Tempe a u e and Cu en Densi y dis ibu ions in a 100 cm
2
PEM Fuel Cell:
E ec s o low ield designs
G.M. Cabello Gonz´
alez
a,*
, Bal asa Toha ias
b
, Felipe Rosa
a,c
, J.J. Gue a
a
,
Al edo I anzo
a,c
a
Depa amen o de Ingenie ía Ene g´
e ica, G upo de Te mo ecnia. Escuela T´
ecnica Supe io de Ingenie ía. Uni e sidad de Se illa. Camino de los Descub imien os s/n.
41092, Se illa, Spain
b
Asociaci´
on de In es igaci´
on y Coope aci´
on Indus ial de Andalucía – AICIA. G upo Te mo ecnia, Camino de los Descub imien os s/n. 41092, Se illa, Spain
c
ENGREEN, Labo a o y o Enginee ing o Ene gy and En i onmen al Sus ainabili y, Uni e sidad de Se illa, Spain
HIGHLIGHTS
•Compa a i e CDM s udy o h ee 100 cm
2
PEM cell low ield con igu a ions.
•As Cu en demand ises, empe a u e and cu en dis ibu ions become mo e une en.
•Tape ed design enhances ope a ion ia uni o m empe a u e and cu en dis ibu ion.
ARTICLE INFO
Keywo ds:
PEM uel cell
Tempe a u e dis ibu ion
Cu en densi y mapping
Dynamic load cycle
ABSTRACT
Elec o- he mal mapping p o ides aluable insigh s in o he pe o mance e alua ion o polyme elec oly e
memb ane uel cells (PEMFCs) by depic ing he spa ial dis ibu ion o cu en densi y and empe a u e. In his
s udy, elec o- he mal maps we e gene a ed o h ee di e en designs o 100 cm
2
PEMFC low ields (con-
en ional se pen ine wi h wo di e en channel dep hs, and se pen ine- ape ed). The pe o mance o each design
was cha ac e ized by analyzing he su ace (in-plane) dis ibu ions o cu en densi y and empe a u e a
di e en cell ol ages. A ele a ed cu en densi ies, a linea inc ease in he non-uni o mi y o empe a u e and
cu en densi y dis ibu ion is obse ed. The cen al egion o he bipola pla e exhibi s highe empe a u es,
whe eas he egion wi h high cu en densi ies is si ua ed nea he hyd ogen inle , g adually diminishing as he
hyd ogen deple es owa ds he ou le . Resul s show ha , in gene al, he ape ed low ield design exhibi s be e
pe o mance wi h a mo e homogeneous empe a u e and cu en dis ibu ion h oughou he en i e ac i e a ea.
This beha io can be a ibu ed o be e wa e managemen and gas di usion owa ds he elec ode due o he
accele a ion and p essu e inc ease o he eac an uel gas along he na owing channel. No el insigh s we e
iden i ied by applying he Cu en Dis ibu ion Mapping (CDM) echnique o analyzing cu en densi y and
empe a u e in-plane dis ibu ions unde dynamic load condi ions, compa ing he di e en channel dep hs o
ape ed designs du ing he dynamic ope a ion o he cell. Du ing dynamic es s, empe a u e inc eased apidly
o inc easing loads bu he dec ease was mo e slowly when load was lowe ed, leading o an o e all g adual
empe a u e ise and less homogeneous dis ibu ion a highe cu en s, while he cu en dis ibu ion adjus ed
almos ins an ly wi h cons an s anda d de ia ion du ing bo h load inc eases and dec eases.
1. In oduc ion
In he pu sui o mi iga ing clima e change, he ene gy sec o s ands
a he o e on o deca boniza ion e o s [1] since, globally, i is he
la ges emi e o global g eenhouse gas emissions [2]. Go e nmen s
wo ldwide a e ge ing in ol ed, implemen ing policies and incen i es o
ansi ion away om ossil uels owa ds cleane al e na i es, limi ing
he ad e se impac s o clima e change. Howe e , achie ing deca bon-
iza ion goals necessi a es no only enewable ene gy sou ces like wind
and sola , bu also e sa ile ene gy ec o s capable o s o ing and
* Co esponding au ho .
E-mail add ess: [email p o ec ed] (G.M.C. Gonz´
alez).
Con en s lis s a ailable a ScienceDi ec
Jou nal o Powe Sou ces
jou nal homepage: www.else ie .com/loca e/jpowsou
h ps://doi.o g/10.1016/j.jpowsou .2025.237625
Recei ed 14 Feb ua y 2025; Recei ed in e ised o m 13 May 2025; Accep ed 7 June 2025
Jou nal o Powe Sou ces 652 (2025) 237625
A ailable online 23 June 2025
0378-7753/© 2025 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY-NC license (
h p://c ea i ecommons.o g/licenses/by-
nc/4.0/ ).
anspo ing ene gy e icien ly [3]. In his ega d, hyd ogen eme ges as a
po en ial ene gy ec o candida e due o i s high ene gy densi y and,
o e all, i s e sa ili y, ha lies in hyd ogen’s abili y o be used in
mul iple applica ions, such as ueling ehicles, p o iding hea o in-
dus ial p ocesses, and se ing as a eeds ock o chemical p oduc ion
[4]. As an ene gy ca ie , hyd ogen can be p oduced h ough elec olysis
using enewable elec ici y, he eby o e ing a pa hway o in eg a e
excess enewable ene gy and s o e i o la e use [5].
