Effect of the dispersion state of minerals on the properties of cellulose nanofiber-based composite films

Applied Clay Science 233 (2023) 106823 0169-1317/© 2023 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by- nc-nd/4.0/). Resea ch Pape E ec o he dispe sion s a e o mine als on he p ope ies o cellulose nano ibe -based composi e ilms Luís Al es a , * , Ana Ramos b , Edua do Fe az c , d , Ped o Sanguino e , Julio San a ´ en , Ma ia G. Ras ei o a , Jos´ e A.F. Gamelas a , * a Uni e si y o Coimb a, CIEPQPF, Depa men o Chemical Enginee ing, Rua Síl io Lima, P´ olo II, Coimb a 3030-790, Po ugal b FibEnTech, Depa amen o Chemis y, Uni e si y o Bei a In e io , Rua Ma quˆ es d’ ´ A ila e Bolama, Co ilh˜ a 6201-001, Po ugal c TECHN&ART, Poly echnic Ins i u e o Toma , Quin a do Con ado , Es ada da Se a, Toma 2300-313, Po ugal d Geobio ec, Uni e si y o A ei o, Geosciences Depa men , Campus Uni e si ´ a io de San iago, A ei o 3810-193, Po ugal e Uni e si y o Coimb a, CEMMPRE, Mechanical Enginee ing Depa men , Rua Luís Reis San os, Coimb a 3030-788, Po ugal Fo me ly TOLSA, SA, Resea ch & Technology o New Businesses, C a. de Mad id a Ri as Ja ama, 35, Mad id 28031, Spain ARTICLE INFO Keywo ds: T ansmission elec on mic oscopy Clay dispe sion Composi es Cellulose nano ib ils T anspa ency Clay mine als ABSTRACT The dispe sion s a e and he e iciency o he mix u e o he di e en componen s in a composi e ilm ha e an impo an impac on i s mechanical and op ical p ope ies. In he p esen wo k, he impac s o di e en dispe sion ea men s on he disagg ega ion s a e o ib ous clay pa icles in wa e , and on he p ope ies o ela ed cellulose nano ibe (CNF)-based composi e ilms, we e e alua ed. X- ay di ac ion s udies, pe o med on samples o sepioli e and palygo ski e, e ealed only mino changes in he di ac ion pa e n when he mine als we e subjec ed o ul asonic ea men , wi h o wi hou he addi ion o di e en chemical dispe sing agen s. Con e sely, mic oscopic s udies e ealed impo an di e ences in he dispe sion s a e o he samples, induced by he addi ion o he di e en dispe san s, showing an imp o emen in he disagg ega ion o he mine al c ys als. The composi e ilms p epa ed wi h sepioli e (and ca boxyme hylcellulose, as chemical dispe san ) dispe sed using ul asonic ea men , and di e en ypes o CNF, showed imp o ed op ical and mechanical p ope ies when compa ed wi h composi es o he same coun e pa s p epa ed wi h sepioli e dispe sed using a high-speed shea dispe se . 1. In oduc ion New bio-based al e na i es o plas ic-based ma e ials, mainly o he single-use ones, a e a ho opic no only in academic esea ch bu also a conce n o he popula ion in gene al. Plas ic ma e ials a e e y well es ablished, due o hei good p ope ies and low cos . Howe e , he eme ging en i onmen al conce ns, ela ed o he mac o- and mic o- plas ic pollu ion in soils, oceans and wa e s eams a e inc easing e e y day (Magalh˜ aes e al., 2020). Thus, i is u gen o de elop ma e ials wi h p ope ies sui able o subs i u e plas ics in mos common applica ions, mainly o single-use pu poses, such as ood packaging. Cellulose nano ibe (CNF)-based ilms a e ega ded as new ma e ials wi h g ea po en ial o he p oposed end. CNFs a e e y a ac i e due o hei high biocompa ibili y and biodeg adabili y, supe io ilm- o ming capabili y and good mechanical p ope ies (Zeng e al., 2021). Ne e heless, CNFs a e expensi e nanoma e ials, being he high cos ela ed o he high ene gy and/o chemical p e- ea men s, usually equi ed o hei p epa a ion. Fu he mo e, he p oduced ilms also p esen some limi a ions, such as low wa e apou ba ie p ope ies, mainly in high-mois u e condi ions. To imp o e he ba ie p ope ies and educe cos s, no el s a egies a e equi ed o o e come he d aw- backs o he CNF-based ilms and po en ia e he indus ializa ion o hese bio-based ma e ials. One o he mos p omising s a egies in ol es he p epa a ion o CNF-mine al composi es (Al es e al., 2019; Gamelas and Fe az, 2015; Gonz´ alez del Campo e al., 2018). Fib ous clay mine als, such as sepioli e and palygo ski e, a e in e - es ing ma e ials ha ind use in a wide ange o applica ions, o ins ance, CO 2 cap u e (Cecilia e al., 2018), clay- ubbe composi es (Takei e al., 2013), nano ib illa ed cellulose-clay nanocomposi es (Al es e al., 2019; Gamelas and Fe az, 2015) o d ug deli e y ca ie s (Wang e al., 2014). Sepioli e and palygo ski e a e na u al ib ous clays whose s uc u e is based on blocks al e na ing wi h ca i ies ( unnels), * Co esponding au ho s. E-mail add esses: [email p o ec ed] (L. Al es), [email p o ec ed] (J.A.F. Gamelas). Con en s lis s a ailable a ScienceDi ec Applied Clay Science jou nal homepage: www.else ie .com/loca e/clay h ps://doi.o g/10.1016/j.clay.2023.106823 Recei ed 31 Augus 