Development and optimization of novel sulfur-containing Ti-based bulk metallic glasses and the correlation between primarily crystallizing phases, thermal stability and mechanical properties

Con en s lis s a ailable a ScienceDi ec Jou nal o Alloys and Compounds jou nal homepage: www.else ie .com/loca e/jalcom Resea ch A icle De elopmen and op imiza ion o no el sul u -con aining Ti-based bulk me allic glasses and he co ela ion be ween p ima ily c ys allizing phases, he mal s abili y and mechanical p ope ies Lucas M. Ruschel a,⁎ , Bas ian Adam a , Oli e G oss b , Nico Neube a , Maximilian F ey a , Hans-Jü gen Wach e c , Ral Busch a a Chai o Me allic Ma e ials, Saa land Uni e si y, 66123 Saa b ücken, Ge many b Amo phous Me al Solu ions GmbH, 66424 Hombu g, Ge many c He aeus AMLOY Technologies GmbH, 63450 Hanau, Ge many a icle in o A icle his o y: Recei ed 6 Decembe 2022 Recei ed in e ised o m 20 Ap il 2023 Accep ed 16 May 2023 A ailable online 18 May 2023 Keywo ds: Bulk me allic glass Ti anium alloys Bioma e ials Mechanical p ope ies The mal s abili y Sul u abs ac The e ec o sul u on he glass o ming abili y, he mal s abili y and mechanical p ope ies o he eu ec ic alloy Ti 33.4 Z 33.3 Cu 33.3 was in es iga ed by con en ional X- ay di ac ion, di e en ial scanning calo ime y and 3-poin lexu al expe imen s. A no el egion o bulk glass o ma ion wi h a c i ical cas ing diame e o up o 4 mm was ound in he qua e na y Ti-Z -Cu-S sys em, howe e , b i le ac u e beha io was p e- dominan . Va ious alloying s a egies we e employed o imp o e mechanical p ope ies and a composi- ional ansi ion om b i le o duc ile ac u e has been iden i ied (e.g. o Ti 36 Z 33.5 Cu 24.5 S 6 ). A change o he p ima y p ecipi a ing phases om a C14 La es o an in e me allic (Ti,Z ) 2 Cu phase can be obse ed, as well as a s abiliza ion o he supe cooled liquid. The o igin o he he mally uns able beha io in Ti-based bulk me allic glasses is aced back o he easy o ma ion o he icosahed al phase upon hea ing, which is s uc u ally close o he supposedly p edominan icosahed al sho - ange o de in he amo phous s a e. The sys ema ic s udy ca ied ou in his wo k indica es a s ong co ela ion be ween p ima y c ys allizing phase and he mal s abili y, bo h poin ing o he ozen sho - ange o de in he amo phous s a e which is p e- de e mining he mechanical p ope ies. The ansi ion om he La es o he in e me allic (Ti,Z ) 2 Cu phase as well as he enla ged supe cooled liquid egion appea o be di ec ly ela ed o a des abiliza ion o he icosahed al sho - ange o de and ul ima ely o he imp o ed mechanical p ope ies. © 2023 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY license (h p:// c ea i ecommons.o g/licenses/by/4.0/). 1. In oduc ion Me allic glasses ha e been subjec o in ense esea ch since hei disco e y in he ea ly 1960s [1]. Unlike c ys alline ma e ials, a oms in amo phous me als a e a anged andomly on a long- ange o de bu exhibi a dis inc sho - and medium ange o de on a omic leng h scales. Compa ed o c ys alline me als o simila composi ion, hey a e cha ac e ized by signi ican ly highe s eng h, exceedingly high ha dness and ou s anding elas ic p ope ies [2]. Due o he high cooling a es equi ed o supp ess c ys alliza ion and enable glass o ma ion, i is pa icula ly challenging o p oduce amo phous me- als e en in he o de o a ew millime e s. Alloy sys ems ha exceed he c i ical cas ing hickness o 1 mm in con en ional cas ing p o- cesses a e called bulk me allic glasses (BMGs). These include, o ins ance, Z - [3], Ti- [4], Pd [5], P - [6,7], Au-[8], Mg-[9]. Fe- [10], and Ni- [11,12] based sys ems. Among hese me al-based BMGs, Ti-based BMGs possess g ea po en ial o p omising applica ions due o hei low densi y, high s eng h and consequen ly high speci ic s eng h [13–15]. The c i ical p oblem o known Ti-based alloys wi h ex- cep ionally good glass- o ming abili y (GFA) combined wi h low densi y is he p esence o he oxic elemen Be, p e en ing hei use in medical applica ions [15,16]. O he good glass o me s wi hou Be include he elemen Pd, an undesi able elemen o a ligh weigh ma e ial due o i s high densi y and cos [16,17]. Recen ly, a p o- mising al e na i e has been disco e ed by ou g oup wi h he new amily o S-con aining alloys, showing high glass- o ming abili y wi hou he use o he men ioned undesi ed elemen s. Based on a ious eu ec ic alloys, di e en bulk glass o ming composi ions ha e been de i ed in he S bea ing Ti-based sys em. S a ing om he e na y eu ec ic Ti 65.5 Ni 22.5 Cu 12 [18], BMGs wi h good GFA and high Ti con en we e de eloped using small amoun s o 4 a % S [19]. In Re . [20], o igina ing om he bina y Ti-Ni and Ti-Cu eu ec ics, Jou nal o Alloys and Compounds 960 (2023) 170614 h ps://doi.o g/10.1016/j.jallcom.2023.170614 0925-8388/© 2023 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/). ]]]] ]]]]]] ⁎ Co esponding au ho . E-mail add ess: [email p o ec ed] (L.M. Ruschel). bulk glass o ma ion up o 3 mm wi h addi ions o Z , Cu, Ni and a S con en o 8 a % we e achie ed. Among hem, he Ni- ee Ti-Z -Cu-S alloys (e.g. Ti 40 Z 35 Cu 17 S 8 ) a e o pa icula in e es o biomedical applica ions, as ew o no me al ions a e eleased in o solu ion due o hei excellen elec ochemical co osion esis ance [17,20]. Sul u i sel is conside ed unc i ical, as i belongs o he mine als and is an impo an componen o se e al amino acids o he human o ganism [21]. This ensu es li le o no in e ac ion wi h human cells and he e o e classi ies hem as biocompa ible. In pa icula , c ys alline Ti alloys a e widely used in o