough hei a gumen s ela ed
o he ole o li hium ions is lawed (as demons a ed in sec ion 4.1), he main concep s o hei model
emain he basis o he heo e ical backg ound on he OMD’s swi ching mechanism. The p esen
chap e deals wi h ele an kine ic and dynamic aspec s o OMD ope a ion. While some ela ed
aspec s ha e been discussed in p e ious wo ks, he a ailable li e a u e does no add ess some ele an
ea u es. Hence, i appea s bene icial o p o ide a summa y and explain said e ec s and hei cause.
One o he undamen al aspec s o OMD’s ope a ion, al eady discussed in his chap e is he
disc epancy be ween he a es o oxida ion and educ ion. The eason o such beha iou is essen ially
he geome y o he OMD. We ha e shown ha in ou h ee- e minal de ice, he on/o s a e is
de e mined by he conduc i i y o he PANI channel. The conduc i i y o PANI can be con olled
h ough edox eac ions ha a e igge ed by he applied elec ical po en ial. This akes place in he
ac i e zone (AZ) o he de ice, which is de ac o an elec oly ic cell. A cen al concep ha needs o
be unde s ood is he dis ibu ion o he ol age along he PANI channel, as i is shown in Figu e 1-16.
A any gi en sou ce-d ain ol age bias, he po en ial p o ile in PANI decays owa ds he g ounded
sou ce elec ode. As a consequence, du ing he anodic scan he oxida ion is igge ed in he sec ion
ha is closes o he d ain once he h eshold alue Vox is eached he e, p og essing owa ds he
sou ce as he bias inc eases. On he con a y, du ing he ca hodic scan, he educ ion in he pa o he
AZ closes o he sou ce will be igge ed a a d ain ol age bias ha is highe han he h eshold alue
(Vbias > V ed).
Fu he mo e, du ing he anodic scan, he PANI-channel as a whole becomes conduc i e only
once he majo i y o PANI-LE in he ac i e zone ansi ions in o PANI-ES. Con e sely, du ing he
ca hodic scan, he conduc i i y o he channel s a s dec easing as soon as a ac ion o he ac i e zone
ansi ions in o he insula ing PANI-LE. In o he wo ds, when compa ing he applied bias wi h he
heo e ical edox po en ials, du ing a s anda d I-V cha ac e iza ion we obse e a somewha delayed
swi ching-on and a p ema u e swi ching-o o he de ice. This e ec may be be e unde s ood om
he so-called kine ic cha ac e iza ions, when he swi ching occu s a a cons an bias. Typical applied
ol age alues a e +0.5 V o +0.6 V o swi ching-on and −0.2 V o swi ching-o . As i was
men ioned abo e, he ansi ion o he OMD om he insula ing o he conduc i e s a e occu s
g adually as he oxida ion p og esses om he d ain owa ds he sou ce elec ode. The ansi ion back
o he insula ing s a e occu s as e because he conduc i i y o he whole channel dec eases
immedia ely a e he educ ion o PANI ini ia es. Fu he mo e, unlike wi h posi i e bias alues, when
applying a nega i e ol age, he whole channel is a a po en ial below V ed, so ha PANI is educed

87
simul aneously in he en i e ac i e zone. This is also he eason o he seemingly pa adoxical
appea ance o he IEL- cu e, displaying posi i e cu en alues a a nega i e applied ol age, as
shown in Figu e 5-1 b). Such beha iou could be obse ed in he kine ic measu emen s om sec ion
4.1.4. As he swi ching-o ini ia es, he S-D-cu en d ops apidly by a ew o de s o magni ude,
while he ga e cu en emains high (in absolu e alues) as he eac ion p og esses. Because he
elec onic cu en is calcula ed as he di e ence be ween he wo, he esul ing alue is posi i e, as i
was he case in Figu e 4-10. In e es ingly, he shape o he cu e o HClO4 esembled he one in
Figu e 5-1 a), which is due o he highe o -s a e cu en s caused by an inhibi ed educ ion p ocess.
The ole o he posi ion and wid h o he ac i e zone
Ano he ac o ha emphasizes he impo ance o he geome y o he de ice is he in luence
o he posi ion and wid h o he ac i e zone, which again is connec ed o he dis ibu ion o he ol age
ac oss he PANI channel. This mos ly a ec s he swi ching-o p ocess, when he en i e ac i e zone
is in he conduc i e s a e. Acco ding o Figu e 5-2 a), he po en ial di e ence wi hin he AZ dec eases
wi h i s wid h, esul ing in a mo e uni o m esponse. Simila ly, posi ioning he ac i e zone close o
he d ain causes PANI's po en ial wi hin i o ma ch mo e closely he applied bias (Figu e 5-2 b). This
leads o a mo e p edic able and esponsi e swi ching beha iou , since he applied ol age equi ed
o swi ching be e ma ches he heo e ical edox po en ials. The swi ching-on is less a ec ed by
such geome ical changes because, in any case, all he ol age d op mos ly occu s in he ac i e zone.
O e all, i becomes clea ha , in p inciple, he posi ion and wid h o he AZ can be used as a ool o
con ol he I-V cha ac e is ics o he OMD.
Howe e , in p ac ice, a p ecise con ol o he posi ion o he ac i e zone becomes di icul
in applica ion-o ien ed, minia u ized de ices as i is mo e con enien o eso o a geome y whe e
he whole PANI channel is co e ed wi h he (poly)elec oly e and is hence ac i e. In his case, ano he
e ec becomes ele an when conside ing he swi ching beha iou a cons an biases. As i was
men ioned in sec ion 0, he swi ching a e also depends on he dis ance be ween he ga e elec ode
and he conduc i e channel. This e ec becomes mo e signi ican wi h an inc easing a io o he
Figu e 5-1: Schema ic ep esen a ion o an OMD’s kine ic esponse a + 0.5 V (oxida ion, swi ching-on) and
−0.2 V ( educ ion,
swi ching-o ). a) Expec ed c
u e shape wi h a nega i e elec onic cu en a a nega i e bias. b) Typically obse ed cu e shape wi h a
posi i e elec onic cu en a a nega i e bias.
88
channel hickness and i s leng h. In a bo om-con ac / op-ga e con igu a ion, he uppe laye s o PANI
eac soone han he lowe ones, as shown in Figu e 5-3. This may be an addi ional sou ce o delay
o bo h he swi ching-on and -o because he conduc i e s a e o he lowes PANI laye s a e he
mos ele an , since hey a e in di ec con ac wi h he sou ce and d ain elec odes. In his espec , a
bo om-con ac /bo om-ga e con igu a ion could be bene icial in e ms o swi ching a es (see Figu e
5-3 b). Such a con igu a ion can only be ealized wi h a solid polyelec oly e ha could be e suppo
a laye o PANI deposi ed on op i . Toge he wi h he ac ha his e ec only applies o sho
channels, his makes he idea o a bo om-ga e OMD con igu a ion be e sui ed o be implemen ed
by high- esolu ion p in ing echniques, which will be discussed in chap e 4.
5.2 The ela ionship be ween elec ode po en ials and he applied bias
I has been es ablished ha he ac ual elec ode po en ial o PANI does no co espond o he
applied d ain ol age bias. Howe e , assuming a cen al posi ion o a na ow ac i e zone in he on-
s a e and an applied bias o +0.4 V, he ue elec ode po en ial o PANI in ha egion would no be
Figu e 5-2: a) The e ec o he wid h o he ac i e zone, a ec ing he po en ial ange
be ween he pa s close o he d ain and sou ce.
b) The e ec o he posi ion o he ac i e zone wi h espec o he sou ce and d ain elec odes, a ec ing he de ia ion o
he ac ual
po en ial alues wi hin he AZ om he ol age bias applied a he d ain (schema ic).
Figu e 5-3: Schema ic ep esen a ion o
he oxida ion and educ ion p ocess o an o ganic mem is i e mic ode ice a) in he s anda d
con igu a ion and b) in he bo om-ga e con igu a ion. The a ows ep esen he p opaga ion o he ansi ion o PANI.
89
Equa ion 14
Equa ion 15
Equa ion 16
+0.2 V, as sugges ed solely by he po en ial dis ibu ion p o ile. As we men ioned abo e, om an
elec ochemical poin o iew, he PANI and he sil e elec ode ha a e in con ac wi h he SPE in
he ac i e zone o m an elec oly ic cell. Hence, he applied bias can be seen as he po en ial di e ence
be ween he anode and ca hode ha is imposed upon he sys em:
In ou case, when a posi i e bias is applied o he d ain, he PANI elec ode ep esen s he anode, and
i s po en ial can be calcula ed as he sum o he applied bias (co ec ed by he ol age dis ibu ion
p o ile) and he elec ode po en ial o he Ag/AgCl ca hode. A his poin , he ac ha he OMD’s
ga e elec ode is no a ue e e ence elec ode becomes ele an , as i s elec ode po en ial is no
cons an h oughou he measu emen . I a he depends on he ac i i y o chlo ide ions whose
concen a ion is also a iable due o he o ma ion o ion g adien s caused by he applied ol age. The
elec ode po en ial can be calcula ed ollowing he Ne ns equa ion (Equa ion 15). Howe e , i would
be necessa y o de e mine he concen a ion and ac i i y coe icien s o he chlo ide ions a di e en
bias alues.


o Ag/AgCl: 
R … uni e sal gas cons an (R = 8.314 J mol
-1
K
-1
)
F … Fa aday cons an (F = 9.6485℅10
4
As/mol)
E°… s anda d elec ode po en ial (E° = 0.222 V o Ag/AgCl [182])
a
i
… ion ac i i y (a
i
=
i
℅c
i
)
i
… ac i i y coe icien
Acco ding o his ela ionship, he OMD swi ching onse a a ound +0.3 V ( heo e ical Vox o PANI)
in some cases is a he coinciden al, and means ha unde he gi en condi ions, his bias equals he
di e ence be ween he elec ode po en ials o PANI and Ag/AgCl. This also explains how he
swi ching may some imes ini ia e a low ol age biases such as ca. + 0.14 V in one o ou endu ance
se ies Figu e 4-23.
Since he chlo ide ion pa akes in bo h hal - eac ions, bo h elec ode po en ials depend on
chlo ide concen a ion. This means ha he dopan concen a ion in PEO gels is no only impo an
om he poin o iew o SPE s abili y bu may also be used as a ool o con ol he swi ching onse
po en ial. I is possible ha he imp o ed kine ics obse ed wi h AlCl3-doped SPEs we e caused by
he inc eased chlo ide concen a ion. I should be no ed ha since chlo ide ions pa ake in bo h hal -
eac ions, he elec ode po en ial o bo h Ag/AgCl and PANI depends on hei concen a ion.
Howe e , due o he applied elec ical ield, he chlo ide concen a ion is highe a he posi i e pole
and lowe a he nega i e one. Acco ding o he Ne ns -equa ion, du ing he swi ching-on wi h a
90
posi i e applied bias, he concen a ion g adien causes a co ec ion o he elec ode po en ial o
Ag/AgCl o sligh ly highe alues, and ha o PANI o sligh ly lowe ones. Consequen ly, a lowe
applied bias should su ice o igge he eac ion acco ding o Equa ion 14. Fo he swi ching-o
eac ion, he si ua ion is e e sed because o he in e sed pola iza ion o he elec odes and o ien a ion
o he concen a ion g adien .