Hyd ogen can be u ilized di ec ly in uel cells o gene a e elec ici y
wi h wa e apo as he only byp oduc , o e ing a clean al e na i e o
combus ion engines. P o on Exchange Memb ane (PEM) uel cells ha e
gained signi ican in e es as powe sou ces o a wide ange o appli-
ca ions, including au omobiles and s a iona y powe sys ems [6]. This is
due o hei nume ous ad an ages, such as as s a -up imes, low noise,
high powe densi y, and high ene gy con e sion e iciency [7]. Fuel cell
echnology de elopmen is equi ing he suppo o s ong es ing ca-
pabili ies and echniques, o accu a ely add ess pe o mance and du a-
bili y o he cells and s acks [8]. Howe e , con en ional uel cell es ing
me hods, like pola iza ion cu e o impedance spec oscopy measu e-
men , p o ide an o e all pe o mance assessmen bu lack localized
in o ma ion along he ac i e a ea. The local dis ibu ions o cu en
densi y and empe a u e a e key in he elec ochemical eac ion o uel
cells because i di ec ly impac s he e iciency, pe o mance, and du a-
bili y o he cell, and hus, unde s anding hei a ia ions along he
ac i e a ea is i al o assessing local ene gy con e sion pe o mance
and po en ial deg ada ion issues [9]. I he cu en densi y is une enly
dis ibu ed, localized egions o high o low cu en densi y may
de elop. This can lead o une en u iliza ion o eac an s, incomple e
elec ochemical eac ions, high empe a u e g adien s, and dec eased
o e all e iciency oge he wi h accele a ed deg ada ion [10].
Non-uni o m dis ibu ion o cu en densi y can also esul in localized
deg ada ion o he elec ode and memb ane ma e ials o ca alys
deg ada ion [11]. High cu en densi y egions may expe ience accel-
e a ed deg ada ion due o inc eased elec ochemical ac i i y, leading o
educed du abili y and li e ime o he uel cell s ack. Non-uni o m dis-
ibu ion o cu en densi y can also lead o une en wa e dis ibu ion
wi hin he cell, causing looding o d ying ou o ce ain egions. This
can impai p o on conduc i i y in he elec oly e memb ane and hinde
he elec ochemical eac ions, leading o dec eased pe o mance. In
spi e o all his, cu en e idence does no conclusi ely demons a e ha
cu en densi y inhomogenei y di ec ly accele a es pe o mance loss o
induces ope a ion a om op imal condi ions. The obse ed deg ada-
ion ends do no show a consis en co ela ion wi h a eas o highe o
lowe cu en densi y [12,13]. I is impo an o no e ha con en ional
cha ac e iza ion echniques will no be able o iden i y such local g a-
dien s and issues. I is well known ha e icien hea dissipa ion is
c ucial o p e en ing empe a u e g adien s wi hin he uel cell s ack,
which a ec s pe o mance and du abili y [14]. Uni o m dis ibu ion o
cu en densi y helps ensu ing mo e uni o m hea gene a ion and
dissipa ion, con ibu ing o s able ope a ing condi ions and p olonged
s ack li e. The local dis ibu ion o cu en densi y depends on ac o s
like he local concen a ion o eac an s a he elec ode su ace, as well
as he po en ial dis ibu ion in he cell, which is in luenced by p essu e,
ela i e humidi y, and empe a u e [15]. S a egies such as elec ode
design op imiza ion, imp o ed ca alys o mula ions, and ad anced low
ield designs a e employed o enhance cu en dis ibu ion and maxi-
mize he o e all pe o mance o uel cell sys ems [16,17].
Since con en ional es ing me hods such as pola iza ion cu e
measu emen o elec ochemical impedance spec oscopy lack o local
in o ma ion, Cu en Densi y Mapping (CDM) p o ides aluable insigh s
in o he pe o mance o a PEM uel cell helping o iden i y a eas o high
and low cu en densi y, which can be used o assess he o e all e i-
ciency and unc ionali y o he uel cell, and may lead o localized issues
such as ca alys deg ada ion, gas s a a ion o memb ane damage o
dehyd a ion [18].
In o de o gain u he knowledge in local dis ibu ions, local
empe a u e measu emen s and i s co ela ion wi h local cu en densi y
can been assessed in oducing he mocouples inside he uel cell. Wil-
kinson e al. [19] p o ed ha local empe a u e measu emen s can
e ec i ely co ela e wi h local cu en densi ies using
mic o- he mocouples, sugges ing a po en ial indi ec me hod o cu en
mapping. Howe e , imp o emen s we e needed in he mocouple
obus ness and op imal placemen . De e mina ion o cu en densi y
dis ibu ion can be assessed by mean o a segmen ed cell whe e cu en
is measu ed indi idually in each segmen by using a Hall-e ec senso .
Hwang e al. [20] employed a specially designed composi e pla e wi h
16 segmen ed cu en collec o s o es a ious ca hodic low- ield pa -
e ns. The s udy e alua ed hei e ec s on cu en dis ibu ion o
a ious ope a ing condi ions and concluded ha he se pen ine low
ield yields he mos uni o m dis ibu ion due o i s supe io mass
ans e and wa e managemen capabili ies.
