Role of In-flight Temperature and Velocity of Powder Particles on Plasma Sprayed Hydroxyapatite Coating Characteristics

Role o In-Fligh Tempe a u e and Veloci y o Powde Pa icles on Plasma Sp ayed Hyd oxyapa i e Coa ing Cha ac e is ics Jan Cizeka,∗ , Khiam Aik Kho b aIns i u e o Ma e ials Science and Enginee ing Facul y o Mechanical Enginee ing B no Uni e si y o Technology, Czech Republic bSchool o Mechanical and Ae ospace Enginee ing Nanyang Technological Uni e si y, Singapo e © 2012. This manusc ip e sion is made a ailable unde he CC-BY-NC-ND 4.0 license (h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/) doi:10.1016/j.su coa .2011.09.058 Final e sion a ailable om h p://www.sciencedi ec .com/science/a icle/pii/S0257897211009455 Abs ac This pape p esen s a sys ema ic esea ch on he p ocess o he mal sp aying o HA encompassing all s ages o laye deposi ion: powde p oduc ion & cha ac e iza ion (op imized p oduc ion led o sphe ical 39.90 ±10.61 µm powde wi h 0.0% con en o i-calcium phospha e [TCP] o e a-calcium phospha e [TTCP] phases), plasma je p ope ies in luence on he in- ligh powde p ope ies (majo in luence o sp ay dis ance ac o ), he in luence o he in- ligh empe a u e and eloci y o hyd oxyapa i e pow- de pa icles ( anges o 2294 K–2708 K and 152 m·s−1– 291 m·s−1) on he inal cha ac e is ics o he deposi ed coa ings. Six combina ions o he sys em pa ame e s leading he low-medium-high in- ligh p ope ies we e selec ed and he espec i e coa ings we e in es iga ed, wi h some da a in con adic ion o p e iously published esul s. I was ound ha he empe a u e o he pa icles plays a c i ical ole o he de elopmen o de imen al CaO (con en inc ease o up o 14.6%) and me as able TTCP (up o 49.5%) phases in he coa ings. Pa icles in- ligh eloci y was ound o in luence he open po osi y o he coa ings (8.8% – 27.9%) and ha e an impac on he mic o-ha dness and moduli o he coa ings ( ela i e di e ences o up o 2.8×). Join incidence o bo h in- ligh p ope ies was ound o signi ican ly in luence he mic os uc u e o he coa ings and i s espec i e su ace oughness le els (Ra= 7.4 – 19.4µm). The mo phology o impinging spla s was ound o be in luenced by bo h he in- ligh eloci y (dominan ac o ) and empe a u e (seconda y ac o ) o he HA pa icles. Keywo ds: plasma sp ay, in- ligh p ope ies, coa ing cha ac e iza ion, hyd oxyapa i e 1. In oduc ion A adi ionally used calcium phospha e ce amics, hyd oxyapa i e has been well p o en du ing he in ensi e esea ch o i s bioac i e p ope ies [1]. Gene ally indica ed as Ca10(PO4)6(OH)2, c ys alline and amo phous HA o ms a majo componen o mammal bones [2] and is he e o e a sough - o compound o e.g. den al and o hopaedic applica ions. Howe e , mechanical p ope ies o HA do no su ice in load-bea ing applica ions. Deposi ion o HA laye s on o me al componen s (e.g. hip join eplacemen s) is he e o e ca ied ou [3]. Bioac i e HA coa ings on me allic implan s ha e been epo ed o [4]: ∗Co esponding au ho Email add ess: [email p o ec ed] (Jan Cizek) P ep in submi ed o Else ie Augus 18, 2015  p omo e ea lie s abiliza ion o he implan in he su ounding bone, suppo ing he bone in-g ow h and imp o ing he implan -bone bond s eng h  dec ease he elease o me allic ions om he implan , educing he possibili y o in oxica ion in ascula sys em and subsequen heal h haza ds  shield he su ace o he p os hesis subs a e om en i onmen al a acks, acili a ing long- e m unc ion pe o mance and lowe ing he isk o subs a e ailu e in load bea ing applica ions  ex end he unc ional li e o an implan as compa ed o cemen ed implan s Nume ous issues ega ding he deposi ion o HA laye s on o me allic subs a e ma e ials ha e a isen (e.g. adhesion o he laye s and p ospec i e delamina ion, coa ings ac u es, changes in phase composi ion due o su ace ea men echnology, selec i e e apo a ion o componen s, he mal shocks o subs a e ma e ials, e c.). Many o he issues we e add essed and e o s o he esea ch s udies ocused on a ious aspec s o he p ocess. P ocedu e o pu e bio-g ade HA powde p epa a ion has been published in e.g. [5] and op imiza ion o i s p ope ies (calcina ion a empe a u es 1100–1300 K [6], mo phology con ol by sp aying in o wa e /ni ogen, amo phous con en de e mina ion [7]) we e desc ibed. The espec i e phenomena o HA powde cha ac e is ics and p ope ies changes du ing he sp ay p ocedu e ha e been in es iga ed wi h espec o he eeding o he powde (in e nal s. ex e nal eeding, HA suspension eeding [8]) and su ounding con olled a mosphe e [9]. The sp ay p ocedu e has also been assessed by nume ical simula ions. The a ious s udies in ol ed compu ing using he plasma je en halpy, pa icle mo phologies and a e age size, su ounding a mosphe e, momen um ans e s due o a ious wo king gases and inal pa icle ajec o ies wi hin he plume and impac upon subs a e (e.g. [10]). Rela ed o he heo e ical s udies, a s udy by Dyshlo enko e al. [11] ocused on he in luence o ou key sp ay pa ame e s o he sys em on o he cha ac e is ics o HA powde and coa ings sp ayed on o subs a es and in o wa e using wo di e en sp ay