2022; Recei ed in e ised o m 3 Janua y 2023; Accep ed 9 Janua y 2023 Applied Clay Science 233 (2023) 106823 2 which gi e ise o highly po ous amewo ks. The ideal o mula o sepioli e is Mg 8 Si 12 O 30 (OH) 4 (OH 2 ) 4 .8H 2 O and (Mg,Al,Fe) 5 Si 8 O 20 (O- H) 2 (OH 2 ) 4 .4H 2 O o palygo ski e. An impo an cha ac e is ic o hese mine als is he high densi y o silanol g oups on he pa icle su ace, much highe han ha o pla ele -like clay mine als. These g oups o igina e om he bonding be ween non-sha ed oxygens om he e ahed al silica shee s (loca ed on channels) and hyd ogen (Guggen- heim and K ekele , 2011). Hyd ogen bonding and Van de Waals in e ac ions con ibu e o he p esence o agg ega es o he acicula indi idual pa icles o ib ous clays in aqueous suspensions, which limi s he quali y o he clay dispe sion (Wang and Wang, 2019). Fo some applica ions, such as bionanocomposi es, a good dispe sion o he clay mine al in a selec ed medium, p e e en ially wa e -based, is essen ial o achie e supe io p ope ies in he inal ma e ials (Al es e al., 2019). To imp o e he dispe sibili y o ib ous mine als, se e al s a egies ha e been e alua ed mainly based on mechanical ea men s, he addi ion o chemical dis- pe san s o he suspension and he chemical modi ica ion o he mine al su ace (Al es e al., 2020b; Fe az e al., 2021; Ga cía e al., 2011). Se e al di e en mechanical ea men s ha e been ied, such as ul asonica ion, high-speed shea ing, o high-p essu e homogeniza ion, among o he s; some o hese ea men s demons a ed he abili y o dis up he bulk bundles in o smalle bundles o ods (Wang and Wang, 2016). Ul asonica ion has been p esen ed as a ela i ely simple app oach o disagg ega e sepioli e ibe s and enhance he s abili y o he suspension, which is no possible o be accomplished using low-powe dispe sing equipmen , such as magne ic s i ing (Al es e al., 2020b). Fib ous clays ea ed in wa e by ul asonica ion we e also demons a ed o ha e hei p ope ies imp o ed. Ul asonica ion has also been demons a ed o be a use ul me hod o gene a e homogeneous and s a- ble dispe sions wi h o he di e en ma e ials in wa e (Lo Dico e al., 2019). The addi ion o dispe sing agen s can also be used o imp o e he s abili y o he dispe sions and ex end he applica ion o ib ous clays (Al es e al., 2020b; Fe az e al., 2021). The added dispe sing agen s can in e ac , mainly, by physicochemical in e ac ions such as hyd ogen bonding and/o hyd ophobic in e ac ions wi h he mine al pa icles (Benli e al., 2012) and change he cha ac e is ics o he ods o he ib ous clay, namely hei cha ge and su ace chemis y, imp o ing he dispe sion quali y. The choice o he dispe san is c ucial o he a ge esul . Phospha es like sodium hexame aphospha e a e he mos popula dispe sing agen s o mine als (Takei e al., 2013). Also, o he agen s ha e been es ed, anging om syn he ic polyme s o bio-based addi- i es, such as cellulose de i a i es (Al es e al., 2020b). In he p esen wo k, he in luence o ul asonica ion and he addi ion o dis inc chemical dispe san s (sodium polyphospha e, hyd ophobi- cally modi ied poly(sodium ac yla e) (HM-PAA), and sodium ca boxy- me hylcellulose (CMC)) on he dispe sion s a e o wo ib ous mine als (sepioli e and palygo ski e) in wa e , and he impac on he p ope ies o CNF-based composi e ilms, we e e alua ed. The capaci y o each dispe sing sys em was s udied by ansmission elec on mic oscopy (TEM), and possible changes in he mine al s uc u e, induced by he ul asonica ion ea men , as well as by he addi ional in e ac ion o he mine al c ys al bundles wi h he dispe sing agen s we e s udied by X- ay di ac ion. The CNF-based composi e ilms we e p oduced using sepi- oli e suspensions p epa ed by ul asonica ion and compa ed wi h com- posi e ilms p oduced wi h sepioli e dispe sed in a high-shea dispe se . In he li e a u e, mos o he epo ed wo ks deal wi h plana mine als (Al es e al., 2019). Few s udies epo ing composi es using ib ous mine als, such as sepioli e, a e a ailable (Ghanadpou e al., 2018; Gonz´ alez del Campo e al., 2018; Ma ín-Samped o e al., 2022). Addi- ionally, o he bes o ou knowledge, no s udy in li e a u e epo s he e ec o di e en dispe sing equipmen on he CNF-mine al composi e ilm p ope ies. 2. Ma e ials and me hods 2.1. Clays and chemicals In his s udy, wo sepioli e samples e e ed o as sepioli e 1 (Sep 1) and sepioli e 2 (Sep 2), ob ained om he deposi o Vallecas-Vic´ al a o (Mad id, Spain), and one palygo ski e sample, om he deposi loca ed in M’bou , egion o Thi` es (Senegal), we e supplied by Tolsa, SA (Mad id, Spain). Sepioli e 1 was p ocessed by a d y mic oniza ion p o- cess using a je mill o b eak he ibe bundles down in o mic on-size pa icles. Sepioli e 2 was mic onized using a we p ocess ha p o- duces an ex ensi