Despi e he in e es ing conclusions ob ained wi h segmen ed cells,
his me hod is limi ed o measu e he cu en densi y dis ibu ion o a
single uel cell o he one a he ends o a s ack, while senso pla es
inse ed in o uel cells can measu e any cell in he s ack since he cu en
passes h ough a p in ed ci cui senso ha measu es cu en dis ibu-
ion by means o a shun . Besides, his kind o senso s can allow highe
esolu ion in he measu emen . In his sense, Geske e al. [18] p esen ed
a measu emen sys em o mapping cu en densi y dis ibu ion u ilizing
mul i-laye echnology o in eg a ing shun esis o s wi hin a p in ed
ci cui boa d ailo ed o he uel cell’s low ield channels. Resul s o a
se pen ine low channels indica e minimal empe a u e in luence du ing
cell wa m-up, wi h maximum local cu en p oduc ion ypically nea he
oxygen inle , hough load cu en induces a ying le els o in-
homogenei y, highligh ing po en ial o op imizing a mo e uni o m
cu en densi y dis ibu ion p o ile. In his sense, Heue e al. [21]
de eloped a CDM senso and inse ed hem be ween he wo cen al cells
o a se pen ine low channels s ack. They ound ha he s oichiome y o
he ca hode was he pa ame e ha in luenced he mos on he cu en
densi y p o ile since i had a la ge impac on he humidi y balance, wi h
lowe o uns able cell ol ages obse ed when he cu en densi y p o ile
was widely sp ead. These esul s a e in line o hose ob ained by Peng
e al. [22] o a comme cial size s ack. They ound ha cu en densi y
dis ibu ion is mo e sensi i e o he low a e o ai han ha o
hyd ogen, and local gas s a a ion occu s when he hyd ogen s oichio-
me ic a io is 1.1. Heue e al. [21] long- e m analysis e ealed an in-
c ease in he sp eading o he cu en densi y p o ile, po en ially ela ed
o aging, unde sco ing he impo ance o CDM o ea ly de ec ion and
eac ion o ope a ional changes compa ed o adi ional cell ol age
measu emen s. Belhadj e al. [9] eached he same conclusion. They
ound ha esh MEA e ealed non-uni o m dis ibu ions p ima ily
caused by p e e en ial gas dis ibu ion in ce ain channels, wi h li le
in luence om cell cu en densi y. Howe e , aging esul ed in signi i-
can ly une en dis ibu ions, wi h e idence o looding a high cu en
densi y in he las sec ion. On a highe scale, Yin e al. [23] combined
Compu a ional Fluid Dynamics (CFD) model wi h cu en mapping
showing ha , in coun e - low hyd ogen and ai ope a ions, he ca hode
ou le has he lowes and he mid po ions he highes local cu en
alues. They epo ed ha cu en dis ibu ion uni o mi y was imp o ed
by inc eased ai s oichiome ic a io and ha c oss- low con igu a ions
enhanced empe a u e uni o mi y mo e e ec i ely han pa allel- low
ones.
In ecen imes, senso s ha e allowed o an elec o- he mal map-
ping, including in o ma ion o bo h, empe a u e and cu en densi y
dis ibu ion. Meye e al. [24] c ea ed o he i s ime a localized
elec o- he mal pe o mance map o a 60 cm
2
PEM uel cell. The s udy
ound ha a low cu en densi ies, he cu en densi y dis ibu ion was
mainly in luenced by he anode and hyd ogen consump ion g adien ,
wi h he highes empe a u e g adien along he ai low di ec ion. On
he o he hand, a high cu en densi ies, empe a u e in luenced he
cu en densi y dis ibu ion. Also, when he senso was coupled wi h a
low- equency impedance spec oscopy de ice, he limi o ope a ion as
G.M.C. Gonz´
alez e al.
Jou nal o Powe Sou ces 652 (2025) 237625
2
well as he op imal pe o mance ange could be ound [25]. The cell
ol ages a he cu en wi h he lowes esis ance we e app oxima ely
0.60 V, which is sligh ly below he usual ope a ing ange o 0.70-0.65 V
o au omo i e applica ions. In con as , he ol age a he cu en wi h
he highes powe densi y is abou 0.50 V. This demons a es ha he
cu en associa ed wi h he lowes esis ance is a mo e app op ia e
single- alue measu e han he poin o maximum powe . Besides, when
elec o- he mal mapping was used in combina ion wi h wa e imaging,
hyd a ion and dehyd a ion e ec s unde a ying load and low condi-
ions could be s udied, inding ha wa e mainly accumula es unde
cooling channels. Using all h ee echniques oge he p o ides a
comp ehensi e unde s anding o wa e managemen in uel cells. A
comple e s udy ca ied ou by Be hapudi e al. [15] compa ing a 25 cm
2
ac al low ield ca hode wi h a con en ional se pen ine low ield
shows he co ela ion be ween cell pe o mance and
cu en - empe a u e dis ibu ion. The ac al design signi ican ly en-
hances pe o mance, pa icula ly a high cu en densi ies, due o highe
cell empe a u es. Elec o- he mal mapping shows he ac al low- ield
achie es a mo e uni o m cu en dis ibu ion and mi iga es looding
be e han he se pen ine design, esul ing in 10–15 % highe segmen
cu en s. He e ogenei y o empe a u e along wi h cu en densi y was
also assessed by Jiang e al. [26]. Using an elec o- he mal mapping
de ice in a se pen ine-pa e n low ield 100 cm
2
ac i e a ea PEM uel
cell, he s udy measu ed his he e ogeneous dis ibu ion du ing pola i-
za ion cu e es s a empe a u es o 50 ◦C and 60 ◦C. They obse ed ha
highe empe a u es and cu en densi ies inc eased he e ogenei y and
o e ed a new way o con e powe -cu en cu es in o
powe -he e ogenei y cu es. This way, i can be ensu ed ha he cell
ope a es wi h simila ou pu pe o mance bu educed cu en densi y
he e ogenei y wi hin a speci ic ange, enhancing i s li espan. Also,
Su´
a ez e al. [27] used his echnology in a 50 cm
2
pa allel se pen ine
Elec oChem Inc uel cell in o de o es di e en con igu a ions and
assess he cell esponse du ing a s anda dized d i ing cycle. They ound
ha in e se hyd ogen low con igu a ion ou pe o med he o he con-
igu a ions in e ms o pola iza ion and powe cu es. Local cu en
densi y dis ibu ion p esen ed signi ican he e ogenei ies showing an
in e se bell-shaped dis ibu ion ac oss all con igu a ions while, in
con as , local empe a u e dis ibu ions we e highly homogeneous.
