Substrate developments for the chemical vapor deposition synthesis of graphene

www.ad ma in e aces.de 1902024 (1 o 10) © 2020 The Au ho s. Published by WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim Re iew Subs a e De elopmen s o he Chemical Vapo Deposi ion Syn hesis o G aphene Qi ao Shi, Klaudia Toka ska, Huy Quang Ta, Xiaoqin Yang, Yu Liu, Sami Ullah, Lijun Liu, Ba ba a T zebicka, Alicja Bachma iuk, Jingyu Sun, Lei Fu, Zhong an Liu, and Ma k H. Rümmeli* DOI: 10.1002/admi.201902024 1. In oduc ion The e a e nume ous app oaches in which single- and bilaye g aphene can be ab- ica ed o isola ed. Cu en ly, he mos popula syn hesis p ocess is chemical apo deposi ion (CVD). The CVD me hod can yield ela i ely high quali y g aphene, has he po en ial o syn hesize g aphene on a la ge scale, is well es ablished and is acile.[1] While he CVD p ocess is o he mos pa s aigh o wa d, some specialis equipmen is equi ed and he p ocess pa ame e s and eac o con igu a ion a e also impo an . CVD, in essence, is a p ocess in which gaseous eac an s a e used o deposi ma e ial on o a subs a e su ace. The use o CVD o he ab ica- ion o g aphene apidly de eloped once g aphene had been demons a ed expe i- men ally. Me al ca alys s, in pa icula ansi ion me als, which had a highly success ul his o y o he g ow h o ca bon nano ubes, whe e he ob ious choice in Since he isola ion o g aphene and nume ous demons a ions o i s unique p ope ies, he expec a ions o his ma e ial o be implemen ed in many u u e comme cial applica ions ha e been eno mous. Howe e , o da e, challenges s ill emain. One o he key challenges is he ab ica ion o g aphene in a manne ha sa is ies p ocessing equi emen s. While ans e o g aphene can be used, his ends o damage o con amina e i , which deg ades i s pe o mance. Hence, he e is an impo an d i e o g ow g aphene di ec ly o e a numbe o echnologically impo an ma e ials, iz., di e en subs a e ma e ials, so as o a oid he need o ans e . One o he mo e success ul app oaches o syn hesis g aphene is chemical apo deposi ion (CVD), which is well es ablished. His o ically, ansi ion me al subs a es a e used due o hei ca aly ic p ope ies. Howe e , in ecen yea s his has de eloped o include many nonme al subs a e sys ems. Mo eo e , bo h solid and mol en subs a e o ms ha e also been demons a ed. In addi ion, he cu en end o p og ess lexible de ices has spu ed in e es in g aphene g ow h di ec ly o e lexible ma e ials su aces. All hese aspec s a e p esen ed in his e iew which p esen s he de elopmen s in a ailable subs a es o g aphene ab ica ion by CVD, wi h a ocus p ima ily on la ge a ea g aphene. Q. Shi, X. Yang, Y. Liu, S. Ullah, P o . A. Bachma iuk, P o . J. Sun, P o . M. H. Rümmeli Soochow Ins i u e o Ene gy and Ma e ials Inno a ions College o Ene gy Key Labo a o y o Ad anced Ca bon Ma e ials and Wea able Ene gy Technologies o Jiangsu P o ince Soochow Uni e si y Suzhou 215006, China E-mail: [email p o ec ed] K. Toka ska, P o . B. T zebicka, P o . A. Bachma iuk, P o . M. H. Rümmeli Cen e o Polyme and Ca bon Ma e ials Polish Academy o Sciences M. Cu ie-Sklodowskiej 34 Zab ze 41-819, Poland The ORCID iden i ica ion numbe (s) o he au ho (s) o his a icle can be ound unde h ps://doi.o g/10.1002/admi.201902024. D . H. Q. Ta, P o . A. Bachma iuk, P o . M. H. Rümmeli Ins i u e o Complex Ma e ials IFW D esden 20 Helmhol z S asse, D esden 01069, Ge many P o . M. H. Rümmeli Ins i u e o En i onmen al Technology VSB-Technical Uni e si y o Os a a 17. Lis opadu 15, Os a a 708 33, Czech Republic X. Yang, P o . L. Liu School o Ene gy and Powe Enginee ing Xi’an Jiao ong Uni e si y No. 28, Xianning Wes Road, Xi’an, Shaanxi 710049, China P o . L. Fu College o Chemis y and Molecula Science Wuhan Uni e si y Wuhan 430072, China P o . Z. Liu Cen e o Nanochemis y Beijing Science and Enginee ing Cen e o Nanoca bons Beijing Na ional Labo a o y o Molecula Sciences College o Chemis y and Molecula Enginee ing Peking Uni e si y Beijing 100871, China © 2020 The Au ho s. Published by WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim. This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion and ep o- duc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. Ad . Ma e . In e aces 2020, 7, 1902024 www.ad ancedsciencenews.com www.ad ma in e aces.de 1902024 (2 o 10) © 2020 The Au ho s. Published by WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim ea ly s udies o he CVD ca aly ic g ow h o g aphene. How- e e , in many cases he di ec g ow h o g aphene o e me al subs a es is undesi able and ans e p ocedu es in oduce damage and so is o he mos pa also unwan ed. This has led o he de elopmen o CVD echniques o he di ec ab- ica ion o g aphene o e di e en ypes o subs a es. Indeed, he a ailabili y o subs a es o g aphene g ow h by CVD has ma u ed signi ican ly in ecen yea s and nume ous me als and nonme als can be used. Mo eo e , hese subs a es can be used in hei solid s a e and in many cases in a mol en s a e. In addi- ion, p og ess now includes di ec g aphene ab ica ion o e lexible subs a es. The e a e nume ous e iews on he CVD syn hesis o g aphene and hey, o he mos pa , end o con- nec a b oad discussion on he syn hesis o g aphene, i s p op- e ies and i s applica ion[2–6] o hey emain b oad in discussing he syn hesis o g aphene and ocus on a mo e speci ic applica- ion, o an excellen example, seconda y me al ion ba e ies[7] o g aphene as a sma ma e ial.