e deagglome a ion o he sepioli e ibe bundles wi hou a ec ing hei aspec a io (Al es e al., 2020b). Palygo ski e (PAL) was p ocessed using a olle mill ea men o a inal pa icle size o less han 45 μ m (Fe az e al., 2021). The dispe sing agen s used we e: sodium polyphospha e (Emplu a g ade, 99%, 106529) pu chased om Sigma-Ald ich (Me ck); hyd o- phobically modi ied poly(sodium ac yla e) (Acusol 820, 30% (solids) emulsion consis ing o 40% me hac ylic acid, 50% e hyl ac yla e and 10% s ea yl oxypoly e hyl me hac yla e) acqui ed om Rohm and Haas, USA; sodium ca boxyme hylcellulose (molecula weigh o 250 kDa, deg ee o subs i u ion o 0.7) ob ained om Sigma-Ald ich (Me ck). All chemicals we e used as ecei ed wi hou any u he pu i ica ion. Cellulose nano ibe s used in he p esen wo k we e ob ained and cha ac e ized ollowing p e iously desc ibed p ocedu es (Al es e al., 2022). Two di e en ypes o CNF we e used in he p esen wo k, namely Mec CNF (mechanical) and TEMPO CNF (TEMPO-oxidised). 2.2. P epa a ion o he mine al suspensions wi h di e en chemical dispe san s Suspensions wi h 1.0 w % o ib ous clay we e p epa ed a oom empe a u e om hei espec i e d y powde s. The clay powde was added o dis illed wa e and he mix u e was s i ed using a magne ic s i e a 200 pm; igh a e , he dispe sing agen was added a a mass a io o 1:10 (dispe san /clay). Then, he suspension p epa a ion was inished using an ul asound p obe (Vib a-cell VC 505, Sonics) wo king a 60% ampli ude and 1 s pulse, o 10 min. Suspensions we e also p epa ed wi hou he addi ion o chemical dispe san . Then, suspensions con aining 0.05 w % clay we e p epa ed o he ansmission elec on mic oscopy analysis by dilu ion o he concen a ed samples. 2.3. Cha ac e iza ion o he mine al samples Fi e samples o each clay mine al, i.e., he s a ing ma e ial and ou suspensions p epa ed by dispe sion o he mine al by ul asonica ion, wi hou and wi h addi i es, we e cha ac e ized in e ms o mine alog- ical cha ac e iza ion by powde X- ay di ac ion using a Siemens D5000 X- ay di ac ome e . A CuK α 1 adia ion sou ce wi h λ =1.54056 Å, ocused by a p ima y Ge c ys al monoch oma o , was used. The de ec o is a s anda d scin illa ion coun e . The Cu ube uns a 40 mA and 40 kV. The s a ing clay powde s and he powde s o he dispe sed samples d ied a 105 ◦C we e placed in p ope suppo . Di ac og ams we e collec ed by he coun ing me hod (s ep 0.02◦and ime 1.0 s) in he 2θ ange o 5–60◦. The mo phology o he pa icles in he d y powde s o he s a ing ma e ials was e alua ed by ield emission scanning elec on mic oscopy (FE-SEM) in a Ca l Zeiss Me lin-61-50 mic oscope. Seconda y elec on mode and accele a ion ol age o 1 kV we e he ope a ional condi ions. The samples we e p e iously spu e -coa ed wi h a hin ilm o Au/Pd. High- esolu ion ansmission elec on mic oscopy (HR-TEM) was ca ied ou in a JEOL JEM-2100 equipmen o in e he dispe sion s a e, mo phology and size o he clay c ys als in he suspensions, wi h and wi hou dispe sing agen , in aqueous medium. A small amoun o each suspension (10 μ L) was deposi ed on a ca bon-coa ed coppe g id, and he samples we e obse ed using an accele a ing ol age o 200 kV. L. Al es e al. Applied Clay Science 233 (2023) 106823 3 2.4. P epa a ion o he CNF-based composi e ilms Sepioli e (sepioli e 1) suspension in wa e (1.0 w %), using CMC as dispe sing agen , was p epa ed ollowing he p ocedu e desc ibed abo e (sec ion 2.2). Fo compa ison, sepioli e suspension (wi hou CMC) was p epa ed using a high-shea dispe se (Dispe ma CV3-PLUS-E, VMA- Ge zmann GmbH, Reichsho , Ge many) a 5000 pm o 15 min. Di e en ilms we e p epa ed using il a ion ollowed by ho p essing, acco ding o a p ocedu e p e iously desc ibed (Al es e al., 2022). B ie ly, suspensions con aining 0.2 w % o he di e en CNF o CNF + mine al we e dispe sed using he high-shea dispe se a 1000 pm o 10 min. A e , he p epa ed suspensions we e dewa e ed using a il a- ion uni pu chased om Kimble Ul a-Wa e Fil a ion Sys ems (DWK Li e Sciences GmbH, Mainz, Ge many) and memb anes o cellulose ac- e a e wi h 0.45 μ m po e and 90 mm diame e (Fil a ech). The dewa- e ed “we cakes” we e inally d ied using a apid d ye o labo a o y shee s (Lo en zen & We e, model 257, Lo en zen & We e GmbH, Munich, Ge many). The d ying was pe o med using a empe a u e o 110 ◦C o 10 min. Fig. 1. HR-TEM mic og aphs o he clay samples a 15,000×magni ica ion: a) sepioli e 1, c) sepioli e 2, and e) palygo ski e suspensions in aqueous medium, ea ed wi h an ul asonic p obe. The scale ba s co espond o 100 nm. F equency dis ibu ions o he hickness (c oss sec ion) o he pa icles ob ained om HR-TEM mic og aphs o b) sepioli e 1, d) sepioli e 2 and ) palygo ski e samples. L. Al es e al. Applied Clay Science 233 (2023) 106823 4 2.5. Cha