Ne e heless, o he bes o he au ho s’ knowledge, he e ha e been
no p io s udies examining he local cu en densi y and empe a u e
dis ibu ions du ing dynamic load cycling es s compa ing di e en
designs o uel cells. This is a signi ican gap in he li e a u e, pa icu-
la ly since he spa ial and empo al a ia ions in hese dis ibu ions
du ing load changes could c ea e challenging g adien s o he cell’s
ope a ion. This s udy aims o add ess his gap by p o iding no el in-
sigh s in o uel cell beha io unde dynamic condi ions, speci ically
add essing how and why di e en low ield designs a ec cu en
densi y and empe a u e in-plane dis ibu ions, which can be also o
in e es o PEM elec olyze s. In pa icula , he no el y o his wo k is
he use o he Cu en Dis ibu ion Mapping (CDM) echnique o
analyzing cu en densi y and empe a u e in-plane dis ibu ions
compa ing di e en channel dep hs o ape ed designs du ing dynamic
condi ions. The analysis was conduc ed o e h ee dis inc single-cell
100 cm
2
low ield con igu a ions: a pa allel se pen ine design wi h
wo di e en channel dep hs, and a ape ed design. In addi ion o hese
compa a i e insigh s, his s udy also examines he ime- a ying
beha io o cu en densi y and empe a u e dis ibu ions du ing load
changes, ollowing a well-es ablished dynamic load p o ocol. The ind-
ings om his analysis p o ide a u he unde s anding on how low ield
design impac s he cell beha io .
2. Expe imen al acili y and me hodology
2.1. Fuel cell design
The expe imen al wo k was conduc ed using a specialized PEM uel
cell s a ion wi h capaci y o es single cells and sho s acks up o 500 W.
The es ing en i onmen included a eac an gas handling uni equipped
wi h mass low con olle s, humidi ie s, back p essu e egula o s, an
elec onic load, and a empe a u e con ol sys em whe e he hea p o-
ided by wo hea ing pads is egula ed, as well as a se o ex e nal ans
o cooling pu poses.
Th oughou he expe imen s, da a on cu en densi y and empe a-
u e dis ibu ion inside he uel cell we e collec ed a 1 Hz acquisi ion
equency using a CDM senso (S++ cu en scan shun , Ge many)
placed be ween he ca hode cu en collec o and he ca hode bipola
pla e, wi h an app op ia e sealing. The ac i e a ea o 100 cm
2
is, his
way, di ided in o elec ically isola ed segmen s in a ma ix o 14x14 o
cu en and 7x7 o empe a u e measu emen . The cell ha dwa e used
in he expe imen s was he P o-RD om P agma Indus ies (F ance),
whe e s a e-o - he a 7-laye MEAs om IRD Fuel Cells (Denma k) we e
used (8
μ
m uel cell memb ane, 0,1 mg/cm
2
in anode and 0.3 mg/cm
2
in
ca hode, wi h Gas Di usion Laye (GDL) in anode o 135
μ
m @1 MPa
and GDL in ca hode o 146
μ
m @1 MPa.
All he expe imen s we e ca ied ou using a single cell. h ee
di e en g aphi e pla es (6.0 mm hick) we e es ed o he ca hode side,
wi h an ac i e a ea o 100 cm
2
. The designs in all cases we e based on a
7-channel pa allel se pen ine, wi h he only di e ence being he dep h
o he channels (Fig. 1). The i s design was a con en ional se pen ine
design wi h a channel dep h o 1 mm, whe eas in he second he dep h
was dec eased down o 0.5 mm. The hi d design was also a 7-channel
pa allel se pen ine wi h a ape ed design whe e he dep h was
educed p og essi ely om 1 mm o 0.5 mm along he channel pa h
(Fig. 1a). The wid h o he channel and he ib emained cons an o he
h ee designs, being 0.8 and 1.0 mm espec i ely. I is impo an o no e
ha he i s design was used a he anode side o all he expe imen ,
and only he bipola pla e o he ca hode side was changed. Reac an
gases en e he pla e om he op le / igh co ne and ci cula e coun-
e cu en in he ho izon al sec ions o he channels and concu en in
he sho e ical pa s, lea ing he pla e h ough he ou le loca ed a
he opposi e co ne (Fig. 1c). The assembly o he cell was done using
wel e bol s and nu s placed in he edges o he pla es, wi h a igh ening
o que o 9.0 Nm, o p e en leaks and p omo e uni o m cu en
collec ion.
2.2. Expe imen al es s
Pola iza ion and powe cu es, along wi h d i ing es s (New Eu o-
pean D i ing Cycle dynamic load es s), we e ob ained o each design
unde wo dis inc ope a ing condi ions (Table 1), ollowing he p o-
ocols es ablished by he EU Commission Join Resea ch Cen e (JRC)
[28]. The naming sys em o he es s is as ollows: PAA_TB-
B_aCCHRDD_cEEHRFFai . In his nomencla u e, AA ep esen s he
ou le gauge p essu e in ba wi h a decimal, BB deno es he ope a ing
empe a u e in Celsius, CC indica es he anode s oichiome y wi h a
decimal, DD signi ies he anode ela i e humidi y in pe cen age, EE
s ands o he ca hode s oichiome y wi h a decimal, and FF indica es
he ca hode ela i e humidi y in pe cen age. The wo d "ai " signi ies ha
he expe imen was conduc ed using ai ins ead o pu e oxygen in he
ca hode.