[8] O he e iews may ocus on a speci ic o m o g aphene such as N doped g aphene[9] o po ous g aphene.[10] Howe e , a e iew ocusing on he sub- s a es o he di ec g ow h o g aphene by CVD is lacking. This e iew add esses he de elopmen s in subs a e sys ems o he g ow h o g aphene by CVD, wi h a ocus on la ge a ea g aphene. 2. The CVD P ocess in B ie CVD g ow h o g aphene is a chemical p ocess in which in ica ely mixed homogeneous gas phase and he e ogeneous su ace eac ions a e in ol ed.[11] In he case o g aphene, he gaseous p ecu so unde goes py olysis o o m ca bon species. These ca bon species a e hen ca ied and deposi ed on o he su ace o he subs a e whe e nuclea ion occu s and hen o ms (g ows) in o he ca bon s uc u e o g aphene. To imp o e he py olysis o decomposi ion o he p ecu so usu- ally a ca alys is used o educe he eac ion empe a u e, bu his is no a p e equisi e. In gene al, he p ocess o CVD g aphene ab ica ion consis s o eigh s eps: 1) mass anspo o he eac an , 2) eac ion o he p ecu so , 3) di usion o gas molecules, 4) adso p ion o he p ecu so , 5) di usion o he p ecu so in o he subs a e (in many cases his is no ue whe e he solubili y is limi ed), 6) su ace eac ion, 7) deso p- ion o p oduc s and 8) emo al o he by-p oduc s.[12,13] O en he mal CVD is used, whe e, as he name sugges s, hea ing is he d i ing sys em o he eac ion. Howe e , he use o a plasma ( iz., plasma enhanced CVD, PECVD) can help educe py olysis empe a u es. In e ms o ope a ion p essu e, wo b oad ca ego ies exis , namely, a mosphe ic p essu e CVD (APCVD) and low p essu e CVD (LPCVD). The eac ion p es- su e, along wi h low a es, empe a u e and g ow h ime can a ec he inal quali y o he as p oduced g aphene. The choice o subs a e is also impo an and di e en subs a es can be used. The mos success ul a e me als, in pa icula Cu as i se es as a ca alys and also, he low ca bon adso p ion p open- si y o Cu allows o easie con ol o o m la ge a ea homo- geneous mono o bilaye g aphene. The main d awback o g aphene ab ica ed o e Cu is ha , usually, he g aphene needs o be ans e ed o he subs a e (e.g., o de ice ab ica ion) and his p ocessing s ep can incu con amina ion and/o damage, which leads o educed pe o mance o he g aphene. Thus, he po en ial o di ec ly g ow g aphene on nonme allic Klaudia Toka ska is a Ph.D. candida e o he Polish Academy o Sciences (CMPW PAN) in Zab ze, whe e she wo ks in Labo a o y o Ca bon and Polyme -Ca bon Ma e ials. She ecei ed he Bachelo o Science deg ee a Silesian Uni e si y o Technology (Zab ze, Poland) in 2016. She joined he P o . Rummeli’s esea ch g oup in 2017. He cu en esea ch ocuses on de elopmen o g aphene coa ings on Si nanopa icles and hei applica ion as high-pe o mance anode ma e ials in seconda y li hium-based ba e ies. Ma k H. Rümmeli heads he elec on mic oscopy and LIN labs a he Soochow Ins i u e o Ene gy and Ma e ials Inno a ions (SIEMIS), Soochow Uni e si y, whe e he is a ull p o esso . He is also di ec o o he cha ac e iza ion cen e a he College o Ene gy and SIEMES. Mo eo e , he is a ull p o esso o he Polish Academy o Sciences (CMPW PAN) in Zab ze and has ull habili a ion igh s. He ob ained his Ph.D. om London Me opoli an Uni e si y and hen wo ked as a pos doc a he Ge man Ae ospace Cen e . His esea ch ocuses on he g ow h mechanisms o 2D nanos uc u es and hei unc ionaliza ion. Qi ao Shi ecei ed his bachelo o science deg ee om he Depa men o physics and op oelec onic ene gy a Soochow Uni e si y, Suzhou, China, in 2016. Cu en ly, he wo ks as doc o al esea che a he Soochow Ins i u e o Ene gy and Ma e ials Inno a ions (SIEMIS) and he College o Ene gy a Soochow Uni e si y China in P o . Ma k H. Rummeli’s g oup. His cu en esea ch ocuses on sol ing he pul e iza ion issues o Si pa icles as anode ma e ials ia space enginee ing o s uc u e op imiza ion. Ad . Ma e . In e aces 2020, 7, 1902024 www.ad ancedsciencenews.com www.ad ma in e aces.de 1902024 (3 o 10) © 2020 The Au ho s. Published by WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim subs a es by CVD is c i ical o expand he applica ion po en ial o g aphene. 3. Subs a e Sys ems o G aphene Syn hesis by CVD The ac i i y o ansi ion me als o ming g aphi ic ma e ial om he e ogeneous ca aly ic eac ions has been known o yea s om indus ial p ocesses. In addi ion, hei success ul implemen a ion o ca bon nano ubes is also well known and so ansi ion me als whe e ini ially used as subs a es o g a- phene g ow h. The choice o ansi ion me al depends on he end goal (e.g., mono laye , bilaye , ew laye g aphene ilms) and conside a ion o he me als s abili y a high empe a u e, and ca bon solubili y a e impo an along wi h hei ca aly ic abili y, which lowe s he ac i a ion ene gy pa hways o he eac ions in he CVD p ocess. In o de o help con ol issues ela ed o high ca bon solubili y, a ious me al alloys ha e been de eloped, and hese a e also discussed in he ollowing subsec ions. Me al subs a es, al hough highly success ul o g aphene g ow h, when using he g aphene o de ice based applica