Anodic TiO2 nanotube layers decorated by Pd nanoparticles using ALD: An efficient electrocatalyst for methanol oxidation

Elec ochimica Ac a 429 (2022) 141044 A ailable online 15 Augus 2022 0013-4686/© 2022 The Au ho s. Published by Else ie L d. This is an open access a icle unde he CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by- nc-nd/4.0/). Anodic TiO 2 nano ube laye s deco a ed by Pd nanopa icles using ALD: An e icien elec oca alys o me hanol oxida ion Bilal Bawab a , Si a amanjaneya M. Thallu i a , b , Jhona an Rod iguez-Pe ei a a , b , Hanna Sopha a , b , Raul Zazpe a , b , Jan M. Macak a , b , * a Cen al Eu opean Ins i u e o Technology, B no Uni e si y o Technology, Pu kyno a 123, 61200 B no, Czech Republic b Cen e o Ma e ials and Nano echnologies, Facul y o Chemical Technology, Uni e si y o Pa dubice, Nam. Cs. Legii 565, 53002 Pa dubice, Czech Republic ARTICLE INFO Keywo ds: TiO 2 nano ube laye s Pd nanopa icles A omic Laye Deposi ion Elec oca alysis Me hanol elec o-oxida ion ABSTRACT He ein, we epo he pe o mance o Pd nanopa icles (NPs) p epa ed by A omic Laye Deposi ion (ALD) as a ca alys o me hanol elec o-oxida ion. Pd NPs we e deco a ed on o anodic TiO 2 nano ube (TNT) laye s as suppo ing ma e ial ha possess a la ge a ailable su ace a ea and di ec elec ical con ac ia he unde lying i anium oil. Di e en Pd loadings (150 – 300 – 450 – 600 ALD cycles) show di e en pa icles sizes anging be ween 7 and 12 nm, as e ealed by ansmission elec on mic oscopy. Coalescence domina ed isibly om 450 ALD cycles, which led o a po ous Pd laye all along he TNT walls a he han he g ow h o indi idual pa icles. Elec oca aly ic pe o mance was in es iga ed by cyclic ol amme y (CV), whe e he ca aly ic ac i i y inc eased p opo ional wi h Pd loading up o he highes alues o 400 and 450 cycles, whe eas a u he inc ease in he numbe o ALD cycles (N ALD ) did no show any addi ional imp o emen in me hanol oxida ion cu en densi ies. TNT laye s deco a ed wi h 400, 450 and 600 Pd ALD cycles show ea u eless cu es sugges ing comple e an i- poisoning abili y o possibly a p oo o a di ec con e sion om CH 3 OH o CO 2 (wi hou any in e media e byp oduc s). The lack o an oxida ion peak du ing he anodic scan and he e o e a educ ion peak du ing he ca hodic scan, con i ms Pd NPs (s abilized by TiO 2 ) e icien ly u ilize OH ads and chemiso bed CH 3 OH in a way ha i s CO poisoning was inhibi ed. As a esul , he uned high su ace a ea TNT laye s exhibi ed excellen pe o mance as a suppo ing ma e ial o Pd NPs agains o ma ion o elec ochemical poisoning species. Finally, he mechanism o he TNT laye s in e ac ion wi h Pd NPs, which led o he p opelling me hanol oxida ion e- ac ion wi hou loss in pe o mance o e cycling is pos ula ed. 1. In oduc ion The e e -inc easing global ene gy demand, along wi h he en i on- men al issues o igina ed om he use o ossil uel, igge ed an in ense sea ch o sus ainable and clean ene gy al e na i es. Di ec me hanol uel cells (DMFCs), in which he chemical ene gy s o ed in me hanol is con e ed o elec ical ene gy, ha e been explo ed in he las yea s. Unlike hyd ogen uel cells, DMFCs use liquid uel ha allows easie handling and anspo a ion. The wo main eac ions ha de e mine he ac i i y o DMFCs a e he anodic me hanol oxida ion eac ion (MOR) and ca hodic oxygen educ ion eac ion (ORR) [1]. The ca aly ic eac- ion kine ics du ing he uel cell ac i i y is dependen on he elec ode su ace mo phology and/o nanopa icles (NPs) sizes. La ge NPs can ela i ely inc ease he NPs su ace exposed o he su ounding elec o- ly e as compa ed o he in e ac ed subs a e su ace [2]. In con a y, smalle NPs a e less likely o ha e CO poisoning specially a high em- pe a u es, due o oxygen bond b eak esul ing in oxygen deso p ion a his le el, which shows mo e esis an o CO poisoning [3]. Noble me al-me al oxide–suppo in e ac ions gene ally play a signi ican ole in elec oca alysis, as hey main ain a high ca aly ic ac i i y o noble me al NPs due o hei ole ance owa ds poisoning om eac ion byp oduc s. The use o me al oxide suppo s, such as MnO 2 [4], SnO 2 [5], o TiO 2 [6], p o ed o be e y bene icial as hey