ac e iza ion o he CNF-based composi e ilms The p epa ed ilms we e cha ac e ized ollowing he p ocedu es desc ibed elsewhe e (Al es e al., 2022). The basis weigh , hickness, anspa ency, and ensile p ope ies ( ensile s eng h, Young’s modulus and elonga ion a b eak) we e ob ained acco ding o he s anda ds a ailable. In b ie , he hickness o he ilm was de e mined using a Adamel Lhoma gy mic ome e (model MI 20, Roissy-en-B ie, F ance), based on 10 measu emen s o 2 eplica e ilms; he epo ed alues co espond o he a e age alues de e mined. The ilm anspa ency (a e age o 2 eplica e ilms) was assessed acco ding o he ISO 22891 s anda d, using a Technidyne Colo Touch 2 spec opho ome e om Technidyne Co po a ion (New Albany, Indiana, USA), and applying he illuminan D65 and he obse e 10◦. The ensile p ope ies we e de e mined acco ding o he ISO 1924-1 s anda d, a 23 ±1 ◦C and 50 ±2% RH, using a ensile es e om Thwing-Albe Ins umen Co. (EJA se ies, Wes Be lin, NJ, USA), wi h he ollowing se up: a ensile a e o 5 mm pe minu e and an ini ial gap (be ween g ips) o 5 cm. The epo ed alues o he ensile p ope ies ( ensile s eng h, Young’s modulus and elonga ion a b eak) co espond o he a e age alues o ou specimens o each ilm ype. 3. Resul s and discussion 3.1. Cha ac e iza ion o he ib ous clay dispe sions h ough elec on mic oscopy The HR-TEM images o suspensions o he wo sepioli e samples and he palygo ski e sample in wa e , ea ed wi h an ul asonic p obe, a e p esen ed in Fig. 1. The HR-TEM images e ealed ha he c ys als we e no ully indi- idualized, using he ul asound p obe ea men , being obse ed ag- g ega es o se e al c ys als, “ ace- o- ace”, in he h ee samples. Bundles wi h hickness (c oss sec ion) in he ange o 13–83 nm o Sep 1, 21–83 nm o Sep 2 and 15–84 nm o he palygo ski e sample, can be obse ed. The equency dis ibu ion o hicknesses (based on 40 mea- su emen s o each sample) shows ha he mode hickness o Sep 1 is 30–40 nm, o he case o Sep 2 is 40–50 nm, and 30–50 nm o he palygo ski e sample. In leng h, all he samples p esen c ys als wi h dimensions up o 1–2 μ m, being obse ed pa icles o 3–4 hund ed nanome e s, which can co espond o pieces o c ys al bundles b oken du ing he ul asonic ea men . The ib ous na u e o hese wo mine als was also obse ed in he FE- SEM images o he s a ing sepioli e and palygo ski e ma e ials. The obse ed ibe s ha e ca. 50–100 nm nanome e s in hickness and a ew mic ome es in leng h (Fig. S1). The ibe s consis o c ys al bundles o small dimensions. The ull disagg ega ion o he c ys al bundles is a challenging s ep and an impo an pa ame e o ob ain composi e ilms wi h he desi ed p ope ies. On he o he hand, an ex ensi e disagg e- ga ion ea men could e en ually lead o c i ical changes in he min- e al’s s uc u e. To access possible changes in he mine al s uc u e, X- Ray di ac ion was pe o med in he mine al samples be o e and a e dispe sion. 3.2. X- ay di ac ion cha ac e iza ion o he clay dispe sions The X- ay di ac og ams o he wo sepioli e samples and he paly- go ski e sample a e p esen ed in Fig. 2. The pa ame e s ob ained o he di e en samples a e esumed in Table 1. The X- ay di ac og ams o he sepioli e samples (d y powde ), e ealed ha he dominan phase o hese samples is sepioli e (2θ =7.3◦, d =12.1 Å o bo h sepioli e samples); on he o he hand, he paly- go ski e sample (d y powde ), showed a dominan peak a 2θ =8.4◦, d =10.5 Å, co esponding o palygo ski e phase, being he peak a 2θ = 26.6◦(d =3.3 Å), co esponding o qua z he second mo e in ense. The p esence o con aminan s was mo e p onounced o he palygo ski e compa ed o he sepioli e samples, as epo ed be o e (Al es e al., 2020b; Fe az e al., 2021). A e dispe sion in aqueous medium, using ul asonica ion as phys- ical ea men , some sligh changes in he X- ay di ac ion pa e ns o he samples we e no iced. Howe e , no ob ious changes in he cha ac- e is ic di ac ion peaks o he sepioli e and palygo ski e phases we e obse ed by ea men wi h he ul asound p obe, indica ing ha he basic s uc u e o sepioli e and palygo ski e was p ese ed. In pa icula , a e he ul asonic ea men he di ac ion peak S 8 (080), cha ac e - is ic o sepioli e, became mo e in ense, as measu ed by i s ela i e in- ensi y no malized o he S 1 (110) peak (Table 1), o bo h he sepioli e samples, which can be a ibu ed o disagg ega ion o he sepioli e i- be s, which enabled he andom o ien a ion o he ibe s in he dispe - sion (Gonz´ alez del Campo e al., 2018). On he o he hand, he S 2 (130) peak o e y low in ensi y disappea ed a e he ul asound ea men (Sep 2) and he S 6 (131) peak is absen in bo h he sepioli e samples a e ul asonica ion; he disappea ance o hese wo peaks can sugges some diso de in he s acking o sepioli e polysomes, induced by he physical ea men (K