ions su e a signi ican d awback, namely, he need o ans e which in oduces de ec s and/o con amina ion which is coun e p oduc i e in ha his diminishes he pe o mance o he g aphene. Hence, conside able e o s ha e been di ec ed o he g ow h o con inuous and homogeneous g ow h o g a- phene di ec ly on dielec ic subs a es such as hBN, Si/SiO2, Al2O3, GaN, MgO, Si3N4, e c. G ow h o g aphene di ec ly on hese subs a es is mo e challenging, none- he-less, g ea s ides ha e been made, and his is also discussed in he sec- ions below. Mo eo e , he g ow h o g aphene o e mol en subs a es and lexible subs a es is also p esen ed. In addi ion, i should be no ed, ha unless explici ly s a ed, mos discussion cen e s on la ge a ea (poly c ys alline) g aphene. We begin wi h discussion on me al su aces. 3.1. CVD o e Me al Subs a es While he e a e a huge numbe o me al ca alys s o choose om in he pe iodic able o elemen s (91 o he 118 elemen s lis ed in he pe iodic able), he mos success ul o he g ow h o g aphene end o be om he ansi ion me als. Ea ly CVD g ow h ocused on solid me als, in pa icula Ni and Cu. How- e e , i soon became clea ha liquid me als can also be used and ha hey may ha e ce ain ad an ages. We now b ie ly look a solid me als i s and hen mol en o liquid me als. 3.1.1. Solid Me al Subs a es Ni was a highly success ul me al in he CVD syn hesis o ca bon nano ubes and, hus, i was no su p ise ha ea ly esea ch o he CVD syn hesis o g aphene exploi ed Ni. Typically, polyc ys alline Ni ilms a e used and p io o syn hesis hey a e annealed in an A /H2 a mosphe e (≈900–1000 °C) o educe he su ace oxide and also inc ease he g ain size. Fo g ow h, he Ni ilm hen equi es a ca bon eeds ock, usually his is CH4. The use o A is no necessa y, bu can be implemen ed. Ni has a ela i ely high C solubili y, so ha as he CH4 decom- poses i p oduces C species (ca aly ically aided by he Ni), which hen dissol e in he Ni ilm a he ele a ed empe a u es.[14] This is ollowed by a cooling down s ep du ing which C a oms di use ou om he Ni–C solid solu ion and p ecipi a e on he Ni su ace and o m g aphene ilms. In o he wo ds, g aphene ilms o m on Ni h ough a ca bon seg ega ion–p ecipi a ion p ocess. Mul i- o single-laye g aphene can be g own on Ni, and can be con olled by he cooling a e.[15–17] Howe e , he g aphene laye numbe s end no o be homogeneous due o excess C dissol ing ou a g ain bounda ies leading o mul i- laye nuclea ion. Fo his eason, a p eannealing ea men is applied o educe he numbe o g ain bounda ies. In addi ion, he g ow h ime and hyd oca bon concen a ion can also a ec he g aphene laye numbe o ma ion.[5] One can also use e y hin Ni ilms which hen by de aul can only abso b a ce ain amoun o ca bon and his limi s C p ecipi a ion. None- he- less la ge a ea homogeneous single-laye g aphene emains di icul o ob ain.[18] A nea ac ic o o e come his issue is o use Mo along wi h Ni, e.g., by deposi ing Ni on a Mo oil o p oduce a bina y ca aly ic subs a e. The echnique yields la ge a ea homogeneous monolaye g aphene.[19] Sys ema ic s udies show ha du ing he APCVD p ocess, he Ni and Mo ilms b eak up, di use and mix as independen islands. G aphene nuclea ion occu s exclusi ely o e he Ni islands while he Mo islands a ound he Ni soak up excess C (see Figu e 1).[20] Thus, C p ecipi a ion is con olled in a use ul way enabling la ge a ea single-laye g aphene o o m. Single-c ys al Ni (111) can be used o homogeneous single-laye g aphene o ma ion,[21] howe e , his is di icul o ansla e o la ge a ea single c ys alline g ow h. Aside om Ni, a numbe o o he me al subs a es wi h di e en ca bon solubili y and ca aly ic e ec can be used. These include Ru,[22] I ,[23] P ,[24] Co,[25,26] Pd,[27] and Re.[28] Va ying deg ees o suc- cess ha e been ob ained wi h hese. A b eak h ough came wi h he use o polyc ys alline Cu oils as a subs a e since i could ela i ely easily deli e high-quali y homogeneous single-laye g aphene a low cos and ela i ely easy ans e .[5,29] In he case o Cu as a subs a e, while ob aining homogenous single- laye g aphene is easie , g ow h pa ame e s such as gas low a e c ucial. Un es ic ed gas lows in APCVD lead o inhomo- geneous laye numbe s. An al e na i e con igu a ion in which one end o he inne ube (in which he sample is placed) is closed so as o es ic he gas low leads o a homogeneous g aphene laye numbe o e la ge a eas.[30] The s udy showed he g aphene laye numbe o depend on he sample place- men , yielding ei he homogeneous monolaye o bilaye g a- phene. The da a showed ha local condi ions play a ole on laye homogenei y such ha unde quasis a ic equilib ium gas condi ions no only is he laye numbe s abilized, bu he quali y o he g aphene imp o es. In ano he sys ema ic s udy, again using Cu as he subs a e in APCVD he low a es and ela i e gas a io o CH4 o H2 we e explo ed and wo e y di e en g ow h windows a e iden i ied. Fo ela i ely high CH4 o H2 a ios, g aphene g ow h is ela i ely apid wi h an ini ial i s ull laye o ming in seconds. The ea e new g a- phene lakes nuclea e and hen g ow on op o he i s laye . Ad . Ma e . In e aces 2020, 7, 1902024 www.ad ancedsciencenews.com www.ad ma in e aces.de 1902024 (4 o 10) © 2020 The Au ho s. Published by WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim The s acking o hese lakes e sus he ini ial g aphene laye is mos ly u bos a ic. This g ow h mode can be likened o S anski−K as ano g ow h. While wi h ela i ely low CH4 o H2 a ios, he g ow h a es a e educed due o a lowe ca bon supply a e. In addi ion, bi-, i-, and ew-laye lakes o m di ec ly o e he Cu subs a e as indi idual islands. E ching s udies we e conduc ed and hey showed ha in his g ow h mode subsequen laye s o m benea h he i s laye p esum- ably h ough ca bon adical in e cala ion. This g ow h mode is simila o ha ound wi h Volme −Webe g ow h and p oduces highly o ien ed AB-s acked g aphene.