hopedic p os he ics and den al im- plan s due o hei high co osion esis ance and good biocompa - ibili y [22–25]. Howe e , hei main d awback is he high elas ic modulus compa ed o ha o a human bone causing e ec s like s ess shielding and loosening o he implan s [22], whe eas he unique s uc u e o amo phous Ti-based BMGs wi h a ela i ely low elas ic modulus in combina ion wi h high enginee ing yield s eng h enables hem o eme ge as a p omising al e na i e o con en ional c ys alline implan s [23]. As o igin o he p esen wo k, he e na y eu ec ic Ti 33.4 Z 33.3 Cu 33.3 (E5) on he coppe - ich side was chosen as new s a ing poin o alloy de elopmen . A oya e e al. epo ed his eu ec ic E5 as he poin wi h he lowes liquidus empe a u e in he Ti-Z -Cu sys em, an ideal s a ing poin , as he liquid s a e is s abi- lized o e y low empe a u es, p omo ing glass o ma ion [26,27]. Molokano and Chebo niko s udied a wide ange o alloys in he qua e na y Ti-Z -Ni-Cu sys em, including he Ti 66.7−x Z x Cu 33.3 e na y alloy se ies wi h a maximum c i ical cas ing hickness o amo phous ibbons o 250 µm o Ti 26.7 Z 40 Cu 33.3 [28]. The c i ical hickness o he eu ec ic composi ion E5 is epo ed as ∼200 µm [28]. Many sys ems ha e shown ha mino addi ions o ano he elemen esul in me allic glasses wi h imp o ed p ope ies and enhanced GFA, i he p ope mino alloying elemen is chosen o he espec i e base alloy [29–31]. In Ti-based sys ems, S has al eady been p o en o be e y bene icial o amo phiza ion by inc easing he opological a iabili y and complexi y o he sys em, hence ol- lowing he empi ical ules o glass o ma ion [19,20,27,32]. Re- cen ly, Wilden e al. ha e shown ha S educes he mel dynamics in Ti-based glass- o ming sys ems, p esumably caused by he o ma- ion o co alen in e ac ions [33]. Apa om me alloid elemen s (e.g. S, Si o B), me als such as Sc, Ni o Al a e equen ly used as mino addi ions [25,34–37]. The la e is o pa icula in e es as i is a ligh elemen ha u he educes he densi y o he sys em, p omo ing ligh weigh applica ions. Wi h espec o alloys no di- ec ly associa ed o medical applica ions, Ni, which is opologically equi alen o Cu, can con ibu e o enhanced glass o ma ion [20,36,38]. In ou p e ious s udy, i was ound ha Ni- ee Ti-based bulk me allic glasses can be o med in he Ti-Z -Cu-S alloy sys em, bu he GFA was limi ed up o only 3 mm [20]. This wo k ocuses p ima ily on di e en alloying s a egies in he Ti-Z -Cu-S sys em and i s e - ec s on GFA, he mal s abili y and mechanical p ope ies. The p i- ma y goal was o enable a la ge applica ion ield by enhancing he GFA, howe e , i was ound ha he mechanical p ope ies (duc ile o b i le ac u e beha io ) a e ema kably sensi i e o small com- posi ional changes. The mechanism causing his dis inc i e change in he mechanical pe o mance is ound o be ela ed o he liquid and glassy s uc u e, which a e known o be domina ed by an ico- sahed al sho - ange o de (ISRO) in Ti-based alloys [39–41]. To deepen he unde s anding o he emb i lemen mechanism in he s udied alloys and e en ually in Ti-based BMGs in gene al, he e- sul s a e u he suppo ed by sys ema ic s udies o Ni and Al ad- di ions, causing simila changes in ac u e beha io . Ul ima ely, his allows he de elopmen o Ti-based me allic glasses wi h imp o ed GFA as well as desi ed mechanical p ope ies o enable hei use in po en ial biomedical applica ions. 2. Expe imen al p ocedu e 2.1. Sample p oduc ion The alloys we e p epa ed om high pu i y aw elemen s: Ti (99.995 w %), Z (99.99 w %), Cu (99.999 w %), Ni (99.99 w %), S (99.9995 w %) and Al (99.99 w %). To add S in o he composi ion, a cus om-made Cu-S p e-alloy was used, syn hesized om high pu i y Cu and S in a qua z ube unde high pu i y a gon a mosphe e wi h he composi ion Cu 67 S 33 (a %). Subsequen ly, he mass o he p e- alloy was checked o iden i y any mass loss compa ed o he weigh ed-in elemen s. The conse a ion o mass allows an accu a e de e mina ion o he inal Cu-S composi ion assuming ha only S (low boiling poin ) and no Cu is los in he p ocess. A de ailed de- sc ip ion o he p oduc ion p ocess is gi en in Re . [42]. Subse- quen ly, he aw elemen s we e alloyed oge he wi h he p e-alloy in an elec ic a c u nace unde a Ti-ge e ed a gon a mosphe e. To ensu e homogeneous dis ibu ion o he elemen s, he ingo s we e lipped and emel ed a leas 5 imes. A e wa ds, he samples we e cas unde a Ti-ge e ed high pu i y a gon a mosphe e in o wa e cooled coppe molds using a cus om-buil suc ion cas ing machine. Fo he poo e glass o me s, pla e-shaped specimens wi h hick- nesses o less han 1 mm we e cas . O he wise, od-shaped samples wi h di e en diame e s (≥ 2 mm) we e ab ica ed o de e mine he c i ical cas ing diame e . 2.2. X- ay di ac ion and calo ime ic analysis The samples used o X- ay di ac ion (XRD) and di e en ial scanning calo ime y (DSC) we e cu om he cylind ical ods a a heigh o 10 mm om he bo om igh nex o each o he . To e i y he amo phous s uc u e as well as iden i y he c ys alline phases, XRD measu emen s we e conduc ed using a PANaly ical X′Pe P o di ac ome e and a D8-A25-Ad ance di ac ome e wi h mono- ch oma ic Cu-Kα adia ion. The c oss sec ion o all specimens, in- cluding he beams o he mechanical es s, we e cha ac e ized o ob ain in o ma ion om he inne egion o he sample, mos c i ical o c ys al o ma ion as i is expe iencing he lowes cooling a e. The mal analyses we e pe o med a hea ing a es o 1 K/s in Cu pans unde a cons an high-pu i y a gon low using a powe -com- pensa ed Pe kin Elme DSC 8000 o de e mine he onse o he glass ansi ion empe a u e, T g , and he onse o he p ima y c ys al- liza ion, T x , o he amo phous samples. 