ekele and Guggenheim, 2008). Fo he palygo ski e sample, he ul asonica ion ea men did no lead o la ge changes in he in ensi y o he palygo ski e cha ac e is ic peaks and did no esul in he disappea ance o hese di ac ion peaks con i ming he p ese a ion o he ib ous mine al s uc u e. 3.3. E ec o addi i es on he dispe sion s a e and s uc u e o he mine al The HR-TEM images o he h ee clay samples, dispe sed using he ul asonic ea men , wi h he addi ion o h ee di e en dispe sing agen s (polyphospha e, HM-PAA and CMC) a e p esen ed in Fig. 3. The addi ion o he dispe sing agen s led, in gene al, o a be e dispe sion o he mine al c ys als in aqueous medium. Compa ing he images ob ained wi hou he addi ion o dispe sing agen s (Fig. 1) wi h he ones shown in Fig. 3 (wi h dispe sing agen s), i is possible o obse e subs an ial di e ences in he agg ega ion/disagg ega ion s a e o he c ys als. Wi h he addi ion o polyphospha e and CMC as dispe sing agen s, good pe o mance in he dispe sion was obse ed o he h ee mine al samples. C ys als wi h ew nanome e s in hickness (Sep 1 +poly- phospha e: 10–30 nm; Sep 1 +CMC: 10–30 nm; Sep 2 +polyphospha e: 20–50 nm; Sep 2 +CMC: 10–40 nm; Pal +polyphospha e: 10–40 nm; Pal +CMC: 10–40 nm), and 300–1000 nm in leng h ( o sepioli e samples) and 250–900 nm ( o palygo ski e) a e obse ed (see Fig. S2 o hickness equency dis ibu ions). Only a ew “ ace- o- ace” c ys als could be obse ed in hese samples. On he o he hand, he use o HM- PAA can also educe he “ ace- o- ace” agg ega es in he h ee mine al samples, bu an ex ensi e “ andom” eagg ega ion is obse ed, due o he lowe s abili y induced by he use o his dispe sing agen . The abili y o his dispe sing agen o disagg ega e he c ys al bundles o mine als was p e iously epo ed (Al es e al., 2020b; Fe az e al., 2021), bu due o he na u e o his polyme , hyd ophobically modi ied, he suspensions possess less s abili y ( o ma ion o gel occu ed wi h ime), and eagg ega ion was obse ed. Some b eaking o nanoc ys als was obse ed, wi h his up u e being induced by he ene gy eleased by he ul asonica o , ope a ing by ca i a ion phenomena. Ca i a ion is e y e icien o disagg ega e he mine al c ys al bundles; howe e , due o i s g ea in ensi y, i can esul in he b eakage and consequen size (leng h) educ ion o indi idualized nanoc ys als (Neaman and Singe , 2000). In he samples wi hou dispe sing agen , his b eaking occu ed o a lesse ex en due o he lowe disagg ega ion achie ed, which enables he bundles o suppo be e he ene gy om he ea men . Fu he mo e, possible s uc u al changes in he mine als, induced by he addi ion o he dispe sing agen s, we e s udied by X- ay di ac ion (Fig. 2 and Table 1). The addi ion o he dispe sing agen s seemed no o p oduce L. Al es e al. Applied Clay Science 233 (2023) 106823 5 Fig. 2. X- ay di ac ion o he sepioli e samples (Sep 1 – op; Sep 2 – middle) and palygo ski e sample (Pal - bo om), wi h he iden i ica ion o he clay mine als. L. Al es e al. Applied Clay Science 233 (2023) 106823 6 subs an ial changes in he main s uc u al ea u es o sepioli e and palygo ski e, al hough some mino di e ences we e no ed, as ollows. The addi ion o CMC led o an inc ease in he in ensi y o di ac ion peak S 8 (080) o bo h he sepioli e samples, s. he ea men wi h only ul asonica ion, e ealing a be e disagg ega ion o he sepioli e ibe s (Gonz´ alez del Campo e al., 2018). Con e sely, he low-in ensi y peaks, S 2 (130) and S 10 (281 o 441), disappea ed a e he ul asound ea - men when CMC was p esen , sugges ing some diso de in he s acking o he polysomes; he same beha iou was obse ed o sepioli e 2 wi h he addi ion o polyphospha e. Fo he palygo ski e sample, i was no ed he disappea ance o he low-in ensi y P 9 peak a e he ul asound ea men ei he wi h he addi ion o polyphospha e o CMC. Howe e , as o he cha ac e is ic palygo ski e di ac ion peaks, he di ac o- g ams a e he ul asonica ion ea men wi h added dispe san s almos ma ched ha o he ini ial palygo ski e sample. The a ou able in e ac ion o he biopolyme CMC wi h he c ys al pa icles o he di e en clays is expec ed o allow a be e dispe sion, as well, as be e s abiliza ion o mine al pa icles in aqueous suspensions. Conside ing he na u e o he biopolyme and sepioli e/palygo ski e, he in e ac ions in ol ed a e hypo hesized o be o wo ypes: hyd ogen bonding be ween he hyd oxyl g oups o he biopolyme and he silanol g oups on he ou e su ace o he mine al pa icles, and ia Van de Waals o ces, due o he amphiphilic na u e o he cellulose de i a i e and sepioli e/palygo ski e, wi h speci ic pa ches o hei s uc u e wi h di e en pola i ies (Benli e al., 2012; Lindman e al., 2021; Wang and Wang, 2016). The FTIR spec oscopy p o ided an addi ional in o ma- ion o his issue. The FTIR spec um o sepioli