[31] Figu e 2 shows exam- ples o he c oss pa ame ic s udies and he wo g ow h modes. 3.1.2. Liquid Me al Subs a es One o he disad an ages o me al subs a es o la ge a ea CVD g ow h o g aphene is he p esence o mic os uc u e de ec s such as g ain bounda ies. Due o di e ing C solubili y a subs a e de ec s (e.g., g ain bounda ies) du ing he g ow h p ocess, uni o m la ge a ea g aphene ilms a e challenging due o excess C p ecipi a ion du ing he cooling down p ocess. In addi ion, subs a e c ys al o ien a ion can a ec g aphene g ow h oo (p edominan ly h ough epi axial conside a ions). In he case o a liquid me al subs a e, hese aspec s can be sig- ni ican ly educed. Liquids end o adop a andom close packed geome y and enable a quasi-a omically smoo h su ace (wi h no g ain bounda ies) h ough he combined ac ion o su ace ension and he he mal mo ion o a oms.[32,33] In e ms o g aphene g ow h, a liquid su ace o e s a simple app oach o o m high quali y uni o m g aphene.[34–36] On liquid su aces he su aces, he he mal mo ion o he a oms means he in e - a omic dis ances luc ua e[37] and hus weaken he in e ac ion be ween he liquid (subs a e) su ace and he adso bed a oms (C species in he case o g aphene g ow h), hus dec easing he mig a ion ba ie ene gy.[38] This means he di usion a e o C a oms is g ea ly accele a ed allowing o he con olled g ow h o single c ys al g aphene islands wi h no el and change- able mo phologies.[39] Mo eo e , H species also mig a e e y e icien ly and so esul s in a unique e ching o de eloping g aphene islands du ing g ow h leading o in e es ing ac al e ching[40] and egula e ching.[41] An ea ly demons a ion was he use o liquid p-block elemen s (e.g., Ga) o he syn hesis o monolaye g aphene using APCVD.[42] The elec on mobili y o single c ys al domains g own on liquid Ga su aces was as high as 7400 cm2 V−1 s−1 unde ambien condi ions indica ing high quali y single c ys al g aphene lakes can be ob ained om his app oach. The echnique is ela i ely simple and does no equi e ilm deposi- ion o acuum sys ems. A la e sys ema ic s udy wi h liquid me als showed hey a e highly sui ed o s ic ly single laye g aphene. Examples a e shown in Figu e 3.[43] This is because du ing cooling om he CVD p ocess, he su ace me al solidi- ies quickly blocking he p ecipi a ion o abso bed ca bon. As a esul , g ow h is a sel -limi ed ca aly ic p ocess and mo eo e , is obus o a ia ions in g ow h pa ame e s. The heological su ace o a liquid me al su ace allows o he o a ion, alignmen and mo emen o g owing g aphene g ains (islands). Fo example, Fu and co-wo ke s[44] showed ha adjacen g aphene islands could me ge wi hou g ain bounda- ies ia a sel -adjus ing o a ion p ocess and hus assemble he c ys als in a supe o de ed and sel -aligned manne .[45] The weak a omic in e ac ion along wi h he high- apo p essu e o liquid me als allows o in e media y- ee g aphene ab ica ion[46] and esul s in e y clean g aphene which is a ac i e o i s p ac- ical applica ion.[36] 3.2. CVD o e Nonme al Subs a es As men ioned ea lie , in o de o a oid he need o ans e g aphene which can damage he g aphene as well as lea e unwan ed su ace con amina ion, he CVD syn hesis o g a- phene o e nonme al subs a es is also impo an , pa icula ly o elec onic de ice ab ica ion whe e i is c ucial o a oid me allic impu i ies and ans e con amina ion which will educe cos s and ime.[47] I is wo h no ing ha g aphene/ Si and g aphene/Ge appea o be p omising candida es o Ad . Ma e . In e aces 2020, 7, 1902024 Figu e 1. STEM in es iga ions o sample c oss-sec ions (lamellas). Fu he mos le : STEM images o sec ions o lamellas. Nex o hese a e elemen al EDS maps (in alse colo ). Rep oduced wi h pe mission.[20] Copy igh 2013, Ame ican Chemical Socie y. www.ad ancedsciencenews.com www.ad ma in e aces.de 1902024 (5 o 10) © 2020 The Au ho s. Published by WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim ansis o s because o he adjus able Scho ky ba ie , which o ms be ween g aphene and semiconduc o .[47] Ea ly wo k wi h nonme al subs a es showed he po en ial o a numbe o oxides, namely, SiO2, Al2O3, MgO,Ga2O3, and Z O as subs a es o g aphene g ow h.[48] The s udy sugges ed ha oxides do ha e a ca aly ic ole o play, and subsequen s udies a he low empe a u e o 325 °C con i med ha oxide base subs a es do p o ide some ca aly ic ole.[49] 3.2.1. Solid Nonme al Subs a es Mos CVD app oaches wi h nonme al subs a es use solid subs a es. One o he mo e impo an subs a es is Si wa e s (Si/SiOx). While a ious s udies ha e been conduc ed o achie e his di ec ly on he SiOx su ace, o achie e la ge a ea homogeneous monolaye g aphene is challenging. One app oach implemen ed he use o O2 o aid an APCVD p ocess which yielded polyc ys alline monolaye g aphene.