dis ances. The main ocus o he s udies ha e been se o he physical, mechanical and chemical p ope ies o he deposi ed HA coa ings [12], hei pos -sp ay ea men (e.g. lase annealing) and biological s udies o he in- i o (usually SBF solu ions) and in- i o esponses o he HA-coa ed specimens [13]. Wi h espec o he biological ole o HA, pu i y and c ys allini y o he laye s a e o u mos impo ance. Syn he ically p oduced HA usually e ains highe c ys allini y a io han HA in biological issues [14]. The a io is de e mined by he p oduc ion p ocess and p ospec i e hea ea men (mainly annealing empe a u e and ime). High le els o c ys allini y o syn he ic HA a e bene icial o long- e m su i abili y and p ope unc ional li e o implan in se ice while high eso babili y o he amo phous phase and a p esence o me as able phases a e use ul o accele a ed implan ixa ion and p omo ion o as e bone emodeling [15]. Among he indus y me hods, plasma sp ay echnology has been widely used owing o i s e sa ili y and he wide ange o applicable ma e ials. Using his me hod, he deposi ion o unc ion pe o ming, p o ec i e o biocompa ible coa ings is e icien , he p ocess e ains high deposi ion a es and he a ge coa ing hickness could be easily con olled [16]. The plasma sp ay p ocess is basically a deposi ion o mol en o semi-mol en ma e ial on o p epa ed subs a es o o m a p o ec i e o unc ion-pe o ming laye . In he p ocess, h ee majo phases could be dis inguished: a) he plasma je gene a ion, b) in e ac ion o he powde pa icles wi h he je du ing ma e ial eeding phase and c) he coa ing build-up on he p epa ed subs a e. The plasma gene a ion p ocess and he plasma je p ope ies is in luenced by a numbe o ac o s. These could be di ided in o wo g oups: pa ially uncon ollable, o en en i onmen -bound and he ac o s selec ed and al e ed in he sys em. The la e pa ame e s in luence e.g. he empe a u e space dis ibu ion wi hin he je o he low o he wo king gases ( hus he momen um o be ans e ed o he pa icles). The in e ac ion o he pa icles du ing he eeding phase is d i en by e.g. he pa icle size dis ibu ion, he momen um o pa icles when en e ing he je (de e mined by he ca ie gas low) o he pa icle dwelling ime wi hin he ho plasma egion [16]. The inal coa ing build-up is d i en by he chemical and physical p ope ies o he subs a e ( o he 1s laye ) and he he modynamic and chemical p ope ies o he deposi ed ma e ials (successi e laye s). 2 Encompassing all he h ee majo phases (a-c), he aim o his s udy was o in es iga e he in lu- ence o he wo mos impo an in- ligh ac o s ( empe a u e, eloci y) o he hyd oxyapa i e powde (“HA”) on selec ed p ope ies o he inal coa ings. Six pa ame e s o he plasma sp ay sys em we e selec ed o ep esen he a iabili y in he phase o plasma je gene a ion [6]. Eigh een combina ions o hese pa ame e s we e in es iga ed acco ding o Taguchi expe imen al design and he espec i e in- ligh p ope ies we e eco ded. Subsequen ly, six combina ions o he sp ay sys em pa ame e s leading o low/medium/high in- ligh p ope ies ( empe a u e, eloci y) o HA pa icles we e selec ed o coa ings deposi ion. Cha ac e iza ion o he deposi ed laye s was hen ca ied ou conside ing he powde in- ligh p ope ies. Wi h espec o he p ospec i e biomedical applica ion o he HA coa ings, mic os uc u e, phase composi ion and po osi y o he coa ings we e hen s udied as well as i s su ace oughness, elas ic moduli and mic oha dness. To aid in ou unde s anding o he deposi ion p ocess (namely he 1s laye phenomenon), single spla collec ion expe imen was ca ied ou . 2. Expe imen al Se up 2.1. Powde Feeds ock In his s udy, we chemical syn hesis p ecipi a ion eac ion (1) was used o p epa e he HA powde . To acili a e he p ecipi a ion, he phospho ic acid was added in a d op-wise manne unde con olled empe a u e o 320 K and he p ocess was e mina ed when pH d opped o 8.9 [17]. 10 Ca(OH)2+ 6 H3(PO4)−−→ Ca10(PO4)6(OH)2+ 18 H2O (1) The p oduced sa u a ed HA-wa e slu y could con ain me as able phases, such as TCP o TTCP. Howe e , hese phases a e suscep ible o con inuous dissolu ion in a non-sa u a ed aqueous solu ion as compa ed o mo e s able HA phase and p ecipi a ion o addi ional HA is likely o occu du ing his p ocess [18]. The e o e, o u he mo e aid in e aining non-con amina ed HA, he slu y was le o se le o 48 hou s, leading o sepa a ion o wa e and he HA p ecipi a e. The deionized wa e was hen eplaced and mixed in o he slu y again. This p ocess was epea ed i e imes. The slu y was hen sp ay d ied using o a y a omize nozzle unde p e iously op imized pa ame e s (10 Hz a omize e olu ion equency, 510 K chambe empe a u e, ∼40 g·min−1slu y eeding). The powde p oduced unde hese condi ions demons a ed he highes yield and a e age pa icle sizes sui able o plasma sp aying while e aining a ela i ely low in insic po osi y. As he in luence o he powde pa icle size on he p ope ies was no in es iga ed in his s udy, i was necessa y o minimize i s po en ial e ec . P ope sie ing esul ed in ob aining a na ow dis ibu ion o he pa icle diame e s, measu ed as 39.90 ±10.61 µm by