Concu en ly wi h he pola iza ion cu e and d i ing es , he
elec o- he mal mapping senso was employed o eco d empe a u e
and cu en densi y da a, wi h an acquisi ion equency o 1 Hz.
P io o ob aining he I-V pola iza ion cu es, he memb ane was
ac i a ed ope a ing he uel cell s ack o 24 h, a P05_T65_a13HR60_
c25HR60ai and cons an cu en densi y o 0.5 A/cm
2
. A e ac i a ion,
he cell was se o Open Ci cui Vol age (OCV) o 120 s. Measu emen s
we e aken in gal anos a ic mode, s a ing om OCV wi h ixed cu en
densi y s eps as pe indica ion o he es ing p o ocol [28]. Fo low
cu en densi ies (up o 0.08 A/cm
2
), measu emen s we e aken e e y
120 s, and o highe cu en densi ies (0.1 A/cm
2
o maximum), e e y
300 s. The es concluded when cell’s ol age d opped below 0.3 V. Da a
G.M.C. Gonz´
alez e al.
Jou nal o Powe Sou ces 652 (2025) 237625
3
we e acqui ed a a a e o one sample pe second, whe e o each cu en
he ol age a e age o he las 30 samples was used o pos -p ocessing.
Th ee epe i ions o each expe imen we e ca ied ou o e i y he
ep oducibili y o he ob ained esul s.
The es s o he Fuel Cell Dynamic Load Cycle (FC-DLC) we e ca ied
ou acco ding o he well-de ined es ing p o ocol [28], whe e 1 es
cycle o 1200 s ep esen s a d i ing dis ance o abou 11 km o e a span
o 20 min.
3. Resul s and discussion
3.1. Pola iza ion pe o mance
The pola iza ion cu es o he h ee designs a wo di e en ope -
a ing condi ions a e compa ed in Fig. 2a while powe s cu en densi y
cu es a e shown in Fig. 2b. As shown, he cu es ob ained unde
P10_T70_a15HR60_ c30HR55ai ope a ing condi ions (con inuous line
in Fig. 2) exhibi be e pe o mance a any cu en densi y. This esul is
consis en wi h p e ious s udies conduc ed on a 7-cell s ack wi h a
simila design (bipola pla es used we e he same as in he i s design)
[29] whe e he e ec s o changes in di e en ope a ing condi ions we e
s udied and such ope a ing condi ion was iden i ied as he op imal one
o he s ack. Fo he 1.0 mm hickness channels, when eaching ce ain
cu en densi ies (abo e 0.6 A/cm
2
a P05_T65_a13HR60_ c25HR60ai
and 0.8 A/cm
2
a P10_T70_a15HR60_ c30HR55ai ), he ol age d ops
clea ly indica ing a pe o mance loss, while he 0.5 mm and he ape ed
design will ope a e a leas un il 1.0 A/cm
2
. The ape ed design p esen s
Fig. 1. a) Pla e designs wi h di e en channel dep h. F om le o igh : 1.0 mm, 0.5 mm, and ape ed (1.0–0.5 mm). b) CDM senso used in he expe imen s (S++,
Ge many). c) Flow con igu a ion o he cell used in he es s. F on iew (le ) and pe spec i e ( igh ).
Table 1
Ope a ing condi ions de ined in he expe imen al es s.
Case Pg
(ba )
T
(◦C)
RHa
(%)
RHc
(%)
λ
a
(−)
λ
c
(−)
Oxidan
P10_T70_a15HR60_
c30HR55ai
1.0 70 60 55 1.5 3.0 Ai
P05_T65_a13HR60_
c25HR60ai
0.5 65 60 60 1.3 2.5 Ai
G.M.C. Gonz´
alez e al.
Jou nal o Powe Sou ces 652 (2025) 237625
4
he highes pe o mance o almos all he cu en densi y ange,
eaching a maximum powe alue o 45 W a 1 A/cm
2
(Fig. 2).
The supe io pe o mance o he ape ed design can be a ibu ed o
he p og essi e accele a ion o he eac an uel gas in he ape ed
channel as well as o he ai p essu e inc ease along he channel pa h,
whe e oxygen is hen o ced in o he GDL enhancing he elec ochemical
eac ion. The p og essi e gas accele a ion in he channel ac ually in-
c eases also he wa e emo al capabili y and hus imp o es cell pe -
o mance. These esul s a e in line wi h hose epo ed in li e a u e,
whe e he posi i e ape ed slope is p o en o pe o m be e , especially
a high cu en densi ies [30]. While i is ine i able ha he use o
ape ed channels incu s a p essu e-loss penal y, his loss inc ease is
gene ally qui e low (a ound 0.018 ba epo ed o a ape ed design wi h
a ape a io o 0.1) [31]. The h ee designs es ed exhibi ed a p essu e
d op below 0.1 ba , which is wi hin he p ecision limi s o he p essu e
senso s ins alled in he es bench.
3.2. Cell empe a u e and cu en densi y dis ibu ion analysis
The empe a u e and cu en dis ibu ion maps ob ained wi h he
CDM senso a low, medium, and high cu en s (0.2, 0.4, and 0.7 A/cm
2
,
espec i ely) o he h ee bipola pla e channel designs a e shown in
Fig. 3. Maps ep esen he local cu en and empe a u e measu ed by
he senso (placed be ween he ca hode bipola pla e and he ca hode
cu en collec o ). In Fig. 3 a), he empe a u e alue is displayed on he
z-axis, while on he x and y axes, he ac i e a ea is ep esen ed (wi h he
segmen numbe as uni s, 7x7 o empe a u e and 14x14 o cu en ).