[50] Ano he APCVD s udy (also wi h CH4 as he p ecu so ) ob ained ew- laye g aphene.[51] Wi h nea equilib ium CVD, esea che s ound single c ys al g aphene lakes wi h hexagonal and dodecagonal shapes o e he SiOx su ace.[52] Howe e , la ge a ea, homogeneous monolaye g aphene is pa icula ly di icul o ob ain. To o e come his issue, a con inemen echnique in which wo Si/SiOx wa e s wi h hei oxide aces in con ac in a sandwich con igu a ion was used o yield homogeneous single- laye la ge a ea g aphene.[53] The g aphene is polyc ys alline and he g ain bounda ies a e ace ed (see Figu e 4) indica ing u he imp o emen s a e needed o single-laye ma e ial, bu his is s ill a posi i e de elopmen . To educe he CVD syn hesis empe a u e, plasma- enhanced CVD (PECVD) can be used. A low empe a u e (550–650 °C) PECVD p ocess was demons a ed by a numbe o in es iga o s.[54–56] PECVD can also be used o he g ow h o g aphene nanowalls.[57] One s udy, showing g ow h Ad . Ma e . In e aces 2020, 7, 1902024 Figu e 2. Rela i e gas low windows used o in es iga e he dependence o g aphene g ow h modes wi h espec o he CH4 pa ial p essu e (es ima ed by he low a e o CH4 di ided by he o al low a e) e sus he o al gas low a) ( o al gas low includes CH4, H2, and a cons an low 1000 SCCM o A ). The small do s indica ed all he measu ed poin s. b–d) A se o SEM images showing SK-like bilaye g aphene g ow h co esponding o b−d) ed squa e spo s in (a). The g aphene lakes i egula in shape in he SK-like mode. e−g) A se o SEM images showing VW-like bilaye g aphene g ow h co esponding o e−g) blue iangle spo s in (a). The g aphene lakes a e egula in shape in he VW-like mode. All scale ba s a e 2 µm. Rep oduced wi h pe mission.[31] Copy igh 2016, Ame ican Chemical Socie y. www.ad ancedsciencenews.com www.ad ma in e aces.de 1902024 (6 o 10) © 2020 The Au ho s. Published by WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim empe a u es e en as low as 400 °C showed mic ome e -sized g aphene c ys als o ming he walls. G aphene can also be di ec ly g own on Al2O3 wa e s, al hough his ypically equi es e y high empe a u es o good quali y (poly-c ys alline) g aphene.[58] Al2O3 is a ac i e in ha i can se e as a de ice subs a e, iz., a de ice may be ab i- ca ed di ec ly on he syn hesized g aphene wi hou he need o in oke a ans e p ocess.[47] The Al si es on γ Al2O3 a e highly eac i e and ac as ca aly ic si es.[59,60] Song e al.[61] p esen ed an APCVD app oach using CH4 as he p ecu so in which hey could ob ain mono-laye g aphene. Ano he g oup employed PECVD and ob ained uni o m g aphene ilms.[54] TiO2 is ano he subs a e explo ed o i s po en ial o g aphene syn hesis by CVD. The di ec syn hesis o g aphene was i s dem- ons a ed o e -TiO2 a e we e ching ollowed by la ening o a omic smoo hness o e he (001), (110), and (100) aces.[62] The g aphene quali y was ound o be supe io o he (110) ace unde he same g ow h pa ame e s. Ano he g oup showed ha g aphene ab ica ion o TiO2 was possible o bo h APCVD and LPCVD o mono- and ew-laye o ma ion.[63] In he case o S TiO3 (STO) g aphene ilm o ma ion by CVD was shown h ough APCVD.[64] The po en ial o g aphene CVD g ow h a low empe a- u es on an oxide was demons a ed using MgO c ys als a he low empe a u e o 325 °C using ace ylene as he p ecu so .[49] Such low empe a u e app oaches a e ele an o main ain he mechanical in eg i y o low-dielec ic cons an (K) in e me al dielec ics in ansis o echnology. Ano he , low empe a u e g aphene ab ica ion, also on a high K dielec ic ma e ial was demons a ed by he same eam on Z O2, again wi h ace ylene and a a low empe a u e o 480 °C wi h ace ylene as he p ecu so .[65] Ano he g oup showed e ically g own g aphene shee s could o m on Z O2 h ough he mal CVD Ad . Ma e . In e aces 2020, 7, 1902024 Figu e 4. TEM cha ac e iza ions o he la ge-a ea syn he ic monolaye g aphene ( om sandwich con igu a ion). a) Low-magni ica ion mic og aph o a g aphene ilm ans e ed on o a holey ca bon TEM g id. b) SAED pa e n o he egion ci cled in (a). The inse p o ile shows he in ensi y p o ile o he di ac ion spo s. c,d) High- esolu ion TEM images showing he honeycomb a omic con igu a ion o g aphene. e) False-colo composi e mic og aph image highligh ing he di e en domain (g ain) o ien a ions and ace ed g ain bounda ies. Rep oduced wi h pe mission.[53] Copy igh 2017, Ame ican Chemical Socie y. Figu e 3. Typical g ow h esul s on liquid o solid me al subs a es. a−c) Op ical mic oscope images o g aphene g own on liquid Cu, In, and Ga, espec i ely, which demons a es he excellen uni o mi y o he single-laye g aphene. d) Op ical mic oscope image o g aphene g own on solid Cu oil unde ambien p essu e, which indica es poo uni o mi y. All he g aphene ilms we e ans e ed on o 300 nm SiO2/Si subs a es o cha ac e iza ion. The scale ba s a e 10 µm. e−h) Laye dis ibu ion de e mined by RGB colo analysis o he co esponding op ical mic oscope images (a−d), espec i ely. The laye hickness is ep esen ed by di e en colo s as seen om he bo om colo codes. Rep oduced wi h pe mission.