a e highly s able and main ain consis en con e sion e iciency in uel cells. Suppo ing ma e ials, deco a ed by NPs as a ca alys , enhance he elec on anspo a he ca alys /elec oly e in e ace. The p esence o oxide as a sup- po ing ma e ial, whe he i is a educible me al oxide, such as TiO 2 o FeO x , o a non educible oxide such as Z O 2 [7], can di ec ly a ec he ca aly ic ac i i y. Fo ins ance, in ca aly ic con e e s used in he au omo i e indus y, CO in e ac s wi h oxygen on he ca alys su ace * Co esponding au ho . E-mail add ess: [email p o ec ed] (J.M. Macak). Con en s lis s a ailable a ScienceDi ec Elec ochimica Ac a jou nal homepage: www.jou nals.else ie .com/elec ochimica-ac a h ps://doi.o g/10.1016/j.elec ac a.2022.141044 Recei ed 20 Ma ch 2022; Recei ed in e ised o m 3 Augus 2022; Accep ed 14 Augus 2022 Elec ochimica Ac a 429 (2022) 141044 2 p oducing CO 2 . Hence, he eac ion can p oceed in wo di ec ions. In case he suppo ing ma e ial p o ides oxygen, as i happens wi h TiO 2 , hese oxygen a oms will subs i u e he missing oxygen on he su ace. O he wise, CO mos likely subs i u es oxygen a oms and deac i a es he su ace [8]. In pa allel, app oaches based on a ia ions in he elec ode mo phology (nano-, meso-, and mac opo ous ma e ials) ha e shown a huge impac o p e en he de elopmen and accumula ion o in e - media e p oduc s on he ac i e elec ode su ace. In addi ion, he use o alkaline elec oly es showed highe ca aly ic ac i i y owa ds he anodic egime compa ed wi h acidic elec oly es [9]. One mus bea in mind ha es ing noble me als, such as Pd, unde in ense anodic condi ions in alkali medium can esul in o al deac i a ion, due o he o ma ion o Pd hyd oxyla es. Also he elec oly e concen a ions can di ec ly in luence he MOR a e [10]. Rega ding he noble me al NPs used as ca alys owa ds me hanol elec o-oxida ion, he inhe en s abili y o P makes i a g ea ca alys choice as an anode in DMFCs. Howe e , aking in o conside a ion he P sca ci y and he co esponding high cos s o la ge scale P p oduc ion, a sea ch o al e na i e ca alys s has been ini ia ed. Alloys o P wi h o he noble me als, such as Pd, a e conside ed o be used on DMFCs elec odes in o de o educe o subs i u e P mass. While a comple e me hanol (CH 3 OH) oxida ion du ing he elec ochemical p ocess in ol es he ans e o 6 elec ons o ob ain CO 2 as an end-p oduc , incomple e oxida ion can esul in o he undesi ed p oduc s such as o maldehydes, o mic acids, and CO [11]. Pd has he endency o bond wi h CO mol- ecules, which leads o poisoning o he ac i e su ace. E en a lowe empe a u es, he p ocess can s ill acili a e CO o eac , o ming Pd (CO) x compounds [12]. Hyd a ion eac ion ha is mo e p edominan on Pd su ace as compa ed o P , can la e o m OH g oups ha suppo he de-poisoning p ocess o he in e media e “CO” du ing me hanol oxida- ion [13], as shown la e in his pape . The impo ance o ha ing NPs is based on he ac , ha i signi i- can ly imp o es he su ace o olume a io, p o iding mo e ca aly ic ac i e si es. The g ow h o uni o m Pd NPs on high speci ic su ace a ea suppo ing ma e ials is o pa amoun impo ance due o he imp o ed NPs coun pe uni a ea o suppo , hus inc easing he ca aly ic ac i i y o he p ocess. Se e al epo s we e epo ed on he employmen o Pd NPs on di e en subs a es [14–16]. The a ailable li e a u e shows ha he deposi ion o Pd NPs on TNT laye s u ilizing di e en echniques such as elec ochemical milling [17], chemical ba h deposi ion [18,19] and elec odeposi ion [20], su e s om non-uni o m deposi ions wi h a ia ions in co e age densi y all along he su ace. Some o he pub- lished li e a u es ha e ocused on ALD based deposi ion o Pd NPs on TNT laye s o hei applica ions on o ganic molecule deg ada ion [2,15, 21,22]. Well dispe sed Pd NPs o e TNT laye s we e ob ained in hese pape s and hei ac i i ies o di e en ca aly ic eac ions we e p o- posed. Howe e , TNT laye s as a s andalone a e no a good ma e ial o he elec oca aly ics pu poses, knowing i s poo elec ical conduc i i y. Howe e , as a suppo ing ma e ial, high su ace a ea TNT laye s can be de imen al o acili a e he oxida ion o eac an molecules (me hanol). This is