2.3. Mechanical es ing Me allic ma e ials a e con en ionally es ed in ensile es s. Howe e , e en me allic glasses ha a e mic oscopically (in- insically) duc ile beha e mac oscopically b i le in his mode due o he o ma ion o one shea band unde app oxima ely 45° o he ensile axis. This single shea band p opaga es and leads o ca a- s ophic ailu e. Fo his eason, he mechanical pa ame e s such as Young’s modulus, yield s eng h, ac u e s eng h and o al s ain we e de e mined in a 3-poin beam bending (3PBB) se up using a Shimadzu uni e sal es ing machine, as i allows he o ma ion o mul iple shea banding ul ima ely enabling he obse a ion o in- insic duc ili y. All beam-shaped specimens ea u e a ec angula c oss-sec ion and we e cas in he dimension o 3 mm × 2 mm x 25 mm using he coppe mold suc ion cas ing echnique and sub- sequen ly sanded wi h up o 1200 g i SiC pape . The suppo span L was 20 mm. Bo h, he applied o ce F and he de lec ion D a he cen e o he beam we e eco ded du ing he es . The enginee ing s ess σ a he sample su ace and he s ain ε a he midpoin o he beam we e calcula ed by simple beam mechanics heo y using he equa ions: L.M. Ruschel, B. Adam, O. G oss e al. Jou nal o Alloys and Compounds 960 (2023) 170614 2 FL wh 3 22 = (1) and Dh L 6 2 = (2) wi h he heigh h and he wid h w o he beam. The de lec ion a e D , i.e. he eloci y o he c osshead, was cons an a 0.3 mm/min. Fo beam specimens wi h hicknesses abou 2 mm and a suppo leng h o 20 mm, Eq. 2 p o ides a s ain a e o 1.5×10 −4 s −1 a he ou e ibe i D is used ins ead o D. To de e mine a meaning s ain, he amp-up beha io a he beginning o he s ess-s ain cu e was co ec ed, as desc ibed in Fig. SI1 o he supplemen a y in o ma ion. Op imized cas ing pa a- me e s allow de ec s such as mic opo es o be la gely elimina ed. Howe e , o exclude andom sca e ing o he esul s in case o such de ec s, h ee es specimens we e es ed o each composi ion. 3. Resul s 3.1. Alloy de elopmen The GFA and he mal s abili y o he no el Ti-based alloys we e in es iga ed by XRD and he mal analysis upon hea ing om he glassy s a e a a a e o 1 K/s. Bo h p ope ies a e undamen al o he p oduc ion and p ocessing o BMGs. Fig. 1a and b summa ize he XRD esul s o he S addi ion se ies (Ti 33.4 Z 33.3 Cu 33.3 ) 100−x S x (x=0, 2, 4, 6, 8, 10, he ea e E5-S x ) o a ious hicknesses, allowing o de- e mine he c i ical cas ing hickness o he s udied alloys. Fo di- mensions below 1 mm, pla e-shaped samples we e es ed ins ead o od-shaped samples. The c i ical cas ing hickness o he composi- ions con aining 2–10 a % S e eal no sha p B agg peaks om any c ys alline phase, bu he ypical di usi e di ac ion maxima e- p esen a i e o monoli hic me allic glasses. Howe e , he alloy wi hou S canno be p oduced as a glass in a con en ional coppe mold cas ing p ocess due i s low GFA, e.g. small c i ical cas ing hickness, d c , o d c ∼200 µm [28]. The composi ions ha u ned ou amo phous we e hen cas in o molds one millime e la ge han he c i ical cas ing hickness (d c +1 mm) o de e mine he p ima y c ys- alline phases, as shown in Fig. 1b. Two domina ing c ys alline phases, he C14 La es phase (F ank-Kaspe s uc u e) and in e - me allic (Ti,Z ) 2 Cu ( e agonal s uc u e) can be iden i ied by XRD. In equilib ium condi ions a bcc phase is addi ionally epo ed by A - oya e e al. o he e na y eu ec ic [26]. To conclude, an app op ia e addi ion o S up o 4 a % signi ican ly enhances he GFA in he Ti-Z - Cu sys em, leading o a d c o up o 4 mm. Fu he addi ions abo e 4 a % lead o a con inuous dec ease o he GFA o a d c o 3, 0.75 and 0.5 mm o E5-S 6 , E5-S 8 and E5-S 10 , espec i ely. The co esponding DSC scans o ully amo phous samples wi h inc easing S-con en a e gi en in Fig. 1c, excep o E5-S 0 , as no amo phous specimen could be p oduced in con en ional coppe mold cas ing due o he limi ed GFA. The c ys alliza ion p oceeds ia mul iple exo he mic c ys alliza ion eac ions o all composi ions, al hough no all o hem ea u e a dis inc endo he mic e en as- socia ed wi h he glass ansi ion. Fo ins ance, no dis inc glass ansi ion could be obse ed o E5-S 2 , as he alloy s a s o c ys- allize immedia ely as he a omic mobili y is egained in he icini y o he glass ansi ion. Towa ds E5-S 10 , he signal o he glass an- si ion e en is ge ing mo e and mo e p onounced, as he onse o c ys alliza ion is shi ed con inuously o highe empe a u es, e- lec ing a s abiliza ion o he supe cooled liquid egion (SCLR), as summa ized in Table 1. The onse empe a u e o he glass ansi ion T g o hose showing one as well as he onse o c ys alliza ion T x is ma ked wi h a ows in each cu e. Each alloy was also hea ea ed un il he i s c ys alliza ion was comple ed and subsequen ly cooled Fig. 1. XRD pa e ns o (Ti 33.4 Z 33.3 Cu 33.3 ) 100−x S x (x=0, 2, 4, 6, 8, 10) a di e en hicknesses wi h (a) showing he c i ical cas ing hickness o as-cas samples, despi e he one wi hou S ha could no be p oduced amo phously and (b) pa ially c ys- allized samples closely abo e he c i ical cas ing hickness. (c) DSC Scans measu ed wi h a hea ing a e o 1 K/s o all amo phous composi ions. L.M. Ruschel, B. Adam, O. G oss e al. Jou nal o Alloys and Compounds 960 (2023) 170614 3 down o de e mine he i s c ys alline phase ex-si u by XRD (see Fig. SI2). The analysis e eals ha he p ima y phase upon hea ing changes om an icosahed al phase (I-phase) o E5-S x (x=2,4,6,8) o he La es phase o E5-S 10 . The e o e, he obse abili y o he glass ansi ion h ough he inc easing