The ai inle is loca ed a coo dina es (1,1) (blue a ow), while he
hyd ogen inle is posi ioned a he co ne (1,7) ( ed a ow). In his
manne , he gases ci cula e coun e cu en h ough he se pen ine,
which is aligned wi h i s long edge pa allel o he y-axis, allowing he ai
o exi a he co ne (7,7) and he hyd ogen a (7,1), as depic ed by he
inse in Fig. 3 a). In all he designs, he empe a u e map p esen s an
in e ed bowl igu e, wi h highe empe a u e alues a he cen al
segmen . This is clea ly due o he ac ha , as he cell is no in e nally
cooled by a e ige an ci cui bu a he by ex e nal ans, he hea
gene a ed is e acua ed only om he cell ex e nal sides om he cen e
ou wa d, c ea ing a adial empe a u e g adien wi h highes empe a-
u es a he cen e . I can be no ed ha he a e age empe a u e o all
es s is sligh ly highe han in ended, as he he mocouple moni o ing
empe a u e is inse ed in o he bipola pla e bu ou side he ac i e a ea.
This o se inc eases as he cu en densi y inc eases, so i is wo h no ing
ha he condi ions ini ially in ended o be es ed a e no p ecisely he
ones occu ing inside he cell (which indica es as well ha he di e -
ences in he empe a u e wi hin he cell ha dwa e a e ele an ). I can be
obse ed ha empe a u e is he pa ame e mos a ec ed by he design
o he channels. In all cases, i is obse ed ha as he cu en inc eases
he empe a u e ises and i s dis ibu ion becomes less homogeneous. A
linea inc ease o he empe a u e s anda d de ia ion wi h cu en
densi y was obse ed (depic ed in Fig. 4). In he case o he 1.0 mm
se pen ine, a P05_T65_a13HR60_c25HR60ai , his design exhibi s he
highes empe a u e alue ac oss he en i e ange o cu en densi y.
Mo eo e , he inc ease in empe a u e wi h cu en densi y is e y sligh
a low cu en alues bu inc eases signi ican ly a high cu en alues
due o he highe hea gene a ion and poo hea managemen o he
ex e nally ai -cooled cell ope a ion. The maximum di e ences in em-
pe a u e and cu en densi y in he measu emen plane a e depic ed in
Fig. 7 o he h ee designs a 0.1, 0.4 and 0.7 A/cm
2
. Tempe a u e
di e ences a e in all cases below 4 ◦C.
Also, he a ia ion in cu en dis ibu ion acco ding o he cu en
Fig. 2. Pola iza ion (a) and powe (b) cu es o he di e en designs. ● ape ed se pen ine design, ■ 0.5 mm dep h se pen ine design and ▴ 1 mm dep h se pen ine
design. Dashed line P05_T65_a13HR60_ c25HR60ai . Solid line P10_T70_a15HR60_ c30HR55ai .
G.M.C. Gonz´
alez e al.
Jou nal o Powe Sou ces 652 (2025) 237625
5

densi y o he h ee channel designs can be obse ed a
P05_T65_a13HR60_c25HR60ai in Fig. 4. As he cu en inc eases, he
he e ogenei y in cu en dis ibu ion wi hin he cell also inc eases, as
e idenced by he ise in s anda d de ia ion wi h cu en densi y. This
e ec is consis en wi h p e iously epo ed li e a u e [22], which
indica es ha when he uel cell is ope a ed a lowe o al cu en s, he
di e ence in local cu en densi ies is smalle , and he dispe sion o
cu en densi y dis ibu ion a highe cu en s is la ge . This cu en
dispe sion is linked wi h high wa e p oduc ion and accumula ion,
indica ing ha he ape ed design, wi h i s amp-shaped channels in he
Fig. 3. Tempe a u e a) and Cu en b) dis ibu ion mapping a 0.2, 0.4 and 0.7 A/cm
2
o he h ee conside ed designs. Ope a ing condi ion:
P05_T65_a13HR60_c25HR60ai . Ai inle /ou le loca ions ma ked wi h blue a ows, hyd ogen in ake/ou ake loca ions ma ked wi h ed a ows. (Fo in e p e a ion
o he e e ences o colou in his igu e legend, he eade is e e ed o he Web e sion o his a icle.)
G.M.C. Gonz´
alez e al.
Jou nal o Powe Sou ces 652 (2025) 237625
6
di ec ion o low, p esen s be e eac an dis ibu ion and wa e man-
agemen as he cu en he e ogenei y is lowe o his design a any
cu en densi y, wi h he maximum he e ogenei y obse ed in he 1.0
mm design. Maximum alues o cu en densi y in all designs a e p esen
in he cen al pa o he ac i e a ea, which co esponds as well o he
zone wi h highe empe a u e. The designs wi h 0.5 mm and 1.0 mm
dep h exhibi a mo e p onounced (highe alues) in he a ea co e-
sponding o he hyd ogen inle and ai ou le . The e is a dec ease in
cu en in ensi y as eac an s a e consumed owa ds he ou le o he
cell, due o he p og essi e educ ion o eac an concen a ion along he
channel pa h oge he wi h he co esponding inc ease in he ela i e
humidi y. Howe e , he ape ed design does no exhibi his end;
ins ead, i s dis ibu ion is mo e pa allel o he (x,y) plane. This indica es
ha he p og essi e na owing o he channel dep h along he low pa h
e ec i ely enhances he uni o m anspo o gases owa ds he elec-
ode, p omo ing he elec ochemical eac ion and imp o ing he o e all
pe o mance o he cell.