5.3 On-line obse a ion o OMD swi ching
Al hough he impo ance o he e ec s connec ed o he po en ial dis ibu ion ac oss he
conduc i e channel has been es ablished in ea ly wo ks on PANI-based OMD, he e is s ill li le
e idence o he p oposed swi ching mechanism. In sec ion 1.2.4, we p esen ed some ea ly
spec oscopic da a aimed o ea i m he alleged pa icipa ion o li hium ions in he swi ching p ocess,
co ec ed by ou own da a in sec ion 4.1. Howe e , hose esul s a e me ely indi ec indica ions o
he in ol emen o one o he o he ionic species. A mo e di ec app oach has been aken by Ba is oni
e al. [183], ea u ing a spec opho ome ic obse a ion o he ansi ion o PANI om he insula ing
o he conduc i e s a e, co ela ed wi h simul aneously eco ded elec ical cha ac e is ics. As a
de elopmen o his app oach, we ha e ca ied ou a s udy o op ically isualize in eal ime he
swi ching be ween he ansien s a es o PANI in he ac i e zone. The esul s o his s udy we e
p esen ed a he MEMRISYS 2019 con e ence in D esden (2019) [184].
The expe imen al equipmen esembled he one used in sec ion 4.1.2 o measu emen s in
liquid elec oly es (see Figu e 4-7), ea u ing a Te lon ough on op o which he OMD was moun ed
“ ace-down”. The anspa en qua z subs a e allows o obse e unde a mic oscope he 52 LS laye s
hick PANI channel om below, wi hou he in e e ence o he ga e elec ode. A sil e wi e o a oil
(125 µm in diame e o hickness) we e used as he ga e elec ode, a ached o he bo om o he Te lon
well, illed wi h a semiliquid PEO gel wi h 0.05 M LiClO4 and 0.1 N HCl o aqueous 0.1 N HCl as
he elec oly e. The ansien s a e o PANI was de e mined by he colou o he channel in he ac i e
zone on which he mic oscope’s digi al came a was ocused. The elec ical cha ac e is ics we e
acqui ed simul aneously in he kine ic mode a cons an applied biases. The p edic able colou
ansi ions o PANI be ween yellow and g een, co esponding o PANI-LE and PANI-ES, a e shown
in Figu e 5-4 b) and e), wi h he p opaga ion o he PANI-ES on om he d ain o he sou ce
elec ode p esen ed in panel ). Meanwhile, panels a), c) and d) display an unexpec ed second
ansi ion om g een o blue and back.
Ini ially, his second ansi ion was a ibu ed o he empo a y oxida ion o PANI-ES o
PANI-PS, wi h an explana ion based on he ol age dis ibu ion ac oss he channel, as schema ically
shown in Figu e 5-5. The ini ial po en ial p o iles (ligh e lines) indica e ha a high enough ol age
biases, he e a e egions o he AZ close o he d ain ha a e a a po en ial exceeding Vox(ES-PS),
igge ing he second oxida ion s ep owa ds he pu ple-blue pe nig aniline o m. As he ol age is
edis ibu ed (da ke lines), e en ually, he po en ial in hese egions d ops below Vox(ES-PS) again,
so ha he g een PANI-ES is e o med. An indica ion o such beha iou was in e p e ed in he
91
Figu e 5-4: a-e) Se ies o snapsho s om digi al on-line
eco ds o he ansi ion o PANI in he ac i e zone o an OMD du ing kine ic
measu emen s a di e en measu ing condi ions. No change was
obse ed a e 20 s in panel e). ) Rep esen a ion o he p opaga ion
o he PANI-ES on om he d ain o he sou ce elec ode. The colou sa u a ion in se ies a) and ) was digi ally edi ed o cla i y.
Figu e 5-6: Reco ded elec ic cha ac e iza ion o he OMD co esponding o he measu emen in Figu e 5-4 c).
Figu e 5-5: Schema ic illus a ion o he p oposed po en ial dis ibu ion ac oss
he PANI channel du ing kine ic measu emen s a high
ol age biases ( he posi ion o he wo oxida ion s eps is chosen a bi a ily). The a ow ep esen s he p opaga ion o
he oxida ion o
PANI-LE. The ed lines a e a simpli ied ep esen a ion o he po en ial p o ile a di e en imes, wi h he espec i e would-
be oxidized
egion o he ac i e zone indica ed by he e ical lines.

92
eco ded I- cha ac e is ics (Figu e 5-6), displaying a nega i e peak in he ionic cu en a a posi i e
applied bias ha was a ibu ed o a educ ion o PANI-PS o PANI-ES. This is accompanied by a
empo a y inc ease o esis i i y, co esponding o he insula ing s a e o PANI-PS. The ac ha no
such ansi ion is obse ed in Figu e 5-4 b) is explained by he lowe applied bias. In panel e), he
measu emen akes place in a liquid elec oly e in which he second ansi ion emains unde ec ed due
o as e kine ics, a ou ed by as e di usion p ocesses.
The easoning behind hese obse a ions was la e econside ed as some o he abo e
a gumen s a e lawed. The i s and o emos sou ce o doub is he colou o he polyme a e he
alleged second oxida ion s ep. Al hough colou s migh appea al e ed due o he supe imposi ion o
di e en ansien s a es o PANI in he same ilm, he blue colou o he polyme co esponds much
a he o eme aldine base han pe nig aniline which should show a shade o pu ple. Secondly,
al hough he ol age ( e)dis ibu ion p o ile is illus a ed schema ically, he insula ing cha ac e o
PANI-PS is no ep esen ed. This ac canno be igno ed because i would signi ican ly change he
pic u e, since he po en ial d op mos ly occu s wi hin he insula ing egions. Las ly, i becomes clea
ha his explana ion does no ake in o accoun he geome y o he de ice wi h a cen al posi ion o
he ac i e zone. This would mean ha PANI’s geome y-co ec ed po en ial alues wi hin he AZ
would be signi ican ly lowe han he heo e ical oxida ion po en ial o he ansi ion om PANI-ES
o PANI-PS, wi h he possible excep ion o Figu e 5-4 d) and e) whe e he applied bias is +1.6 V.
We hence assume ha he occu ence o he blue colou a high biases and he inc ease o
esis ance in Figu e 5-6 co espond o he appea ance o eme aldine base as a consequence o PANI
dedoping. The nega i e peak o he ionic cu en is a ibu ed o he pola iza ion o he elec ode a he
han o a chemical eac ion. I is easonable o assume ha such dedoping was caused by he s ong
channel elec ic ield expelling he p o ons om he posi i ely cha ged PANI elec ode. This also
explains he absence o a g een-blue ansi ion in he cases wi h lowe applied ol ages, whe e he
elec ic ield is weake . The e u ning g een colou , meaning he e-doping o PANI, is due o he
g adien o chlo ide ions ha is in e se o he one o p o ons. The accumula ing nega i e cha ge is
g adually educed h ough he ep o ona ion o p og essi ely o med imine ni ogen si es, p o iding
he possibili y o he o ma ion o ionic couples wi h chlo ide. This assump ion is consis en wi h
he ac ha a sligh blue colou appea s in panel a) o Figu e 5-4 while none is obse ed in panel b),
al hough he applied bias is equal in bo h cases. The eason is ha in he i s case, he s eng h o he
elec ic ield is ampli ied by he su ace a ea and pa allel o ien a ion o he sil e oil as opposed o
he wi e. Las ly, no ansi ion o PANI-EB is obse ed in panel e) because o he inc eased ionic
mobili y o p o ons ( ollowing he G o huss mechanism) and chlo ide ions in an aqueous elec oly e.
This e ec is also likely he eason o he shape o he hys e esis cu e in Figu e 4-19 b). The slope
o he cu e sligh ly dec eases a e +0.4 V, indica ing a lowe conduc i i y o he channel ha may
ha e been caused by dep o ona ion d i en by he elec ic ield. The in luence o he coun e ion, pH
and he ing ess/eg ess o H+ o anions in and ou o he polyme on he swi ching p ocess and on he
conduc i i y o PANI ha e been discussed in he wo ks by Focke e al. [83] and Kalaji e al. [185].
93
The easoning behind he beha iou o he OMD’s ac i e zone is con i med by simila esul s by Xia
e al. [186] who ha e s udied a NiO/PANI composi e ma e ial ha esponded in a simila manne o
he applica ion o di e en ol age biases. In a combina ion o ou wo k and ha o Ba is oni e al.
[183] Lapkin e al. [187] p esen ed a s udy whe e he abso bance o he PANI laye in he ac i e zone
o he OMD has been egis e ed by he CCD ma ix o an op ical mic oscope. These esul s a e in
acco dance wi h he ones p esen ed in ou wo k, al hough no in e media e ansi ion was obse ed
he e, since he measu emen s we e ca ied ou a lowe biases.
In conclusion, he concep s in oduced in Chap e 4 add o he unde s anding o he ac o s
a ec ing he s abili y and endu ance o OMDs. Speci ically, he demons a ed dependence o he
pe o mance o he de ices on he pH and he concen a ions o chlo ide ions gi es ise o specula ion
as o whe he mo e ex ensi e modi ica ion o he ma e ials could u he inc ease he s abili y.
Modi ied polyanilines such as poly(N-e hyl aniline) [158],[159] could pa ially sol e he pH
dependency because he isubs i u ed ni ogen a oms do no equi e p o ona ion o he o ma ion o
pola ons upon oxida ion. Besides, he suscep ibili y o chain clea age h ough hyd olysis would be
lowe ed because he ni ogen a oms would be s abilized by he +I-e ec o ano he subs i uen .
Consequen ly, he de elopmen o mo e s able SPEs would also be acili a ed, since he ocus could
be comple ely shi ed owa ds he s abili y o he PEO gel, wi h no ega d o he addi ion o acidic
dopan s. On he o he hand, sul ona ed [80] o o he wise sel -doping polyanilines wi h a ca e ully
con olled a io o dopan g oups and ni ogen a oms would elimina e he channel’s dependency om
anion g adien s wi hin he elec oly e. Thus, a sul ona ed, N-alkyla ed polyaniline (see Figu e 5-7)
seems o be an in e es ing candida e o u he esea ch.
Figu e 5-7: Chemical s uc u e o a sul ona ed, N-
alkyla ed polyaniline as a po en ial
candida e as a conduc i e polyme o u u e OMDs wi h inc eased endu ance.
94
6 Towa ds au oma ed OMD manu ac u ing –
p in ed neu omo phic de ices
In he p e ious chap e , we explo ed ways o p olong he li e ime o he OMD by imp o ing
he solid polyelec oly e as one o i s mos labile componen s and by sealing he de ices wi h NEA
o p o ec hem om en i onmen al e ec s. These app oaches allowed us o signi ican ly inc ease he
endu ance and cycle- o-cycle ep oducibili y o indi idual OMDs wi h espec o ea lie s anda ds.
So a , he de ice assembly has mos ly been done manually, inhe en ly b inging a ce ain deg ee o
a iabili y wi h each manu ac u ing s ep. Fo ins ance, al hough he Langmui −Schae e echnique
enables a p ecise con ol o e he hickness o he deposi ed polyaniline laye , he quali y o each
laye may a y and depends on he expe ience and p o iciency o he ope a o . Ano he sou ce o
a iabili y du ing manual ab ica ion is he posi ioning and wid h o he ac i e zone, as well as he
hickness o he SPE laye ha a ec he kine ics, as we ha e shown ea lie . These and o he ac o s
lead o a ela i ely poo de ice- o-de ice ep oducibili y o elec ical cha ac e is ics, e en a e
imp o ing each indi idual de ice’s endu ance. The s a is ics in ou case gi es he peak ou pu cu en
anging be ween 5 and 30 µA, wi h mos ecu ing alues o 10−20 µA. While i is mos ly equi ed a
s able kine ics, among o he ea u es he ep oducibili y o he on/o a io alue also plays an
impo an ole inso a as i s con ol ep esen s a key o implemen ing mo e complica ed,
sophis ica ed sys ems such as ANNs and sensing in e aces.