liquid suspension lase analysis. The powde was calcined a 1220 K o 120 mins o u he aid in he pu i ica ion p ocess and o inc ease he coalescence o he nanome ic size c ys alli es wi hin he pa icles [5], [19]. SEM in es iga ion o he HA powde (Fig. 1) e ealed a sphe ical na u e o pa icles sui able o plasma deposi ion. Powde c oss-sec ion analysis indica ed a po ous s uc u e o he powde . The in insic po osi y was speci ied as 8.0% by helium pycnome ic me hod. The hea - ea men o he powde esul ed in inc eased c ys allini y o he HA powde . This was con i med by na owing o he HA peaks and hei ela i e heigh inc ease in he measu ed XRD pa e n (Fig. 2). Fo compa ison pu poses, a heo e ical calcula ed p edic ion o pu e HA pa e n is included (HA ICSD #26205 s anda dized pa e n was used in ATOMS so wa e o calcula e he hkl planes, dspacing alues, symme y and space g oup. The calcula ed in o ma ion was hen used in comme cial Powde Cell so wa e o he spec a calcula ion). Rie eld analysis [20] was hen used o de e mina ion o he powde phase con en (Table 1). I could be seen ha he HA powde con ained a 5% amoun o calcium ca bona e phase and no aces o TCP and TTCP phases we e de ec ed. The de ec ed CaCO3could ha e o igina ed in he ma e ial by a eac ion o he o iginal Ca(OH)2wi h a mosphe ic ca bon dioxide (2) du ing he long- unning syn hesis. Wa e , he only by-p oduc o his eac ion, was hen emo ed du ing he subsequen sp ay d ying p ocess, which CaCO3unde ook oge he wi h he in oduced HA powde . Ca(OH)2+ CO2−−→ CaCO3+ H2O (2) 3 2.2. Plasma Sys em In his s udy, P axai plasma sys em (P axai Inc., Danbu y, USA) equipped wi h P axai SG-100 gun was used. Pu i ied a gon was used as a main a c o ming gas and ca ie gas and helium was used as an auxilia y gas. The a iabili y o he sys em was ep esen ed by six selec ed ac o s: ne powe , low o he main and auxilia y gases, e olu ions o he powde eede wheel, low o he ca ie gas and sp ay dis ance, i.e. he dis ance be ween he gun and he a ge (Table 2). In he sys em, he six ac o s ep esen he mos obus pa ame e s in luencing he p ope ies o he plasma je and he subsequen in e ac ion o he powde du ing i s pene a ion in o he plume. The ne powe [kW] unc ion egula es he closed loop con ol unc ion o he plasma ene gy by changing he DC powe inpu ins an ly. I eac s o he changing physical condi ions, au oma ically adjus ing he sp ay pa ame e s o main ain a cons an powde hea ing egion. Gene ally, inc ease in ne powe gi es ise o je s o highe empe a u e and eloci y. The ela i e inc ease o he wo espec i e p ope ies is impo an : highe plasma je empe a u es ans e mo e hea o he powde pa icles, howe e , he highe eloci ies esul in sho e dwelling imes in he je and, subsequen ly, educed hea ans e . Main a c gas low [l·min−1] is an impo an ac o in luencing bo h he plasma je empe a u e and eloci y. Highe main gas low is expec ed o gi e ise o highe eloci ies o he pa icles. Main gas is pa ially used o cooling he SG-100 gun. Auxilia y a c gas low [l·min−1], used o addi ional s abiliza ion o he plasma a c and ex ended hea ans e , is usually lowe han he low o main gas and is he e o e expec ed o ha e a lesse impac on he in- ligh pa icle p ope ies. The ca ie gas low [l·min−1] is a c i ical ac o o de e mining he ajec o ies o he pa icles wi hing he plasma je . Too low ca ie gas low could esul in pa icle bouncing o he je a he han pene a ing i while oo high alue o ca ie gas could cause he pa icles o in e sec he plume and a el on he o he pe iphe y. Also, he low o a ca ie gas con ibu es o he o al low o gases in he plasma gun o i ice. Linked o he ca ie gas low pa ame e , powde hoppe e olu ions pe minu e [1] in luence he amoun o powde en e ing he plasma je . An inc ease in he amoun o powde en e ing he je could lead o dec ease in a e age empe a u e and eloci y o he pa icles, espec i ely (iden ical amoun o hea and momen um dis ibu ed o a highe numbe o pa icles). Las ac o selec ed o in es iga ion is no di ec ly se by he sys em con ol module. Howe e , due o a p esence o s eep space g adien s in empe a u e and eloci y o he je , sp ay dis ance [mm] is one o he key ac o s in luencing he powde pa icles dwelling ime in he je and de e mines he amoun o hea ing and accele a ion he pa icles unde go. 2.3. In- ligh P ope ies 2.3.1. Taguchi design o Expe imen The ull- ac o ial design o in- ligh p ope ies measu emen expe imen o he selec ed six ( h ee- le el) ac o s would en ail 36= 729 measu emen s. Fu he mo e, such p ocessing does no ake in o conside a ion he po en ial in e ac ions be ween he ac o s, a phenomena likely o occu du ing plasma sp ay p ocess (e.g. a mu ual in e ac ion be ween he low o he pa icula gases). Taguchi design o expe imen [21] was he e o e chosen as a sui able echnique o accomplish he ask. The s a is ical concep o he Taguchi’s design e ec i ely combines he indi idual ac o le els, esul ing in high e iciency o he expe imen al design, i.e. i main ains he p ima y in o ma ion yield while educing he numbe o ial uns signi ican ly. Fo he case o six h ee-le el ac o s (used in