Superparamagnetic [sic] nanofibers by electrospinning

Supe pa amagne ic [sic] nanofibe s by elec ospinning† R. L. Ande sson, a L. Cabedo, b M. S. Hedenq is , a R. T. Olsson a and V. S ¨ om* c The p epa a ion o supe pa amagne ic hin fibe s by elec ospinning dispe sions o nanosized magne i e (Fe 3 O 4 , SPIO/USPIO) in a PMMA/PEO polyme solu ion is epo ed. The sa u a ion magne iza ion and coe ci i y we e no affec ed by he concen a ion (0, 1, 10, 20 w %) o fibe o ien a ion, showing hys e esis loops wi h high magne iza ion (64 A m 2 kg 1 @ 500 kA m 1 ) and eco d low coe ci i y (20 A m 1 ). AC suscep ibili y measu emen s s. empe a u e a equencies om 60 o 2 kHz confi med supe pa amagne ism. The mechanical p ope ies we e only sligh ly dependen on he pa icle concen a ion because he nanopa icles we e sepa a ely encapsula ed by he polyme . A uni o m fib e ac u e c oss sec ion was ound a all he in es iga ed pa icle con en s, which sugges s a s ong in e ac ion a he polyme /pa icle in e ace. A heo e ical alue o he magne ic low field suscep ibili y was calcula ed om he Lange in unc ion and compa ed wi h measu ed alues. The esul s show a dis inc bu concen a ion-independen aniso opy, a o ing magne iza ion along he fibe o ien a ion wi h no sign o exchange in e ac ion, explained by comple e nanopa icle sepa a ion. Supe pa amagne ism canno be in e ed om pa icle size alone, so a ele an in e p e a ion and c i e ion o supe pa amagne ism is p esen ed, in acco dance wi h Neel's o iginal defini ion. F om he measu emen s, i can be concluded ha magne ic cha ac e iza ion can be used o elucida e he ma e ial mo phology beyond he esolu ion o a ailable mic oscopy echniques (TEM and SEM). In oduc ion Applica ions o supe pa amagne ic magne i e include con as agen s o Magne ic Resonance Imaging (MRI) o a senic emedia ion o d inking wa e . 1–3 O he p oposed applica ions o magne ic b e ma e ials include l e s wi h adjus able pa icle size disc imina ion, emo ely ac ua ed/ope a ed cell g ow h empla es o mechanical ac ua o s e.g. loudspeake memb anes o magne o heological liquids. 4,5 Many o hese applica ions need e y somagne ic cha ac e is ics. This is gene ally ealised wi h supe pa amagne ic i on oxide (SPIO) o ul a-small supe pa amagne ic i on oxide (USPIO) pa icles. These pa icles may indeed be supe pa amagne ic, bu sup- po ing magne ic da a a e seldom epo ed. Ins ead, a common app oach is o s a e ha he pa icle size is smalle han a c i - ical size and indi ec ly claim supe pa amagne ism. The c i ical size is he uppe limi o an indi idual pa icle o be supe - pa amagne ic, and can in heo y be calcula ed om da a on he magne ic ma e ial. 6,7 Fo magne i e a oom empe a u e his size is ca. 30 nm. 6 Howe e , he shape o he pa icles, exchange in e ac ion due o e y small pa icle–pa icle dis ances and al e ed magne ic p ope ies due o hei mo phology, such as su ace aniso opy, can p e en he ma e ial om being uly supe - pa amagne ic. 8 I mus also be emembe ed ha he ansi ion be ween supe pa amagne ism and e omagne ism is e y sha p. 7 Pa icles only ma ginally la ge han he c i ical size will no be supe pa amagne ic, c ea ing a e o/ e imagne ic con ibu ion, whe eas an en i ely supe pa amagne ic sample will show bo h a anishing emanence (hal heigh o hys e esis loop a ze o eld) and coe ci i y (hal wid h o hys e esis loop a ze o magne iza ion), acco ding o Ne´ el's deni ion. 9,10 The absence o coe ci i y will cons i u e a conse a i e con ma ion o he absence o a e omagne ic con ibu ion, in con as o e.g. he de e mina ion o he blocking empe a u e ha only de ec s he p esence o a supe pa amagne ic con ibu ion. 11 This si ua ion can lead o inco ec conclusions. The objec i e o his wo k is o show ha i is qui e possible o in eg a e supe pa amagne ic nanopa icles in o ma s o elec ospun b es whe e nei he he magne ic unc ionali y no he mechanical p ope ies ha e been comp omised. The la ge elonga ion o ces de eloped du ing he b e s e ching a oid p oblems conce ned wi h adi ional p epa a ion me hods such as mel p ocessing o sol en cas ing. 12 Ano he bene is ha a Depa men o Fib e and Polyme Technology, KTH-Royal Ins i u e o Technology, S ockholm, SE 100 44, Sweden b Polyme s and Ad anced Ma e ials G oup (PIMA), Uni e si a Jaume I, Campus de Riu Sec, 12071 Cas ell´ on de la Plana, Spain c Depa men o Ma e ials Science and Enginee ing, KTH-Royal Ins i u e o Technology, S ockholm, SE 100 44, Sweden. E-mail: al e @k h.se †Elec onic supplemen a y in o ma ion (ESI) a ailable: Fig. S1 o S3. See DOI: 10.1039/c5 a27791d Ci e his: RSC Ad .,2016,6, 21413 Recei ed 26 h Decembe 2015 Accep ed 16 h Feb ua y 2016 DOI: 10.1039/c5 a27791d www. sc.o g/ad ances This jou nal is © The Royal Socie y o Chemis y 2016 RSC Ad .,2016,6, 21413–21422 | 21413 RSC Ad ances PAPER Open Access A icle. Published on 16 Feb ua y 2016. Downloaded on 24/05/2016 