due o an eno mous amoun o su ace hyd oxyl g oups ha a e well known o oxidize in e media e p oduc s, majo ly CO gene a ed du ing me hanol elec ooxida ion [23]. The aim o he cu en s udy is o con o maly deco a e TNT laye s by homogeneously dis ibu ed Pd NPs using ALD and o e alua e hei ca aly ic ac i i y p o iles o me hanol elec o-oxida ion. The e o e, he abili y o Pd NPs on TNT (Pd/TNT) laye elec odes o wi hs and agains su ace poisoning is elabo a ed. Pd NPs we e deco a ed by ALD on o 5 µm hick ana ase TNT laye s. The mo phology and he chemical composi ion o he esul ing Pd/TNT laye s we e in es iga ed in de ail by scanning elec on mic oscope (SEM), ansmission elec on mic o- scope (TEM), X- ay di ac ion (XRD), and X- ay pho oelec on spec- oscopy (XPS). The elec oca aly ic ac i i y o Pd/TNT laye s owa ds me hanol elec o-oxida ion was e alua ed by cyclic ol amme y (CV) and ch onoampe ome y (CA) as a unc ion o he numbe o ALD cycles. 2. Expe imen al pa 2.1. TNT laye p epa a ion The de ailed p o ocol o he TNT laye was published in ou p e i- ous wo k [24]. B ie ly, Ti oils (Sigma-Ald ich, 0.127 mm, 99.7% pu i y) we e deg eased and anodized a oom empe a u e using a high- ol age po en ios a (PGU-200 V, IPS Elek oniklabo GmbH) o de elop TNT laye s o ~5 µm hickness and ~230 nm diame e in an e hylene glycol-based elec oly e con aining 10% wa e and 0.15 M NH 4 F a 100 V o 4 hou s. Ti oils and TNT laye s we e annealed in ai o 1 h a 400 ᵒ C. 2.2. A omic laye deposi ion o Pd Pd NPs we e deposi ed on o TNT laye s and on o annealed Ti oils using TFS 200 ALD eac o (Beneq) p o ided wi h s op- low con igu a- ion. Palladium(II)hexa luo oace ylace ona e Pd(C 5 HF 6 O 2 ) 2 (95%, S em Chemicals) and o malin (37% o maldehyde in wa e wi h 10–15% o me hanol, Sigma-Ald ich) we e used as he Pd p ecu so and co- eac an , espec i ely. Pd p ecu so was hea ed up o 65◦C, while he deposi ion p ocess empe a u e was 200◦C. One ALD cycle (N ALD =1) was de ined by he ollowing sequence: Pd pulse (2 s)-exposu e (10 s)-N 2 pu ge (20 s)- o malin pulse (1 s)-exposu e (10 s)-N 2 pu ge (20 s). A p elimina y s ep o enhance he densi y o hyd oxyl unc ional g oups on he su ace subs a e was applied o imp o e he Pd nuclea ion. Such p elimina y s ep consis ed o 5 ALD cycles (wa e pulse (0.5 s)-exposu e (10 s)-N 2 pu ge (20 s)) ollowed by 20 cycles (Pd (0.25 s)-exposu e (10 s)-N 2 pu ge (20 s)) was employed wi h pulse, exposu e, and pu ge du- a ions p og ammed as “5 * (0.5:10:20 s) and 20 * (0.25:10:20 s)” o H 2 O and Pd p ecu so espec i ely p io o he ac ual deposi ion o Pd. The sole necessi y o H 2 O pulse is o c ea e -OH unc ional g oups on he subs a e so ha Pd can nuclea e e icien ly. 2.3. Cha ac e iza ion me hods Blank and Pd deco a ed TNTs (Pd/TNTs) and Pd/Ti oils we e cha ac e ized by scanning elec on mic oscope (SEM) FEI Ve ios 460L. Pd/TNTs we e also cha ac e ized bya high- esolu ion ansmission elec on mic oscope (HRTEM) The mo Fishe Scien i ic Ti an Themis 60-300, ope a ed a 300 kV and equipped wi h a C s image abe a ion co ec o , a high angle annula da k ield de ec o o scanning ans- mission elec on mic oscopy (STEM-HAADF) imaging, and Supe -X en- e gy dispe si e X- ay (EDX) spec ome e wi h ou 30 mm 2 windowless de ec o s o STEM-EDX analysis o check he mo phology and Pd NPs size/dis ibu ion. The c ys alline s uc u e o Pd/TNT laye s was assessed using X- ay di ac ion XRD (Rigaku Sma lab 3 kW di ac- ome e equipped wi h Cu-K α 1 adia ion sou ce). Su ace chemical s a e o Pd NPs be o e and a e elec ochemical measu emen s we e s udied using X- ay pho oelec on spec oscopy (XPS) K a os Analy ical Axis Sup a ins umen , wi h Al-K α monoch oma ic X ay sou ce (h ν = 1486.69 eV). The spec a we e decon olu ed using CasaXPS so wa e and e e enced o he Femi Le el cu o , which means a 0 eV on he binding ene gy scale. Pd 3d spec a we e decon olu ed wi h asymme ic Lo en zian unc ion LA (1.5, 4, 45) o me allic s a e (Pd 0 ), mixed Gaussian-Lo en zian unc ions GL (30) o Pd oxides o hyd oxides and Pd plasmon loss. 2.4. Elec ochemical analysis The elec ochemical measu emen s we e ca ied ou in a no mal h ee elec odes