Hence, om a echnological s andpoin , he e has long been a need o s eps owa ds he
au oma ion o he manu ac u ing p ocess. In ac , besides minimizing a iabili y among di e en
de ices, o he ad an ages o au oma ed manu ac u ing a e he imesa ing (i.e. he so-called apid
p o o yping) and ease o downscaling ou es, he la e being a p e equisi e o highe pe o mance,
as e kine ics and mass p oduc ion. While OMD downscaling has been a emp ed ea lie by Lapkin
e al. [157] and Ba is oni e al. [181] (bo h wo ked wi h channel leng hs o 200 µm ins ead o a ew
millime es), hose wo ks s ill ea u ed a manual deposi ion o PANI wi h he LS echnique. The
p esen chap e is hence de o ed o de eloping me hods o au oma ed deposi ion o he OMD’s
indi idual componen s by means o manu ac u ing p o ocols based on 3D p in ing echniques.
6.1 P in ing echniques (s a e o he a )
A ough o e iew o he a ious p in ing echniques is shown in Figu e 6-1. Because o hei
e sa ili y and e iciency, we ocused on non-con ac , di ec w i ing (DW) echnologies, such as
Inkje (IJP) and Ae osol Je P in ing (AJP), ha a e conside ed as p omising me hods in o ganic
mic oelec onics manu ac u ing. Di ec w i ing echniques s and ou as pu ely addi i e me hods
al e na i e o s anda d echniques such as pho oli hog aphy widely employed o he manu ac u ing
o s anda d (e.g. me al/me al oxide junc ions) and o ganic-based elec onic de ices. In pa icula , hey
compa e a ou ably o o he p in ing echniques, e.g. sc een p in ing, nanoimp in ing o g a u e
p in ing, due o hei gen leness and e sa ili y in e ms o compa ible subs a es, including la ,
95
lexible and e en agile ones, such as op ical ib e glass [188]. Despi e he ela i e simplici y
combined wi h he indus ially scalable cha ac e , hese echniques a e highly e icien in ha he
ma e ial was e is minimized and he ab ica ion cos is educed, while s ill allowing o gene a e
complex pa e ns. Majo d awbacks come om he di icul y o achie e as mass-p oduc ion o
de ices and sys ems. The basic p inciples o he Inkje and Ae osol Je P in ing me hods, as well as
hei ad an ages and limi a ions, a e p esen ed in he ollowing sec ions.
6.1.1 Inkje P in ing
Inkje P in ing (IJP) is a con ac less, di ec -w i ing me hod based on he ejec ion o
mic od ople s con aining he unc ional ma e ial o be deposi ed. The ink (a mix u e o sol en s and
he unc ional ma e ial) is ed in o he deposi ion head as a solu ion o colloidal suspension and i is
di ec ly pa e ned on o he subs a e, hea ed o assis sol en e apo a ion. The ink is eleased h ough
a nozzle in he o m o mic od ople s, con olled by a piezoelec ic, he mal o elec ohyd odynamic
ac ua o [189], as shown in Figu e 6-2. The d ople je is p oduced due o p essu e pulses, gene a ed
h ough de o ma ion o he luid ca i y, h ough he mal expansion o he ink and he sudden
o ma ion ollowed by a collapse o apou bubbles, o by gene a ing an elec ic ield be ween he
nozzle and he subs a e, espec i ely. The ejec ion o d ople s can occu in ei he o he wo ollowing
modes. The con inuous ink je (CIJ) mode in ol es he applica ion o a DC ield be ween he nozzle
and he subs a e pla e, esul ing in an unin e up ed elease o cha ged d ople s in a s eam ha b eaks
down in o a column o indi idual d ople s, due o Rayleigh ins abili y [190],[191]. The ink is ecycled
by di ec ing he je owa ds a gu e o de lec ed owa ds he subs a e by applying an elec ic ield
be ween wo me allic pla es. The d ople -on-demand (DOD) mode is cha ac e ized by a pulse-like
elease o d ople s when eques ed by he pa e n o be p in ed, hus sa ing ink ma e ial when he
nozzle is no ac ua ed. Ejec ion equencies o 1−20 kHz can be achie ed wi h piezoelec ic
ansduce s [190]. Elec ohyd odynamic ac ua o s enable he nozzle ope a ion in bo h he CIJ and
DOD mode, depending on he ype o applied cu en (DC o AC, espec i ely) [189],[192].
Figu e 6-1
: O e iew o common p in ing echniques employed
o o ganic ma e ials (adap ed om [189]).
102
d ying o he d ople s caused by he shea h gas low. The op imal NMP/xylene/IPA a io was ound
o be 1:1:2 / , since lowe pe cen ages o NMP led o PANI p ecipi a ion. This new ink was p in ed
on qua z and polyimide ape subs a es wi h he same CGF and ShGF se ings, whe eas he highe
PANI concen a ion allowed a as e deposi ion a a pla e speed o 2 mm/s. The new ink o mula ion
in combina ion wi h he UV/ozone ea men o he subs a e (which is no sui ed o ib oin
subs a es) allowed o ob ain well-de ined, 160 µm wide lines wi hou any no able o e sp ay o
spo ing, as shown in Figu e 6-4 c, d).
O he p epa a ions, such as ink il e ing wi h a 0.2 µm PTFE sy inge il e , we e excluded
because i was de imen al o he p in ing p ocess, causing spo ing. The eason is ha due o he low
po ion o NMP, some o he PANI p ecipi a ed, o ming a colloidal suspension ha was emo ed by
il a ion, lowe ing oo much he ma e ial load o he ink. Hence, il e ing was omi ed in he ollowing
es s and ul asound ea men was ound o be he only p ocess allowing a be e PANI dispe sion in
he chosen sol en mix u e. Wi h his app oach, six s acking laye s o PANI we e p in ed on o a
15x7 mm qua z subs a e equipped wi h C elec odes de ining he OMD channel. La ge-a ea ilms
on qua z ha e p o en p oblema ic as we obse ed a poo laye adhesion as well as gaps be ween he
p in ed lines, as shown in Figu e 6-5 a). Hence, high esis ance alues, in he o de o MΩ, we e
ound a e he ilm doping wi h HCl. Fo compa ison, LS-deposi ed channels (30 laye s) o
Figu e 6-4: Uppe panels: mic og aphs o ae osol je p in s o a p is ine 0.1 mg/ml PANI ink in
NMP/IPA (1:1 / ) on a) qua z and
b) ib oin, wi h line wid hs o ca. 200 µm. P in ing on qua z esul ed i
n inhomogeneous d ying and he o ma ion o PANI islands
while he ea u es we e be e de ined on ib oin. Lowe panels: mic og aphs o ae osol je p in s wi h
a 1.0 mg/ml PANI ink in
NMP/xylene/IPA (1:1:2 / ) on c) qua z and d) polyimide ape, wi h line wid hs o ca. 160 µm.
The sha p lines coming o he p in ed
line in panel d) a e sc a ches on he ape su ace.

103
equi alen dimensions displayed esis ance alues be ween 5 and 50 kΩ. The same ink was p in ed
on o Si/SiO2 subs a es (1−2 µm he mal oxide laye ) wi h gold elec odes p e ab ica ed by
pho oli hog aphy (dSD = 200, 10 o 5 µm; w = 6 mm). The line de ini ion was la gely imp o ed
because o he be e we abili y o he subs a e, as shown in Figu e 6-5 b). A e doping in HCl, he
measu ed channel esis ances we e in he anges o 10−50 kΩ o 200 µm channels, 450−1500 Ω o
10 µm channels and 230−1000 Ω o 5 µm channels. Hence, esis ance alues scaling was consis en
wi h he scaling o channel leng h, indica ing he ilms’ quasi-ohmic beha iou . These alues a e
compa able o s anda d OMDs, albei lowe alues a e o be expec ed conside ing he geome y o
he channel. Despi e he imp o ed co e age, na ow gaps be ween he p in ed lines we e s ill
ecognizable upon op ical mic oscopy. Fu he mo e, he SEM images and mic og aph in Figu e 6-7
a, b, e) e ealed ha he deposi ed laye was again qui e inhomogeneous, wi h c acks and po es up o
1 µm (wi h he deposi ion o a ew la ge ones up o 10 µm) in diame e .
Using he abo e PANI ilms, he i s e e made OMD wi h a p in ed PANI laye was
ab ica ed on es pa e ns wi h dSD = 200 µm and using P30-Alli-0113 as he SPE. The scheme and
I-V cha ac e is ics o he de ice a e shown in Figu e 6-6, displaying an o dina y mem is i e
cha ac e is ics, ea u ing a hys e esis and ec i ica ion o he I-V cu e and p o ing he concep o
p in ed o ganic mem is i e de ices.
Figu e 6-5: Mic og aphs o PANI
p in ed wi h he 1.0 mg/ml ink in NMP/xylene/IPA (1:1:2 / ) on o a) a s anda d qua z subs a e
(6 p in ed laye s) and b) a silicon subs a e (1−2 µm oxide laye ) wi h p e ab ica ed gold elec odes (dSD = 200 µm, 10 p in ed laye s).
Figu e 6-6: a) Schema ic ep esen a ion and b) I-V cha ac e is ics o an OMD ab ica ed wi h P30-Alli-0113 as he SPE, ea u
ing an
AJ p in ed PANI laye on Si/SiO2 wi h Au elec odes (dSD = 200 µm). Scanning was done wi h a dwell ime o 2 s pe s ep o 0.02 V.
104
Figu e 6-7: SEM images o ae osol j
e p in ed p is ine PANI on Si subs a es wi h he 1.0 mg/ml ink in NMP/xylene/IPA (1:1:2 / ):
a, b) 10 p in ed laye s wi hou plasma p e- ea men o he subs a e a 1K and 30K magni ica
ion, espec i ely; c, d) 4 p in ed laye s
wi h plasma p e-
ea men o he subs a e a 10K and 30 K magni ica ion, espec i ely. No ewo hy is he imp o ed co e age and
educed po e amoun and a e age diame e in c, d). Panel e) shows a mic og aph o he
sample om panels a, b). No e he p o ile
di e ence be ween he bulk and edges o he lines as well as he c acks wi hin he p in ed laye . Fo compa ison, panel ) s
hows a SEM
image o an LS-deposi ed PANI channel (24 LS-laye s) om e e ence [109].
The LS laye has a g anula mo phology while he
p in ed s uc u es a e a he ib ous.
105
Ne e heless, such de ec i e ilms imply he need o a new p in ing pa e n, consis ing o an
enhanced o e lapping be ween adjacen p in ed lines, as demons a ed in Figu e 6-8. This pa e n has
been adop ed in all he ollowing deposi ion uns, aiming o educe he gaps be ween adjacen lines
and he po osi y o he laye . Addi ionally, o u he imp o e he we abili y o he subs a e, i was
subjec ed o a O2-plasma p e- ea men , aimed a bo h emo ing e en ual esidual o ganic
con aminan s and making i mo e hyd ophobic by elimina ing su ace-bound wa e molecules. Using
his app oach, he esis ance o he p in ed laye s was u he lowe ed o abou 340−6500 Ω and
30−500 Ω o subs a es wi h dSD = 200 and 10 µm, espec i ely. As shown by he SEM images in
Figu e 6-7 c, d), he mo phology o he deposi ed laye is signi ican ly imp o ed, gi ing a much be e
conduc ing channels made o ewe p in ed laye s.