s abiliza ion o he SCLR o 72 K o E5-S 10 is di ec ly connec ed o he des abiliza ion o he quasi- c ys alline I-phase. Howe e , he maximum in GFA and SCLR does no coincide wi h each o he and show con a y beha io . Among he alloys s udied, he supe cooled liquid o he bes glass o me E5- S 4 beha es he mally e y uns able wi h almos no SCLR, in con as o he mos s able composi ion E5-S 10 which shows a poo GFA. As will be shown la e , he bes alloy E5-S 4 wi h espec o GFA is mechanically b i le, whe eas he amo phous alloy Ti 40 Z 35 Cu 17 S 8 de eloped by Kuball e al. in he same quasi- e na y Ti-Z -Cu-S sys em beha es duc ile in 3-poin lexu al es s [20]. The objec i e o u he alloy op imiza ion is o scan he composi ional space in be- ween o obse e and unde s and he ansi ion om b i le o duc ile ac u e beha io . The egion o in e es is depic ed in he quasi- e na y (Ti,Z )-Cu-S diag am in Fig. 2 and is called “connec ion line” in he ollowing. The specimens a e labeled “CL” acco dingly. These include he composi ions Ti 32 Z 32 Cu 32 S 4 , Ti 34 Z 32.8 Cu 28.2 S 5 , Ti 36 Z 33.5 Cu 24.5 S 6 , Ti 38 Z 34.3 Cu 20.7 S 7 and Ti 40 Z 35 Cu 17 S 8 , which a e u he e e ed o as E5-S 4 (as be o e), CL-S 5 , CL-S 6 , CL-S 7 and CL-S 8 , espec i ely. The composi ion E5-S 4 ( ed squa e) esul s om he alloy de elopmen shown in Fig. 1, while CL-S 8 (blue squa e) is he bes glass o ming alloy desc ibed in Re . [20]. The in e ening alloys CL-S 5 o CL-S 7 (g ey squa es) a e calcula ed acco ding o a linea equa ion. Mo e de ails on he de e mina ion a e p o ided in he supplemen a y in o ma ion. The XRD esul s o he “connec ion line” o bo h bulk glass o ming egions a e shown in Fig. 3a and b o he cas ing hick- nesses o 4 mm and 5 mm, espec i ely. The bes composi ion in e ms o c i ical cas ing hickness om he S se ies E5-S 4 is plo ed again o illus a e he end in GFA and he change in he p ima y p ecipi a ing phases. Re e ing o Fig. 3a, a wide GFA ange o 4 mm was ound anging om E5-S 4 o CL-S 6 . CL-S 7 also exhibi s a mos ly amo phous XRD pa e n bu shows i s signs o p ima y c ys alliza ion a he main peak o he amo phous halo. The p ima y p ecipi a ing phase (Ti,Z ) 2 Cu is mo e p ominen in CL-S 8 , which exhibi s a GFA o 3 mm acco ding o Re . [20], hence is no expec ed o solidi y wi hou c ys alliza ion a 4 mm. The change o he p i- ma y c ys alline phases in cooling om he equilib ium mel can be iden i ied mo e p ecisely in Fig. 3b, whe e he La es phase in E5-S 4 changes o a coexis ence o he La es and in e me allic (Ti,Z ) 2 Cu phase in CL-S 5 , ollowed by quasi-bina y (Ti,Z ) 2 Cu in CL-S 6 o CL-S 8 . The alloy wi h he highes GFA, CL-S 5 , exhibi s a mos ly amo phous pa e n in addi ion o a ew B agg peaks, condensing o a c i ical cas ing hickness sligh ly below 5 mm. In e es ingly, he maximum in GFA is exac ly loca ed whe e he p ima y c ys alline phases a e changing. By mo ing composi ionally close o CL-S 8 , he GFA d ops con inuously due o he p ima y o ma ion o he less complex in- e me allic (Ti,Z ) 2 Cu. The co esponding DSC measu emen s wi h an inc easingly he mal s abili y owa ds CL-S 8 is shown in Fig. 3c wi h he he mal p ope ies summa ized in Table 1. All u he alloy de elopmen s a egies we e subsequen ly pe o med on CL-S 6 (Ti 36 Z 33.5 Cu 24.5 S 6 ) a he han he bes composi ion in e ms o GFA, CL-S 5 , as i exhibi s supe io mechanical p ope ies and no b i le ac u e ailu e, as will be shown la e (Fig. 5). Fig. 4a and b show he XRD da a o 4 mm and 5 mm ods o Ti 36 Z 33.5 Cu 24.5−y Ni y S 6 (y=2, 4, 6, 8, 10, 12) whe e Cu is con inuously subs i u ed wi h Ni. Rega ding a s uc u al poin o iew, Ni and Cu a e conside ed opologically equi alen due o hei simila a omic size [43]. The e o e, bo h a e equen ly conside ed in e changeable [44]. Low Ni con en s do no lead o any no iceable change in he GFA compa ed o he Ni ee a ian , as bo h alloys (y=2 and y=4) s ill comple ely i i y in 4 mm ods and show c ys alliza ion o a dia- me e o 5 mm. Howe e , highe Ni con en s lead o a con inuous decline in GFA due o he o ma ion o he I-phase, ollowed by he C14 La es phase. A simila ansi ion om an amo phous s uc u e o he o ma ion o quasic ys als and La es phase was also epo ed in ano he S-bea ing BMG sys em based on he e na y Ti-Ni-Cu eu ec ic [19]. In gene al, bo h c ys alline phases a e common and equen ly epo ed in Ti-based BMGs [15,19,36,41,45–47]. Nex o Ni addi ion, he in luence o Al as ligh elemen was in es iga ed by equia omic subs i u ion, keeping he elemen al a ios o he me- chanically bes composi ion CL-S 6 cons an as shown in Fig. 4c and d. Table 1 Calo ime ic p ope ies o ully amo phous samples s udied in his wo k. T g e- p esen s he onse o he glass ansi ion empe a u e, T x de e mines he onse o he p ima y c ys alliza ion, and ΔT x de e mines he wid h o he SCLR (ΔT x =T x -T g ). The c i ical diame e d c co esponds o XRD amo phous samples p oduced by coppe mold cas ing. Composi ion (a %) T g (K) T x (K) ΔT x (K) d c (mm) Ti 33.4 Z 33.3 Cu 33.3 (E5-S 0 ) - - - ∼ 0.2 [28] (Ti 33.4 Z 33.3 Cu 33.3 ) 98 S 2 (E5-S 2 ) - 673 - 0.75 (Ti 33.4 Z 33.3 Cu 33.3 ) 96 S 4 (E5-S 4 ) 660 712 52 4* (Ti 33.4 Z 33.3 Cu 33.3 ) 94 S 6 (E5-S 6 ) 699 751 52 3 (Ti 33.4 Z 33.3 Cu 33.3 ) 92 S 8 (E5-S 8 ) 721 782 61 0.75 (Ti 33.4 Z 33.3 Cu 33.3 ) 90 S 10 (E5-S 10 ) 740 812 72 0.5 Ti 34 Z 32.8 Cu 28.2 S 5 (CL-S 5 ) 663 717 54 4* Ti 36 Z 33.5 Cu 24.5 S 6 (CL-S 6 ) 668 723 55 4 Ti 38 Z 34.3 Cu 20.7 S 7 (CL-S 7 ) 675 729 54 4 Ti 40 Z 35 Cu 17 S 8 (CL-S 8 ) 677 730 53 3 [20] Ti 36 Z 33.5 Cu 22.5 Ni 2 S 6 