elec ochemical cell a oom empe a u e. Au olab VIONIC po en ios a suppo ed by INTELLO so wa e was used o his pu pose. Two solu ions o 1M KOH and 1M KOH/CH 3 OH we e used as elec oly es. The e e ence elec ode was Ag/AgCl 3M KCl (E ֯ =0.210 V s RHE). A P oil was used as coun e elec ode, whe e Pd deco a ed Ti B. Bawab e al. Elec ochimica Ac a 429 (2022) 141044 3 oils (Pd/Ti) and Pd/TNT laye s we e used as wo king elec odes. Be o e e e y measu emen , degassing o all elec oly es was ca ied ou by bubbling wi h ul a-high pu i y ni ogen o a leas 15 minu es, his o ensu e he emo al o he dissol ed unwan ed oxygen ha can in e e e du ing he elec ochemical oxida ion p ocess. All po en ials gi en in his pape we e ecalcula ed e sus RHE. Cyclic ol amme ic (CV) mea- su emen s we e ca ied a a scanning a e o 10 mV s −1 . Elec ochemical impedance spec oscopy (EIS) was conduc ed using he same se up as abo e, using me hanolic KOH as a solu ion a 0.8 V s. RHE. In o de o check and compa e he s abili y o a ious Pd/TNT laye s and Pd/Ti oils, ch onoampe ome ic (CA) measu emen s we e pe o med a an applied po en ial o 0.9 V s RHE o 2 hou s. The geome ical su ace a ea (0.636 cm 2 ) was used o de e mine he cu en densi ies. In addi- ion, he elec ochemical ac i e su ace a ea (ECSA) has been de e - mined and compa ed wi h he li e a u e esul s. 3. Resul s and discussion Fig. 1 shows SEM op- iew images o TNT laye s (5 µm hick and wi h an inne diame e o ≈250 nm) deco a ed wi h Pd NPs using di e en Figu e 1. SEM op iews o 5 µm hick TNT laye s deco a ed wi h Pd NPs using di e en N ALD : (a) 0, (b) 150, (c) 300, (d) 350, (e) 400, ( ) 450, (g) 600. The scale ba s in all images ep esen 300 nm. B. Bawab e al. Elec ochimica Ac a 429 (2022) 141044 4 numbe o Pd ALD cycles (N ALD ): 0, 150, 300, 350, 400, 450 and 600. Homogenous dis ibu ion o Pd NPs all along he inne and ou e TNT walls was e i ied. As one can expec , an inc ease in he N ALD esul s in a highe densi y o Pd NPs on he TNT laye su ace. F om N ALD ≥450 cycles (Fig. 1d), Pd NPs s a o coalesce esul ing in o a po ous Pd laye . Figu e S1 shows SEM op- iew images o Ti oils deco a ed wi h Pd NPs using di e en numbe o Pd ALD cycles (N ALD ): 0, 150, 300, 350, 400, 450 and 600. The Pd NPs a e al eady coalesced a 300c, which is much ea lie han o he TNT laye s, which o e sligh ly di e en nuclea ion densi y and much highe su ace a ea o hos highe loading o Pd. Conside ing he coalescence, he analyses o NPs size om Pd/Ti has no o limi ed sense. Mo eo e , TEM analysis o hese specimens is no possible, as hey a e oo hick o elec ons o pass h ough. Co esponding EDX analyses o Pd/TNTs p o ided he weigh pe - cen age (w %) dis ibu ion o C, Ti, O, and Pd. Figu e S2 shows an in- c ease in Pd con en on Pd/TNT laye s wi h inc easing N ALD . EDX spec a shown in Figu e S3 app o e he inc easing Pd along wi h inc easing N ALD . The mo phological ea u es o Pd/TNT laye s we e u he cha ac- e ized by HRTEM and compa ed wi h he non-deco a ed Pd laye (0 N ALD ). TEM images in Fig. 2 p o ide an o e iew o high- esolu ion images showing he size and dis ibu ion o Pd NPs on TNTs. The Pd Figu e 2. TEM images o single TNT wi h Pd NPs g own a e 0 (a), 150 (b), 300 (c), 450 (d), and 600 (e) Pd N ALD , espec i ely. ( ) A e age NPs size as unc ion o N ALD ( o he same samples as in a) o e)). B. Bawab e al. Elec ochimica Ac a 429 (2022) 141044 5 NPs we e uni o mly dis ibu ed all along he walls o TNT laye s. As men ioned abo e, Pd NPs coalescence s a ed om 450 ALD cycles (Fig. 2d). Rega ding he Pd NPs size dependency wi h he numbe o ALD cycles, Fig. 2 shows he a e age Pd NPs size o each sample wi h di e en numbe o ALD cycles, de e mined using he s a is ical analyses om TEM images. A e age Pd NPs sizes o 150, 300, and 450 ALD cycles we e 7.5, 10.5, and 12.5 nm espec i ely. A highe N ALD (600) ende ed he coalescence o Pd NPs, hus ini ia ing he p ocess o o m a po ous Pd laye along he TNT walls. XRD pa e ns o he blank TNT laye in Fig. 3a show ypical peaks o ana ase TiO 2 phase wi h he majo peaks loca ed a 25.2◦and 48◦ ep- esen ing (101) and (200) ana ase peaks. The addi ional peaks