[43] Copy igh 2014, Ame ican Chemical Socie y. www.ad ancedsciencenews.com www.ad ma in e aces.de 1902024 (7 o 10) © 2020 The Au ho s. Published by WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim using me hane o e hanol as he p ecu so .[66] This wo k was also in e es ing because ypically PECVD is equi ed o e ical g aphene g ow h. Ano he echnologically impo an subs a e o g aphene g ow h is silicon ni ide (Si3N4). This was accom- plished wi h a wo-s age CVD p ocess. The g aphene shee s a e seen o o m independen ly and hen la e me ge o o m a polyc ys alline ilm.[67] G aphene can also o m on aluminum ni ide. Di ec g ow h o g aphene was shown on ano he ni ide, namely, AlN (AlN/Si (111)). In his case, p opane se ed as he p ecu so and, as ound wi h a numbe o non-me al ca alys s, empe a u e is ele an in e ms o he g aphene quali y, such ha i imp o es as he empe a u e inc eases. In his case, 1350 °C was needed o high quali y g aphene.[68] G aphene can also be g own on a numbe o adi ional glasses and his could be impo an o a numbe o daily li e applica ions. An APCVD app oach demons a ed o he i s ime he di ec well-con olled g ow h o high quali y g aphene on insula ing solid glasses was possible. Mo eo e , he laye hickness could also be uned.[69] PECVD can also be employed o g ow g aphene ilms di ec ly o e qua z (see Figu e 5). Nanog aphene ilms wi h good uni o mi y we e g own o e 4 in. wa e s.[54] Ve ical g aphene g ow h o e qua z by CVD has also been shown o e qua z subs a es[66] and sugges s i is a a he e sa ile subs a e o g aphene o ma ion. In he case o e ical g aphene, his can also occu o e ca bon bu e laye s.[70] Ad . Ma e . In e aces 2020, 7, 1902024 Figu e 5. Ca alys - ee APCVD g ow h o uni o m g aphene on a ious solid glasses. a) Schema ic diag am o he ca alys - ee APCVD g ow h me hod. b) Pho og aph o he bo osilica e glass subs a es be o e (le mos ) and a e g aphene g ow h wi h di e en CH4 low a es a 2, 5, 7.5, and 10 sccm. c) Demons a ion o he hyd ophobic and hyd ophilic na u e o g aphene/qua z glass ( he le pa ) and he ba e qua z glass, espec i ely. d) Pho og aph showing he di e ences in wa e con aining beha io s be ween he g aphene-coa ed and p is ine qua z es ubes. e) Pho og aph o g aphene/sapphi e glass pla e displaying a good anspa ency. Scale ba : 4 cm. ) Rep esen a i e Raman spec a o di ec ly g own g aphene on di e en ypes o solid glasses. g) T ans e cu e o he g aphene FET; he inse shows an op ical mic oscope image o an indi idual de ice. The g aphene was g own on qua z glass unde he ollowing APCVD condi ion: A /H2/CH4, 100/50/8 sccm a 1020 °C o 3 h. Scale ba : 100 µm. h) Shee esis ance and UV– is ansmi ance spec a in he wa eleng h ange o 350−800 nm o he g aphene/qua z glass. Rep oduced wi h pe mission.[69] Copy igh 2015, Ame ican Chemical Socie y. www.ad ancedsciencenews.com www.ad ma in e aces.de 1902024 (8 o 10) © 2020 The Au ho s. Published by WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim 3.3. Mol en Nonme al Subs a es I is belie ed ha by using a mol en liquid insula ing subs a e, he uni o m nuclea ion and accele a ed g ow h o g aphene can occu . This was i s shown on soda-lime glass (which has a low so ening poin o 620 °C). Soda lime glass is cheap and inexpensi e and is used in a ious applica ions.[71] In his case an APCVD app oach was implemen ed. The appli- ca ion po en ial o he as-p oduced g aphene on soda-lime glass was also demons a ed o sma hea ing-de ices such as anspa en de ogge s and he moch omic displays, and o biocompa ible cell cul u e mediums. The same eam ex ended he g ow h o uni o m g aphene (o e soda-lime glass) o show la ge a ea g ow h o e 12 in. a eas.[72] 3.4. Flexible Nonme allic Subs a es As o igid nonme al subs a es, a key ad an age o g owing g aphene di ec ly on lexible subs a es is he a oidance o ans e induced damage (e.g., c acks, ea s, and w inkles). Mo eo e , he di ec g ow h o g aphene o e lexible subs a es has signi ican p omise o lexible and s e chable elec onics. Examples include e-skin and heal h moni o ing on humans.[73–75] Howe e , he e a e cons ain s wi h some o hese ma e ials due o low empe a u es being a p e equisi e o a oid mel ing, de o ming o damaging he subs a e (e.g., wi h polyimide subs a es). Fo some subs a es, his is no so, e.g., mica and hBN, which we i s look a . Nanog aphene ilms we e g own on mica using a PECVD app oach wi h CH4 as he p ecu so a he ela i ely low empe a u e o 525 °C.[54] The e is conside - able in e es in g aphene g ow h o e hBN because i s la ice pa ame e is he same as o g aphene.[8,76] The CVD g ow h o g aphene o e hBN as shown o hBN ini ially g own o e a Cu subs a e (and he Cu base subs a e emained o he CVD g ow h o g aphene, so s ic ly he subs a e was hBN/Cu).[77] Ding e al.[78] de eloped a me al ee CVD app oach o ew-laye g aphene.[77] Simila ly, Liu e al.[79] ob ained ew-laye g aphene o e hBN using cyclohexane as he p ecu so using an APCVD app oach. Ano he app oach is o g ow g aphene di ec ly o e hBN using seed-assis ed g ow h. The use o he well-de ined poly(me hyl me hac yla e) seeds enabled e ec i e con ol o e he nuclea ion densi ies and loca ions o he g aphene domains on p edeposi ed h-BN monolaye s. This allowed he o ma ion o pa e ned G/h-BN a ays o con inuous ilms.