666 713 47 4 Ti 36 Z 33.5 Cu 20.5 Ni 4 S 6 663 709 46 4 Ti 36 Z 33.5 Cu 18.5 Ni 6 S 6 - 704 - ∼3** Ti 36 Z 33.5 Cu 16.5 Ni 8 S 6 - 706 - ∼3** Ti 36 Z 33.5 Cu 14.5 Ni 10 S 6 - 711 - ∼3** Ti 36 Z 33.5 Cu 12.5 Ni 12 S 6 - 706 - ∼3** (Ti 36 Z 33.5 Cu 24.5 S 6 ) 99 Al 1 670 720 50 4 (Ti 36 Z 33.5 Cu 24.5 S 6 ) 98 Al 2 676 726 50 4* (Ti 36 Z 33.5 Cu 24.5 S 6 ) 97 Al 3 677 727 50 4 (Ti 36 Z 33.5 Cu 24.5 S 6 ) 96 Al 4 680 729 49 4 (Ti 36 Z 33.5 Cu 24.5 S 6 ) 95 Al 5 695 736 41 4 * Highes GFA o he espec i e se ies ** 2 mm × 3 mm amo phous beams Fig. 2. Quasi- e na y (Ti,Z )-Cu-S diag am wi h he bes - ound glass o me s Ti 32 Z 32 Cu 32 S 4 o his wo k as well as Ti 40 Z 35 Cu 17 S 8 om Re . [20]. The g ey squa es indica e he ”connec ion line” composi ions. L.M. Ruschel, B. Adam, O. G oss e al. Jou nal o Alloys and Compounds 960 (2023) 170614 4 Aluminum was conside ed a he as mic oalloying elemen and was added in small amoun s, up o a maximum o 5 a % Al. All Al-con- aining composi ions solidi y glassy in 4 mm molds, indica ing no de e io a ion in GFA. In case o 5 mm cas s, Al addi ion o jus 1 a % is no su icien o imp o e he GFA, yielding a simila c i ical cas ing hickness as he "base" alloy CL-S 6 (compa e z=1 o Fig. 4d o CL-S6 o Fig. 3b). Fu he Al alloying o abou 2–3 a % esul in he bigges imp o emen wi h a XRD pa e n indica ing a mos ly glassy sample supe imposed by a ew (Ti,Z ) 2 Cu/La es B agg peaks. The declining GFA wi h inc easing Al con en goes again hand in hand wi h a change o he p ima y c ys allizing phase om he (Ti,Z ) 2 Cu in e - me allic o a C14 La es phase. Unlike he Ni se ies, he phase change does no occu h ough he in e media e s ep o he icosahed al phase o ma ion. This e ec o imp o ed GFA a mino addi ion o elemen s wi h u he de e io a ion a la ge amoun s is o en ob- se ed in li e a u e o he so-called mic o-alloying echnique [30]. Fig. 4e and depic he co esponding DSC scans, wi h he Ni se ies showing a con inuous des abiliza ion o he glass ansi ion un il no T g is obse able (Ni ≥ 6 a %), and he Al se ies showing only a sligh educ ion o he SCLR. The cha ac e is ic empe a u es a e sum- ma ized in Table 1. 3.2. Mechanical p ope ies Moni o ing he mechanical p ope ies and hei e olu ion is pa icula ly impo an o he de elopmen o applica ion-o ien ed alloys, especially as S-con aining Ti-based BMGs allow he usage o indus ial-g ade aw ma e ial wi hou signi ican de e io a ion o GFA, as shown by Kuball e al. o Ti 40 Z 35 Cu 17 S 8 in Re . [20]. A ce - ain amoun o duc ili y is o pa icula in e es o p e en sudden ailu e in case he applied load unexpec edly exceeds he yield s eng h. P io o es ing, he amo phous s uc u e o he lexu al beams was e i ied by XRD. Fig. 5a, b, and c show he 3PBB en- ginee ing s ess-s ain cu es o h ee alloy se ies de eloped in his wo k, he “connec ing line”, he Ni se ies as well as he Al se ies. The no el glass o ming alloy E5-S 4 based on he Ti-Z -Cu eu ec ic E5 wi h a GFA o 4 mm ails b i le in 3PBB expe imen s wi h a o al s ain below 2%. Such p ema u e ailu e is likely ela ed o a low ac u e oughness, esul ing in poo esis ance o de ec s such as mic opo es o su ace sc a ches. These allow c ack p opaga ion mo e easily in he ensile loaded egion o he beams, u ning 3PBB ex- pe imen s in o an elabo a e way o s udy he in insic mechanical p ope ies. Mo ing composi ionally owa ds Ti 40 Z 35 Cu 17 S 8 , he ansi ion om b i le o duc ile ac u e beha io occu s exac ly a he in e media e alloy composi ion Ti 36 Z 33.5 Cu 24.5 S 6 (CL-S 6 ). This composi ion also ea u es he la ges o al s ain o ailu e o a ound 4.2% in bending o he s udied alloys. The e o e, i was chosen as new o igin o alloy de elopmen o he Ni and Al se ies, al hough i was no he composi ion wi h he highes GFA. The s epwise Ni addi ion shown in Fig. 5b led o a dec ease in duc ili y, ollowed by comple e emb i lemen a Ni con en s abo e 2 a %. Despi e hei amo phous s uc u e, he la e do no e en each hei yield s eng h. Simila , bu less ca as ophic emb i lemen is also ob- se ed h oughou he Al se ies in Fig. 5c. Howe e , he bes com- posi ion (Ti 36 Z 33.5 Cu 24.5 S 6 ) 98 Al 2 in e ms o GFA s ill eaches i s yield s eng h and exhibi s mino duc ili y be o e a al ac u e oc- cu s. Al con en s abo e 2 a % ul ima ely led o p ema u e, b i le ailu e. The Young’s moduli o all di e en composi ions s udied in his wo k ange om 83 o 89 GPa in bending. A de ailed summa y o he mechanical p ope ies σ yield (0.2%), σ ac u e , E bending and ε o al co esponding o he yield s eng h a 0.2% s ain, he ac u e s eng h, he Young’s modulus in bending and he maximum achie ed s ain, espec i ely, is p o ided in Table 2. Fig. 3. (a) and (b) shows he X- ay di ac ion pa e ns o 4 mm and 5 mm as-cas samples o he “connec ion line” composi ions. (c) The co esponding DSC scans we e acqui ed on ully amo phous samples a a a e o 1 K/s. L.M. Ruschel, B. Adam, O. G oss e al. Jou nal o Alloys and Compounds 960 (2023) 170614 5 Fig. 4. XRD pa e ns o 4 mm and 5 mm ods o he e alua ion o he c i ical cas ing hickness o he Ni se ies Ti 36 Z 33.5 Cu 24.5−y Ni y S 6 (y=2, 4, 6, 8, 10, 12) in (a) and (b) and o he Al se ies (Ti 36 Z 33.5 Cu 24.5 S 6 ) 100−z Al z (z=1, 2, 3, 4, 5) in (c) and (d). (e) and ( ) depic s he co esponding DSC scans o bo h se ies acqui ed a a a e o 1 K/s. L.M. Ruschel, B. Adam, O. G oss e al. Jou nal o Alloys and Compounds 960 (2023) 170614 6 4. Discussion The XRD analyses o he in es iga ed qua e na y Ti-Z -Cu-S as well as quin ena y Ti-Z -Cu-S-Al alloys e eal he C14 La es phase as well as he in