obse ed a 38.4◦and 40.2◦ ep esen he (002) and (101) c ys alline planes o Ti subs a e. The (111) Pd plane di ac ion peak obse ed o highe Pd deco a ed TNTs (450 and 600 N ALD ) should appea a 40.1◦bu due o o e lapping wi h Ti (101) peak i is di icul o disce n. Ha ing a close look a 600 N ALD 40.1ᵒ peak, he peak becomes b oade (compa ed o o he peaks) wi h a lowe in ensi y co e ing bo h 40.1◦and 40.2◦, his is due o he dec easing size o Pd NPs c ys als. A b oad peak de ec ed a ound 46.2◦could be asc ibed as Pd (200), which was isible only in TNT laye s deco a ed wi h 450 and 600 Pd N ALD . This could be due o o ma ion o po ous Pd s uc u e unlike he samples deposi ed wi h a lowe N ALD (150 and 300) e ealing mo e dispe sed Pd NPs wi hin TNT walls, p e en ing X- ay di ac ion de ec ion. Fig. 3b show high esolu ion Pd, Pd 3d XPS spec a ob ained be o e and a e he CA. Bo h spec a show he co esponding spin-o bi spli - ing Pd 3d 5/2 / Pd 3d 3/2 wi h simila shapes. Se en componen s we e used o decon olu e bo h spec a; six associa ed o he spin-o bi spli - ing, which co espond o h ee di e en chemical species, and he le o e o he cha ac e is ic plasmon loss ha appea s as Pd is mos ly me allic. The i s double ( ed peaks) a e cen e ed a 335.3 / 340.6 eV con i ming he p esence o me allic Pd (Pd 0 ) [25]. The binding ene gies co esponding o he second double (blue peaks) a 336.9 / 340.5 eV, we e ela ed o Pd 2+ [26], while hose om he hi d double (g een peaks), loca ed a 338.7 / 344.0 eV we e assigned o Pd 4+ [27]. Pd 2+ and Pd 4+ could be ela ed o co esponding oxides o hyd oxides [28,29]. Since no majo di e ences we e ound om chemical species and/o binding ene gies shi s in Pd spec a be o e and a e he elec ochemical measu emen s, quan i ica ion o he oxida ion s a es was pe o med, as shown in Fig. 3c. Acco ding o hese esul s, i is e iden ha a e CA measu emen s he amoun o me allic Pd dec eases sligh ly (a ound 4.3%), while he con en o Pd 2+ inc eases. The su ace chemical s a e o TNT laye s deco a ed wi h 600 Pd N ALD was in es iga ed be o e and a e he elec ochemical measu emen s by XPS. The su ey spec a (Figu e S4a) in bo h cases e eal he p esence o C, O, Ti, F and Pd. The C signal comes om he ad en i ious ca bon. The p esence o oxygen can be iden i ied om he Auge O KLL signal, since he O 1s has a s ong o e lapping wi h he Pd 3p 3/2 signal. Ti peaks a e due o he subs a e (TNT laye s), while F is a emnan om he Ti oil anodiza ion p ocess. In o de o e i y he chemical na u e o TiO 2 a e he ALD p ocess, addi ional high esolu ion XPS analysis was conduc ed o Pd/TNTs sample p oduced wi h 600c o Pd ALD p ocess (i.e. be o e any elec ochemical expe imen ). This sample spen he longes ime in he ALD educ i e a mosphe e. Figu e S4b shows Ti 2p high esolu ion spec um o he co esponding sample. The spec um e eals only he Ti 4+ s a e, since Ti 2p3/2 is a ~459 eV. Thus, i is clea ha no educ ion o Ti 4+ o Ti 3+ occu ed du ing he ALD p ocess. Fig. 4a exhibi s he elec oca aly ic ac i i y p o ile o Pd/Ti oils in elec oly e o 1 M KOH solu ion. The blank oil shows no esponse along he CV in he po en ial ange (0 o 1.2 V), indica ing no oxida ion e- ac ion occu s. Inc easing Pd N ALD om 150 o 600 enhanced he cu en densi y a lowe po en ials, showing sha p peak a a ound 0.1 V. This peak ep esen s he deso p ion o H on he Pd NPs su ace due o he elec oly e (1M KOH) en iched wi h dissocia ed hyd oxyl g oups. While i is well known ha he Pd su ace ge s g adually oxidized as he anodic scan mo es owa ds highe po en ial egion o E >0.7 V [30], he e e se scan shows a educ ion peak a a ound 0.7 V ep esen ing he deso p ion o oxygen ( educ ion o PdO o Pd), adso bed du ing he p e ious o wa d scan be ween 0.75 and 1.2 V. Fig. 4b depic s he CVs ob ained om he se o Pd/Ti oils es ed in 1 M KOH/CH 3 OH. Figu e 3. (a) XRD pa e ns o 5 µm hick Pd/TNT laye s deco a ed wi h N ALD =0, 150, 300, 450, and 600 Pd. b) Pd 3d XPS high esolu ion spec a o 600c Pd/TNT laye s be o e and a e ch onoampe ome ic (CA) measu emen s. c) A omic concen a ion o Pd deduced by XPS. B. Bawab e al. Elec ochimica Ac a 429 (2022) 141044 6 No iceable