As he o e all cu en inc eases, he maximum cu en and empe -
a u e di e ence inc eases o all con igu a ions (Fig. 5). The ape ed
con igu a ion gene ally shows lowe cu en di e ences and gene ally
exhibi s lowe o simila empe a u e di e ences, indica ing ha he
ape ed design may esul in mo e uni o m cu en and he mal dis i-
bu ion, especially a medium cu en s.
3.3. CDM s udy unde dynamic condi ions (load cycling es s)
The h ee designs pe o med no mally du ing he load cycling es s
simula ing d i ing condi ions (ca ied ou acco ding o es ing p o ocol
[28]), esponding swi ly o changes in he se poin , wi h minimal
ol age oscilla ions du ing pe iods o cons an in ensi y. Fig. 6 shows he
ol age a ia ions o he cell o he h ee designs unde he wo ope -
a ing condi ions conside ed, du ing he comple e load cycle. Fo bo h
condi ions, he 1.0 mm design exhibi s a lowe ol age compa ed o he
o he designs as i was he case du ing he pe o mance pola iza ion
cu e es s (sec ion 3.1).
CDM da a was also collec ed du ing dynamic es s. Fo he sake o
simplici y, only he CDM images o he ape ed design es a e shown o
one o he condi ions in Fig. 7, while Fig. 8 will ga he he quan i a i e
esul s o all h ee es s a he wo ope a ing condi ions analyzed. Fi e
ep esen a i e poin s we e selec ed. The i s and second ones a e se on
a amp wi h low cu en ascen and descen , espec i ely; he hi d is on
a medium cu en ascen ; he ou h is on a high cu en ascen ; and,
inally, he i h is in a s eady s a e a high cu en . Also, he es esul s
can be isualized in ideo o ma as Supplemen a y Ma e ial.
Supplemen a y ideo ela ed o his a icle can be ound a h ps://
doi.o g/10.1016/j.jpowsou .2025.237625
Du ing he es , i was obse ed ha he empe a u e esponds
apidly o posi i e cu en changes, inc easing i s alue. Howe e , when
he cu en demand dec eases, he empe a u e d ops a a much slowe
a e han i ises due o he limi a ions in he hea ans e a e, and a
s eady s a e is no eached du ing pe iods o cons an demand. As a
esul , he o e all e ec du ing he i s 800 s o he es is a g adual
inc ease in empe a u e a each poin wi hin he cell. In he second
phase, a highe cu en demands, he empe a u e dis ibu ion expe i-
ences g ea e changes, wi h a less homogeneous dis ibu ion and he
appea ance o ho spo s in he cen e o he cell, which ex end owa ds
he sides. On he o he hand, he cu en dis ibu ion esponds almos
ins an aneously o he demands o he d i ing cycle, quickly eaching a
Fig. 4. Tempe a u e a) and Cu en b) mean (M; ba s) and s anda d de ia ion
(SD; lines) a 0.1, 0.4 and 0.7 A/cm
2
o he h ee conside ed designs. Ope a ing
condi ion: P05_T65_a13HR60_ c25HR60ai .
Fig. 5. Maximum empe a u e a) and cu en b) di e ence a 0.1, 0.4 and 0.7
A/cm
2
o he h ee conside ed designs. Ope a ing condi ion:
P05_T65_a13HR60_ c25HR60ai .
G.M.C. Gonz´
alez e al.
Jou nal o Powe Sou ces 652 (2025) 237625
7
s eady s a e.
As in he p e iously shown cases, i can be obse ed ha empe a u e
inc eases as mo e cu en is demanded, and, in he same way, he ho-
mogenei y in he cu en dis ibu ion dec eases a high densi y cu en s
(Fig. 8). Also, he 0.5 mm design is always p esen ing he lowes a e age
empe a u e while he ape ed design is he one wi h he highes em-
pe a u e excep o low cu en densi y condi ions. Howe e , i is wo h
no ing ha i we compa e he ise and all (poin 1 and poin 2 in Fig. 7)
a he same cu en densi y, al hough he a e age empe a u e emains
almos cons an , he s anda d de ia ion is highe in he case o he
descen cu en , and hus a less homogeneous empe a u e dis ibu ion
is ea u ed o ansien ope a ion while he cu en is dec eased. This
phenomenon is obse ed a any cu en densi y, wi h empe a u e di -
e ences be ween he load inc ease and load dec ease being mo e p o-
nounced in he cen al a ea o he cell, whe e hea dissipa ion is mo e
challenging. Fo example, a a cu en demand o 10 A, he s anda d
de ia ion o he empe a u e dis ibu ion du ing he amp-up phase is
0.26, while o he amp-down phase i is 0.32. A a demand o 45 A, he
s anda d de ia ion du ing he amp-up phase inc eases o 0.53, which is
highe han a 10 A, as expec ed based on p e iously desc ibed ends.
Fo he amp-down phase a 45 A, he s anda d de ia ion is 0.58.
The same in e ed bowl shape obse ed a he a ious poin s in he
pola iza ion cu e is also e iden du ing he dynamic es s. The non-
homogenei y o cu en dis ibu ion, on he o he hand, ollows a
linea end as cu en densi y inc eases. O he h ee designs es ed, he
ape ed design exhibi s he mos uni o m cu en dis ibu ion ac oss he
en i e a ea, ega dless o he cu en densi y alue o ope a ing condi-
ion. Unlike he beha io obse ed wi h empe a u e, in he case o
cu en dis ibu ion, he s anda d de ia ion emains cons an du ing
bo h inc eases and dec eases.
4. Conclusions
Elec o- he mal mapping was used o de e mine he pe o mance o
Fig. 6. D i ing es esul s o he h ee design a a) P05_T65_a13HR60_
c25HR60ai and b) P10_T70_a15HR60_ c30HR55ai .