e me allic compound (Ti,Z ) 2 Cu o be he p ima y c ys allizing phases. Conside ing he composi ion o he espec i e alloy, he la e esul s om con inuous solubili y o Ti and Z in he e agonal (Ti,Z ) 2 Cu phase h oughou he e na y phase diag am om Ti 2 Cu o Z 2 Cu [26]. The mix u e o Ti and Z a oms in he e agonal sub-la ices leads o changes in he la ice pa ame e s (Vega d’s law), esul ing in a shi o he B agg peaks as exempla ily shown o CL-S 6 in Fig. SI3 in he supplemen a y in o ma ion. A si- mila change in he p ima y c ys als is obse ed upon Ni addi ion o he qua e na y Ti-Z -Cu-S sys em, al hough an addi ional quasi- c ys alline I-phase is o med as in e media e s ep. A s abiliza ion o he I-phase by Ni has been obse ed se e al imes, o example in he Ti-Ni-S [20] and Ti-Z -Ni-Cu [48] sys em. Since Cu and Ni a e conside ed o be opologically equi alen , he subs i u ion o bo h elemen s should no a ec he s uc u e, ye S iehle e al. ound signi ican s uc u al di e ences in he sho - and medium- ange o de o Z -based alloys, which we e mainly a ibu ed o di e si ied local elec onic in e ac ions [49]. In combina ion wi h he s ong nega i e en halpy o mixing o Ni wi h Ti and Z ( Hmix Ti Ni =−35 kJ/mol, Hmix Z Ni =−49 kJ/mol [44]), an icosahed al s uc u e seems o be p omo ed in he liquid s a e, which is suppo ed by high-ene gy X- ay di ac ion expe imen s o Goldman e al., who obse ed a p o- g essi e inc ease o an ISRO by he addi ion o Ni o bina y Ti-Z [40]. Saida e al. also ound ha elemen s ha ing a posi i e o weak chemical a ini y o one o he cons i u ional elemen s ( Hmix Cu Ni =4 kJ/ mol [44]) in Z - o H -based glass- o ming sys ems a e bene icial o he p ecipi a ion o he I-phase [50]. Mo eo e , he I-phase com- monly o ms as p ima y phase upon hea ing om he glassy s a e, causing he he mally uns able beha io in mos o he alloys s u- died and in gene al in Ti-based BMGs [15,45,51,52]. Some o hem, mainly he Ni se ies, do no e en show a p onounced glass ansi- ion, as T g is supe imposed by p ima y c ys alliza ion. This he mal ins abili y is expec ed o be ela ed o s uc u al simila i ies be ween he p edominan ISRO in he liquid/amo phous s a e and he quasi- c ys alline I-phase, as equen ly epo ed o Ti-based BMGs [19,39–41,53]. Those s uc u al simila i ies esul in a educed in- e acial ene gy and consequen ly, acco ding o classical nuclea ion heo y, in a educed nuclea ion ba ie o he I-phase [41,54–57]. The impo ance o he in e acial ene gy o he GFA o me allic glasses has been shown o P -P-based liquids [58]. In hese liquids, a high in e acial ene gy is able compensa e a high d i ing o ce o c ys alliza ion and a agile liquid beha io . In he case o he Ti- based glasses, a low in e acial ene gy due o a p onounced ISRO esul s in an immedia e c ys alliza ion as soon as a omic mobili y is es o ed in he icini y o he glass ansi ion. Simila phenomena we e also obse ed in Al-based BMGs wi h high Al-con en s, whe e he glass ansi ion is supe imposed by he apid o ma ion o cc α- Al nanoc ys als [59–61]. Ano he indica ion o a p edominan ISRO is e lec ed in a change o he p ima y phase upon cooling om he e agonal (Ti, Z ) 2 Cu o an icosahed al o La es phase, as he la e con ains a high p opo ion o icosahed al clus e s [40,62,63]. This means ha elemen s s abilizing an ISRO in Ti-based BMGs esul in educed he mal s abili y. Ou s udy sugges s ha he addi ion o S is an e ec i e way o e a d/supp ess he o ma ion o he I-phase du ing hea ing, as de- mons a ed o he E5-S x se ies wi h inc easing S con en . Ul ima ely S addi ion leads o a dis inc glass ansi ion and s able SCLR (Fig. 1c) p io o he eme gence o c ys alliza ion. The posi i e in luence o S on he mal s abili y seems o be an almos uni e sal ea u e, as i has al eady been demons a ed in a ious BMG sys ems, like Cu-, Z -, Ni and Pd-based ones [32,42], [64]. In e es ingly, he maximum GFA in he Ti-Z -Cu-S sys em (Ti 32 Z 32 Cu 32 S 4 ) coincides wi h ha epo ed in he Ti-Z -Ni-Cu-S sys em (Ti 58 Z 7.5 Ni 18.5 Cu 12 S 4 ) a a sul u con en o 4 a % [19]. Cha ac e is ic a e he elemen s Ti and Z as well as Cu and Ni, which each o m a comple e solid solu ion in hei bina y phase diag ams [65,66]. Fu he mo e, Ti and Z ha e simila elec on con- igu a ions, hence hey can be conside ed chemically equi alen [67]. Ni and Cu, on he o he hand, a e o en conside ed opologically equal Fig. 5. Enginee ing s ess-s ain cu es o di e en bulk glass o ming alloy compo- si ions measu ed in a 3PBB se up. A o al s ain o less han 2% indica es a b i le ac u e beha io . L.M. Ruschel, B. Adam, O. G oss e al. Jou nal o Alloys and Compounds 960 (2023) 170614 7 in me allic glass- o ming sys ems due o hei simila a omic adius [43]. Due o hose simila i ies o Ti and Z as well as Cu and Ni, he sys ems Ti 32 Z 32 Cu 32 S 4 and Ti 58 Z 7.5 Ni 18.5 Cu 12 S 4 can be conside ed quasi- e na y (Ti,Z )-(Cu,Ni)-S. F om his poin o iew, he a io o (Ti,Z ) o (Cu,Ni) is app oxima ely equal in bo h sys ems wi h wo o one, which seem o be c ucial o he ideal sul u con en . The alloys s udied by Kuball e al. suppo his wo king hypo hesis, as hey e- qui e a sul u con en o 8 a % o bulk glass o ma ion (e.g. Ti 40 Z 35 Cu 17 S 8 ) due o a highe amoun o Ti and Z in he sys em [20]. Thus, he (Ti, Z ) o (Cu, Ni) a io seems o de ine he op imal S- con en , despi e di e en p ope ies (chemical and opological) a e compa ed wi h each o he . Fig. 6 shows he ideal S-con en equi ed o ob ain he highes glass o ming abili y o di e en (Ti,Z )/(Cu, Ni) a ios. The highe he Cu/Ni con en , he less S is needed o