oxida ion peaks appea s a ing om Ti oils deco a ed wi h 300 Pd N ALD . The esul s also indica e ha inc easing Pd N ALD deposi ed on Ti oils leads o an inc ease in he me hanol oxida ion peak cu en densi y, and he e o e in he ca hodic scan he CO oxida ion peak cu - en densi y. The highes cu en densi y alue o 8.9 mA cm −2 co e- sponds o he Ti oils deco a ed wi h 450 Pd N ALD , while hose deco a ed wi h 300 and 600 Pd N ALD show compa a i ely lowe alues o 5.8 and 7.8 mA cm −2 a ~0.8 V (no ed om he hi d CV cycle). The elec oca aly ic ac i i y o Pd NPs on Ti oils in me hanolic KOH elec oly e shown in Fig. 4b ollows he end o anodic oxida ion o MeOH and ca hodic oxida ion o CO [31]. The ob ained CV scans a e explained by he ac ha du ing he anodic scan, he sample su ace is being ac i a ed wi hin a ce ain po en ial ange, a e which i becomes deac i a ed by oxide species (such as CO), a ached on Pd NPs su ace p e en ing u he elec o-oxida ion o me hanol. Oxida ion eac ion ini ia es a a ound 0.7 V and he esul ing cu en densi y inc eases o each he elec o-oxida ion limi a 0.8 V. A his poin (0.8 V), he ca aly ic ac i e si es o Pd NPs su ace we e poisoned due o adso p ion o CO in e media e o med du ing me hanol oxida ion, dec easing he cu en densi y un il he su ace is comple ely inac i a ed. Howe e , Figu e 4. CVs showing h ee consecu i e cycles o Pd/Ti oils (a, b) and Pd/TNT laye s (c, d) deco a ed wi h di e en N ALD (0, 150, 300, 350, 400, 450, and 600), eco ded in 1 M KOH (a, c) and me hanolic (1 M KOH +1 M CH 3 OH) elec oly e by applying a scanning a e o 10 mV s −1 . B. Bawab e al. Elec ochimica Ac a 429 (2022) 141044 7 eac i a ion o he Pd NPs su ace occu s du ing he e e se ca hodic scan wi hin a po en ial alue anging be ween 0.7 and 0.6 V, due o oxida ion o he adso bed CO molecules ( o CO 2 ) and educ ion o PdO o me allic Pd. The me allic Pd and Pd hyd oxides a e esponsible o he inc ease in he cu en densi y among samples o Pd/Ti oils. Along he po en ial inc ease in he anodic scan he numbe o neighbo ing OH g oups on Pd also inc eases. Thus, such inc emen o OH g oups a highe po en ials acili a es an addi i e e ec on he emo al o he CO molecules, adso bed du ing he anodic scan. Fig. 4c-d shows he elec oca aly ic pe o mance ob ained om Pd/ TNT laye s, which beha es di e en ly om Pd/Ti oils, when i comes o me hanol oxida ion. TNT laye s es ed in non-me hanolic elec oly e (Fig. 4c) show simila peaks compa ed wi h Ti oils. In he oxygen adso p ion ange (0.7 – 1.2 V), Pd-OH, PdO, and/o PdOx could be o med, while a deso p ion educ ion peak is mo e p onounced in Pd/ TNT laye deco a ed wi h 400, 450 and 600 Pd N ALD compa ed wi h Ti oils due o highe Pd loadings. Fu he mo e, Fig. 4d shows Pd/TNT laye s pe o mance in me hanolic KOH, showing cu en densi ies o 66.7, 64.8 and 62.6 mA cm −2 ob ained a 1.2 V o 400, 450 and 600 Pd N ALD espec i ely. The CVs show no isible anodic and ca hodic peaks, hus ep esen ing a ea u eless pola iza ion cu e wi h an onse po en- ial o 0.5 V. The TNT laye is an oxyphilic oxygen a oms p o ide . I appea s o be an e ec i e oxidize o mi iga e elec o-oxida ion o me hanol on he Pd/TNT laye su ace, ci cum en ing ei he CO poisoning o Pd NPs su ace o a oiding o ma ion o in e media e o ganic species. [8,23,32] Du ing he elec ode deac i a ion, obse ed o Pd/Ti oils, CO species, in addi ion o o he in e media es o med du ing he p ocess o me hanol elec o-oxida ion, a e belie ed o be he main elec ode poisoning species. As a ma e o ac , based on he p e ious li e a u e [33] CO is no he only sou ce o poisoning he ca alys su ace, and emo al o CO is a di usion con olled p ocess, which explains he slow deso p ion p ocess. The oxida ion o CO ads is conside ed as he a e de e mining s ep acco ding o he ollowing eac ion, and i is well known ha he comple e MOR can yield CO 2 as end p oduc : CH3OHads +4OH−→COads +4H2O+4e−(Eq. 1) COads +2OH−→CO2+H2O+2e−(Eq. 2) In a ypical anodic scan o MOR (Fig. 4b) he oxida ion s a s and ca ies on un il a ce ain po en ial whe e he adso bed OH g oups can no longe eac wi h he a ailable CO molecules adso bed on he su ace (CO ads ), leading o