e a e ul asonica ion (wi hou added biopolyme ) was easonably simila o ha o he o ig- inal mine al sample (Fig. S3, o he egion o he OH s e ching in he Si- O-H, Mg-O-H and Mg-OH 2 bonds), showing he bands expec ed o he mine al. Howe e , he spec um o sepioli e a e ul asonica ion wi h added CMC (Fig. S3) showed a ema kable dec ease o he in ensi y o all he bands in he men ioned egion and, in pa icula , he e y weak band due o he OH s e ching in Si-O-H bonds a 3717 cm −1 disappea ed. The la e obse a ion is an indica ion o he exis ence o in e ac ions by hyd ogen bonding be ween he silanol g oups o he mine al su ace and he hyd oxyl g oups o he CMC (Ma ín-Samped o e al., 2022). B eaking o Mg-O bonds in he oc ahed al laye on he su ace o he sepioli e c ys als may also ha e occu ed. On he o he hand, he use o HM-PAA, an amphiphilic polyme , as dispe sing agen , also showed a low impac on he s uc u al pa ame e s o he mine als, when dispe sed in aqueous suspensions using ul a- sonica ion. The ype o in e ac ions o his polyme wi h he clays is expec ed o be simila o he case o CMC, bu he inc eased hyd o- phobici y (in oduced by he p esence o long hyd ophobic chains) which delays he expansion o he polyme (Al es e al., 2015) led o poo s abiliza ion o he suspensions, being obse ed agg ega es o “ andom” o ien ed c ys als in all he samples using his dispe san (Al es e al., 2020b; Fe az e al., 2021). In sum, he ob ained esul s indica e ha he addi ion o dispe sing agen s, such as polyphospha e and CMC ep esen s a good ou e o ob ain ib ous clay suspensions wi h supe io disagg ega ion and s a- bili y, wi hou changing he ib ous mine al s uc u e, when used as a complemen o ul asonic ea men . 3.4. E ec o dispe sion s a e o sepioli e on he CNF-based composi e ilm p ope ies The dispe sion s a e and mix u e o he di e en componen s in composi e ilms play a key ole on he mechanical and op ical p ope ies o he composi e (Al es e al., 2019). In Fig. 4 a e p esen ed pho os o TEMPO oxidised CNF-based ilms, nea and wi h he addi ion o 20% sepioli e, wi h he mine al p e iously dispe sed using di e en ap- p oaches: a high-shea dispe se (TEMPO 55 +SEP) and applying ul asonica ion and a chemical dispe san (TEMPO 55 +SEP Son). CMC was chosen as dispe sing agen , ins ead o he classical dispe san polyphospha e, due o i s bio-based na u e and good dispe - sion abili y. Mo eo e , polyphospha e has in i s composi ion phospho- ous which can always induce unwan ed wa e eu ophica ion. On he o he hand, sepioli e 1 was chosen, due o he sligh ly be e dispe sion appa en ly ob ained (i.e., sligh ly lowe hicknesses o he dispe sed c ys als, as shown in sec ion 3.3). The anspa ency o he ilms con aining 20% sepioli e is dependen on he dispe sion me hod used, being he ilm p epa ed wi h sepioli e dispe sed using ul asonica ion mo e anspa en han he ilm p oduced om he mine al suspension p epa ed wi h he high-shea dispe se . The be e dispe sion s a e ob ained using he i s dispe sion me hod allows a be e mix u e o bo h he componen s, being he size o he cellulose ib ils and he sepioli e c ys als in he same o de o magni ude (Al es e al., 2020a; Al es e al., 2020b), con a y o he case whe e he sus- pension o sepioli e was dispe sed wi h he high-shea dispe se , which p o ided pa icles o highe size, ha can be in e ed by he phase Table 1 Pa ame e s o Sep 1, Sep 2 and Pal om X- ay di ac ion da a. Sample Peak hkl D y powde Wi hou addi i es Wi h polyphospha e Wi h HM-PAA Wi h CMC d obse ed (Å) I ela i e d obse ed (Å) I ela i e d obse ed (Å) I ela i e d obse ed (Å) I ela i e d obse ed (Å) I ela i e Sep 1 / Sep 2 S 1 110 12.1 / 12.1 100 / 100 12.0 / 11.9 100 / 100 12.0 / 11.8 100 / 100 12.1 / 12.0 100 / 100 11.8 / 11.9 100 / 100 S 2 130 7.4 / 7.5 3 / 2 7.5 / — 1 / — 7.4 / — 1 / — 7.5 / 7.5 2 / 1 — / — — / — S 3 200 o 040 6.7 / 6.7 4 / 5 6.6 / 6.6 9 / 10 6.6 / 6.6 12 / 14 6.7 / 6.7 9 / 8 6.6 / 6.6 13 / 11 S 4 150 5.0 / 5.0 2 / 1 5.0 / 4.9 2 / 2 — / 5.0 — / 1 5.0 / 5.0 2 / 2 5.0 / 5.0 2 / 3 S 5 060 4.5 / 4.5 5 / 3 4.4 / 4.4 5 / 6 4.4 / 4.4 7 / 8 4.4 / 4.4 6 / 5 4.4 / 4.4 14 / 11 S 6 131 4.3 / 4.3 3 / 2 — / — — / — — / — — / — — / — — / — — / — — / — S 7 260 3.7 / 3.7 6 / 4 3.7 / 3.7 4 / 3 3.7 / 3.7 2 / 2 3.7 / 3.7 4 / 3 3.7 / 3.7 2 / 1 S 8 080 3.3 / 3.3 26 / 18 3.3 / 3.3 48 / 51 3.3 / 3.3 49 / 57 3.3 / 3.3 46 / 37 3.3 / 3.3 90 / 76 S 9 510 o 0,10,0 2.7 / 2.7 3 / 4 2.7 / 2.7 7 / 8 2.7 / 2.7 8 / 10 2.7 / 2.7 7 / 6 2.7 / 2.7 14 / 11 S 10 281 o 441 2.6 / 2.6 4 / 3 2.6 / 2.6 2 / 1 2.6 / — 1 / — 2.6 / 2.6 2 / 1 — / — — / — S 11 312 o 062 o 2,10,1 2.2 / 2.2 3 / 3 2.2 / 2.2 4 / 5 2.2 / 2.2 5 / 6 2.2 / 2.2 4 / 4 2.2 / 2.2 9 / 7 Pal P 1 110 10.5 100 10.5 100 10.5 100 10.5 100 10.5 100 P 2 200 6.4 15 6.4 17 6.4 19 6.4 17 6.4 19 P 3 130 