6.2.2 Ae osol je p in ing o a chi osan:PANI composi e ma e ial
Al hough ou app oach p oduces PANI ilms wi h educed de ec s and lowe esis ance
alues, i seemed o be oo much dependen on he p ocessing ou e. So, i appea ed o be mo e
con enien o change he s a egy and de elop a new PANI-based ink ha would allow a be e
co e age o he subs a e. The dewe ing issues connec ed o he ink composi ion a e di icul o
o e come e en by means o su ace ea men , wi h he main p oblem being he p esence o NMP.
As a non ola ile sol en , NMP is gene ally no well-sui ed o p in ing applica ions. Besides, i poses
se e e heal h isks being classi ied as oxic and e a ogenic. Fo his eason, we p oceeded by
sea ching o modi y he p in ed ma e ial i sel since p is ine PANI is insoluble in o he sol en s.
We ook inspi a ion om he wo k by Ra uchne e al. [236], desc ibing a chi osan:PANI
composi e ma e ial ha is epo ed o be mo e conduc i e han polyaniline (due o a highe su ace
a ea). Acco ding o ha wo k, his ma e ial is also mo e he mally and elec ochemically s able,
making i a p omising ma e ial o elec och omic de ices and capaci o s. Mos impo an ly o he
p esen hesis wo k, he ma e ial is epo ed o be wa e -soluble, making i an in e es ing candida e as
a PANI su oga e o p in ing applica ions. The wa e -solubili y is impa ed o he blend by chi osan
– a biocompa ible polyme ha has al eady been used in OMDs as an SPE [237].
Figu e 6-8: Schema ic ep esen a ion o he new adop ed p in ing pa e n, a) op iew, b) side iew.
In one case each
nex laye is p in ed on op o he p e ious one and in he o he , he e is an o se o hal a s ep wi h each nex laye .
106
Acco ding o he p ocedu e om li e a u e [236], he chi osan:PANI blend (CPA, see Figu e
6-9) was syn hesized om aniline in hyd ochlo ic medium in he p esence o p e-dissol ed, low
molecula chi osan, ollowing an oxida i e polyme iza ion mechanism. The ull syn hesis ins uc ion
is gi en in sec ion 2.3. A da k g een p oduc was ob ained, indica ing he eme aldine sal o m o
PANI. The IR spec um in Figu e 6-10 ea u es he cha ac e is ic abso p ion bands: δop(C−H) and
δip(C−H) in 1,4-disubs i u ed ings a 828 cm-1 and 1106−1164 cm-1, espec i ely; bipola on ν(=NH+−)
a 1140 cm-1 (as a shoulde band); pola on ν(C−N•+) a 1238 cm-1; ν(C−N) o seconda y a oma ic
amines a 1307 cm-1; ν(C=N) in he icini y o quinoid ings a 1379 cm-1; benzenoid ν(C=C) a 1495
cm-1; quinoid ν(C=C) a 1586 cm-1 [17]. The absence o a sha p band a ound 3610−3645 cm-1 indica es
ha no ee −OH g oups a e p esen , sugges ing ha hey pa icipa e in in e molecula hyd ogen
bonds, cha ac e ized by a b oad band a ound 3200−3400 cm-1. The same applies o a b oad band
a ound 3000−3200 cm-1 which can be a ibu ed o H-bound NH g oups. The band a 3385 cm-1
co esponds o ee NH g oups o PANI ha do no pa icipa e in H-bonding. The small band a 1009
cm-1 co esponds o ν(C-O) o he seconda y cyclic alcohol moie ies in chi osan. O e all, he chi osan-
ela ed signals a e weak wi h espec o hose o PANI, indica ing ha he blend mos likely consis s
mainly o polyaniline.
Figu e 6-9: Chemical s uc u e o he chi osan:PANI composi e (on he le ) and a pho og aph o he ob ained p oduc (on he igh ).
Figu e 6-10: FTIR spec um o he syn hesized CPA composi e (KB pelle ).
107
Addi ional s uc u al in o ma ion could be gained om MALDI-TOF measu emen s,
al hough no p ecise in o ma ion could be gi en abou he molecula weigh dis ibu ion o he o med
PANI chains. The analysis has p o en di icul because he composi e is insoluble in he mos
common sol en s like wa e , NMP, isop opanol, e hanol, DMSO, ace oni ile (ACN), chlo o o m o
THF. Dispe sions on he o he hand adhe ed oo weakly on he subs a e/ma ix, o en making an
e ec i e ioniza ion di icul . As exempli ied in Figu e 6-11 a), he e was a g oup o h ee signals
a ound 3400 u ha eoccu ed in he majo i y o all he acqui ed spec a. I appea s as hough his
pa icula o ma ion is especially abundan in ou ma e ial and/o mo e s able unde he ioniza ion
condi ions. Al hough in o ma ion could be gained abou he s uc u e o his species, i is easonable
o assume ha he signals a ise om chain-like o cyclic polyaniline s uc u es. An indica ion o his
a e he di e ences in he m/z o he indi idual peaks, being ca. 71.3 u be ween he second and he
i s , and ca. 115.8 be ween he hi d and he i s one. These a omic weigh s co espond o a doping
o imine ni ogen a oms by wo HCl uni s (ca. 72.9 u) o wo CH3COOH uni s (ca. 120.1 u, emaining
sol en aces om he syn hesis). The pe iodically occu ing, weake signals in Figu e 6-11 a) also
ha e a di e ence o ca. 72 be ween hem, u he indica ing he p esence o PANI chains wi h a ious
le els o HCl doping. Simila esul s can be seen in Figu e 6-11 b), exhibi ing he same pe iodici y
o signals in he inse and indica ing chain leng hs o up o 8 kDa. Howe e , conside ing he solubili y
issues, i is likely ha he majo i y o he p oduc has a signi ican ly highe molecula weigh , since
only a ac ion o i could be dissol ed in HFIP. In he u u e, a mo e de ailed analysis o he p oduc
(e. g. deposi ed as a solu ion o di ec ly p in ed by AJP) could deli e mo e insigh abou i s s uc u e.
Figu e 6-11: MALDI-TOF mass spec a o he syn hesized CPA composi e deposi ed on he subs a e a) as a dispe sion
in
H2O/ACN (1:1 / ), b) as a solu ion in H2O/HFIP (1:1 / ). The baselines a e sub ac ed, he ma gin o e o is 1000 ppm.

108
Despi e he solubili y issues o CPA, i could easily be dispe sed wi hin 10 s by means o
ul asound ea men in wa e /IPA (8:2 / ), o ming a da k g een suspension ha emains s able o e
long pe iods o ime, wi h minimal ep ecipi a ion a e weeks. The ac ha he ma e ial is ully
dispe sed and no dissol ed is p o en by il a ion wi h a 0.45 µm PTFE sy inge il e , esul ing in a
clea , colou less il a e. We a ibu e such beha iou o he ela i ely small ac ion o chi osan in
he blend, as sugges ed by IR, as well as i s p esumably high molecula weigh . In he o iginal pape
[236], he molecula weigh o he u ilized chi osan was no speci ied, bu he epo ed pa icle size
and pla e-like mo phology o he ma e ial ob ained by Ra uchne e al. (see SEM image in Figu e 6-12
d) sugges ha i migh ha e been la ge han in ou case (low molecula weigh , 20−50 kDa). Fo
compa ison, Figu e 6-12 a, b) shows SEM images o CPA p in ed wi h he wa e /IPA-based, 2.0
mg/ml ink (8:2 / ) on Si/SiO2 subs a es. The mo phology esul s o be ib ous and a he simila o
he deposi ed p is ine PANI samples om he p e ious sec ion. The co e age o he subs a e is
la gely imp o ed, e en hough he su ace has no been p e- ea ed wi h plasma, sugges ing a much
be e we ing o he subs a e and e apo a ion a e o he ink. The esul ing p in ed lines we e e y
well de ined, wi h wid hs down o 120 µm. The SEM images show ha he p in ed laye is qui e
po ous and ough, wi h andomly occu ing si es whe e some agg ega ion o he ma e ial occu s. The
hickness o he laye was es ima ed om ATM measu emen s o single-laye ilms on he edge o
Figu e 6-12: a, b) SEM images o single laye p in ed CPA on a Si/SiO2
subs a e a 1K and 20K magni ica ion, espec i ely. c) P o ile
o he p in ed laye measu ed by means o AFM. d) SEM image o he blend om e e ence [236], o compa ison.
109
he p in ed laye (see Figu e 6-12 c). Typically, alues a ound 1 µm pe laye we e ob ained. Despi e
he excellen de ini ion o ea u es, he adhesion o he laye was qui e poo as i could be
mechanically emo ed wi h ease. Plasma ea men o he subs a e could possibly sol e his issue in
u u e expe imen s. Resis ance as low as 14 Ω could be achie ed a e doping in HCl. Conside ing a
channel wid h o 4 mm, leng h dSD o 10 µm and es ima ed hickness o 2 µm (1 µm pe laye ), his
co esponds o a conduc i i y o ca. 1.8 S/cm. Fo compa ison, p is ine PANI ilms deposi ed wi h
he LS echnique (30 laye s) on s anda d 15x7 mm qua z subs a es show conduc i i y le els a ound
50 S/cm. In i s doped, eme aldine sal s a e, he laye displays nea ly ideal ohmic beha iou , as shown
in he inse o Figu e 6-13 a).
A mem is i e de ice has been ab ica ed by p in ing wo CPA laye s as he conduc i e
channel on a Si/SiO2 subs a e equipped wi h gold elec odes (dSD = 10 µm), using he gel P30-Alli-
01521 as he SPE and a sil e wi e as he coun e elec ode. The I-V cha ac e iza ion a e shown in
Figu e 6-13 a). The de ice displays a e y high pe o mance wi h a well-de ined, ep oducible
hys e esis o he I-V cu e. Fu he es s ha e shown ha he OMD emains ully unc ional a scan
a es as high as 100 mV/s, wi h mino widening o he hys e esis, al hough he oxida ion eac ion
s a s o no ably lag behind he applied bias a scan a es abo e 40 mV/s (Figu e 6-13 b).
Figu e 6-13: a) 40 consecu i e I-V cycles
measu ed wi h a dwell ime o 2s pe s ep o 0.02 V (10 mV/s scan a e) on an OMD
ab ica ed wi h 2 p in ed CPA laye s as he conduc i e channel and P30-Alli-01521 as he SP
E (on Si subs a e wi h Au elec odes and
dSD = 10 µm). In he inse , he almos ideal ohmic beha iou o he laye is demons a ed. b) I-
V cha ac e iza ion o he de ice a
di e en scan a es. c) Pho og aph o he de ice (only he middle le laye is in ol ed).