ob ain a maximum in he GFA, as shown o a Cu-based BMG Cu 46.3 Ti 33.5 Z 10.8 Ni 7.9 S 1.5 (Vi 101S 1.5 ) o he same sys em [68]. Fu - he mo e, he maximum achie able S con en (app ox. 8.4 a %) o high (Ti,Z ) con en s is ma ked in he diag am. Assuming ha he GFA maximum ollows he linea co ela ion, a simple Cu-S p e-alloy is no longe su icien , and an addi ional Ti-S o Z -S p e-alloy is necessa y o achie e BMGs wi h (Ti,Z ) con en s highe han ∼75 a %. The op- imal sul u con en as a unc ion o he (Ti,Z )/(Cu,Ni) a io could be ela ed o he size misma ch o he cons i uen a oms, which migh exhibi an op imal dense andom packing o he a oms a ce ain a ios. Lu e al. al eady poin ed ou ha a de ined a omic size dis- ibu ion ins ead o an excessi e size misma ch can lead o he highes packing densi y in he liquid and hus o an op imum in he GFA [69]. Howe e , his does no explain why he op imum S con en does no change signi ican ly upon exchange o Ti and Z , despi e hei sig- ni ican a omic adius di e ence o a ound 10% [43]. In o he wo ds, he obse ed linea end canno be a ibu ed exclusi ely o opology. Hence, he chemical con ibu ion seems o be o equal impo ance, as he s udies by Wilden e al. a he sugges ed o sul u i sel o ma- nipula e he chemis y o Ti-Ni-S liquids, by educing he mel dy- namics, p esumably caused by he o ma ion o co alen in e ac ions [33]. The mel iscosi y was mo e han doubled in he e na y Ti-Ni-S sys em in di ec compa ison o bina y Ti-Ni. The co ela ion be ween he maximum GFA and he (Ti,Z )/(Cu,Ni) o S con en appea s o be decisi e and is wo h no ing, al hough u he wo k is necessa y o sys ema ically alida e his co ela ion. A mo e gene al explana ion o he imp o ed GFA in he sys em s udied can be a ibu ed o he p edominan o ma ion o he I-phase as p ima y p ecipi a ing phase upon hea ing om he glassy s a e (Fig. SI1), sugges ing an ISRO in he liquid as epo ed in many alloy sys ems [53,70–72]. An ISRO likely con ibu es o educed a omic mobili y coupled wi h an inc ease in mel iscosi y [73–75]. A simila slowdown in mel dynamics is also expec ed he e, esul ing in an imp o ed GFA. The composi ion Ti 40 Z 35 Cu 17 S 8 is epo ed o exhibi a highe o al s ain o ∼4.3% compa ed o he specimen o same composi ion es ed in his wo k wi h only 3.2% (Table 2) [20]. Howe e , he highe e- po ed s ain is no su p ising, since he measu ed sample size in his wo k was signi ican ly la ge and he o ma ion o shea bands s ongly depends on specimen size [76,77]. In addi ion, smalle spe- cimens possess a highe ic i e empe a u e due o he as e cooling a e and he e o e a highe deg ee o ee olume, p omo ing mul iple shea band o ma ion [78–80]. Jiang e al. epo ed a simila depen- dence o o al s ain on specimen dimensions [64]. Fo his eason, all beams s udied in his wo k ha e he same dimensions allowing a obus compa abili y among hem. In se e al o he S con aining sys- ems, inc easing S con en led o an emb i lemen o he ma ix as in Table 2 Mechanical p ope ies o he di e en alloy composi ions in es iga ed in his wo k: Enginee ing yield s eng h a 0.2% s ain σ yield , ac u e s eng h σ ac u e , he Youngs modulus E bending and he o al s ain ε o al calcula ed om he lexu al s ess-s ain cu es. Composi ion (a %) σ yield (GPa) σ ac u e (GPa) E bending (GPa) ε o al (%) Ti 32 Z 32 Cu 32 S 4 (E5-S 4 ) - 1.32 ± 0.4 87 ± 2 1.48 ± 0.4 Ti 34 Z 32.8 Cu 28.2 S 5 (CL-S 5 ) - 1.85 ± 0.2 89 ± 1 2.09 ± 0.2 Ti 36 Z 33.5 Cu 24.5 S 6 (CL-S 6 ) 2.59 ± 0.1 2.85 ± 0.1 88 ± 1 4.23 ± 0.3 Ti 38 Z 34.3 Cu 20.7 S 7 (CL-S 7 ) 2.64 ± 0.1 2.79 ± 0.1 89 ± 1 3.59 ± 0.4 Ti 40 Z 35 Cu 17 S 8 (CL-S 8 ) 2.58 ± 0.1 2.63 ± 0.1 87 ± 1 3.25 ± 0.1 Ti 36 Z 33.5 Cu 22.5 Ni 2 S 6 2.69 ± 0.1 2.76 ± 0.1 88 ± 1 3.44 ± 0.1 Ti 36 Z 33.5 Cu 20.5 Ni 4 S 6 - 1.40 ± 0.2 88 ± 2 1.61 ± 0.2 Ti 36 Z 33.5 Cu 18.5 Ni 6 S 6 - 1.45 ± 0.1 88 ± 1 1.66 ± 0.1 Ti 36 Z 33.5 Cu 16.5 Ni 8 S 6 - 1.03 ± 0.5 85 ± 3 1.36 ± 0.3 Ti 36 Z 33.5 Cu 14.5 Ni 10 S 6 - 1.35 ± 0.2 87 ± 2 1.59 ± 0.2 Ti 36 Z 33.5 Cu 12.5 Ni 12 S 6 - 1.18 ± 0.4 88 ± 2 1.39 ± 0.5 (Ti 36 Z 33.5 Cu 24.5 S 6 ) 99 Al 1 2.51 ± 0.2 2.59 ± 0.3 88 ± 1 3.54 ± 0.4 (Ti 36 Z 33.5 Cu 24.5 S 6 ) 98 Al 2 - 2.39 ± 0.1 88 ± 1 2.77 ± 0.1 (Ti 36 Z 33.5 Cu 24.5 S 6 ) 97 Al 3 - 1.53 ± 0.1 83 ± 2 1.82 ± 0.1 (Ti 36 Z 33.5 Cu 24.5 S 6 ) 96 Al 4 - 1.32 ± 0.3 86 ± 2 1.55 ± 0.3 (Ti 36 Z 33.5 Cu 24.5 S 6 ) 95 Al 5 - 0.65 ± 0.4 86 ± 3 0.86 ± 0.3 Fig. 6. (Ti+Z )/(Cu+Ni) a io as a unc ion o he S con en o di e en BMGs in he (Ti,Z )-(Cu,Ni)-S sys em. The composi ions Ti 40 Z 35 Cu 17 S 8 , Ti 58 Z 7.5 Ni 18.5 Cu 12 S 4 and Cu 46.3 Ti 33.5 Z 10.8 Ni 7.9 S 1.5 o igina e om Re . [20], [19] and [68], espec i ely. The e o ba s esul om he sc eening inc emen s o S con en pe o med o each alloy sys em. When he S con en was sc eened in 1 a % inc emen s, he op imum S con en is na owed down o abou ± 0.9 a %. By sc eening in s eps o 0.5 a %, he GFA max- imum was de e mined wi h an accu acy o abou ± 0.4 a % a ound he ideal S con en . The solid ed line ep esen s a linea i wi h a R 2 o 0.98. L.M. Ruschel, B. Adam, O. G oss e al. Jou nal o Alloys and Compounds 960 (2023) 170614 8 he case o Cu-Z -Al-S and Pd-Ni-S alloys [42,64]. Fo Ti-based BMGs, his seems o be di e en , as he mechanical p ope ies s ongly de- pend on he ac ions o he esidual cons i uen s by how hey s a- bilize o des abilize he ISRO. Ob iously, he mechanical p ope ies a e di ec ly ela ed o he in insic amo phous s uc u e, i.e. he sho - ange and medium- ange o de as well as ee olume in he glass (see Fig. SI4) [38,56]. Fig. 7 schema ically depic s how a s abiliza ion/ des