his s udy), he co esponding s anda dized o hogonal a ay is L-18 (L18). Following he L-18 design, he numbe o necessa y ial uns is educed om 729 o 18. Wi h espec o he numbe o ac o s, numbe o le els pe each ac o , he an icipa ed mu ual bonds, esul s eliabili y and a ia ions o epea abili y o he p ocess, he s anda dized a ay had o be modi ied acco ding o a desc ibed p ocedu e [21] and he a ay columns we e hen assigned acco dingly. The espec i e ac o s le els acco ding o he modi ied a ay could be seen om Table 2. To u he mo e inc ease he esul s eliabili y, wo uns pe combina ion we e ca ied ou and all uns we e ca ied ou in andom o de . Analysis o a iance (“ANOVA”) o he esul s ob ained om Taguchi design was ca ied ou . ANOVA is a s a is ical, decision-making ool o de ec ing changes in a e age pe o mance o measu ed da a. By compa ing he a ia ions o he da a measu ed a di e en ac o le els wi h he o e all a ia ion changes, ANOVA de e mines he c edibili y o measu ed da a se s and espec i e ac o s in luences. To asce ain 4 he in luences o pa icula ac o s on he in- ligh empe a u e and eloci y o HA powde pa icles, ANOVA esul s we e used o de e mine a pe cen con ibu ion o e e y ac o o he o e all changes a he selec ed anges. I has o be no ed ha despi e i s ad an ages o e o he gene al designs, he de e mina ion o single pa ame e ’s in luence is less demons able in Taguchi design. 2.3.2. In- ligh Measu emen Tools To de ec he pa icles and de e mine hei espec i e p ope ies du ing he deposi ion, a as -shu e CCD lase -assis ed came a was used (Osei L d., Tampe e, Finland). O hogonal pa e n o he se up was used, i.e. he axis o he plasma je was pe pendicula o he came a lens and lase emission axes (Fig. 3). Du ing he s udy, i was ound ha he au oma ic mode o he came a sys em has an in luence on he measu ed da a (measu ed a e age alues we e ound o be linked o he se ings o he came a exposu e ime). As in he au oma ic mode he came a pa ame e s a e changed equen ly in o de o cap u e he maximum numbe o pa icles, he da a ob ained wi hin one measu emen se expe ience o se shi s due o he se ings changes. To ensu e inc eased c edibili y o he esul s wi h espec o he ela i e di e ences in he measu ed p ope ies, manual de ec ion mode wi h a cons an shu e ime o 15 µs was he e o e used o all measu emen s. The manual se ings o he de ec ion pa ame e s esul ed in a lowe numbe o de ec ed pa icles pe 1 second. The da a o de ec ed pa icles we e he e o e collec ed o a pe iod o a leas 1 minu e o each un. The o al s a is ics (numbe o cap u ed pa icle o each un) was no lowe han 200 pa icles pe un o he sys em. Fu he mo e, he came a ocus dis ance was se u he om he plume o 235 mm, in o de o de ec he en i e olume o he plasma je a any gi en sp ay dis ance. 2.4. Coa ings P oduc ion and Cha ac e iza ion Based on he ob ained esul s o he in- ligh p ope ies, six combina ions we e selec ed o subsequen coa ing deposi ion (deno ed as A-F, Table 3). The objec i e o he selec ion was o ep esen he ex eme alues o he HA powde in- ligh p ope ies, i.e. o deposi coa ings unde low-medium-high in- ligh empe a u e/ eloci y (Fig. 4). Coa ings we e deposi ed using 15 passes o a obo ic a m mo ing a a ans e se speed o 100 mm·s−1. The subs a es used in his s udy we e pla ele s (29 ×24 ×2.5) mm o Ti-6Al-4V alloy con o ming o ASTM F136 G ade 5 s anda d. The subs a es we e chemically deg eased and g i blas ed using s anda d- ized uni o m p ocedu e wi h cons an angle, dis ance and dwelling ime ( oughness Ra= 5.07 ±0.32µm) The sp ayed samples we e embedded in o a wo-componen esin, g ound and polished and hei mic os uc u e was in es iga ed using SEM. X- ay di ac ome y spec a o he coa ings we e ob ained and assessed using Rie eld analysis [20]. This allowed a quan i a i e analysis o he phase composi ions o he p oduced coa ings. To ob ain unbiased in o ma ion, a ious echniques we e used o assess he po osi y o he p oduced plasma sp ayed coa ings wi h espec o he po osi y o he ini ial HA powde eeds ock. So wa e image analysis o he polished SEM samples and helium pycnome y me hods we e used o in e nal po osi y e alua ion, while me cu y in usion po osime y p o ided in o ma ion on he open po osi y. Su ace oughness indica o s we e measu ed acco ding o ISO: 4287 – 1984 s anda d. Mean de ia ion o he p o ile (Ra=1 nP|yi|) and oo - o-mean squa e de ia ion (Rq=q1 nPy2 i) indica o s we e mea- su ed in wo pe pendicula di ec ions o a ec he po en ial e ec o he sp ay gun mo emen di ec ion. Fi e loading and unloading cycles (sample unloaded o 90% o he cu en load) we e used o one mic o-ha dness inden a ion. Ten inden a ion pe sample we e ca ied ou and p o ided da a on mic o- ha dnesses and elas ic moduli o he p oduced coa ings. In-house-made shu e -appa a us was used o cap u e he single spla s o he HA powde on o polished Ti-6Al-4V su aces. The condi ions o deposi ion o he spla s we e se iden ically o he six combina ions selec ed o coa ings deposi ion. The size o he p epa ed polished subs a es was selec ed so as o collec pa icles om ho middle o he plasma je as well as om i s pe iphe y o comp ehend he a ia ions in empe a u es and eloci ies wi hin he je . Spla ypes wi h he highes occu ences o he gi en sample se we e in es iga ed (shape, mo phology and splashing beha io ). 