10:30:50. This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial 3.0 Unpo ed Licence. View A icle Online View Jou nal | View Issue he composi e ma e ial is “locked”,i.e. solidied on a imescale o he o de o milliseconds, effec i ely elimina ing agg ega- ion. 13 The chosen magne ic ma e ial was nanosized magne i e pa icles (Fe 3 O 4 ), which is enowned o i s good sa u a ion magne iza ion (ca. 90 A m 2 kg 1 in bulk a ambien condi ions) and non- oxici y. 14 The ca. 10 nm sized pa icles we e p epa ed using he ‘ apid mixing’me hod, which esul ed in a uni o m pa icle size, high sa u a ion magne iza ion and e y low coe - ci i y, all cha ac e is ics o supe pa amagne ism. 15 The measu ed coe ci i y o he be s was less han 20 A m 1 (only ca. hal o he ea h's magne ic eld), which is smalle han p e iously epo ed. 16,17 The b es we e spun on o a o a ing d um, which aligned he b es pa allel o i s ci cum e ence, i.e. o he p ope assessmen o he mechanical and magne ic p ope ies along and pe pendicula o he b e di ec ion. 18 This made i possible o assess he mechanical p ope ies o he b es and o shed ligh on he pa icle dis ibu ion/ o ganiza ion, enabling a quan i a i e co ela ion be ween magne ic aniso opy and b e o ien a ion. The aligned b es wi h dispe sed magne ic nanopa icles may se e as a model sys em o supe pa amagne ically unc ionalized composi e b es, whe e bo h he mechanical and he magne ic aspec s ha e been add essed. P e iously, only wo a icles ha e epo ed he nume ical alue o he magne ic coe ci i y ae elec ospinning. 16,17 Ma e ials and me hods Ma e ials Fo he supe pa amagne ic nanopa icle (SPIO) syn hesis, i on(II) chlo ide e ahyd a e (FeCl 2 $4H 2 O, $98.0%, Fluka) and i on(III) chlo ide hexahyd a e (FeCl 3 $6H 2 O, 97% ACS eagen , Sigma-Ald ich) we e used. An aqueous solu ion o 2 M ammo- nium hyd oxide (Al a Aesa ) was used as p ecipi a ing agen . The polyme s used we e polyme hyl-me hac yla e (PMMA) wi h aM w o 410 kDa (Al a Aesa ) and polye hylene oxide (PEO) wi h aM w o 600 kDa (Ac os O ganics), bo h in powde o m. Dime hyl o mamide (DMF) (99.8%, BDH P olabo) was used as he elec ospinning sol en . Syn hesis o magne i e nanopa icles The SPIO pa icles we e p epa ed by aqueous co-p ecipi a ion in which a ‘ apid mixing’o he eac an s occu ed in he ime- ame o milliseconds. 15 B iey, wo solu ions loaded in wo sepa a e sy inges we e simul aneously injec ed as con e ging je s in o a single 0.5 mm hin ube. The wo sy inges con ained he ammonia and he i on ion solu ion ( he Fe 3+ o Fe 2+ concen a ion a io was 2 : 1 acco ding o he s oichiome y in Fe 3 O 4 ), and his esul ed in a s able colloidal suspension o supe pa amagne ic magne i e nanopa icles. 19 The pa icles we e washed h ee imes wi h highly pu ied ype 1 wa e (18.2 MUcm) acco ding o ASTM D1193. P io o mixing wi h he polyme solu ion, a sol en exchange in o DMF was pe o med (3 imes cen i uga ion ollowed by eplacemen wi h DMF and e-dispe sion). P epa a ion o he b e solu ions The solu ions o be elec ospun we e p epa ed by mixing he supe pa amagne ic pa icles suspended in DMF wi h he poly- me s. The nominal quan i ies o nanopa icles (SPIO), PMMA, PEO and sol en (DMF) o a o al amoun o ca. 10 mL spinning solu ion a e lis ed in Table 1. The a io o PMMA o PEO was 75/25 by weigh o all he b es p epa ed, since his p opo ion imp o es he PMMA b e oughness. 20,21 The sol en con en was adjus ed o yield a cons an polyme concen a ion o 10 w %. The componen s o he b e solu ions we e added o a ial and gen ly s i ed a oom empe a u e o 24 hou s. The suspensions we e hen hea ed o 70 C and kep a his empe a u e unde cons an s i ing o one hou be o e spinning. Elec ospinning and collec ion o b e ma s The solu ions we e con inuously ed om a 5 mL polyp opylene sy inge a a a e o 20 mL min 1 , ia a PTFE ube, o an 18-gauge needle wi h a a ip. The needle ip was posi ioned 220 mm e ically abo e he collec o . The elec ic po en ial applied be ween he needle and he collec o was 10 kV du ing he elec ospinning. Two diffe en collec o s we e used: a s a iona y aluminium pla e and a o a ing aluminium d um 50 mm in diame e o he p epa a ion o espec i ely andomly deposi ed and aligned b e ma s, see Fig. 1. The aligned b e ma s we e ob ained by deposi ing he elec ospun b es on he aluminium d um o a ing a 2000 pm. Cha ac e iza ion echniques Scanning elec on mic oscopy (SEM) obse a ions we e made in a Hi achi S-4800 cold-eld-emission scanning elec on mic o- scope. A ca. 8 nm coa ing o pla inum–palladium was spu e ed on o he su ace o he samples (40 s a 80 mA) in a C essing on 208HR high- esolu ion spu e . T ansmission elec on mic o- scope (TEM) mic og aphs we e acqui ed in a Hi achi HT-7700 high- esolu ion mic oscope ope a ed a 100 kV. Fo he nano- pa icle analysis, d ops o a dilu e aqueous nanopa icle suspension we e d ied o e he ca bon-coa ed g id (200 mesh o m a -ca bon, Ted Pella, USA). The samples o TEM we e di ec ly elec ospun o e he ca bon-coa ed coppe g id. The wide-angle X- ay diff ac ion (WAXS) measu emen s we e pe o med using a PANaly ical X'Pe P o diff ac ome e . The samples we e scanned a oom empe a u e in a B agg–B en- ano geome y, using Cu Ka adia ion (l¼154 pm). The da a was collec ed o e a ange o sca e ing angles 2q:2–40. The mo-g a ime ic analysis (TGA) was conduc ed using a Me le Toledo TGA/DSC 1 wi h 70 mLAl 2 O 3 c ucibles Table 1 Composi ion o ca. 10 g o fib e solu ions Sample SPIO (g) PMMA (g) PEO (g) DMF (g) 0% 0 0.750 0.250 9.00 1% 0.012 0.854 0.285 10.3 10% 0.092 0.621 0.207 7.45 20% 0.184 0.552 0.184 6.62 21414 |RSC Ad .,2016,6, 21413–21422 This jou nal is © The Royal Socie y o Chemis y 2016 RSC Ad ances Pape Open Access A icle. Published on 16 Feb ua y 2016. Downloaded on 24/05/2016 10:30:50. This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial 3.0 Unpo ed Licence. View A icle Online con aining ca. 10 mg o ma e ial. The hea ing a e was 10 C min 1 un il a maximum empe a u e o 550 C and wi h a ni ogen gas ow o 10 mL min 1 . The ensile measu emen s we e ca ied ou on he aligned b e ma s in he di ec ion o he b e axis in acco dance wi h a p e iously epo ed echnique. 18 The sample size was 10 by 5 mm 2 and he measu emen s we e made a a s ain a e o 0.5 mm min 1 (10% o sample leng h pe minu e). The s ess alues we e calcula ed by di iding he measu ed o ce by he c oss sec ional a ea o he b e ma . This c oss-sec ional a ea was calcula ed om he mass pe uni a ea o he elec ospun b e ma and he densi y o he b e ma e ial de e mined by he ule o mix u es. In o de o p e en damage o he b e ma be o e es ing, he b es we e xed o an aluminium empla e using alkoxy-e hyl-cyanoac yla e (Loc i e 460, Henkel AG & Co. KGaA, Ge many) p io o being clamped in he ensile es e . Magne ic cha ac e iza ion A ib a ing sample magne ome e (VSM), (EG&G P ince on Applied Resea ch model 155), was used o acqui ing hys e esis loops be ween 500 kA m 1 . In o de o ob ain he pa allel and pe pendicula magne ic esponses o he aligned elec ospun supe pa amagne ic b es, he b es we e collec ed on o a hin polye hylene lm which was a ached o he su ace o he d um p io o spinning. This allowed ec angula pieces o be cu and olled in o small ubes (10 by 2 mm 2 ). Depending on he olling di ec ion ela i e o he b e o ien a ion, he b es could be o ien ed pa allel o o pe pendicula o he ube axis, which was kep pa allel o he magne ic eld. Ae magne ic cha ac e - iza ion, he samples we e analysed by he mog a ime y so ha he co ec amoun o magne ic ma e ial could be assessed. Resul s and discussion Supe pa amagne ic be ab ica ion Fig. 1a and b show ansmission elec on mic og aphs o he supe pa amagne ic pa icles p io o he elec ospinning. The pa icles we e highly c ys alline nanosized magne i e de e - mined om hei inge pa e ns c ossing he en i e pa icles (HRTEM, Fig. 1b). The olume a e aged pa icle diame e was 8 2 nm, de e mined om manual measu emen s o mo e han 500 pa icles in TEM (his og am a ailable in ESI Fig. S1†), and he X- ay diff ac ion spec a was iden ied as magne i e spinel phase s uc u e (JCPDS 19-629), Fig. 1e. Fig. 1d shows ha a high loading o hese magne i e nanopa icles (20 w %) esul ed in a somewha co uga ed be su ace, whe eas unloaded be s show smoo h be s, see Fig. 1c. Howe e , he o e all o ma ion o hese 1–2mm hick and uni o m elec o- spun be s was no signican ly affec ed by he inco po a ion o he magne i e pa icles. This was achie ed as a esul o a success ul sol en exchange o he aqueous phase used o he p epa a ion o he magne i e pa icles in o DMF, which is miscible wi h wa e and se ed as an op imal ca ie o he magne i e and dissol ed PMMA phase du ing he elec ospinning. Fig. 1 (a) TEM mic og aph shows he magne i e nanopa icles wi h an a e age size o 8 2 nm. (b) HRTEM o he same pa icles wi h clea inge pa e ns. Scanning elec on mic og aphs showing ep esen a i e mic os uc u es o he elec ospun ma s wi h (c) andom o ien a ion and (d) uniaxially aligned fib es. (e) WAXS spec a o he fib es wi h diffe en nanopa icle con en s, he peaks co esponding o magne i e a e indica ed wi h he app op ia e c ys al planes. This jou nal is © The Royal Socie y o Chemis y 2016 RSC Ad .,2016,6, 21413–21422 | 21415 Pape RSC Ad ances Open Access A icle. Published on 16 Feb ua y 2016. Downloaded on 24/05/2016 10:30:50. This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial 3.0 Unpo ed Licence. View A icle Online In ac , all he samples wi h magne i e con en s a ying om 1 o 20 w % could be spun in o uni o m be s and aligned as unidi ec ional be ma s. The wide-angle X- ay spec oscopy (WAXS) da a o he composi e b e ma s con med ha he nanopa icles inside he b es emained in hei non-oxidized magne i e phase, see Fig. 1e. The peaks we e om he (220), (311), (400), (422), (511) and (440) magne i e c ys alline planes. The small peaks a 2q:19and23 (Fig. 1e) we e due o poly- e hylene oxide (PEO) c ys als, indica ing ha some phase sepa- a ion occu ed in all he samples. P e iously, his phase sepa a ion was demons a ed o acili a e b e necking unde a ensileload,whichsignican ly imp o ed he ac u e ough- ness. 21 In he 20 w % sample, hese wo peaks a e much smalle , sugges ing ha he magne i e p e en ed he PMMA/PEO phase sepa a ion and/o induced a s uc u e wi h much smalle PEO c ys als. This mechanism esembles a p e iously demons a ed effec ha silica nanopa icles es ic PEO c ys al nuclea ion and g ow h when he pa icle con en app oaches 20 w %. 