Fig. 7. CDM measu emen s du ing he d i ing es o he ape ed design a P05_T65_a13HR60_ c25HR60ai .
G.M.C. Gonz´
alez e al.
Jou nal o Powe Sou ces 652 (2025) 237625
8
h ee di e en 100 cm
2
low ield designs in a PEM uel cell. The pe -
o mance o each design was assessed by analyzing he su ace dis i-
bu ion o cu en and empe a u e a di e en cell ol ages unde
s a iona y and dynamic ope a ion. The s udy unco e ed ela ionships
be ween local cu en densi ies and local empe a u es as well as he
in luence o he channel design in he cell pe o mance. A high cu en
densi ies, he non-uni o mi y in empe a u e and cu en densi y dis-
ibu ion inc eases linea ly. The a ea o highes empe a u es is loca ed
a he cen e o he cell due o he ex e nal ai -cooling design, while he
a ea o high cu en densi ies is obse ed nea he hyd ogen inle and
p og essi ely dec eases as he hyd ogen is consumed along i s pa h o-
wa ds he ou le .
The ape ed design exhibi s supe io pe o mance compa ed o he
o he designs, pa icula ly a cu en densi ies abo e 0.5 A/cm
2
. Addi-
ionally, i demons a es he leas a ia ion in bo h su ace empe a u e
dis ibu ion and cu en densi y unde any ope a ing condi ion. This
demons a es he supe io capabili y o a p og essi ely educed passage
a ea in enhancing gas anspo owa ds he elec ode, as well as an
imp o ed wa e managemen wi hin he cell as p og essi e gas accel-
e a ion p omo es wa e emo al along he channels.
Con e sely, he 1.0 mm channel hickness design pe o ms he wo s ,
wi h a mo e p onounced ol age d op and highe mass anspo po-
la iza ion egion, obse ed also du ing he dynamic load es s. When
compa ing he ise and all a he same cu en densi y, he s anda d
de ia ion o empe a u e is highe du ing he cu en dec ease, leading
o a less homogeneous empe a u e dis ibu ion in ansien ope a ion.
Fo cu en dis ibu ion, he s anda d de ia ion emains cons an du ing
bo h inc eases and dec eases.
CRediT au ho ship con ibu ion s a emen
G.M. Cabello Gonz´
alez: W i ing – e iew & edi ing, W i ing –
o iginal d a , Visualiza ion, Valida ion, So wa e, Me hodology,
In es iga ion, Fo mal analysis, Da a cu a ion. Bal asa Toha ias:
W i ing – e iew & edi ing, Visualiza ion, Valida ion, So wa e, In es-
iga ion, Fo mal analysis, Da a cu a ion. Felipe Rosa: Supe ision, Re-
sou ces, P ojec adminis a ion, Funding acquisi ion, Concep ualiza ion.
J.J. Gue a: Supe ision, Resou ces, P ojec adminis a ion, Funding
acquisi ion. Al edo I anzo: Supe ision, Resou ces, P ojec adminis-
a ion, Me hodology, In es iga ion, Concep ualiza ion.
Decla a ion o compe ing in e es
The au ho s decla e ha hey ha e no known compe ing inancial
in e es s o pe sonal ela ionships ha could ha e appea ed o in luence
he wo k epo ed in his pape .
Acknowledgemen
G an TED2021-130706B-I00 unded by MCIN/AEI/10.13039/
501100011033 and by Eu opean Union “Nex Gene a ionEU”/PRTR,
and g an PID2023-146745OB-I00 unded by MICIU/AEI/10.13039/
501100011033 and by ERDF/EU. Expe imen al in as uc u e unded
by UNSE15-CE2962 and EQC-2018-004258-P by MCIN/AEI/10.13039/
501100011033, co- unded wi h ERDF unds. Con ibu ion om Bal asa
Toha ias Funded by CDTI, wi h he suppo om Minis e io de Ciencia e
Inno aci´
on. Financiado po la Uni´
on Eu opea- Nex Gene a ion EU
(p ojec H2ENRY CER-20231027). Au ho s hank Pablo I anzo o he
ealiza ion o he ideo in Supplemen a y Ma e ial.
Da a a ailabili y
Da a will be made a ailable on eques .
Re e ences
[1] H. Liu, I. Khan, A. Zaka i, M. Alha hi, Roles o ilemma in he wo ld ene gy sec o
and ansi ion owa ds sus ainable ene gy: a s udy o economic g ow h and he
en i onmen , Ene gy Policy 170 (2022) 113238, h ps://doi.o g/10.1016/j.
enpol.2022.113238.
[2] H. Ri chie, Sec o by sec o : whe e do global g eenhouse gas emissions come om?
Ou Wo ld Da a (2020). Published online: h ps://ou wo ldinda a.o g/ghg-
emissions-by-sec o .
[3] E. Papadis, G. Tsa sa onis, Challenges in he deca boniza ion o he ene gy sec o ,
Ene gy (Calg.) 205 (2020) 118025, h ps://doi.o g/10.1016/j.
ene gy.2020.118025.
[4] D.A. Cullen, K.C. Neye lin, R.K. Ahluwalia, R. Mukundan, K.L. Mo e, R.L. Bo up, e
al., New oads and challenges o uel cells in hea y-du y anspo a ion, Na .
Ene gy 6 (2021) 462–474, h ps://doi.o g/10.1038/s41560-021-00775-z.
Fig. 8. (a) P05_T65_a13HR60_ c25HR60ai (b) P10_T70_a15HR60_ c30HR55ai . Poin s 1 o 5 co espond o hose ma ked in he d i ing es in Fig. 7.
G.M.C. Gonz´
alez e al.
Jou nal o Powe Sou ces 652 (2025) 237625
9