110
6.3 Towa ds AJP-compa ible polyelec oly e ma e ials
Al hough he qui e good and no el esul s ob ained om de ices ab ica ed wi h he CPA-
based ink om he p e ious sec ion, he e is a swi ching kine ics ha is no ma ching he expec a ion
based on he educed channel dimensions. Such a beha iou can be explained conside ing he
hickness o he p in ed CPA laye . Wi h a hickness o 1 µm, on a scale o a channel leng h o only
10 µm, he laye a ains bulk p ope ies in ha i s conduc i i y s ongly depends on he cha ge ans e
be ween he single deposi ed laye s, o in he case o p in ed CPA, a he be ween he single ib e
mic os uc u es. This is especially ele an o he bo om-con ac / op-ga e geome y used so a ,
whe e he swi ching o he channel ini ia es close o he ga e elec ode, bu only comes in o e ec
once he conduc i e connec ion is es ablished wi h he sou ce and d ain e minals. As was discussed
in sec ion 5.1, his issue could be ci cum en ed by adop ing a di e en kind o de ice a chi ec u e
wi h a bo om-con ac /bo om-ga e a angemen , whe e he bulk p ope ies o he channel play a
mino ole.
The e sa ili y o AJP p o ides he necessa y ools o app oach he sough -a e goal o
ab ica ing ully p in ed de ices, including all he componen s o an OMD. Wi h he abili y o p in
conduc i e me al nanopa icle inks, AJP allows us o ab ica e a sil e ga e [218] and gold sou ce and
d ain elec odes [238], wi h he inal s ep o in e connec ing hem wi h he conduc i e CPA ink
de eloped abo e. The missing link is he deposi ion o he polyelec oly e ma e ial o c ea e he
junc ion be ween he ga e and he conduc i e channel. By consequence, i would be equi ed o p in
CPA on o he la e , howe e i is no expec ed o be an issue since AJP is sui able o he deposi ion
on a a ie y o subs a es, including lexible, une en, and so ones. The ollowing sec ion is dedica ed
o making s eps owa ds de eloping p in able polyelec oly e ma e ials.
6.3.1 Low- iscosi y PEO-based hyd ogels
Ou de elopmen o p in ed polyelec oly es elies on he knowledge acqui ed when
imp o ing he PEO-based SPE composi ion in chap e 4. Since he p ope ies o he gel emain
unchanged when main aining he same dopan :PEO a ios, i seems possible o p epa e a less
concen a ed gel wi h a low enough iscosi y o being p in ed. Howe e , gels ab ica ed wi h high-
molecula PEO (Mw = 8 MDa) ha e p o en oo iscous o be p ocessed a easonable concen a ions,
e en down o 5 mg/ml. We hence p oceeded o p epa e PEO gels wi h lowe a e age chain leng h
(Mw = 1 MDa), yielding mo e luid solu ions o su icien ly low iscosi y a 30 mg/ml. The du abili y
o he gels has been assessed op ically as in he p e ious cases. Ou o all he p epa ed gels, he h ee
mos p omising candida es we e P1M30-Alli-01521 (0.1 M AlCl3, 0.05 M LiClO4), P1M30-Alli-01532
(0.09 M AlCl3, 0.06 M LiClO4) and P1M30-ACG-0231 (0.15 M AlCl3, 0.05 M GndCl). The P1M30-
ACG-0231 gel has p o en he mos du able in compa ison o he LiClO4 doped coun e pa s, as shown
in Figu e 6-14 a, b), whe eas bo h ypes o doping we e on pa in analogous gels wi h 8 MDa PEO.
We assume ha in he case o lowe molecula , 1 MDa PEO, he o e all hyg oscopici y o he dopan
111
ensemble becomes less c i ical because he chains become easie o hyd a e, while he bene i o he
s ong chao opic cha ac e o he guanidinium ca ion becomes mo e signi ican and comes in o e ec
mo e.
Despi e he low iscosi y o he gel, he a omiza ion o he SPE ink could no be achie ed
and hence i could no be used in he AJP. This is likely due o he s ill e y high mean molecula
weigh o PEO chains, which p e en s hem om e ec i ely en e ing he ae osol c ea ed by he
ul asonic a omize . On he o he hand, u he lowe ing o he molecula weigh o PEO also was
unsuccess ul as he p epa ed gels o compa able concen a ions (30−150 mg/ml) did no solidi y e en
o e nigh , po en ially making i unsui able o he deposi ion a oom empe a u e. Assis ing he wa e
e apo a ion h ough mild hea ing a 40 °C esul ed in apid and i e e sible d ying ou o he samples,
as shown in Figu e 6-14 c). I appea s ha he sys em becomes less e ec i e a con aining humidi y
wi h dec easing PEO Mw, while he in e ac ions be ween he sho e PEO chains a e oo s ong o
in e he o ma ion o agg ega es by inc easing he humidi y.
While he applica ion o P1M gels was unsuccess ul by means o AJP, i is possible ha
p in ing could succeed wi h he o he wise less ad anced IJP echnique, hanks o i s mo e s aigh -
o wa d mechanism o he o ma ion o he ink je ha equi es no p o isional a omiza ion o he
ma e ial. Issues migh a ise om he need o hea he deposi ion pla e o a be e deposi ion and
de ini ion o p in ed ea u es, ega ding he empe a u e and humidi y sensi i i y o he SPE sys em.
This does no conce n he deposi ion p ocess o he PEO gel i sel , since in his case, he hea ing o
he pla e in pa icula and he p ecision o p in ing in gene al is no equi ed, since i is applied as a
bulk and only se es o p o ide he medium o he elec ochemical eac ions be ween he channel
and he ga e. Much a he , i is he las manu ac u ing s ep, being he deposi ion o CPA in a bo om-
con ac /bo om-ga e geome y, which is mos likely o cause d ying ou and c ys alliza ion o PEO.
Howe e , i has been shown in sec ion 4.2.3.4 ha gels based on guanidine chlo ide can be quickly
egene a ed in a humid a mosphe e a e d ying, making P1M30-ACG-0231 i a p omising candida e
o u u e a emp s o p in ing wi h IJP.
Figu e 6-14: Pho og aphs o lowe molecula PEO gels wi h Mw = 1 MDa, 6 h a e deposi ion. a) Gel P1M30-Alli-
01521 (30 mg/ml
PEO, 0.1 M AlCl3, 0.05 M LiClO4) wi h incipien d ying in he cen e o he cas d op. b) Gel P1M30-ACG-
0231 (30 mg/ml PEO,
0.15 M AlCl3, 0.05 M GndCl). The agg ega es on he pe ime e did no g ow conside ably o e
ime. c) An e en lowe molecula
PEO gel wi h Mw = 12.5 kDa a e d ying a 40 °C. The c ys allized PEO could no be egene a ed in a humid a mosphe e o by
di ec ly adding wa e o he sys em.
118
s udy o abo emen ioned noise- ela ed and o he e ec s, bu also make ad ances in he
bio echnological, biomedical, compu e science, neu oscience and obo ics ields.
In conclusion, wi h ou esea ch wo k we belie e we achie ed impo an , undamen al s eps
in his di ec ion by de eloping ma e ials and me hod o ab ica ions ha allow o p oduce much be e
pe o ming, mo e s able, eliable and ep oducible OMDs wi h ma e ials ha make he in e acing
wi h he bio-wo ld iable.

119
Bibliog aphy
[1] Y. Li, Z. Wang, R. Midya, Q. Xia, J. J. Yang. “Re iew o mem is o de ices in
neu omo phic compu ing: ma e ials science and de ice challenges.” J. Phys. D: Appl. Phys.
51, 503002 (2019).
[2] D. O. Hebb. “The O ganiza ion o Beha io : A Neu opsychological Theo y.” (Wiley, 1949).
[3] N. Roches e , J. H. Holland, L. H. Haib , W. L. Duda. “Tes s on a cell assembly heo y o
he ac ion o he b ain, using a la ge digi al compu e .” IRE T ans. In o m. Theo . 2, 80–93
(1956).
[4] F. Rosenbla . “The pe cep on: A p obabilis ic model o in o ma ion s o age and
o ganiza ion in he b ain.” Physiolog. Re . 65, 386–408 (1958).
[5] B. Wid ow, W. H. Pie ce, J. B. Angell. “Bi h, li e, and dea h in mic oelec onic sys ems.”
IRE T ans. Mil. Elec on. MIL–5, 191–201 (1961).
[6] L. O. Chua. “Mem is o − The missing ci cui elemen .” IEEE T ans. Ci cui Theo . 18,
507–519 (1971).
[7] D. B. S uko , G. S. Snide , D. R. S ewa , R. S. Williams. “The missing mem is o ound.”
Na u e Le . 453, 80–84 (2008).
[8] A. Asami su, Y. Tomioka, H. Kuwaha a, Y. Toku a. “Cu en swi ching o esis i e s a es
in magne o esis i e mangani es.” Na u e 388, 50–52 (1997).
[9] H. Pagnia, N. So nik. “Bis able swi ching in elec o o med me al–insula o –me al de ices.”
Phys. S a us Solidi A 108, 11–65 (1988).
[10] Y. B. Nian, J. S ozie , N. J. Wu, X. Chen, A. Igna ie . “E idence o an oxygen di usion
model o he elec ic pulse induced esis ance change e ec in ansi ion-me al oxides.”
Phys. Re . Le . 98, 146403 (2007) (2007).
[11] S. Q. Liu, N. J. Wu, A. Igna ie . “Elec ic-pulse-induced e e sible esis ance change e ec
in magne o esis i e ilms.” Appl. Phys. Le . 76, 2749–2751 (2000).
[12] A. Beck, J. G. Bedno z, C. Ge be , C. Rossel, D. Widme . “Rep oducible swi ching e ec
in hin oxide ilms o memo y applica ions.” Appl. Phys. Le . 77, 139–141 (2000).
[13] V. E okhin, T. Be zina, M. P. Fon ana. “Hyb id elec onic de ice based on polyaniline-
polye hyleneoxide junc ion.” J. Appl. Phys. 97, 064501 (2005).
[14] J. Yao, Z. Sun, L. Zhong, D. Na elson, J. M. Tou . “Resis i e swi ches and memo ies om
silicon oxide.” Nano Le . 10, 4105–4110 (2010).
[15] A. Mehonic, S. Cue , M. Wojdak, S. Hudziak, O. Jambois, C. Labbé, B. Ga ido, R. Rizk,
A. J. Kenyon. “Resis i e swi ching in silicon suboxide ilms.” J. Appl. Phys. 111, 074507
(2012).
[16] J. W. Han, M. Meyyappan. “Coppe oxide esis i e swi ching memo y o e- ex ile.” AIP
Ad . 1, 032162 (2011).
120
[17] L. Aleksee a, T. Naba ame, T. Chikyow, A. Pe o . “Resis i e swi ching cha ac e is ics in
mem is o s wi h Al2O3/TiO2 and TiO2/Al2O3 bilaye s.” Jap. J. Appl. Phys. 55, 08PB02
(2016).
[18] Y. E. Syu, T. C. Chang, J. H. Lou, T. M. Tsai, K. C. Chang, M. J. Tsai, Y. L. Wang, M. Liu,
S. M. Sze. “A omic-le el quan ized eac ion o H Ox mem is o .” Appl. Phys. Le . 102,
172903 (2013).
[19] J. Sun, E. Lind, I. Maximo , H. Q. Xu. “Mem is i e and memcapaci i e cha ac e is ics o a
Au/Ti-H O 2-InP/InGaAs diode.” IEEE Elec on. De ice Le . 32, 131–133 (2011).
[20] W. Gee Kim e al. “Dependence o he swi ching cha ac e is ics o Resis ance Random
Access Memo y on he ype o ansi ion me al oxide; TiO2, Z O2, and H O2.” J.
Elec ochem. Soc. 158, H417–H422 (2011).
[21] A. P akash, D. Jana, S. Maikap. “TaOx-based esis i e swi ching memo ies: P ospec i e and
challenges.” Nanoscale Res. Le . 8, 418 (2013).