5 3. Resul s & Discussion 3.1. In- ligh P ope ies The measu ed a e age in- ligh empe a u es and eloci ies o he HA powde pa icles could be seen om Table 2and Fig. 4, including hei espec i e measu ed s anda d de ia ions. I was ound ha he empe a u es o he pa icles du ing deposi ion (wi hin a ange o 2294 K–2708 K) we e abo e he incong uen decomposi ion poin o hyd oxyapa i e (1843 K). The eloci ies o he HA pa icles eached om 152 m·s−1 o 291 m·s−1, i.e. o an a e age diame e o 39.90 µm, he kine ic ene gy o impinging pa icles could di e by ∼270%. Analysis o a iances o he measu ed p ope ies acco ding o Taguchi design allowed o de e mine pa icula in luences o he six selec ed pa ame e s. The ela i e s eng hs o he six ac o s in luences on he in- ligh p ope ies could be seen om Fig. 5. I could be seen ha o he selec ed applied anges o he pa icula pa ame e s, sp ay dis ance is he mos in luen ial ac o o bo h empe a u e and eloci y o he pa icles. To quan i y he e ec , a single measu emen was ca ied ou : he sys em was se and un acco ding o he combina ion 18 om Taguchi design (Table 2) and he in- ligh p ope ies we e measu ed a 75 mm o 125 mm sp ay dis ance ange. I was ound ha he empe a u e d opped by 265 K while he eloci y d opped by 83 m·s−1wi h he inc ease om 75 mm o 125 mm posi ion. Dyshlo enko e el. [11] ca ied ou a ull- ac o ial expe imen al wo k on he in luence o ou selec ed pa ame e s (elec ic a c powe , composi ion o A +H2 o ming gas, sp aying dis ance and mode o sp ay- ing - wa e s. subs a e). The coa ings we e hen sp ayed using a cons an 100 mm dis ance and i was epo ed ha he mos in luen ial pa ame e wi h espec o he c ys alline phase and phase con en o he coa ings is he ne powe . This is in good ag eemen wi h he phase changes obse a ions desc ibed in 3.2.2 Based on he esul s, six combina ions o he sys em pa ame e s we e selec ed o coa ing p oduc ion (deno ed as A - low empe a u e, low eloci y o F - high empe a u e, high eloci y, Figu e 4). The maximum di e ences o he measu ed in- ligh p ope ies we e 355 K and 98 m·s−1(samples A and F), expec ed o gi e ise o di e en coa ing p ope ies. To acili a e easie o ien a ion in he ollowing ex , a summa y o he in- ligh p ope ies o he six selec ed samples is p o ided in Table 3. 3.2. Coa ing P ope ies 3.2.1. Mic os uc u e The op laye mo phology and c oss-sec ion o he p oduced coa ings could be seen om Fig. 6 and Fig. 7, espec i ely. The analysis o he images e ealed ha he mic os uc u e o he coa ings is dependen on bo h he in- ligh p ope ies o he pa icles. The in es iga ion o he coa ings c oss- sec ions e ealed a change in he mo phology ( om used- ype owa ds he indi idual spla s s uc u es) wi h inc easing empe a u e and dec easing eloci y o he pa icles. Con a y o gene al acco dance [12], he highe empe a u es o he impinging HA pa icles (samples B, D, F) esul ed in indi idual spla s uc u e o he coa ings sepa a ed by he ypical oid chains while he lowe empe a u e pa icles (A, E) ga e ise o bulk- ype coa ings wi h no dis inc spla s obse ed and ela i ely low po osi y o sphe ical na u e. This seeming disag eemen could be explained by he dec easing in- ligh eloci y o he pa icles: in gene al, he lowe he pa icle eloci y, he less used mo phology o he p oduced coa ing. Sample C (in- ligh empe a u e in-be ween ha o samples E, D) composed o used coa ing wi h some indi idual spla s s uc u e in he icini y o he in e ace, in acco dance o he ob ained esul s and sugges ed join in luence o in- ligh empe a u e and eloci y. A ne wo k o ela i ely sho e ical mic o-c acks was ound in low empe a u e sample coa ings (A, E). Addi ionally, ho izon al mac o-c acks we e obse ed in he coa ing E. As he ho izon al mac o-c acks we e obse ed in he middle o he coa ing laye s deposi ed by indi idual gun passes, hey a e mos likely no he consequence o he pauses be ween he passes, bu could be a consequence o coalescence o mic o-c acking o s ess alle ia ion du ing he cooling phase ins ead. 3.2.2. Phase Composi ion The Rie eld analyses o he measu ed XRD spec a indica ed ha he phase con en changed sub- s an ially a e he deposi ion p ocess. As could be seen om Table 1, signi ican phase con en changes we e ound in he coa ings, wi h he loss o o iginal HA phase o up o 76% as compa ed o o iginal HA 6 eeds ock (sample A). No dis inc ela ion be ween in- ligh empe a u e o eloci y and he pe cen age loss o he HA phase was ound. Rec ys alliza ion o he ini ial HA phase p esumably occu ed in he coa ings as he calcula ed p e e ed o ien a ions o he c ys allog aphic planes o hyd oxyapa i e wi hin he samples we e calcula ed as (0 2 1) and (0 0 4) as compa ed o ini ial (0 1 0) and (0 0 2) planes. Clea end was, howe e , obse ed o he con en o CaO phase in he coa ings: wi h inc easing in- ligh empe a u e o he pa icles, he con en o CaO was dec easing (Fig. 8). De imen al CaO phase con en inc eased o up o 14.6% in sample A, yielding such coa ing unaccep able o biomedical applica ions. Low CaO con en (≤2.6%) was ob ained in high in- ligh empe a u e samples (B, F). In acco dance wi h Lugscheide e al. [22], me as able i-calcium and e a-calcium phospha e phases (α-TCP, β-TCP, TTCP) de eloped du ing he sp ay p ocess, al e ing he unc ional eliabili y o he coa ings o biomedical use. TCP con en a ied om 13.3% o 19.4% and did no ollow any end wi h espec o he in- ligh p ope ies. The con en o TTCP in he p oduced coa ings was he highes among o he calcium phospha e phases and a guably he phase change om HA o TTCP is he mos a o ed du ing he sp ay p ocess when eaching empe a u e necessa y o ini ia e such change (T1). Ma kedly lowe con en o TTCP phase was de ec ed in he wo high- empe a u e samples (B, F; Fig. 8). In he wo samples, calcium py ophospha e [CPP] phases p esence was de ec ed ins ead (13.2% and 35.6%, espec i ely). I is he e o e sugges ed ha he empe a u es abo e ce ain le el T2induce u he ans o ma ion (3). HA abo e T1 −−−−−−→ TTCP abo e T2 −−−−−−→ CP P (3) Using he ob ained da a, we can s ipula e ha he empe a u es T1≤2350 K 2350 K≤T2≤2520 K(4) in (3). Li e a u e indica es he lowes empe a u e o phase ans o ma ion om HA o TTCP as 1820 K, i.e. lowe han he alues measu ed du ing he sp aying p ocess [18]. Acco ding o B own [23] and he CaO-P2O5-H2O e na y diag ams de i ed by Ha is [18], he CPP phases a e “CaO de icien ” as compa ed o TTCP phase (≤54w % CaO as compa ed o ≥57w % CaO; a apo p essu e 67 kPa, Fig. 9). Re aining he CPP phases upon cooling o he oom empe a u e would he e o e be accompanied by a d op in CaO con en , a phenomenon obse ed in he wo coa ings. I is necessa y o no e ha he esul s a e based on he selec ion o he nine conside ed calcium-phospha e phases and po en ial exis ence o ano he phase no in ol ed in he pe o med Rie eld analysis could in luence he phase ans o ma ion p ocesses. Also, gi en he phase changes unde he selec ed sp ay condi ions, he p oduced coa ings would no be accep able o issue in eg a ion and u iliza ion o p os heses mainly due o he signi ican loss o HA con en . 3.2.3. Po osi y Fo plasma sp ay echnology coa ings, in e nal po osi y le els o 2 – 20% a e indica ed in he li e a u e [16]. The measu ed in e nal po osi y o he coa ings p oduced in his s udy as assessed using wo me hods (Table 4) was 7 – 17%. Such po osi y le els all wi hin he 5 – 20% ange indica ed as sui able o os eoblas cells p oli e a ion [15]. The measu ed coa ing in e nal po osi ies a e sligh ly highe han he po osi y de e mined o he powde eeds ock (8.0%), indica ing u he o ma ion o in e nal po osi y (sphe ical po es o oid chains) in he coa ings. No signi ican end be ween he in- ligh p ope ies and he in e nal po osi y le els was de e mined. Howe e , he SEM image analysis indica ed ha he samples p oduced unde low in- ligh empe a u e (A, 6.9%; E, 8.3%) con ained lowe po osi y le els as compa ed o hei highe - empe a u e coun e pa s. This is p obably a consequence o di e en mo phology o he coa ings: samples A, E exhibi ed used-like ype o coa ings, esul ing in isola ed sphe ical po osi y (yielding low in e nal po osi y alues) as compa ed o he oid channels in he o he coa ings. This end was no con i med by he subsequen helium pycnome ic me hod. The p inciple o he helium pycnome ic me hod consis s in illing he well-de ined olume con aining he powde sample wi h helium gas and measu ing he loss o gas olume in an in e connec ed chambe . Ob ained alue p o ides an in o ma ion on ue densi y o he sample ma e ial and i s espec i e po osi y (based on he heo e ical alue o bulk densi y o HA 7 ρ= 3.156 g.cm−3). Gi en he measu ed accu acy o he me hod (po osi y o one sample measu ed by se e al epe i i e uns exhibi ed maximum di e ence o 0.66%), he di e ence in he measu ed alues as compa ed o SEM de e mina ion was p obably caused by he he inabili y o he pycnome ic me hod o di e en ia e be ween he sphe ical po osi y o bulk-like coa ings A, E and he in e connec ed chains o oids in he spla s s uc u e o samples B, F and he in low o helium in o he open mic o-c acks o he coa ings. I was ound ha he open po osi y o he coa ings (as measu ed by me cu y in usion me hod) is signi ican ly highe o he low in- ligh eloci y samples (A 22.4%, B 27.9%), i.e. he highe kine ic ene gy o as -mo ing pa icles p obably con ibu ed o a dec ease in he open po osi y con en . Simila ly, he lowes open po osi y con en was measu ed o he wo highes in- ligh empe a u e samples (E 9.8%, F 8.8%). The SEM image analysis me hod allowed speci ica ion o he a e age po e sizes. F om he esul s, a e age po e diame e s we e subsequen ly calcula ed assuming ci cula geome y o he po es. The ob ained da a sugges ha wi h inc easing in- ligh empe a u es o he pa icles, he a e age po e size dec eased in he inal coa ings. Howe e , he absolu e di e ences a e no signi ican and he measu emen could be in luenced by he ac ha he geome ical app oxima ion may no be co ec in he case o oid chains obse ed in coa ings B, D and F. The li e a u e iews on he op imal po e size o success ul bone in-g ow h a y, howe e , he measu ed alues o he po e diame e s o he p oduced HA coa ings appea unde sized o such applica ion, yielding he coa ings unsui able o bone issue in-g ow h [15]. 