a cu en densi y d op. CO and OH g oups a e belie ed o be e ol ing du ing he same po en ial window. Howe e , he e ec o highe po en ial excels he ac i i y o OH g oups. To simpli y, CO bonding on he su ace can only lead o a loss o su ace ac i i y as explained be o e and shown in (Fig. 4b), while he CO oxida ion eac ion occu ing wi h he neighbo ing OH g oups, bonded in o he su ace o high su ace a ea TNTs, plays an impo an ole in clea ing CO molecules and he o he adso bed species. The ob ained ea u eless cu es (Fig. 4d) show no decline in he cu en densi y du ing consecu i e CVs wi hin he applied po en ial ange (0 o 1.2 V). In o de o s udy such e ec , he po en ial ange is spli in o wo pa s. A lowe po en ial egion below he peak po en ial alue a ound 0.8 V (Fig. 4b), and a highe egion abo e his alue. A he lowe po en ial egion, adso p ion o me hanol akes place, and Eq. (1) applies esul ing in CO adso bed on he su ace. Following, Eq. (2) ap- plies a highe po en ial, suppo ed by he oxophilic na u e o TiO 2 and i s elec on dono p ope ies, which enhance he CO ads oxida ion ki- ne ics enough o keep he CH 3 OH oxida ion wi hou su ace blockage by in e media es. Thus, CO ads oxida ion eac ion is no conside ed as he a e de e mining s ep anymo e wi h he Pd/TNT laye s. The elec ochemical su ace a ea (ECSA) was de e mined using he in eg a ed educ ion peak o PdO. The alue o cha ge o he educ ion o a PdO monolaye was aken om he li e a u e [34,35] wi h QM = 424 μ C/cm2. The ECSA alue was de e mined using he ollowing o mula. ECSA =Qin QM Since he alue o Q in is scan a e dependen , i is co ec o use di e en scan a es o de e mining he inal alue o cha ge in eg a ed om he educ ion peak. The ECSA o Pd/TNT laye s deco a ed wi h 450 Pd and 600 Pd N ALD we e calcula ed o be 176 cm 2 and 213 cm 2 , a highe as compa ed o 0.636 cm 2 ob ained as he co esponding geome ical a ea (de ined as he dimensional a ea exposed o he elec oly e, wi hou aking he su ace po osi y in o conside a ion). The ECSA alue ob ained o he TNT laye deco a ed wi h 600 Pd N ALD is abou 7 imes la ge han he ECSA alues epo ed in he li e a u e, which we e es ima ed o be 22.3, 29.6, and 7.1 cm 2 o Pd/TNT laye s (500, 700, and 900 Pd N ALD ) [2], and 14.2 cm 2 o palladium suppo ed TiO 2 nano ube a ay (Pd-TNTA) de eloped by elec ochemical milling and ace ing (ECMF) me hod [17]. Elec ochemical impedance spec oscopy (EIS) was used as a powe ul ool o s udy he elec on- ans e kine ics o he me hanol oxida ion eac ion. EIS measu emen s we e ca ied ou a a equency ange (10 −2 and 10 5 Hz), and an AC ampli ude o 10 mV a a po en ial o 0.8 V s RHE in 1M KOH/CH 3 OH elec oly e. In gene al, he cha ge ans e esis ance alue is ep esen ed by he diame e o he semici cle. A smalle diame e co esponds o a smalle cha ge ans e esis ance, which in u n ep esen s a high cha ge ans e a e p o ided he e al- ua ion pe o med o all he samples a he same po en ial. Fig. 5 ep- esen s he Nyquis plo s o Pd/TNT laye s deco a ed wi h di e en N ALD o Pd. The Nyquis plo s o Ti oils and a compa a i e assessmen o cha ge ans e esis ance alues o bo h oils and TNTs wi h Pd a e p o ided in Figs. S5 and S6. The dec ease in cha ge ans e esis ance o TNTs deco a ed wi h N ALD Pd is e iden and no signi ican di e ence among samples wi h 400, 450 and 600 N ALD was obse ed, which aligns wi h obse ed cu en densi ies om CV o he samples. In o de o e alua e he s abili y o Pd NPs in 1M KOH elec oly e wi hin he po en ial ange o 0 – 1.4V, 22 consecu i e CV cycles we e pe o med o TNTs deco a ed wi h Pd NPs using N ALD =600, as shown in Fig. 6. The esul ing CV cu es showed se e al meaning ul egions in he anodic as well as in he ca hodic scan, as desc ibed ahead. The lowe po en ial egion (1) co esponds o hyd ogen deso p ion o adso p ion in addi ion o K + ca ion species [36]. This p ocess akes place as a esul o a p e ious adso p ion p ocess happened du ing he e e se scan a po- en ial anges be ween (0 – 0.2V). The peak in ensi y inc ease wi h cycling could be due o accumula ion o he adso bed species a he su ace du ing cycling. In egion (2), he lowe cu en densi y co e- sponds o he double