5.4 9 5.4 6 5.4 5 5.4 6 5.4 5 P 4 040 4.5 16 4.5 11 4.5 7 4.5 9 4.5 8 P 5 310 4.1 8 4.1 8 4.1 4 4.1 6 4.1 3 P 6 240 3.7 7 3.7 6 3.7 5 3.7 5 3.7 5 P 7 400 3.2 16 3.2 20 3.2 32 3.2 24 3.2 42 P 8 440 2.7 8 2.7 8 2.7 2 2.7 8 2.7 3 P 9 530 2.3 6 2.3 4 – – 2.3 3 – – hkl indexes we e ob ained om (Nagy and B adley, 1955) o sepioli e and om (B adley, 1940) o palygo ski e. L. Al es e al. Applied Clay Science 233 (2023) 106823 7 sepa a ion obse ed o his suspension a e ew hou s (Al es e al., 2020b). The smalle size o he mine al pa icles educes he ligh sca e ing which esul ed in ilms wi h highe anspa ency. The imp o emen in he anspa ency o he ilms (less opaque) when he mine al suspension was p e iously p epa ed by ul asonica ion wi h added CMC occu ed o he wo es ed CNF (see esul s o Mec CNF and TEMPO CNF in Fig. 5). The smalle pa icle size o he sepioli e pa icles and he be e mix u e allowed, no only had an impac on he op ical p ope ies o he composi e ilms bu also on hei mechanical pe o mance. The me- chanical p ope ies ob ained o he s udied composi e ilms a e shown in Table 2 and Fig. 6. In gene al, i was obse ed an imp o emen in he p ope ies o he CNF-based composi es wi h a be e dispe sion s a e o sepioli e. As i can be seen, he addi ion o sepioli e o he ilms p oduced wi h Mec CNF led o a educ ion in he mechanical p ope ies, being his deple ion Fig. 3. HR-TEM images a 15,000×magni ica ion o suspensions con aining 0.05 w % o ib ous clay ea ed wi h ul asonic p obe. a) sepioli e 1 wi h poly- phospha e; b) sepioli e 1 wi h polyac yla e; c) sepioli e 1 wi h CMC; d) sepioli e 2 wi h polyphospha e; e) sepioli e 2 wi h polyac yla e; ) sepioli e 2 wi h CMC; g) palygo ski e wi h polyphospha e; h) palygo ski e wi h polyac yla e; i) palygo ski e wi h CMC. The scale ba s ep esen 100 nm. L. Al es e al. Applied Clay Science 233 (2023) 106823 8 mi iga ed wi h a be e dispe sion o sepioli e p e iously o he com- posi e ilm p epa a ion. In ac , he ensile s eng h, he maximum suppo ed o ce, and he elonga ion a b eak o he composi e ilms, p epa ed using sepioli e and Mec CNF, a e imp o ed wi h he be e dispe sion s a e o he sepioli e c ys als, which no only allows a be e mix u e o he di e en componen s, bu also esul s in a mo e compac s uc u e, and hus in a lowe dec ease o he mechanical p ope ies ela i ely o he co esponding ilm wi h nea CNF. The esul s ob ained o TEMPO oxidised CNF-based composi e ilms a e somehow su p is- ingly posi i e, being possible o imp o e he mechanical p ope ies o he composi e ilms, compa ed wi h he nea ilm, e en wi hou p e ious sonica ion o he mine al. Howe e , he ul asonica ion o sepioli e suspension, beyond an addi ional end o a sligh imp o emen in he ensile s eng h, allowed o keep he anspa ency o he ilm con aining 20% o sepioli e wi h a alue close o ha o he nea TEMPO CNF ilm. The mechanical p ope ies o he composi e ilms ob ained in he p esen wo k a e in line o hose o ela ed ilms epo ed in he li e a u e (Gonz´ alez del Campo e al., 2018; Ma ín-Samped o e al., 2022). Fo example, Ma ín-Samped o e al. epo ed he p epa a ion o hyb id nanopape s o TEMPO-oxidised CNF and sepioli e (20%) wi h a ensile s eng h o 79.0 MPa and Young’s modulus o 4.7 GPa; ou composi e ilms p esen ed compa able alues o hese p ope ies (74.8 MPa and Fig. 4. Digi al pho os o TEMPO oxidised CNF-based ilms. Le – ilm ob ained using nea CNF, wi hou sepioli e; middle – composi e ilm p epa ed wi h TEMPO- oxidised CNF and 20% sepioli e wi h he mine al dispe sed using he high-shea dispe se ; igh – composi e ilm p epa ed wi h TEMPO-oxidised CNF and 20% sepioli e wi h he mine al dispe sed using ul asonica ion in he p esence o CMC. Fig. 5. T anspa ency o he CNF-based composi e ilms p epa ed using sepioli e (Sep 1) dispe sed using di e en dispe sing equipmen (Sep, using he high-shea dispe se and Sep Son, using ul asonica ion in he p esence o CMC). Table 2 S uc u al and mechanical p ope ies o he p epa ed CNF-based composi e ilms. Film Basis weigh (g/m 2 ) Thickness ( μ m) Max. Fo ce (N) Elonga ion a b eak (%) Mec a 40.5 ±0.5 34 ±4 71.5 ±4.9 10.5 ±1.7 Mec +Sep b 38.0 ±1.2 55 ±15 44.4 ±3.4 4.3 ±0.6 Mec +Sep Son c 38.9 ±0.8 45 ±5 49.0 ±2.8 7.6 ±0.5 TEMPO 55 a 40.8 ±0.8 39 ±5 25.6 ±13.5 2.0 ±1.3 TEMPO 55 +Sep b 40.2 ±1.1 37 ±2 34.9 ±8.0 4.1 ±2.3 TEMPO 55 +Sep Son c 39.8 ±0.3 37 ±3 41.3 ±2.8 3.6 ±0.8 a nea CNF ilms. b composi e ilms wi h 20% sepioli e wi h he mine al dispe sed using he high-shea dispe se . c Composi e ilms wi h 20% sepioli e wi h he mine al dispe sed using ul asonica ion in he p esence o CMC. L. Al es e al. Applied Clay Science 233 (2023) 106823 9 4.0 GPa, espec i ely), e en hough hey we e p oduced using a CNF om a di e en sou ce (bleached eucalyp us pulp in he p esen s udy s. bleached