3.2.4. Su ace Roughness The esul s o he oughness measu emen s o HA coa ings could be seen om Table 5. No in luence o he in- ligh eloci y was con i med and no clea end could be dis inguished, i.e. he kine ic ene gy o he pa icles did no con ibu e o he o e all su ace oughness. I was ound ha he highes in- ligh empe a u e sample (F) e ained he lowes le els o oughness (7.4 µm). The in e io alue o he su ace oughness could be explained by he spla o ma ion beha io due o he apid solidi ica ion om he high empe a u e. This phenomena esul s in a o ma ion o smoo h la ened spla su ace wi h adial ails and is u he discussed in he sec ion 3.2.6. 3.2.5. Elas ic Moduli & mic oha dness Mic o-ha dness o HA plasma deposi ed coa ings was assessed. Ten inden a ions we e ca ied ou pe sample, each consis ing o i e loading and unloading cycles unde inc easing o ce. A e age alues o he en inden a ions we e conside ed as he measu ed ha dness o he sample. The inden a ions p o ided in o ma ion on he espec i e coa ings moduli. The esul s could be seen om Table 6. The samples deposi ed unde low in- ligh eloci ies (A, B) exhibi ed he highes le els o ma e ial ha dness (1.16 GPa and 0.86 GPa, espec i ely). This esul in no in acco dance wi h he p e alen idea o inc eased ha dness coa ings deposi ed by means o build-up o high kine ic ene gy pa icles [24]. Fu he mo e, no clea end be ween he mic o-ha dness o he coa ings and hei espec i e po osi y le els was obse ed. I should be no ed ha he measu ed le els o mic o-ha dness a e lowe han he alues gene ally indica ed o plasma sp ayed ce amic coa ings. In co ela ion wi h he mic o-ha dness measu emen s, i was ound ha he highes moduli among he HA coa ings exhibi ed he low in- ligh eloci y samples (A, 35.5 GPa and B, 26.8 GPa). No dependence o bo h coa ing cha ac e is ics on he in- ligh empe a u e o he HA pa icles was con i med as he alues did no ollow any clea end. 3.2.6. Single spla s In he mal sp ay p ocesses, mo phology, shape, sp eading and solidi ica ions beha io o single spla s subs an ially in luence he cha ac e is ics o he inal coa ings [25]. A numbe o cha ac e is ics a e p e- dominan ly in luenced by he i s deposi ed laye . Single spla s ( om mol en o semi-mol en d ople s) we e deposi ed on o mi o -polished Ti-6Al-4V su ace unde condi ions iden ical o he coa ings deposi- ion se up (A – F, Fig. 10). The SEM analysis showed signi ican di e ences in a e age shapes and mo phologies o he spla s sp ayed unde di e en sp ay condi ions. I was ound ha he splashing beha io o he pa icles is in luenced mainly by he in- ligh eloci y. Con a y o he expec ed in luence o he kine ic ene gy o 8 he pa icles, he lowe in- ligh pa icle eloci y samples (A 152 m·s−1, B 183 m·s−1and D 205 m·s−1) esul ed in splashed shapes. Among he h ee, he in ensi y o splashing was ound o be in luenced by he in- ligh empe a u e: he highe he in- ligh empe a u e o he pa icles, he highe he deg ee o splashing wi hin he h ee samples subse . The splashed ails anged om ela i ely sho (sample A) o a ne wo k o long, non-in e connec ed ails spanning he en i e deposi ion a ea (sample B). The high in- ligh eloci y samples (C 242 m·s−1, F 250 m·s−1and E 279 m·s−1) o m a second subse o non-splashed, p edominan ly ound spla s wi h no (C, E) o limi ed (F) splashing a i s pe iphe ies. As in he case o he o me subse , he in- ligh empe a u e was he seconda y ac o de e mining he inal mo phology o he spla s. The low in- ligh empe a u e sample E con ained mos ly ound spla s wi h no able im a ound hei pe iphe y while he high empe a u e sample F con ained mino adial splashed ails. The sample F (high eloci y, high empe a u e) con ained mic o-c acks wi hin he deposi ed spla s. Dhiman e al. [26] sugges ed a dimensionless solidi ica ion pa ame e Θ de ined as he a io o solidi ied pa o o e all spla hickness du ing he solidi ica ion p ocess. The pa ame e inco po a es a numbe o ac o s. Among o he s, he impac eloci y o pa icles and he mal con ac esis ance be ween spla s and su ace a e conside ed. I was sugges ed ha he e a e h ee di e en condi ions de e mining he beha io du ing impac and solidi ica ion: 1. o Θ 1, i.e. low impac eloci ies, he spla s will sp ead in o a hin shee liquid and up u e in e nally, leading o ex ensi e agmen a ion, p oducing a small cen al spla su ounded by a ing o deb is; in ou s udy, he h ee low eloci y samples (A, B, D) mee hese c i e ia and a e cha ac e ized by he said spla mo phology 2. o Θ ∼0.1 o 0.3, i.e. high impac eloci ies, he signi ican g ow h o solid laye es ic s he spla s om sp eading oo a ; in ou s udy, spla s o he samples C, E ollow his end 3. o Θ ∼1 he apid solidi ica ion obs uc s he sp eading liquid and ound spla s wi h adial ails a e c ea ed; as his was obse ed in ou s udy o sample F (high eloci y, high empe a u e), i is sugges ed ha he pa icle empe a u e aided in a u he inc ease o he alue Θ, which esul ed in a o ma ion o he adial ails a he im o he spla s In ag eemen wi h he s udy [26], he obse ed mo phologies o he collec ed single spla s a e mos likely in luenced mainly by he impac eloci y o he HA pa icles, he solidi ica ion beha io is howe e addi ionally in luenced by he in- ligh empe a u e o he pa icles and he he mal con ac esis ance o he wo ma e ials (HA, Ti-6Al-4V). 