22 The specic su ace a ea o he magne i e nanopa icles in his wo k was ca. 120 m 2 g 1 . This la ge alue is consis en wi h he p e ious sugges ion ha he pa icles es ic he o ma ion o PEO c ys als. Pa icle dis ibu ion wi hin he elec ospun b es The SEM mic og aphs o he diffe en ly pa icle-loaded b es (Fig. 2, lecolumn) show ha he polyme en i ely encapsu- la ed he nanopa icles du ing elec ospinning. These mic o- g aphs use low ene gy seconda y elec ons o c ea e he images and hus show images o he su ace opology wi h li le pene- a ion dep h. Fig. 2, igh column, shows mic og aphs o he same imaged a eas acqui ed using backsca e ed elec ons, which ha e almos he same ene gy as he inciden elec on beam, bu pene a e much deepe , and he sca e ed in ensi y is oughly p opo ional o he a omic numbe o he cons i uen elemen s in he sample. 23 The i on (i.e. magne i e) in he sample is he e o e clea ly shown e en i i is bu ied in he in e io o he b es. A signican numbe o he pa icles we e p esen as clus e s wi h an a e age size o 100 nm. La ge ca. 1mm clus e s we e only appa en in he 20 w % b es. A highe magnica ion o one o he la ge clus e s (Fig. 2h and j) shows ha hese la ge en i ies we e also well encapsula ed loose agglome a es. 24 The we ing o he pa icles appea ed o be good, as indica ed by he absence o any delamina ion o he pa icle/polyme Fig. 2 (a–h) Scanning elec on mic og aphs o he fib es wi h inc easing nanopa icle con en s ( op o bo om: 0, 1, 10 and 20 w %) ob ained simul aneously wi h seconda y (le ) and backsca e ed ( igh ) elec on de ec o a he same magnifica ion. (i) Dependence o he a e age fib e diame e on nanopa icle con en . (j) SEM de ail o he p o uding nanopa icles in he fib es wi h 20 w % magne i e. Fib e ma (1 2cm 2 ) a es (k) and a ac ed by a pe manen magne (l). 21416 |RSC Ad .,2016,6,21413–21422 This jou nal is © The Royal Socie y o Chemis y 2016 RSC Ad ances Pape Open Access A icle. Published on 16 Feb ua y 2016. Downloaded on 24/05/2016 10:30:50. This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial 3.0 Unpo ed Licence. View A icle Online in e ace du ing mechanical es ing (see he Mechanical sec ion). A gene al cha ac e is ic wi h he highe lle con en s was ha he la ge clus e s oen p o uded om he cylind ical shape o he b es, Fig. 2h and j. The b e diame e depended on he pa icle con en (Fig. 2i and ESI Fig. S2†), and showed a  s a s eep inc ease om 1.2 mm o unloaded b es o 2 mm o he 1 w % pa icle con en . This was ollowed by a g adual dec ease in diame e o ca. 1mm wi h inc easing pa icle con en (10 and 20 w %). I is sugges ed ha he diame e inc ease om 0 o 1 w % is due o a signican ly s onge elec ic eld. This can be explained by conside ing ha he unloaded polyme liquid is an excellen elec ic insula o so ha e en a minu e ac ion o conduc i e elemen s om pa icle inclusions (o an impe ec washing/sol en exchange) can gi e ise o a highe conduc i i y. In his si ua ion, wi h a conduc ing je , he ol age ac s o e a sho e dis ance leading o a s onge eld (be ween he spinne e and he collec o ). 25 The dec ease in b e diame e wi h highe pa icle con en s (abo e 1 w %) is explained by an inc ease in he ze o-shea iscosi y o he polyme solu ion p io o spinning due o he highe lle con en . This is in ag eemen wi h p e iously e- po ed obse a ions o he size wi h inc easing pa icle con en s in elec ospun cellulose-lled PMMA nanob es. 26 This inhe en phenomenon is use ul when p epa ing ul ima ely hin elec ospun b es wi h high ino ganic lle con en s, since i acili a es apid e apo a ion o he liquid phase and he eby p e en s he usion o he b es isible in Fig. 2c and d. Fig. 2k and l show he exible polyme b e ma s wi h he mac o- scopically e enly dis ibu ed Fe 3 O 4 phase (20 w %) and i s esponse o a magne ic eld applied by a s ong pe manen magne . The mal s abili ies o magne ic b es Fig. 3a and b shows he he mal deg ada ion o he elec ospun b es unde ni ogen and hei de i a i e (DTGA). The mass emaining abo e 500 C co esponded well o he nominal Fe 3 O 4 con en s wi h espec o he polyme in he elec o- spinning solu ions, see Fig. 3a. The wo peaks a app oxima ely 280 and 380 C we e a ibu ed o he wo cons i uen s o he polyme blend. This is consis en wi h he in eg a ed DTGA a eas o hese wo peaks, which co espond o he polyme composi ion, i.e. 75 w % PMMA and 25 w % PEO. Fig. 3c shows a 20 C inc ease in he mal s abili y wi h inc easing ino ganic con en , a common phenomenon seen in nanocomposi es due o su ace passi a ion du ing he deg a- da ion p ocess. 