[22] X. Cao, X. Li, X. Gao, W. Yu, X. Liu, Y. Zhang, L. Chen, X. Cheng. “Fo ming- ee colossal
esis i e swi ching e ec in a e-ea h-oxide Gd2 O3 ilms o mem is o applica ions.”
Jou nal o Applied Physics 106, (2009).
[23] A. Younis, D. Chu, S. Li. “Oxygen le el: The dominan o esis i e swi ching cha ac e is ics
in ce ium oxide hin ilms.” J. Phys. D Appl. Phys. 45, 355101 (2012).
[24] A. Kuma , Y. Rawal, M. S. Baghini. “Fab ica ion and cha ac e iza ion o he ZnO-based
mem is o .” In e na ional Con e ence on Eme ging Elec onics 2012, P oceedings 4–6
(2012) doi:10.1109/ICEmElec.2012.6636244.
[25] F. M. Simanjun ak, D. Panda, K. H. Wei, T. Y. Tseng. “S a us and p ospec s o ZnO-based
esis i e swi ching memo y de ices.” Nanoscale Res. Le . 11, 368 (2016).
[26] G. Milano, L. Boa ino, C. Riccia di. “Junc ion p ope ies o single ZnO nanowi es wi h
asymme ical P and Cu con ac s.” Nano echnology 30, 244001 (2019).
[27] V. Jousseaume e al. “Back-end-o -line in eg a ion app oaches o esis i e memo ies.”
IEEE In e na ional In e connec Technology Con e ence P oceedings 41–43 (2009)
doi:10.1109/IITC.2009.5090335.
[28] Q. Wang e al. “Reliable Ge2Sb2Te5 based phase-change elec onic synapses using ca bon
doping and p og ammed pulses.” J. Ma e iomics in p ess, (2021).
[29] S. I. Oh, I. H. Im, C. Yoo, S. Y. Ryu, Y. Kim, S. Choi, T. Eom, C. S. Hwang, B. J. Choi.
“E ec o elec ode ma e ial on he c ys alliza ion o GeTe g own by a omic laye deposi ion
o phase change andom access memo y.” Mic omachines (Basel) 10, 281 (2019).
[30] D. Ielmini, A. L. Lacai a. “Phase change ma e ials in non- ola ile s o age.” Ma e . Today
14, 600–607 (2011).
[31] G. Be i Bene en i e al. “Ca bon-doped GeTe: A p omising ma e ial o Phase-Change
Memo ies.” Solid-S a e Elec on. 65–66, 197–204 (2011).
121
[32] B. S. Lee, R. M. Shelby, S. Raoux, C. T. Re e , G. W. Bu , S. N. Bogle, K. Da mawika a,
S. G. Bishop, J. R. Abelson. “Nanoscale nuclei in phase change ma e ials: O igin o di e en
c ys alliza ion mechanisms o Ge2Sb2Te5 and AgInSbTe.” J. Appl. Phys. 115, 063506
(2014).
[33] G. Liu, Y. Chen, C. Wang, W. Zhang, R.-W. Li, L. Wang. “Polyme mem is o o
in o ma ion s o age and neu omo phic applica ions.” Ma e . Ho iz. 1, 489–506 (2014).
[34] L. Yuan, S. Liu, W. Chen, F. Fan, G. Liu. “O ganic memo y and mem is o s: om
mechanisms, ma e ials o de ices.” Ad . Elec on. Ma e . 7, 2100432 (2021).
[35] T. Zhang e al. “Nega i e di e en ial esis ance, memo y, and econ igu able logic unc ions
based on monolaye de ices de i ed om Gold nanopa icles unc ionalized wi h
elec opolyme izable TEDOT uni s.” J. Phys. Chem. C 121, 10131–10139 (2017).
[36] S. Goswami e al. “Robus esis i e memo y de ices using solu ion-p ocessable me al-
coo dina ed azo a oma ics.” Na u e Ma e . 16, 1216–1224 (2017).
[37] G. Zhou, Z. Ren, L. Wang, B. Sun, S. Duan, Q. Song. “A i icial and wea able albumen
p o ein mem is o a ays wi h in eg a ed memo y logic ga e unc ionali y.” Ma e . Ho iz. 6,
1877–1882 (2019).
[38] R. Kuma , R. G. Pillai, N. Pekas, Y. Wu, R. L. McC ee y. “Spa ially esol ed Raman
spec oelec ochemis y o solid-s a e poly hiophene/ iologen memo y de ices.” J. Am.
Chem. Soc. 134, 14869–14876 (2012).
[39] J. Q. Yang e al. “Fluo enone/ca bazole based bipola small molecules o non- ola ile
memo y de ices.” O g. Elec on. 78, 105584 (2020).
[40] A. A. Minnekhano , A. . Emelyano , D. A. Lapkin, K. E. Niki uy, B. S. Sh e so , A. A.
Nesmelo , V. . Rylko , V. A. Demin, V. . E okhin. “Pa ylene based mem is i e de ices
wi h mul ile el esis i e swi ching o neu omo phic applica ions.” Sci. Rep. 9, 10800
(2019).
[41] A. N. Ma suka o a e al. “Resis i e swi ching kine ics and second-o de e ec s in pa ylene-
based mem is o s.” Appl. Phys. Le . 117, (2020).
[42] A. N. Ma suka o a, A. A. Minnekhano , V. . Rylko , V. A. Demin, A. . Emelyano .
“Resis i e swi ching kine ics o pa ylene-based mem is i e de ices wi h Cu ac i e
elec odes.” J. Phys. Con . Se ies 1758, 012025 (2021).
[43] N. Raeis Hosseini, J. S. Lee. “Resis i e swi ching memo y based on bioinspi ed na u al solid
polyme elec oly es.” ACS Nano 9, 419–426 (2015).
[44] M. K. Ho a, M. K. Be a, B. Kundu, S. C. Kundu, C. K. Mai i. “A na u al silk ib oin p o ein-
based anspa en bio-mem is o .” Ad . Func . Ma e . 22, 4493–4499 (2012).
[45] M. K. Ram, M. Adami, M. Sa o e, M. Sale no, S. Paddeu, C. Nicolini. “Compa a i e s udies
on Langmui -Schae e ilms o polyanilines.” Syn h. Me . 100, 249–259 (1999).
122
[46] R. C. San elice, D. T. Balogh, F. Lede le, J. Adams, S. Beue mann. “S udies o Langmui
and Langmui -Schae e Films o Poly(3-Hexyl hiophene) and Poly(Vinylidene Fluo ide).”
J. Phys. Chem. B 124, 7037–7045 (2020).
[47] M. Na a o, J. Beni o, L. Pase a, I. Gascón, J. Co onas, C. Téllez. “Thin-Film
Nanocomposi e Memb ane wi h he Minimum Amoun o MOF by he Langmui -Schae e
Technique o Nano il a ion.” ACS Appl. Ma e . In e aces 10, 1278–1287 (2018).
[48] R. Y. N. Gengle , A. Veligu a, A. Eno iadis, E. K. Diaman i, D. Gou nis, C. Józsa, B. J. an
Wees, P. Rudol . “La ge-yield p epa a ion o high-elec onic-quali y g aphene by a
Langmui -Schae e app oach.” Small 6, 35–39 (2010).
[49] Z. U. Khan, A. Kausa , H. Ullah, A. Badshah, W. U. Khan. “A e iew o g aphene oxide,
g aphene buckypape , and polyme /g aphene composi es: p ope ies and ab ica ion
echniques.” J. Plas . Film Shee . 32, 336–379 (2016).
[50] A. . Kazak, M. A. Ma chenko a, K. S. Kho ko , D. A. Kochue , A. . Rogache , I. .
Kholodko , N. . Usol’ se a, M. S. Sa elye , A. Y. Tolbin. “Ul a hin Langmui –Schae e
ilms o slipped-co acial J- ype ph halocyanine dime : Sup amolecula o ganiza ion,
UV/Vis/NIR s udy and nonlinea abso bance o em osecond lase adia ion.” Appl. Su .
Sci. 545, 148993 (2021).
[51] J. S anley, Y. Dai, L. Bouche on, B. Lin, M. Me on, O. Shpy ko. “No el compa ison o
mic oscopy and di ac ion echniques on he s uc u e o i on oxide nanopa icle
monolaye s ans e ed by Langmui -Schae e me hod.” Re . Sci. Ins um. 86, 063704
(2015).
[52] S. E okhina, V. So okin, V. E okhin. “Polyaniline-based o ganic mem is i e de ice
ab ica ed by laye -by-laye deposi ion echnique.” Elec on. Ma e . Le . 11, 801–805
(2015).
[53] A. G. G een, A. E. Woodhead. “Aniline-black and allied compounds. Pa I.” J. Chem. Soc.
97, 2388–2403 (1910).
[54] A. G. G een, A. E. Woodhead. “Aniline-black and allied compounds. Pa II.” J. Chem. Soc.
101, 1117–1123 (1912).
[55] E. T. Kang, K. G. Neoh, K. L. Tan. “Polyaniline: a polyme wi h many in e es ing in insic
edox s a es.” P og. Polym. Sci 23, 277–324 (1998).
[56] J. G. Mas e s, Y. Sun, A. G. Macdia mid, A. J. Eps ein. “Polyaniline: allowed oxida ion
s a es.” Syn h. Me . 41, 715–718 (1991).
[57] J. G. Mas e s, J. M. Ginde , A. G. MacDia mid, A. J. Eps ein. “The moch omism in he
insula ing o ms o polyaniline: Role o ing- o sional con o ma ion.” J. Chem. Phys. 96,
4768–4778 (1992).
[58] T. Fukuda, H. Takezoe, K. Ishikawa, A. Fukuda, H. S. Woo, S. K. Jeong, J. Oh, J. S. Suh.
“IR and Raman s udies in h ee polyanilines wi h di e en oxida ion le els.” Syn h. Me . 69,
175–176 (1995).
123
[59] L. B ožo á, P. Holle , J. Ko ářo á, J. S ejskal, M. T cho á. “The s abili y o polyaniline in
s ongly alkaline o acidic aqueous media.” Polym. Deg ad. S ab. 93, 592–600 (2008).
[60] G. E. As u ias, A. G. Macdia mid, R. P. Mccall, A. J. Eps ein. “The oxida ion s a e o
‘eme aldine’ base.” Syn h. Me . 29, E157–E162 (1989).
[61] K. S ejskal, R. G. Gilbe . “Polyaniline. p epa a ion o a conduc ing polyme (IUPAC
Technical Repo ).” Pu e Appl. Chem 74, 857–867 (2002).
[62] F. Lux. “P ope ies o elec onically conduc i e polyaniline: a compa ison be ween well-
known li e a u e da a and some ecen expe imen al indings.” Polyme (Guild ) 35, 2915–
2936 (1994).
[63] J. S ejskal, P. K a och il, A. D. Jenkins. “The o ma ion o polyaniline and he na u e o i s
s uc u es.” Polyme (Guild ) 37, 367–369 (1996).
[64] J. W. Jeon, Y. Ma, J. F. Mike, L. Shao, P. B. Balbuena, J. L. Lu kenhaus. “Oxida i ely s able
polyaniline:polyacid elec odes o elec ochemical ene gy s o age.” Phys. Chem. Chem.
Phys. 15, 9654–9662 (2013).
[65] G. D’Ap ano, M. Lecle c, G. Zo i. “S abiliza ion and cha ac e iza ion o pe nig aniline sal :
he ‘acid-doped’ o m o ully oxidized polyanilines.” Mac omol. 25, 2145–2150 (1992).