4. Conclusions Bioac i e HA coa ings we e deposi ed on o Ti-6Al-4V subs a es unde di e en a e age in- ligh empe a u es and eloci ies. Rela i e di e ences in he mic os uc u e, phase composi ion, po osi y le els, su ace oughness, mic o-ha dness and coa ing moduli we e in es iga ed wi h espec o he in- ligh p ope ies. F om he esul s i can be concluded ha : 1. The in- ligh cha ac e is ics a e subs an ially in luenced by he se ing o he sys em pa ame e s; he dominan ac o was ound o be he sp ay dis ance - wi h a change om 75 mm o 125 mm, he empe a u e and eloci y d opped by 265 K and 83 m.s−1, espec i ely 2. The mic os uc u e o he deposi ed coa ings is in luenced by he in- ligh empe a u e and eloci y o he pa icles; con a y o o he li e a u e, he high empe a u e pa icles (≥2500 K) esul ed in o indi idual spla mo phology wi h in e connec ed oid chains in he coa ings while low empe a u e pa icles (≤2470 K) ga e ise o a bulk ype well used coa ings; his ac could be explained by a join incidence o he in- ligh kine ic ene gy o HA pa icles ( eloci y) 3. The phase composi ion o he coa ings di e s subs an ially om he ini ial powde eeds ock; he loss o HA phase could accoun o 76%; he amoun o de imen al CaO phase in he coa ings is in e sely p opo ional o he in- ligh empe a u e o HA pa icles; majo i y o HA phase was ans o med in o TTCP phase, p esumably he a o ed HA phase change du ing plasma sp aying; he highes in- ligh empe a u e samples con ained some CPP phase and he amoun o TTCP phase was lowe ed - i has been sugges ed ha some o he TTCP could ha e changed in o he CPP phases unde ce ain condi ions 9 Figu e 7: SEM images o p oduced coa ings c oss-sec ions 16 5,4 8,5 1,3 14,6 3,7 2,6 49,5 36,7 35,4 43,8 14,6 3,9 35,6 13,7 0,0 0,00,0 13,2 0 2 4 6 8 10 12 14 16 18 20 2300235024002450250025502600265027002750 In- ligh empe a u e [K] CaO con en [%] 0 10 20 30 40 50 60 TTCP and CPP con en [%] CaO TTCP CPP Figu e 8: CaO, TTCP and CPP con en in p oduced HA coa ings 5055606570 1200 1300 1400 1500 1600 1700 HA β1-TCP + TTCP β1-TCP + HA liq. + TTCP liq. liq. + β1-TCP liq. liq. + β1-TCP α-TCP + liq. α-TCP + α-CPP α-TCP + HA TTCP + HA CaO + HA Tempe a u e [°C] CaO con en [w %] CaO + TTCP Figu e 9: Phase diag am o CaO - P2O5a apo p essu e o 66.7 kPa 17 Figu e 10: Plasma deposi ed HA spla s on polished subs a es Table 1: Phase con en o HA powde & p oduced coa ings Phase Con en [%] ICSD Powde Coa ings no. A B C D E F HA 26205 93.0 18.6 56.3 41.0 33.4 39.3 28.6 CaCO373446 5.0 - - . . - - CaO 75785 1.0 14.6 2.6 8.5 3.7 5.4 1.3 α-TCP 00923 - 2.1 . - 7.3 4.0 12.1 β-TCP 97500 - 11.2 14.1 15.1 11.9 14.4 7.3 TTCP 02631 - 49.5 13.7 35.4 43.8 36.7 14.6 α-CPP 22225 1.0 3.9 1.9 - - . 18.8 β-CPP 14313 1.0 - 11.3 - - - 16.8 Ca(OH)215471 - - . - - - 0.5 No e: Phase con en <0.5% indica ed as aces ( .) 18 Table 2: Taguchi design o in- ligh p ope ies measu emen s Run Ne Main Auxilia y Feed Ca ie Sp ay Measu ed powe gas gas a e gas dis ance empe a u e eloci y [kW] [l·min−1] [l·min−1] [RPM] [l·min−1] [mm] [K] [m·s−1] 1 7 20 10 2 3 75 2532 174 2 7 34 24 3 5 100 2490 187 3 7 48 38 4 7 125 2486 212 4 11 20 10 3 5 125 2353 152 5 11 34 24 4 7 75 2469 279 6 11 48 38 2 3 100 2592 261 7 15 20 24 4 3 100 2434 222 8 15 34 38 2 5 125 2344 236 9 15 48 10 3 7 75 2502 242 10 7 20 38 3 7 75 2517 183 11 7 34 10 4 3 100 2539 176 12 7 48 24 2 5 125 2708 250 13 11 20 24 2 7 125 2294 170 14 11 34 38 3 3 125 2636 291 15 11 48 10 4 5 75 2513 205 16 15 20 38 4 5 100 2449 227 17 15 34 10 2 7 100 2415 196 18 15 48 24 3 3 125 2430 226 No e: Feed a e indica ed in e olu ions o powde eede wheel pe minu e Table 3: In- ligh p ope ies o he six selec ed combina ions Sample Tempe a u e Veloci y [K] [m·s−1] A 2353 152 B 2517 183 C 2502 242 D 2513 205 E 2469 279 F 2708 250 The o de ed sequence o samples wi h espec o in- ligh empe a u e is: TA< TE< TC< TD< TB< TF The o de ed sequence o samples wi h espec o in- ligh eloci y is: A< B< D< C< F< E Table 4: Po osi y o p oduced coa ings Sample Po osi y [%] A e age Image Pycnome ic Me cu y po e size analysis me hod in usion [µm] A 6.9 9.4 22.4 3.36 B 12.8 12.8 27.9 2.50 C 12.0 9.5 10.1 2.72 D 17.0 8.8 11.5 3.17 E 8.3 10.7 9.8 3.28 F 15.2 9.1 8.8 2.53 19 Table 5: Roughness le els o plasma sp ayed HA coa ings Sample Roughness [µm] RaRq A 11.3 14.9 B 9.8 12.4 C 19.4 23.8 D 12.4 15.7 E 13.9 17.2 F 7.4 9.4 Table 6: Mic o-ha dness & elas ic moduli o plasma sp ayed HA coa ings Sample Mic o-ha dness Modulus [GPa] [GPa] A 1.16 35.5 B 0.86 26.8 C 0.06 12.8 D 0.30 17.3 E 0.26 18.4 F 0.24 20.3 20