[80] Fo empe a u e sensi i e lexible subs a es such polyimide (PI) and polydime hylsiloxane (PDMS) me al capping laye s a e used o p o ide a ca alys -assis ed CVD sys em and hus enable low empe a u es o be used. Cu deposi ed on PI allowed a low empe a u e PECVD ou e a 300 °C.[81] While o PDMS Ni was deposi ed o e he base PDMS.[7] 4. Ou look Fo g aphene (and o he an de Waals 2D ma e ials) o ul ill hei p omise in applica ions, i will be c ucial ha hey can be syn hesized unde app op ia e condi ions, such as, wi hin ce ain empe a u e windows, and ha his, in many cases, can be achie ed di ec ly on he su ace o he ma e ial in ques- ion, iz., on any subs a e. This is a g and ask. CVD is also a well-es ablished echnique and is e sa ile. As has been dem- ons a ed in his e iew, esea che s ha e been making signi i- can s ides o de elop he me hod o g aphene syn hesis o e a b oad ange o subs a es unde a a ie y o condi ions and, mo eo e , ha e been pushing he bounda ies o achie e his on a ema kably low empe a u es. S ill he e a e challenges. As was highligh ed, he di ec syn hesis o g aphene on polyme s, which will be c ucial o lexible based de ices, is an example whe e imp o ed CVD app oaches a e needed. This is likely an a ea whe e impo an de elopmen s will be seen in he nea u u e. In addi ion, di ec ly g own pa e ned g aphene, o example o de ices, will become e e mo e ele an and i is concei able ha CVD me ged wi h ano he echnique could be a key playe . One could imagine CVD eac ions ope a ing sligh ly below he nuclea ion and g ow h h esholds in combina ion, wi h, o example, an elec on beam o lase beam ha can hen o e come he h esholds a speci ic local egions and in his way yield pa e ned g aphene g ow h. This same app oach could be applied o o he 2D ma e ials and pa e he way o pa e ned he e o 2D ma e ials ab ica ions bo h in s acked and la e ally s i ched con igu a ions. Ano he impo an aspec o g aphene g ow h likely o be seen in he u u e is la ge a ea and wa e scale single c ys al g ow h o g aphene. This emains chal- lenging, bu he pace o change in his ield sugges s b eak- h oughs will eme ge soon, and a key aspec o his is likely o be ela ed o a mix o he igh CVD eac ion/ eac o con igu- a ions and cle e choice/implemen a ion o subs a e. Ano he aspec one can an icipa e in he u u e is high speed g ow h o la ge a ea single c ys al g aphene by CVD. Mo eo e , u u e de elopmen s wi h subs a e use in CVD will minimize o e en emo e he need o ans e and hus, enable he bes g aphene pe o mance o be ex ac ed in which e e applica ion i is used. 5. Conclusion The changes seen in he g ow h o g aphene by CVD ha e been apid since he i s e idence o single-laye g aphene. CVD is a echnologically es ablished echnique, and is ela i ely e sa ile and, is highly success ul a p oducing g aphene. Al hough ini- ially me al subs a es we e equi ed, in ecen yea s his is no longe ue and nume ous dielec ic subs a es can be used o he success ul g ow h o g aphene by CVD. Mo eo e , signi i- can in oads a e being made in e ms lexible subs a es which will be a key echnology ield in he nea u u e. In addi ion, i is likely ha he pa e ned di ec g ow h o g aphene and la ge a ea single c ys al g aphene will be demons a ed soon. In all hese de elopmen s (pas , p esen , and u u e), subs a e a ailabili y is c ucial o he applica ions success no only o g a- phene, bu also o he 2D ma e ials and hei combina ions o yield he e o 2D sys ems. Acknowledgemen s Q.S. and K.T. con ibu ed equally o his wo k. This wo k was suppo ed by he Na ional Science Founda ion China (NSFC, P ojec 51672181), Ad . Ma e . In e aces 2020, 7, 1902024 www.ad ancedsciencenews.com www.ad ma in e aces.de 1902024 (9 o 10) © 2020 The Au ho s. Published by WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim he Na ional Science Cen e , Poland o he inancial suppo wi hin he ame o he Opus p og am (G an ag eemen 2015/19/B/ST5/03399), he Czech Republic om he ERDF “Ins i u e o En i onmen al Technology - Excellen Resea ch” (No. CZ.02.1.01/0.0/0.0/16_019/00 00853). M.H.R. and L.F. hank he Sino-Ge man Resea ch Ins i u e o suppo (p ojec : GZ 1400). Con lic o In e es The au ho s decla e no con lic o in e es . Keywo ds chemical apo deposi ion, g aphene, subs a e, suppo , syn hesis Recei ed: No embe 29, 2019 Re ised: Janua y 12, 2020 Published online: Feb ua y 16, 2020 [1] J. H. Wa ne , F. Schs el, A. Bachma iuk, M. H. Rümmeli, G aphene, Else ie , UK 2013. [2] R. S. Edwa ds, K. S. Coleman, Nanoscale 2013, 5, 38. [3] P. A ou is, C. Dimi akopoulos, Ma e . Today 2012, 15, 3. [4] W. Feng P. Long, Y. Feng, Y. Li, Ad . Sci. 2016, 3, 1500413. [5] Y. Zhang, L. Zhang, C. Zhou, Acc. Chem. Res. 2013, 46, 2329. [6] M. J. Allen, V. C. Tung, R. B. Kane , Chem. Re . 2010, 110, 132. [7] G. Li, B. Huang, Z. Pan, X. Su, Z. S. , L. An, Ene gy En i on. Sci. 2019, 12, 2030. [8] B. Aïssa, N. K. Memon, A. Ali, M. K. Kh aisheh, F on . Ma e . 2015, 2, 58. [9] H. Wang, T. Maiyalagan, X. Wang, ACS Ca al. 2012, 2, 781. [10] P. Russ, A. Hu, G. Compagnini, Nano-Mic o Le . 