elm pulp in he p e ious li e a u e s udy). The elonga ion a b eak was sligh ly highe in he p esen wo k (3.6% s. 1.9%). Howe e , when compa ing he composi e ilms wi h he nea CNF ilms, an imp o emen in he ensile s eng h om ca. 21 o 79 MPa was obse ed in he p e ious s udy, while in he p esen wo k his imp o emen was lowe ( om ca. 44 o 75 MPa). In ano he s udy, Gonz´ alez del Campo e al. epo ed nanopape s o TEMPO CNF ( om eucalyp us pulp) and sepioli e wi h a Young’s modulus o 3.4 GPa and a elonga ion a b eak o 4.3%, which a e also alues compa able o hose o he p esen wo k. Howe e , in ha s udy, he au ho s ound a ema kable dec ease in he elonga ion a b eak going om nea TEMPO CNF (19.5%) o TEMPO CNF +20% sepioli e (4.3%), which we ha e no obse ed in he p esen wo k (elonga ion a b eak a ied om ca. 2 o 3.6%). Some di e ences obse ed could be due mainly o he di e en p ocedu es used o p e- pa e he ilms (sol en cas ing o il a ion me hod) and he di e en ca boxyl con en o he CNFs used. In sum, ou esul s con i m ha in addi ion o he good dispe sion s a e o he mine al, he size and he shape o he used mine al ha e simila i ies wi h he size and shape o TEMPO CNF, bo h ac o s wi h a high impac on he op ical and me- chanical p ope ies o he p oduced composi e ilms. 4. Conclusions The s a ing sepioli e and palygo ski e samples showed a ib ous na u e, as expec ed, consis ing o agg ega es/c ys al bundles and he HR-TEM images e ealed ha he c ys als we e no ully indi idualized when dispe sed in aqueous medium using he ul asound p obe ea - men , being obse ed agg ega es, “ ace- o- ace”, o se e al c ys als. The X- ay di ac ion s udies e ealed only e y mino changes in he di ac ion pa e n o he samples when subjec ed o he ul asonic ea men , being hese sligh changes a ibu ed o he disagg ega ion o he c ys als o he mine als. The u he addi ion o dispe sing agen s did no lead o ele an changes in he ib ous mine al s uc u e. On he o he hand, he mic oscopic s udies e ealed impo an imp o emen s in he dispe sion s a e o he mine al samples induced by he addi ion o he di e en dispe san s, mainly polyphospha e and CMC; when he clay samples we e p epa ed using polyac yla e, an ex ensi e “ andom” eagg ega ion was obse ed. The dispe sion s a e o he mine al, sepioli e, showed an impo an impac on he op ical and mechanical p ope ies o he composi e ilms p oduced wi h nanocelluloses. The use o he same con en (20%) o sepioli e dispe sed wi h ul asonica ion and CMC as dispe san led o a lowe dec ease in he mechanical p ope ies and anspa ency o he ilms p epa ed using Mec CNF ( s. he nea CNF ilm) when compa ing wi h he ilm p epa a ion using sepioli e dispe sed wi h high-shea dispe se . The ilm p oduced using TEMPO CNF and sepioli e, wi h he mine al p e iously dispe sed wi h ul asonica ion and CMC, p e- sen ed mechanical p ope ies supe io o hose o he co esponding nea TEMPO CNF ilm, and a high anspa ency, close o ha o he nea CNF ilm. These ilms could ind applica ions in p in ed elec onics, whe e he anspa ency o he ilms is a ele an issue. The inco po a ion o highly dispe sed sepioli e no only can educe he p ice o he ilms bu a he same ime imp o e he mechanical p ope ies o he ilms. The p esen wo k in ends o be he basis o de elop a new gene a ion o composi e ilms, in which i will be possible o une he p ope ies o he ilm, acco ding o he desi ed applica ion, by changing he o mula ion and/o he dispe sion s a e o he mine als used. Funding The p esen esea ch was suppo ed by he R&D P ojec “FILCNF- New gene a ion o composi e ilms o cellulose nano ib ils wi h mine al pa icles as high s eng h ma e ials wi h gas ba ie p ope ies” (PTDC/ QUI-OUT/31884/2017, CENTRO 01–0145-FEDER-031884), and he S a egic Resea ch Cen e P ojec s UIDB/00102/2020 and UIDB/ 05488/2020 unded by he Fundaç˜ ao pa a a Ciˆ encia e Tecnologia (FCT) and FEDER. CRediT au ho ship con ibu ion s a emen Luís Al es: Concep ualiza ion, In es iga ion, Me hodology, W i ing – o iginal d a , W i ing – e iew & edi ing. Ana Ramos: In es iga ion, W i ing – e iew & edi ing. Edua do Fe az: Me hodology, W i ing – o iginal d a , W i ing – e iew & edi ing. Ped o Sanguino: In es iga- ion, W i ing – e iew & edi ing. Julio San a ´ en: W i ing – e iew & edi ing. Ma ia G. Ras ei o: W i ing – e iew & edi ing, Funding acquisi ion. Jos´ e A.F. Gamelas: Concep ualiza ion, In es iga ion, Funding acquisi ion, P ojec adminis a ion, W i ing – e iew & edi ing. Fig. 6. Mechanical p ope ies o he CNF-based composi e ilms p epa ed using sepioli e (Sep 1) dispe sed using di e en dispe sing equipmen (Sep, using he high- shea dispe se and Sep Son, using ul asonica ion in he p esence o CMC). The ensile s eng h co esponds o he black ba s and he Young’s modulus o he g ey ba s. L. Al es e al.