27,28 The passi a ion mechanism is in his case sugges ed o be due o he accumula ion o ino ganic lle a he su ace o he mol en sample when he ini ial e apo a ion o he ma ix occu s, i.e. when he p ima y PEO deg ada ion occu s (a he  s peak in he DTGA) a laye /c us o ino ganic ma e ial is o med on he su ace, which e a ds he PMMA deg ada ion. Mechanical cha ac e iza ion S ain–s ess cu es o all he ma e ials s udied a e plo ed in Fig. 4a, oge he wi h ha o a b i le PMMA sample o compa ison. The addi ion o 25 w % PEO o he PMMA esul ed in a d ama ic inc ease (ca. 1500%) in he o al wo k equi ed o ac u e he b es, which made he mechanical p ope ies o he composi es di ec ly dependen o he lle con en . The ac u e ene gy equi ed o b eak he p is ine PMMA b es was 0.13 MJ m 3 compa ed o 2.14 MJ m 3 wi h 25% PEO. The addi ion o 10 w % magne i e pa icles led o an inc ease in ensile s eng h (+41%) and in Young's modulus (+33%), bu a dec ease in he elonga ion a b eak (41%), see Table 2. The oughness did no howe e change signican ly, since he inc ease in s eng h compensa ed o he dec eased in elonga- ion a b eak. This ein o cemen effec is due pu ely o he addi ion o he s onge and s iffe magne i e pa icles. A u he addi ion o pa icles up o 20 w % ga e an elonga ion a b eak and modulus simila o he alues achie ed wi h 10 w %, bu wi h a lowe s eng h (31%) and lowe oughness (32%), sugges ing ha he pa icle dispe sion o ma ix adhesion was no as good wi h 20 w % as wi h 10 w %, also indica ed by he la ge p o usions seen in Fig. 2g, h and j. The 1 w % b es had he lowes s eng h and modulus, see Table 2. This is explained by he sligh using phenomenon seen in Fig. 2c and d, since he b es a e no comple ely solidied du ing collec ion. O e all, i was appa en ha he addi ion o PEO o PMMA main ains la ge amoun s o magne i e in he elec ospun b es while gi ing mechanical p ope ies supe io o hose o he p is ine PMMA. I is sugges ed ha his is due o a s ong in e ac ion in he pa icle/ma ix in e ace, supp essing he o ma ion o oids/ delamina ion, see Fig. 4b–d, which ha e p e iously been e- po ed o occu in elec ospun b es and nanocomposi es. 29,30 The dec ease in elonga ion a b eak wi h highe pa icle loading Fig. 3 (a) TGA cu es o he decomposi ion o he fib es du ing hea ing unde a ni ogen a mosphe e wi h (b) he co esponding de i a i e (DTGA). (c) The empe a u e a which he peak mass loss a e occu s in he PMMA/PEO fib es as a unc ion o he magne i e con en . This jou nal is © The Royal Socie y o Chemis y 2016 RSC Ad .,2016,6,21413–21422 | 21417 Pape RSC Ad ances Open Access A icle. Published on 16 Feb ua y 2016. Downloaded on 24/05/2016 10:30:50. This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial 3.0 Unpo ed Licence. View A icle Online is in e p e ed as being due o diffe en abili ies o he b es o allow b e necking o occu , see Fig. 4b which shows he necking in he case o he p is ine polyme blend. This necking phenomenon is g adually es ic ed wi h inc easing pa icle con en , see Fig. 4c and d. Magne ic cha ac e iza ion In o de o de e mine he magne iza ion – he magne ic momen pe weigh o he magne ic ma e ial, which is an in insic p ope y making a compa ison meaning ul – he weigh o he magne ic ac ion is equi ed. The nominal b e composi ions a e gi en in Table 1, bu , since he sol en exchange p ocess and u he mixing may lead o a loss o some ma e ial, he esul s o he TGA expe imen s ha e been used o assess he ino ganic (i.e. magne i e) con en wi h g ea e condence. The TGA measu emen s we e conduc ed ae he magne ic cha ac e iza ion on he same samples, he eby elimina ing possible quan i a i e e o s. The magne ic esponses we e cha ac e ized by being ‘Lange in- like’wi h e y small coe ci i y, which is oen ega ded as he signa u e o supe pa amagne ism, see Fig. 5. The concep o supe pa amagne ism, o iginally in oduced by Ne´ el and B own, has as a p e equisi e a anishing coe ci i y when he Fig. 4 (a) Typical s ain–s ess cu es o he aligned PMMA/PEO fib es con aining diffe en amoun s o nanopa icles (0 o 20 w %), and o pu e PMMA (dashed line). (b) Scanning elec on mic og aph o he necking and ac u e su ace o a fib e wi hou magne i e pa icles a e ensile es ing. (c) Scanning elec on mic og aph o a fib e con aining 20 w % magne i e and (d) an enla gemen o he ac u e su ace. Table 2 The ensile p ope ies o he aligned supe pa amagne ic fib es a Fe 3 O 4 (w %) S eng h (MPa) Modulus (GPa) Elonga ion (%) Toughness (MJ m 3 ) 0 29.7 3.1 0.85 0.08 14.1 4.1 2.14 0.28 1 24.0 3.9 