laye cha ge, bu he cu en densi y inc eases upon cycling. I is shown in e . [37] ha i is unlikely ha he ise o he anodic peak du ing cycling is ela ed o oxide o ma ion. A he same ime, he dec ease in oxygen adso p ion cu en densi y ( egion 3) may indica e he lowe quan i ies o he adso bed OH g oups on he su ace and PdO could be o med al e na i ely. The de ailed oxida ion p ocess is no ye ully unde s ood. The o med OH g oups eac wi h monoxides ob ained du ing dehyd ogena ion p ocess leading o a comple e eac- ion. Then he sligh inc ease in cu en densi y o he sha p ca hodic peak ( egion 4) ep esen s he educ ion o PdO in o Pd, whe e he in- ensi y inc eases wi h inc easing he palladium oxide on su ace. In o de o exploi he s abili y o Pd NPs on TNTs in 1M KOH +1M CH 3 OH elec oly e, 22 consecu i e CV cycles we e pe o med o TNTs deco a ed wi h Pd NPs using N ALD =450, wi hin he po en ial ange o 0 – 1.2V. The esul ing CV cu es shown in Figu e S7 p o e e y good s abili y o he sample es ed. In o de o gain u he insigh in o he s abili y o he Pd/Ti oils and Pd/TNT laye s´elec oca alys , ch onoampe ome ic (CA) measu e- men s we e ca ied ou o 2 hou s a 0.9 V. CA cu es ob ained o Pd/ TNT laye s a e shown in Fig. 7. CA cu es o Pd/Ti oils a e shown in Figu e S8. Pd/Ti oils show a d as ic dec ease in he cu en densi ies B. Bawab e al. Elec ochimica Ac a 429 (2022) 141044 8 du ing he CA analysis con i ming he e ec o CO poisoning ha led o a loss o ac i e si es esponsible o me hanol elec o-oxida ion. Whe eas he esul s ob ained om Pd/TNT laye s con i m he bene i s o noble me al - me al oxide suppo in e ac ion, by p o iding signi ican ly highe ole ance o CO poisoning. The CA esul s depic ela i ely s able cu en densi ies along he CA es o he Pd/TNT laye s deco a ed wi h 600 Pd N ALD , de eloping a cu en densi y o 28.5 mA cm −2 (Fig. 7). In con as , he Pd/TNT laye s deco a ed wi h 450, 400, 350, 300, and 150 Figu e 5. EIS Nyquis plo s o Pd/TNT laye s in me hanolic (1 M KOH +1 M CH 3 OH) elec oly e using di e en N ALD . Figu e 6. CV cu es (22 consecu i e cycles) ob ained in 1M KOH o 5µm TNTs deco a ed wi h Pd NPs using N ALD =600. B. Bawab e al. Elec ochimica Ac a 429 (2022) 141044 9 Pd N ALD show compa a i ely a highe cu en densi y d op (compa ed o he ini ial cu en densi y) a e 2h es , wi h cu en densi ies o 24.8, 16.7, 0.5, 1.07, and 0.07 mA cm −2 , espec i ely. In e es ingly, he d op in cu en densi ies was a enua ed o a highe Pd N ALD . Thus, he pe cen ages o cu en densi y loss we e 16.1%, 23.7%, 32.1%, 42.8%, 49%, and 75.8% o 600, 450, 400, 350, 300 and 150Pd N ALD , espec i ely. The e o e, he esul s indica e ha he Pd NPs size plays a signi ican ole in he elec oca aly ic pe o mance. Thus, he domi- nance o he OH co e age on small Pd NPs will dec ease he ca aly ic ac i e Pd su ace esponsible o me hanol adso p ion, leading o a highe d op in he cu en densi ies unde long e m pe o mance [38]. Hence, based on he esul s ob ained in his wo k, i is sugges ed ha he Figu e 7. Ch onoampe ome ic (CA) cu es demons a ing he s abili y o Pd/TNT laye s wi h di e en N ALD (150, 300, 350, 400, 450, and 600). CA cu es we e eco ded in 1M KOH +1M CH 3 OH a 0.9V RHE o 2 h. Table 1 O e iew o cu en densi ies ob ained om he published li e a u e and compa ed wi h densi y acqui ed in his wo k. All alues de e mined o calcula ed a 0.9V s RHE, i no deno ed o he wise. F om CV F om CA Subs a e Scan a e mV. s ¡1 Cu en densi y (mA cm −2 ) Onse po en ial A ea conside ed omcu en densi y (mA cm −2 ) @ ime@ V Re . E hanol oxida ion SnO 2 /Pd 500 cycles 25 1.19 ECSA -0.465 (ECSA) 0.4 @0s 0.19 @3500s [39] TiO 2 TNTs/Pd 500 cycles 25 0.63 ECSA -0.465 (ECSA) 0.2375 @0s 0.175 @3500s [2] Pd/TiO 2 C 50 18 0.424 Ch onopo en iome ic 3mA/cm 2 @[email p o ec ed] 3mA/cm 2 @28800s@ 0.72V [19] Me hanol oxida ion Smoo h P elec ode 50 3.84×10 −9 A 0.5 NA [40] Pd/C 20 3.1 0.6 NA [30] PdRh/C 20 5 0.6 NA Pd/TiO 2 NTs in H 2 SO 4 50 8.66 A 0.436 NA [41] 60-65% po ous Pd 25 710 0.15 NA [10] Pd/TiO 2 -C 50 2.24 (ECSA) 0.44 (ECSA) 1.83 @7200s [42] TiO 2 /Pd 450 cycles 10 32 ~0.5 (Geome ic) 24.8 @7200sec @0.9V This wo k B. Bawab e al.