2013, 5, 260. [11] V. P. Pham, Di ec G ow h o G aphene on Flexible Subs a es owa d Flexible Elec onics: A P omising Pe spec i e, In echOpen 2018, h ps://doi.o g/10.5772/in echopen.73171. [12] J. E. Kim, D. H. Lee, T. H. Han, J. W. Hwang, S. Jeon, S.-Y. Choi, S. H. Hong, W. J. Lee, R. S. Ruo , S. O. Kim, Ad . Ma e . 2010, 22, 1247. [13] D. W. Hess, K. F. Jensen, T. J. Ande son, Chemical Vapo Deposi ion: A Chemical Enginee ing Pe spec i e, Vol. 3, Wal e de G uy e GmbH, Ge many 2011. [14] T. B. Massalski, H. Okamo o, P. R. Sub amanian, L. Kacp zak, ASM Handbook: Alloy Phase Diag ams, Vol. 3, ASM In e na ional, Ma e ials Pa k, OH 2002. [15] Q. Yu, J. Lian, S. Si ipongle , H. Li, Y. Chen, S.-S. Pei, Appl. Phys. Le . 2008, 93, 113103. [16] L. G. De A co, Y. Zhang, A. Kuma , C. Zhou, IEEE T ans. Nano echnol. 2009, 8, 135. [17] A. Reina, X. Jia, J. Ho, D. Nezich, H. Son, V. Bulo ic, M. S. D esselhaus, J. Kong, Nano Le . 2009, 9, 30. [18] A. J. Polla d, R. R. Nai , S. N. Sabki, C. R. S addon, L. M. A. Pe digao, C. H. Hsu, J. M. Ga i , S. Gangopadhyay, H. F. Gleeson, A. K. Geim, P. H. Be on, J. Phys. Chem. C 2009, 113, 16565. [19] B. Dai, L. Fu, Z. Zou, M. Wang, H. Xu, S. Wang, Z. Liu, Na . Commun. 2011, 2, 522. [20] M. H. Rümmeli, M. Zeng, S. Melkhano a, S. Go an la, A. Bachma iuk, L. Fu, C. Yan, S. Oswald, R. G. Mendes, D. Maka o , O. Schmid , J. Ecke , Chem. Ma e . 2013, 25, 3880. [21] Y. Zhang, L. Gomez, F. N. Ishikawa, A. Mada ia, K. Ryu, C. Wang, A. Badmae , C. Zhou, Chem. Le . 2010, 1, 3101. [22] P. W. Su e , J.-I. Flege, E. A. Su e , Na . Ma e . 2008, 7, 406. [23] J. Co aux, A. T. N′Diaye, C. Busse, T. Michely, Nano Le . 2008, 8, 565. [24] P. Su e , J. T. Sadowski, E. Su e , Phys. Re . B 2009, 80, 245411. [25] A. Va ykhalo , O. Rade , Phys. Re . B 2009, 80, 035437. [26] S. M. Wang, Y. H. Pei, X. Wang, H. Wang, Q. N. Meng, H. W. Tian, X. L. Zheng, W. T. Zheng, Y. C. Liu, J. Phys. D: Appl. Phys. 2010, 43, 455402. [27] S.-Y. Kwon, C. V. Ciobanu, V. Pe o a, V. B. Shenoy, J. Ba en, V. Gambin, I. Pe o , S. Kodambaka, Nano Le . 2009, 9, 3985. [28] E. Miniussi, M. Pozzo, A. Ba aldi, E. Vesselli, R. R. Zhan, G. Comelli, T. O. Men es, M. A. Nin, A. Loca elli, S. Lizzi , D. Al e, Phys. Re . Le . 2011, 106, 216101. [29] X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, R. Pine , A. Velamakanni, I. Jung, E. Tu uc, S. K. Bane jee, L. Colombo, R. S. Ruo , Science 2009, 324, 1312. [30] M. H. Rümmeli, S. Go an la, A. Bachma iuk, J. Phiele , N. Geißle , I. Ib ahim, J. Pang, J. Ecke , Chem. Ma e . 2013, 25, 4861. [31] H. Q. Ta, D. J. Pe ello, D. L. Duong, G. H. Han, S. Go an la, V. L. Nguyen, A. Bachma iuk, S. V. Ro kin, Y. H. Lee, M. H. Rümmeli, Nano Le . 2016, 16, 6403. [32] G. K esse, J. Ha ne , Phys. Re . B 1993, 47, 558. [33] J. L. Finney, P oc. R. Soc. A 1970, 319, 479. [34] H.-B. Li, A. J. Page, C. He ich, B. A adi, C. Köhle , T. F auenheim, S. I le, K. Mo okuma, Chem. Sci. 2014, 5, 3493. [35] D. Geng, B. Wu, Y. Guo, L. Huang, Y. Xue, J. Chen, G. Yu, L. Jiang, W. Hu, Y. Liu, P oc. Na l. Acad. Sci. USA 2012, 109, 7992. [36] J. Liu, L. Fu, Ad . Ma e . 2019, 31, 1800690. [37] J. Chen, X. Zhao, S. J. R. Tan, H. Xu, B. Wu, B. Liu, D. Fu, W. Fu, D. Geng, Y. Liu, W. Liu, W. Tang, L. Li, W. Zhou, T. C. Sum, K. P. Loh, J. Am. Chem. Soc. 2017, 139, 1073. [38] M. Einax, W. Die e ich, P. Maass, Re . Mod. Phys. 2013, 85, 921. [39] B. Wu, D. Geng, Z. Xu, Y. Guo, L. Huang, Y. Xue, J. Chen, G. Yu, Y. Liu, NPG Asia Ma e . 2013, 5, e36. [40] D. Geng, B. Wu, Y. Guo, B. Luo, Y. Xue, J. Chen, G. Yu, Y. Liu, J. Am. Chem. Soc. 2013, 135, 6431. [41] D. Geng, H. Wang, Y. Wan, Z. Xu, B. Luo, J. Xu, G. Yu, Ad . Ma e . 2015, 27, 4195. [42] J. Wang, M. Zeng, L. Tan, S. Wang, L. Peng, J. Ecke , B. Dai, Y. Deng, M. H. Rümmeli, L. Fu, Sci. Rep. 2013, 3, 2670. [43] M. Zeng, L. Tan, J. Wang, L. Chen, M. H. Rümmeli, L. Fu, Chem. Ma e . 2014, 26, 3637. [44] M. Zeng, L. Tan, L. Wang, R. G. Mendes, Z. Qin, Y. Huang, T. Zhang, L. Fang, Y. Zhang, S. Yue, M. H. Rümmeli, L. Peng, Z. Liu, S. Chen, L. Fu, ACS Nano 2016, 10, 7189. [45] M. Zeng, L. Wang, J. Liu, T. Zhang, H. Xue, Y. Xiao, Z. Qin, L. Fu, J. Am. Chem. Soc. 2016, 138, 7812. [46] W. Lu, M. Zeng, X. Li, J. Wang, L. Tan, M. Shao, J. Han, S. Wang, S. Yue, T. Zhang, X. Hu, R. G. Mendes, M. H. Rümmeli, L. Peng, Z. Liu, L. Fu, Ad . Sci. 2016, 3, 1600006. [47] A. Khan, S. M. Islam, S. Ahmed, R. R. Kuma , M. R. Habib, K. Huang, M. Hu, X. Yu, D. Yang, Ad . Sci. 2018, 5, 1800050. [48] M. H. Rümmeli, C. K ambe ge , A. G üneis, P. Ayala, T. Gemming, B. Büchne , T. Pichle , Chem. Ma e . 2007, 19, 4105. [49] M. H. Rümmeli, A. Bachma iuk, A. Sco , V. Ho man, J.-H. Lin, G. Cunibe i, F Bö ne , J. H. Wa ne , B. Büchne , ACS Nano 2010, 4, 4206. [50] J. Chen, Y. Wen, Y. Guo, B. Wu, L. Huang, Y. Xue, D. Geng, D. Wang, G. Yu, Y. Liu, J. Am. Chem. Soc. 2011, 133, 17548. [51] H. Bi, S. Sun, F. Huang, X. Xie, M. Jiang, J. Ma e . Chem. 2012, 22, 411. [52] J. Chen, Y. Guo, L. Jiang, Z. Xu, L. Huang, Y. Xue, D. Geng, B. Wu, W. Hu, G. Yu, Y. Liu, Ad . Ma e . 2014, 26, 1348. [53] J. Pang, R. G. Mendes, P. S. W obel, M. D. Wloda ski, H. Q. Ta, L. Zhao, L. Giebele , B. T zebicka, T. Gemming, L. Fu, Z. Liu, J. Ecke , A. Bachma iuk, M. H. Rümmeli, ACS Nano 2017, 11, 1946. Ad . Ma e . In e aces 2020, 7, 1902024