0.75 0.12 11.4 2.1 1.82 0.80 10 42.0 5.5 1.13 0.24 8.3 3.0 2.15 0.56 20 28.8 6.2 1.12 0.19 9.8 2.2 1.47 0.55 a S eng h: maximum s ess o he sample, modulus: calcula ed om he ini ial slope, elonga ion: alue when s ess eaches ze o o when sample ac u es, oughness: calcula ed om he o al a ea unde he s ess–s ain cu e. 21418 |RSC Ad .,2016,6, 21413–21422 This jou nal is © The Royal Socie y o Chemis y 2016 RSC Ad ances Pape Open Access A icle. Published on 16 Feb ua y 2016. Downloaded on 24/05/2016 10:30:50. This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial 3.0 Unpo ed Licence. View A icle Online pa icle size is below he c i ical ma e ial-dependen alue. 6 The idea is ha he mal agi a ion, i.e. he ambien empe - a u e, is sufficien o con inuously al e he di ec ion o magne iza ion o he magne ic en i ies. 9,10 The e o e i is no s ic ly jus iable o claim supe pa amagne ism i a ni e coe ci i y has been de ec ed. Since i is in p inciple no possible o ule ou an a bi a y small coe ci i y, he se ious app oach is o epo an expe imen ally de e mined low alue o he coe ci i y. E enso,ap oo o supe pa amagne ismshouldalso in ol ebo h imeand empe a u e.Oneapp oachis heze o- eld-cooled/eld-cooled magne iza ion measu emen , which de e mines he empe a u e (so called ‘blocking empe a u e’) whe e a andomly o ien ed bu e omagne ic ma e ial align wi h a (weak) ex e nal eld upon hea ing due o ha i becomes supe pa amagne ic. P io o his magne iza ion s. empe a u e measu emen , he sample has been cooled wi hou eld,whichincaseo asupe pa amagne ic ma e ial ende he sample, e.g. a collec ion o non-in e ac ing pa icles magne ized in andom di ec ions, an a e age equal o ze o. An ex apola ion o his idea is o measu e in- and ou -o -phase AC suscep ibili y s. empe a u e a diffe en equencies. When he ma e ial unde goes a ansi ion om he e o-/ e i- o he supe pa amagne ic s a e, a peak in bo h he in- and ou -o -phase suscep ibili y is expec ed. The empe a u e a which hein-phasepeaksisusually ega ded as he blocking empe a u e, al hough he ou -o -phase peak is oen mo e dis inc . 31 This empe a u e is dependen on measu emen ime, which in case o AC suscep ome y ansla es in o ecip ocal equency. Highe equency means a highe blocking empe a u e. 32 In Fig. 6 we show an AC suscep ibili y measu emen s. empe a u e om 100 o 300 K a 4 equencies (60–2 kHz), whe e he peak is shied o highe empe a u e wi h inc easing equency (see ESI Fig. S3†). Howe e , hese app oaches canno ule ou e o-/ e imagne ic impu i ies, whe eas a es o coe ci i y is e y sensi i e o he de ec ion o hose and a low alue ensu es a small ac ion. To make his easoning mo e quan i a i e, he coe ci i y can be ela ed o he eld equi ed o each sa u a ion. We sugges ha a use ul c i e ion is ha he coe ci i y should be no g ea e han 1/1000 o he eld equi ed o each 90% o he sa u a ion magne iza ion. This eld is commonly e e ed o as H 90 , and is also he poin whe e he suscep ibili y has dec eased o exac ly 3% o i s ze o eld alue. 33,34 Hence, by his p ac ice a ma e ial is only supe - pa amagne ic i he coe ci i y is less han H 90 /1000. Howe e , in a plo showing 90% o he magne iza ion, he wo aces o inc easing and dec easing eld s eng hs will be indis in- guishable a any meaning ul line wid h. To show he coe ci i y, a magnica ion o a leas 100 is needed. This will help o make meaning ul compa isons be ween epo ed da a and elimina e hose ha a e no en i ely supe pa amagne ic. In he p esen case he coe ci i y was ca. 20 A m 1 ega dless o b e o ien a ion and magne ic con en (below he 1/1000 coe ci i y c i e ion), see Fig. 5c. This alue is e y small, only abou hal o ea hs' eld, and o ou knowledge lowe han any p e iously epo ed alue. An impo an implica ion o his obse a ion o a e y minu e coe ci i y is ha ‘exchange in e - ac ion’due o e y close pa icle–pa icle dis ances is no effec i e since his would p omo e e omagne ism and b eak he supe pa amagne ic beha iou . This in u n means ha all indi idual nanopa icles a e sepa a ed om each o he by a dis ance o a leas ca. 0.1–0.2 nm. In ou polyme -based composi e, he easoning is ha he polyme has encapsu- la ed essen ially all he pa icles, due o a s ong adhesion a he Fig. 5 (a) Hys e esis loops o all samples. (b) 100magnifica ion (x-axis) o he loops o all samples showing dis inc o ien a ion dependence essen ially independen o he o e all magne i e nanopa icle concen a ion. (c) 1000magnifica ion (x-axis) showing he 18 (blue) and 19.5 ( ed) Am 1 coe ci i y in he 10 w % sample. This jou nal is © The Royal Socie y o Chemis y 2016 RSC Ad .,2016,6,21413–21422 | 21419 Pape RSC Ad ances Open Access A icle. Published on 16 Feb ua y 2016. Downloaded on 24/05/2016 10:30:50. This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial 3.0 Unpo ed Licence. View A icle Online polyme –pa icle in e ace isible in he ac u e su aces. The