[66] W. Lu, A. G. Fadee , B. Qi, B. R. Ma es. “Fab ica ing conduc ing polyme elec och omic
de ices using ionic liquids.” J. Elec ochem. Soc. 151, H33–H39 (2004).
[67] F. F. C. Bazi o, L. T. Sil ei a, R. M. To esi, S. I. Có doba de To esi. “On he s abiliza ion
o conduc ing pe nig aniline sal by he syn hesis and oxida ion o polyaniline in
hyd ophobic ionic liquids.” Phys. Chem. Chem. Phys. 10, 1457–1462 (2008).
[68] A. Ray, A. F. Rich e , A. G. Macdia mid, A. J. Eps ein. “Polyaniline:
p o ona ion/dep o ona ion o amine and imine si es.” Syn h. Me . 29, E151–E156 (1989).
[69] E. T. Kang, K. G. Neoh, K. L. Tan, B. T. G. Tan. “P o ona ion o he amine ni ogens in
eme aldine-e idence om X- ay pho oelec on spec oscopy.” Syn h. Me . 46, 227–233
(1992).
[70] P. Snauwae , R. Lazza oni, J. Riga, J. J. Ve bis , D. Gonbeau. “A pho oelec on
spec oscopic s udy o he elec ochemical p ocesses in polyaniline.” J. Chem. Phys. 92,
2187–2193 (1990).
[71] K. G. Neoh, E. T. Kang, K. L. Tan. “P o ona ion o leucoeme aldine in he solid s a e and
in solu ion.” J. Polym. Sci. B Polym. Phys. 31, 395–401 (1993).
[72] S. S a s öm, J. L. B édas, A. J. Eps ein, H. S. Woo, D. B. Tanne , W. S. Huang, A. G.
MacDia mid. “Pola on la ice in highly conduc ing polyaniline: heo e ical and op ical
s udies.” Phys. Re . Le . 59, 1464–1467 (1987).
[73] A. J. Eps ein, J. M. Ginde , F. Zuo, R. W. Bigelow, H.-S. Woo, D. B. Tanne , A. F. Rich e ,
W.-S. Huang, A. G. Macdia mid. “Insula o - o-me al ansi ion in polyaniline.” Syn h. Me .
18, 303–309 (1987).

124
[74] A. J. Eps ein, J. M. Ginde , F. Zuo, H.-S. Woo, D. B. Tanne , A. F. Rich e , M.
Angelopoulos, W.-S. Huang, A. G. Macdia mid. “Insula o - o-me al ansi ion in
polyaniline: e ec o p o ona ion in eme aldine.” Syn h. Me . 21, 63–70 (1987).
[75] S. Kuma Gup a, V. Lu h a, R. Singh. “Elec ical anspo and EPR in es iga ions: a
compa a i e s udy o d.c. conduc ion mechanism in mono alen and mul i alen ions doped
polyaniline.” Bull. Ma e . Sci 35, 787–794 (2012).
[76] M. B. Inoue, K. W. Nebesny, Q. Fe nandoa, M. Lnoueb. “X-Ray pho oelec on spec oscopy
o new soluble polyaniline pe chlo a es: e idence o he coexis ence o pola ons and
bipola ons.” J. Ma e . Chem. 1, 213–216 (1991).
[77] T. Hagiwa a, T. Demu a, K. Iwa a. “Syn hesis and p ope ies o elec ically conduc ing
polyme s om a oma ic amines.” Syn h. Me . 18, 317–322 (1987).
[78] E. T. Kang, K. G. Neoh, S. H. Kho , K. L. Tan, B. T. G. Tan. “S uc u al de e mina ion o
polyaniline by X-Ray pho oelec on spec oscopy.” J. Chem. Soc. Chem. Commun. 695–697
(1989).
[79] K. L. Tan, B. T. G. Tan, E. T. Kang, K. G. Neoh. “X- ay pho oelec on spec oscopy s udies
o he chemical s uc u e o polyaniline.” Phys. Re . B 39, 8070–8073 (1989).
[80] J. Yue, A. J. Eps ein. “Syn hesis o Sel -Doped Conduc ing Polyaniline.” J. Am. Chem. Soc.
112, 2800–2801 (1990).
[81] A. G. MacDia mid, J. C. Chiang, A. F. Rich e , A. J. Eps ein. “Polyaniline: a new concep
in conduc ing polyme s.” Syn h. Me . 18, 285–290 (1987).
[82] W. W. Focke. “Conduc ion mechanisms in polyaniline.” (1987).
[83] W. W. Focke, G. E. Wnek, Y. Wei. “In luence o oxida ion s a e, pH, and coun e ion on he
conduc i i y o polyaniline.” J. Phys.Chem. 91, 5813–5818 (1987).
[84] G. B. Appe ecchi, F. Alessand ini, M. Ca ewska, T. Ca uso, P. P. P osini, S. Scaccia, S.
Passe ini. “In es iga ion on li hium-polyme elec oly e ba e ies.” J. Powe Sou ces 97,
790–794 (2001).
[85] Z. Xue, D. He, X. Xie. “Poly(e hylene oxide)-based elec oly es o li hium-ion ba e ies.”
J. Ma e . Chem. A 3, 19218–19253 (2015).
[86] S. Zhao, Q. Wu, W. Ma, L. Yang. “Polye hylene oxide-based composi es as solid-s a e
polyme elec oly es o li hium me al ba e ies: a mini e iew.” F on . Chem. 8, 640 (2020).
[87] T. Be zina, V. E okhin, M. P. Fon ana. “Spec oscopic in es iga ion o an elec ochemically
con olled conduc ing polyme -solid elec oly e junc ion.” J. Appl. Phys. 101, 024501
(2007).
[88] T. Be zina, S. E okhina, P. Camo ani, O. Kono alo , V. E okhin, M. P. Fon ana.
“Elec ochemical con ol o he conduc i i y in an o ganic mem is o : A ime- esol ed X-
ay luo escence s udy o ionic d i as a unc ion o he applied ol age.” ACS Appl. Ma e .
In e aces 1, 2115–2118 (2009).
125
[89] V. E okhin, T. Be zina, A. Sme ie i, P. Camo ani, S. E okhina, M. P. Fon ana. “Bio-inspi ed
adap i e ne wo ks based on o ganic mem is o s.” Nano Commun. Ne w. 1, 108–117 (2010).
[90] E. M. Geniès, M. Lapkowski, J. F. Penneau. “Cyclic ol amme y o polyaniline:
in e p e a ion o he middle peak.” J. Elec oanal. Chem 249, 97–107 (1988).
[91] E. M. Geniès, M. Lapkowski. “Spec oelec ochemical e idence o an in e media e in he
elec opolyme iza ion o aniline.” J. Elec oanal. Chem 236, 189–197 (1987).
[92] S. P uneanu, E. Ve ess, I. Ma ian, L. Oniciu. “Cha ac e iza ion o polyaniline by cyclic
ol amme y and UV-Vis abso p ion spec oscopy.” J. Ma e . Sci. 34, 2733–2739 (1999).
[93] A. Sme ie i, V. E okhin, M. P. Fon ana. “O igin o cu en oscilla ions in a polyme ic
elec ochemically con olled elemen .” J. Appl. Phys. 103, 094517 (2008).
[94] V. A. Demin, V. . E okhin, P. K. Kashka o , M. . Ko alchuk. “Elec ochemical model o
he polyaniline based o ganic mem is i e de ice.” J. Appl. Phys. 116, 064507 (2014).
[95] N. S. Sa ici ci, H. Kuzmany, H. Neugebaue , A. Neckel. “S uc u al and elec onic
ansi ions in polyaniline: A Fou ie ans o m in a ed spec oscopic s udy.” J. Chem. Phys.
92, 4530–4539 (1990).
[96] S. I. Zhelude a, M. . Ko alchuk, S. Lagoma sino, N. N. No iko a, I. N. Bashelkhano , V.
. E okhin, L. A. Feigin. “Obse a ion o e anescen and s anding X- ay wa es in egion o
o al ex e nal e lec ion om molecula Langmui -Blodge ilms.” JEPT Le . 52, 170–175
(1990).
[97] S. H. Jo, T. Chang, I. Ebong, B. B. Bhad iya, P. Mazumde , W. Lu. “Nanoscale mem is o
de ice as synapse in neu omo phic sys ems.” Nano Le . 10, 1297–1301 (2010).
[98] L. Deng, G. Li, N. Deng, D. Wang, Z. Zhang, W. He, H. Li, J. Pei, L. Shi. “Complex lea ning
in bio-plausible mem is i e ne wo ks.” Sci. Rep. 5, 1–10 (2015).
[99] Z. Wang e al. “Mem is o s wi h di usi e dynamics as synap ic emula o s o neu omo phic
compu ing.” Na . Ma e . 16, 101–108 (2017).
[100] S. Ba is oni, V. E okhin, S. Ianno a. “F equency d i en o ganic mem is i e de ices o
neu omo phic sho e m and long e m plas ici y.” O g. Elec on. 65, 434–438 (2019).
[101] A. Sme ie i, T. Be zina, V. E okhin, M. P. Fon ana. “Polyme ic elec ochemical elemen o
adap i e ne wo ks: Pulse mode.” J. Appl. Phys. 104, 1–7 (2008).
[102] E. Juzekae a, A. Nas e dino , S. Ba is oni, T. Be zina, S. Ianno a, R. Khazipo , V.
E okhin, M. Mukh a o . “Coupling co ical neu ons h ough elec onic mem is i e
synapse.” Ad . Ma e . Technol. 4, 1800350 (2019).
[103] A. . Emelyano e al. “Fi s s eps owa ds he ealiza ion o a double laye pe cep on based
on o ganic mem is i e de ices.” AIP Ad ances 6, 111301 (2016).
[104] S. Ba is oni, V. E okhin, S. Ianno a. “O ganic mem is i e de ices o pe cep on
applica ions.” J. Phys. D Appl. Phys. 51, 284002 (2018).
126
[105] V. E okhin, T. Be zina, P. Camo ani, A. Sme ie i, D. Va oulis, J. Feng, M. P. Fon ana.
“Ma e ial mem is i e de ice ci cui s wi h synap ic plas ici y: lea ning and memo y.”
Bionanoscience 1, 24–30 (2011).
[106] M. Ziegle , R. Soni, T. Pa elczyk, M. Igna o , T. Ba sch, P. Meu els, H. Kohls ed . “An
elec onic e sion o Pa lo ’s Dog.” Ad . Func . Ma e . 22, 2744–2749 (2012).
[107] G. B. E men ou , R. F. Galán, N. N. U ban. “Reliabili y, synch ony and noise.” T ends
Neu osci. 31, 428–434 (2008).
[108] J. D. Hun e , J. G. Mil on, P. J. Thomas, J. D. Cowan. “Resonance e ec o neu al spike
ime eliabili y.” J. Neu ophysiol. 80, 1427–1438 (1998).
[109] K. Go shko , T. Be zina. “On he hys e esis loop o o ganic mem is i e de ice.”
BioNanoSci. 1, 198–201 (2011).
[110] S. Ba is oni, R. Sajapin, V. E okhin, A. Ve na, M. Cocuzza, S. L. Ma asso, S. Ianno a.
“E ec s o noise sou cing on o ganic mem is i e de ices.” Chaos, Soli ons F ac als 141,
110319 (2020).