abiliza ion o he ISRO could be unde s ood. The schema ic on he le side exempli ies he s uc u e o he composi ion E5-S 4 wi h he highes GFA de i ed om he e na y Ti-Z -Cu eu ec ic. A s abilized ISRO is indica ed by an inc eased numbe o icosahed al mo i s, which is also expec ed o he s udied alloys con aining Ni and Al. The o me is a well-known elemen s abilizing an icosahed al s uc u e in Ti-based sys ems [40]. Howe e , small addi ions o Al also end o inc ease he s abili y o icosahed al s uc u es by o ming s able Al- cen e ed icosahed a ins ead o Cu-cen e ed icosahed a, domina ing in Cu-Z based alloys [81]. A mo e des abilized ISRO is coupled wi h a educed numbe o icosahed al mo i s as ou lined on he igh side o Fig. 7. The idea is ha a ge ed alloy de elopmen s a egies (as done o he „connec ion line“ shown in Fig. 2) lead o educed numbe o closed-packed icosahed al mo i s, esul ing in "less densely packed" egions wi h highe deg ee o ee olume. This p omo es he o - ma ion o shea ans o ma ion zones and ul ima ely shea bands [78,82], culmina ing in a ansi ion om b i le o duc ile as obse ed in Fig. 5a om E5-S 4 o CL-S 6 . The p esen wo k sugges s ha he p ima y o ma ion o c ys alline phases du ing cooling p o ides in- o ma ion abou he ISRO s uc u e in he liquid s a e and hus also in he glassy s a e i nuclea ion has been supp essed. Once quasi-c ys- alline o c ys alline phases a e o med ha con ain icosahed al uni s (e.g. I-phase and La es phase obse ed in all alloy se ies, see Fig. 1, Fig. 3 and Fig. 4), i is likely ha he ISRO is s abilized, esul ing in low he mal s abili y wi hou a dis inc SCLR [63]. Conside ing ha he maximum s a e o an ISRO is he I-phase and ha he La es phase also con ains a high p opo ion o icosahed al clus e s, i is no su p ising ha Ti-based BMGs wi h a s abilized ISRO beha e mac oscopically b i le, analogous o he complex c ys alline phases [38,63,83–85]. In con as , he composi ions ha p ima ily o m quasi-bina y (Ti,Z ) 2 Cu in e me allic wi h a e agonal s uc u e ins ead o a complex s uc- u e exhibi a co espondingly less p onounced, des abilized ISRO, which ul ima ely ansla es in o imp o ed he mal s abili y and me- chanical p ope ies (e.g. CL-S 6 , CL-S 7 and CL-S 8 in Fig. 3 and Fig. 5a). Thus, he o ma ion o simple phases in Ti-based BMGs appea o indica e a s uc u e ha a o s highe duc ili y in he amo phous s a e. Fo ins ance, Wang e al. ound Ti-based BMGs wi h high duc ili y in comp ession by in en ionally choosing a eu ec ic sys em con aining a duc ile B2 phase ins ead o a complex phase wi h a F ank-Kaspe s uc u e [38]. Howe e , a simple c ys alline s uc u e is usually accompanied by educed GFA due o he highe cooling a es equi ed o supp ess c ys alliza ion o his simple phase. He e, he alloys wi h he highes GFA Ti 34 Z 32.8 Cu 28.2 S 5 and (Ti 36 Z 33.5 Cu 24.5 S 6 ) 97 Al 3 always p ecipi a e he complex La es phase (see Fig. 3b, Fig. 4d), coupled by poo mechanical p ope ies. This b i le beha io is al e ed i he p ima y c ys alline phase changes by u he alloy op imiza ion. Ou sys ema ic s udy demons a es ha a judicious manipula ion o he sho - ange o de h ough alloying s a egies allows he syn hesis o composi ions wi h supe io me- chanical p ope ies coupled wi h imp o ed GFA by analyzing he p ima y c ys als as well as he mal s abili y. Ideally, a single XRD/DSC measu emen is su icien o judge whe he o no alloy de elopmen is p oceeding owa ds p omising mechanical p ope ies as he p i- ma y c ys alline phases as well as he he mal s abili y mi o s he p e ailing glassy s uc u e. In o he wo ds, i he I- o La es phase a e p ima ily o med, which is ela ed o a low he mal s abili y, he e is no need o an ex ensi e in es iga ion o he mechanical p ope ies, since a p edominan ly b i le ac u e beha io is o be expec ed. Howe e , i he XRD measu emen indica e a simple c ys alline phase, such as (Ti,Z ) 2 Cu, i is wo h o in es iga e he mechanical pe o mance, as good p ope ies wi h a ce ain duc ili y a e expec ed. To summa ize and g aphically elucida e ou indings, schema ic ime- empe a u e- ans o ma ion (TTT) diag ams o he di e en alloying s a egies shown in Fig. 8a o c we e cons uc ed om he XRD c ys alliza ion da a a di e en hicknesses as well as he knowledge o he i s c ys allizing phase upon hea ing (Fig. SI2). Since no TTT diag ams we e measu ed expe imen ally in his wo k, he g aphs a e simply a p oposed schema ic isualiza ion o he indi idual alloying s a egies and hei in luence on he c ys al- liza ion beha io . The e ec o S on he posi ion o he c ys alliza ion cu es o he La es phase and (Ti,Z ) 2 Cu phase is illus a ed in Fig. 8a (solid blue and o ange lines). The wo lines a e d awn igh nex o each o he , as bo h phases a e de ec ed in he XRD pa e ns. An in- c easing S con en pushes bo h c ys alliza ion “noses” o longe imes due o he highe GFA and lowe c i ical cooling a e equi ed o glass o ma ion, which is e lec ed in he maximum d c o 4 mm a an op imum S con en o 4 a %. E en highe S con en s in u n lead o a educ ion in GFA o 500 µm o S 10 and hus o sho e "nose" imes, indica ed by dashed lines (Fig. 1). The DSC measu emen s e ealed a con inuous s abiliza ion o he SCL om ini ially no isible T g due o he o ma ion o he I-phase o a b oad SCLR o 72 K Fig. 7. Schema ic isualiza ion o a s abilized and des abilized ISRO. The le and igh side depic s exempla ily he s uc u e o an alloy wi h an inc eased and dec eased numbe o icosahed al mo i s, espec i ely. L.M. Ruschel, B. Adam, O. G oss e al. Jou nal o Alloys and Compounds 960 (2023) 170614 9