magne iza ion was also e y simila o he diffe en magne i e concen a ions and diffe en b e o ien a ions. A he maximum p obed eld o 500 kA m 1 , whe e he magne iza- ion e sus eld esponse s ill has a conside able slope, he magne iza ion eached 64 A m 2 kg 1 , implying ha he ue sa u a ion magne iza ion was dis inc i ely highe . This esul , ha bo h he sa u a ion magne iza ion (pe weigh o magne ic ma e ial) and he coe ci i y a e independen o pa icle concen a ion and b e o ien a ion, is consis en wi h he iew ha he samples beha e as ensembles o non-in e ac ing supe pa amagne ic pa icles. The ma s we e howe e dis inc i ely easie o magne ize along he b e axis a he han pe pendicula o i , i.e. he ma s showed a p onounced di ec ional aniso opy o he suscep i- bili y bu wi hou concen a ion dependence, see Fig. 5b. The suscep ibili y, which is he slope o he hys e esis loop a he o igin, can also be de i ed heo e ically, and he measu ed da a can be compa ed wi h he heo e ical. Paul Lange in, who  s de i ed he dependence o he magne iza ion o non-in e ac ing (supe )pa amagne ic en i ies on a magne ic eld a i ed a his exp ession o e a hund ed yea s ago: M Msa ¼co hðaÞ1 a whe e ais gi en by: a¼m0mH kBT which a low elds con e ges o: M Msa ¼a 3 whe e Mand M sa a e espec i ely he magne iza ion and sa u- a ion magne iza ion, m 0 is he pe meabili y o acuum (4p 10 7 kg m C 2 ), mis he magne ic momen o a single pa icle (m¼M sa Vwhe e Vis he pa icle olume), His he magne ic eld, k B is he Bol zmann cons an and Tis he absolu e empe a u e. 34 This in u n means ha he slope, i.e. suscep i- bili y c, o he plo o M e sus H is equal o: c¼m0mMsa 3kBT Wi h M sa ¼70 A m 2 kg 1 ( om measu ed a he han li e a u e da a), he densi y o magne i e 5200 kg m 3 , he in insic suscep ibili y o cuboid pa icles wi h ca. 10 nm side leng h ( om TEM mic og aphs) a oom empe a u e (293 K) is c¼14. Acco ding o magne os a ic heo y, he measu ed suscep ibili y is always less han he in insic suscep ibili y due o a demagne izing effec o he geome y and he o ien a ion o he magne ic en i ies. In gene al, his demagne izing effec is low in slende geome ies magne ized along hei long axis. 7 Ou samples we e cylinde -shaped olls o hin shee s o ien ed wi h hei symme y axis pa allel o he magne ic eld in he VSM, and his means ha he magne ic eld was always applied in he shee planes, which has a e y small demagne izing effec . The e o e, he aniso opy was due only o whe he he eld was applied pa allel o o pe pendicula o he b e axis. The measu ed alues we e c¼12–13 o he pa allel and c¼8 o he pe pendicula o ien a ion wi h li le o negligible dependence on he o al pa icle concen a ion, Table 3. The e o e, we conclude ha he magne ic ma e ial was a anged in elonga ed geome ies wi h he long axis pa allel o he b e axis. The o e all pa icle concen a ion had li le effec on his aniso opy, sugges ing ha he size and shape o hese elonga ed en i ies we e independen o he pa icle Fig. 6 In- and ou -o phase suscep ibili y s. empe a u e a equencies om 60 o 2000 Hz o he 20 w % fib e ma aligned pa allel wi h he magne ic field. The measu ing field ampli ude was H¼ 100 A m 1 . Table 3 Measu ed magne ic da a o he supe pa amagne ic fib es a Fe 3 O 4 (w %) Fib e-o ien a ion H C (A m 1 ) M * S (A m 2 kg 1 ) c mass (10 3 m 3 kg 1 ) c (-) ck c 1 Pa allel 20.5 64.3 2.32 12.0 1.54 10 Pa allel 19.5 65.8 2.60 13.5 1.54 20 Pa allel 16.0 63.8 2.36 12.2 1.50 1 Pe pendicula 19.0 63.9 1.51 7.8 — 10 Pe pendicula 18.0 65.6 1.68 8.7 — 20 Pe pendicula 16.5 63.5 1.58 8.2 — a H C : magne ic coe ci i y, M * S : magne iza ion @ 500 kA m 1 no malised wi h espec o he mass o Fe 3 O 4 ,c mass : mass magne ic suscep ibili y a ze o eld, c : olume suscep ibili y, c k /c : a io o he suscep ibili y o he b es aligned pa allel and pe pendicula o he applied eld. 21420 |RSC Ad .,2016,6,21413–21422 This jou nal is © The Royal Socie y o Chemis y 2016 RSC Ad ances Pape Open Access A icle. Published on 16 Feb ua y 2016. Downloaded on 24/05/2016 10:30:50. This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial 3.0 Unpo ed Licence. View A icle Online concen a ion. This uni e sal concen a ion independence means ha i is p obable ha hese en i ies exis ed al eady be o e he spinning e en . This explana ion o he aniso opy o suscep ibili y in combina ion wi h he absence o any sign o exchange in e ac ion implies: (1) Tha he pa icles we e sepa a ed om each o he by a leas a hin ca. 0.1–0.2 nm polyme laye so ha exchange in e ac ion is ineffec i e, and (2) ha he magne ic ma e ial appea ed in elonga ed geome ies wi h hei long axis p e e en ially aligned along he b e axis. Nei he he o ien a ion no he size and shape we e dependen on he o al pa icle concen a ion. Conclusions Elec ospun b e ma s o PMMA/PEO blends ha e been p epa ed wi h up o 20 w % nanosized magne i e. The composi e b e ma s showed no