[111] D. A. Be na ds, G. G. Mallia as. “S eady-s a e and ansien beha io o o ganic
elec ochemical ansis o s.” Ad . Func . Ma e . 17, 3538–3544 (2007).
[112] X. S akosas, M. Bongo, R. M. Owens. “The o ganic elec ochemical ansis o o
biological applica ions.” J. Appl. Polym. Sci. 132, 41735 (2015).
[113] J. Ri nay, S. Inal, A. Salleo, R. M. Owens, M. Be gg en, G. G. Mallia as. “O ganic
elec ochemical ansis o s.” Na . Re . Ma e . 3, 17086 (2018).
[114] A. Ci a elli, T. Be zina, A. Pa isini, S. Ianno a. “Mem is i e esponse and elec ochemical
p ocesses in polyaniline based o ganic de ices.” O g. Elec on. 83, 105757 (2020).
[115] D. O a a, D. A. Bu y. “De e mina ion o ion popula ions and sol en con en as unc ions
o edox s a e and pH in polyaniline.” J. Am. Chem. Soc. 109, 3574–3581 (1987).
[116] N. Agmon. “The G o huss mechanism.” Chemical Physics Le e s 244, 456–462 (1995).
[117] E. Ven osa, W. Schuhmann. “Scanning elec ochemical mic oscopy o Li-ion ba e ies.”
Phys. Chem. Chem. Phys. 17, 28441–28450 (2015).
[118] H. Ha, J. Paye . “The e ec o sil e chlo ide o ma ion on he kine ics o sil e dissolu ion
in chlo ide solu ion.” Elec ochim. Ac a 56, 2781–2791 (2011).
[119] V. I. Bi ss, C. K. Smi h. “The anodic beha io he o ma ion and educ ion hin sil e
chlo ide ilms.” Elec ochim. Ac a 32, 259–268 (1987).
[120] S. Jaya, T. P. Rao, G. P. Rao. “Mono- and mul ilaye o ma ion s udies o sil e iodide on
sil e elec ode om iodide-con aining solu ions.” J. Appl. Elec ochem. 17, 997–1001
(1987).
[121] B. M. Jo ić, V. D. Jo ić, D. M. D ažić. “Kine ics o chlo ide ion adso p ion and he
mechanism o AgCl laye o ma ion on he (111), (100) and (110) aces o sil e .” J.
Elec oanal. Chem 399, 197–206 (1995).
127
[122] P. B o, N. Ma incic. “The high a e oxida ion o sil e elec odes in chlo ide solu ions.” J.
Elec ochem. Soc. 116, 1338 (1969).
[123] C. Cache , I. Epelboin, M. Keddam, R. Wia . “The Ag/Ag+ sys em: an impedance model
o nuclea ion and g ow h.” J. Elec oanal. Chem 100, 745–757 (1979).
[124] S. B. Aziz, T. J. Woo, M. F. Z. Kadi , H. M. Ahmed. “A concep ual e iew on polyme
elec oly es and ion anspo models.” J. Sci. Ad . Ma e . De . 3, 1–17 (2018).
[125] R. Sajapin, T. Be zina, R. Bu gano a, S. Ianno a. “New insigh in he ope a ion mechanism
o o ganic mem is i e de ices: The ole o PEO-based polyelec oly e solu e ions.” O g.
Elec on. 94, 106173 (2021).
[126] R. Bu gano a, A. Pa isini, S. Van aggio, R. Sajapin, T. Be zina. “The ole o he
polyelec oly e composi ion in kine ic beha iou o o ganic mem is i e de ice.”
Mic oelec on. Eng. 239–240, 111527 (2021).
[127] S. H. Lee, J. C. Rasaiah. “Molecula dynamics simula ion o ion mobili y. 2. Alkali me al
and halide ions using he SPC/E model o wa e a 25 °C.” J. Phys. Chem. 100, 1420–1425
(1996).
[128] R. E. A. Dillon, D. F. Sh i e . “Ion anspo in c yp and and c own e he li hium sal
complexes.” Chem. Ma e . 11, 3296–3301 (1999).
[129] M. L. Kaplan, E. A. Rie man, R. J. Ca a, L. K. Hol , E. A. Chand oss. “C own e he
enhancemen o ionic conduc i i y in a polyme -sal sys em.” Solid S a e Ion. 25, 37–40
(1987).
[130] M. Jóźwiak, L. Madej. “Complex o ma ion o c own e he s and ca ions in wa e -o ganic
sol en mix u es: The he modynamic unc ions o complex o ma ion o benzo-15-c own-
5 wi h Na+ in wa e + e hanol a 298.15 K.” J. Chem. Eng. Da a 55, 1965–1970 (2010).
[131] L. C. Manege, T. Takayanagi, M. Oshima, S. Mo omizu. “Ion associa ion eac ions in
aqueous solu ions be ween non-UV-abso bing c own e he -alkali me al complexes wi h
pic a e ion by capilla y zone elec opho esis.” Analys 125, 1928–1932 (2000).
[132] M. Rashid, A. A. Rahim, M. J. Noo din. “Elec opolyme iza ion o p- oluene sul onic acid
doped polyaniline on coppe and i s applica ion as a co osion inhibi o .” An i-Co os.
Me hod. Ma e . 58, 131–139 (2011).
[133] M. Babazadeh. “Aqueous dispe sions o DBSA-doped polyaniline: one-po p epa a ion,
cha ac e iza ion, and p ope ies s udy.” J. Appl. Polym. Sci. 113, 3980–3984 (2009).
[134] L. F. Nims. “The second dissocia ion cons an o phospho ic acid om 20 o 50°.” J. Am.
Chem. Soc. 55, 1946–1951 (1933).
[135] P. H. Ba y, J. W. Lynch. “Liquid junc ion po en ials and small cell e ec s in pa ch-clamp
analysis.” J. Memb ane Biol. 121, 101–117 (1991).
[136] U. W. G umm , A. P on, M. Zago ska, S. Le an . “Polyaniline based op ical pH senso .”
Anal. Chim. Ac a 357, 253–259 (1997).
134
[231] G. Ta abella, D. Vu o, S. Lai, P. D’Angelo, L. Asca i, S. Ianno a. “Ae osol je p in ing o
PEDOT:PSS o la ge a ea lexible elec onics.” Flex. P in . Elec on. 5, 014005 (2020).
[232] J. Fan, A. A. Fo e o Pico, M. Gup a. “A unc ionaliza ion s udy o ae osol je p in ed o ganic
elec ochemical ansis o s (OECTs) o glucose de ec ion.” Ma e ials Ad ances 2, 7445–
7455 (2021).
[233] D. Majak, J. Fan, S. Kang, M. Gup a. “Del a-9- e ahyd ocannabinol (Δ9-THC) sensing
using an ae osol je p in ed o ganic elec ochemical ansis o (OECT).” J. Ma e . Chem. B
9, 2107–2117 (2021).
[234] K. Pa a e e al. “Ae osol-je -p in ed g aphene immunosenso o label- ee cy okine
moni o ing in se um.” ACS Appl. Ma e . In e aces 12, 8592–8603 (2020).
[235] C. Fishe , B. J. Wa mack, Y. Yu, L. N. Skol ood, K. Li, P. C. Joshi, T. Sai o, T. Ay ug. “All-
ae osol-je -p in ed highly sensi i e and selec i e polyaniline-based ammonia senso s: a
ou e owa d low-cos , low-powe gas de ec ion.” J. Ma e . Sci. 56, 12596–12606 (2021).
[236] F. Ra uchne, M. Danczuck, E. G. de Cas o. “Enhanced s abili y and conduc i i y o
(polyaniline/chi osan) composi es.” O bi al 10, 239–246 (2018).
[237] A. Ci a elli, A. Pa isini, T. Be zina, S. Ianno a. “O ganic mem is i e elemen wi h chi osan
as solid polyelec oly e.” Mic oelec on. Eng. 193, 65–70 (2018).
[238] A. Lesch, D. Momo enko, F. Co és-Salaza , I. Wi h, U. M. Te ashe, F. Meine s, B. Vaske,
H. H. Gi aul , G. Wi s ock. “Fab ica ion o so gold mic oelec ode a ays as p obes o
scanning elec ochemical mic oscopy.” J. Elec oanal. Chem 666, 52–61 (2012).
[239] A. Ci a elli, T. Be zina, A. Pa isini, V. E okhin, S. Ianno a. “Polysacca ides-based gels and
solid-s a e elec onic de ices wi h mem esis i e p ope ies: Syne gy be ween polyaniline
elec ochemis y and biology.” AIP Ad ances 6, 111302 (2016).
[240] A. Ci a elli, A. Pa isini, S. Ianno a, T. Be zina. “O ganic mem is i e de ices based on
pec in as a solid polyelec oly e.” Mic oelec on. Eng. 185–186, 55–60 (2018).
[241] L. D. Koh e al. “S uc u es, mechanical p ope ies and applica ions o silk ib oin
ma e ials.” P og. Polym. Sci. 46, 86–110 (2015).
[242] D. N. Rockwood, R. C. P eda, T. Yücel, X. Wang, M. L. Lo e , D. L. Kaplan. “Ma e ials
ab ica ion om Bombyx mo i silk ib oin.” Na u e P o oc. 6, 1612–1631 (2011).
[243] D. H. Kim e al. “Dissol able ilms o silk ib oin o ul a hin con o mal bio-in eg a ed
elec onics.” Na . Ma e ials 9, 511–517 (2010).
[244] C. Vepa i, D. L. Kaplan. “Silk as a bioma e ial.” P og. Polym. Sci. 32, 991–1007 (2007).
[245] X. Wang, E. Wenk, A. Ma sumo o, L. Meinel, C. Li, D. L. Kaplan. “Silk mic osphe es o
encapsula ion and con olled elease.” J. Con ol. Release 117, 360–370 (2007).
[246] M. A. Tomeh, R. Hadianam ei, X. Zhao. “Silk ib oin as a unc ional bioma e ial o d ug
and gene deli e y.” Pha maceu ics 11, 494 (2019).

135
[247] H. J. Kim, H. S. Kim, A. Ma sumo o, I. J. Chin, H. J. Jin, D. L. Kaplan. “P ocessing windows
o o ming silk ib oin bioma e ials in o a 3D po ous ma ix.” Aus . J. Chem. 58, 716–720
(2005).
[248] Y. Wang, H. J. Kim, G. Vunjak-No ako ic, D. L. Kaplan. “S em cell-based issue
enginee ing wi h silk bioma e ials.” Bioma e ials 27, 6064–6082 (2006).
[249] U.-J. Kim, J. Pa k, C. Li, H.-J. Jin, R. Valluzzi, D. L. Kaplan. “S uc u e and p ope ies o
silk hyd ogels.” Biomac omol. 5, 786–792 (2004).
[250] C. H. Wang, C. Y. Hsieh, J. C. Hwang. “Flexible o ganic hin- ilm ansis o s wi h silk
ib oin as he ga e dielec ic.” Ad . Ma e . 23, 1630–1634 (2011).
[251] Y. Xiao, K. Kalai zidou, D. Yao, W. H. Yeo, T. A. L. Ha is. “Challenges and ad ances in
Ae osol Je P in ing o egene a ed silk ib oin solu ions.” Ad . Ma e . in e aces 7, 1902005
(2020).
[252] J. W. Seo, H. Kim, K. H. Kim, S. Q. Choi, H. J. Lee. “Calcium-modi ied silk as a
biocompa ible and s ong adhesi e o epide mal elec onics.” Ad . Func . Ma e . 28,
1800802 (2018).