dependence on concen a ion o o ien a ion o he magne iza ion o coe ci i y. Sa u a ion magne iza ion was g ea e han 64 A m 2 kg 1 (no malized wi h espec o magne i e) and he coe ci i y was ex emely low, ca. 20 A m 1 , consis en wi h supe pa amagne ism, which was u he e ied by AC suscep ibili y s. empe a u e measu e- men s a diffe en equencies. I is he e sugges ed ha a low measu ed coe ci i y (<1/1000) in ela ion o he eld needed o each 90% o sa u a ion should be a use ul c i e ion o supe - pa amagne ism. A anishing coe ci i y con ms he absence o a e omagne ic impu i y. This is a mo e s ingen c i e ion, in line wi h Ne´ el's deni ion, han e.g. he de e mina ion o he blocking empe a u e o a e e ence o pa icle size. I also makes i possible o compa e esul s om diffe en wo ke s. The di ec ional dependence o he low eld suscep ibili y is explained by he magne ic ma e ial being geome ically aligned p e e en ially along he b e axis. This di ec ional dependence is independen o he o al magne ic concen a ion. The absence o any sign o exchange in e ac ion (i.e. ul a- low coe ci i y) is explained by he indi idual pa icles being sepa a ed om each o he by a leas 0.1–0.2 nm, which sugges s ha he polyme encapsula e all he pa icles e en i hey appea in clus e s. The mechanical p ope ies o he PMMA/PEO polyme ma e ial we e essen ially main ained wi h nanopa icle loading up o 20 w %, which is p esumably due o s ong nanopa icle/ polyme in e ac ion and he absence o oids and delamina ion, u he suppo ed by he absence o exchange in e ac ion/ magne ic coe ci i y. Due o he good mechanical p ope ies e en a ela i ely high pa icle loadings, i is clea ha he pa icle loading can be signican ly inc eased o gi e e en s onge magne ic unc ionali y. Acknowledgemen s The au ho s acknowledge P o KV Rao o he in oduc ion in o he ascina ing wo ld o magne ic ma e ials and magne ic measu emen s. The nancial aid p o ided by “P og ama “Jos´ e Cas illejo”pa a es ancias de mo ilidad en el ex anje o de j´ o enes doc o s CAS14/00241” o D Cabedo's s ay in S ock- holm, Sweden, au umn 2014 is also acknowledged. No es and e e ences 1 L. X. Tie enaue , A. Tschi ky, G. K¨ uhne and R. Y. And es, Magn. Reson. Imaging, 1996, 14, 391–402. 2 J. T. Mayo, C. Ya uz, S. Yean, L. Cong, H. Shipley, W. Yu, J. Falkne , A. Kan, M. Tomson and V. L. Col in, Sci. Technol. Ad . Ma e ., 2007, 8,71–75. 3 Q. A. Pankhu s , J. Connolly, S. K. Jones and J. Dobson, J. Phys. D: Appl. Phys., 2003, 36, R167. 4 S. Galland, R. L. Ande sson, M. Salajko a, V. S om, R. T. Olsson and L. A. Be glund, J. Ma e . Chem. C, 2013, 1, 7963–7972. 5 C.-L. Zhang and S.-H. Yu, Chem. Soc. Re ., 2014, 43, 4423– 4448. 6 D. J. Dunlop, J. Geophys. Res., 1973, 78, 1780–1793. 7 B. D. Culli y and C. D. G aham, In oduc ion o magne ic ma e ials, John Wiley & Sons, 2011. 8 G. F. Goya, T. S. Be qu´ o, F. C. Fonseca and M. P. Mo ales, J. Appl. Phys., 2003, 94, 3520–3528. 9L.N ´ eel, Ann. Geophys., 1949, 5,99–136. 10 W. F. B own J , J. Appl. Phys., 1963, 34, 1319–1320. 11 C. P. Bean and J. D. Li ings on, J. Appl. Phys., 1959, 30, S120– S129. 12 H. Al hues, J. Henle and S. Kaskel, Chem. Soc. Re ., 2007, 36, 1454–1465. 13 I. G een eld, K. Fezzaa, M. H. Ra ailo ich and E. Zussman, Mac omolecules, 2012, 45, 3616–3626. 14 C. J. Goss, Phys. Chem. Mine ., 1988, 16, 164–171. 15 V. S ¨ om, R. T. Olsson and K. V. Rao, J. Ma e . Chem., 2010, 20, 4168–4175. 16 B. W. Ahn, M. H. Hahn and T. J. Kang, 18 h In .Con . on composi e m ls, Ko ea, 2011. 17 Y. K. Sung, B. W. Ahn and T. J. Kang, J. Magn. Magn. Ma e ., 2012, 324, 916–922. 18 R. L. Ande sson, M. Salajko a, P. E. Mallon, L. A. Be glund, M. S. Hedenq is and R. T. Olsson, J. Polym. En i on., 2012, 20, 967–975. 19 F. Mei, S. Val e , T. O. Richa d, B. Lyubo and K. V. Rao, Nano echnology, 2012, 23, 145601. 20 R. L. Ande sson, A. Ma ´ ınez-Abad, J. M. Laga on, U. W. Gedde, P. E. Mallon, R. T. Olsson and M. S. Hedenq is , In . J. Mol. Sci., 2014, 15, 15912–15923. 21 R. L. Ande sson, V. S ¨ om, U. W. Gedde, P. E. Mallon, M. S. Hedenq is and R. T. Olsson, Sci. Rep., 2014, 4, 6335. 22 J. Khan, S. E. Ha on, P. Akco a, B. C. Benicewicz and S. K. Kuma , Mac omolecules, 2009, 42, 5741–5744. 23 P. J. Goodhew, J. Humph eys and R. Beanland, Elec on mic oscopy and analysis, CRC P ess, Thi d edn, 2000. 24 A. D. McNaugh , IUPAC Compendium o Chemical Te minology: The Gold Book, In e na ional Union o Pu e and Applied Chemis y, 2005. 25 I. N. S ain, Q. Wu, A. M. Pou ahimi, M. S. Hedenq is , R. T. Olsson and R. L. Ande sson, J. Ma e . Chem. A, 2015, 3, 1632–1640. This jou nal is © The Royal Socie y o Chemis y 2016 RSC Ad .,2016,6, 21413–21422 | 21421 Pape RSC Ad ances Open Access A icle. Published on 16 Feb ua y 2016. Downloaded on 24/05/2016 10:30:50. This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial 3.0 Unpo ed Licence. View A icle Online