In vitro safety evaluation of rare earth-lean alloys for permanent magnets manufacturing

1
Vol.:(0123456789)
Scien i ic Repo s | (2021) 11:12633 | h ps://doi.o g/10.1038/s41598-021-91890-0
www.na u e.com/scien i ic epo s
In i o sa e y e alua ion o a e
ea h‑lean alloys o pe manen
magne s manu ac u ing
Ca los Rumbo1*, C is ina Cancho Espina1, Jü gen Gassmann2, Oli ie Tosoni3,
Rocío Ba os Ga cía1, Sonia Ma el Ma ín1 & Juan An onio Tamayo‑Ramos1*
Due o hei excep ional physico‑chemical and magne ic cha ac e is ics, a e ea h (RE) pe manen
magne s a e applied in mul iple c i ical echnologies. Howe e , se e al en i onmen al and
economic di icul ies a ising om ob aining RE elemen s ha e p omp ed he sea ch o al e na i es
wi h accep able magne ic p ope ies bu con aining a lowe pe cen age o hese elemen s in hei
composi ion. The aim o his wo k was o pe o m a p elimina y oxicological e alua ion o h ee o ms
o newly de eloped RE‑lean alloys (one NdFeTi and wo NdFeSi alloys) applying di e en in i o
assays, using as a benchma k a comme cial NdFeB alloy. Thus, he e ec s o he di ec exposu e o
powde suspensions and o hei de i ed leacha es we e analysed in wo model o ganisms ( he A549
human cell line and he yeas Saccha omyces ce e isiae) applying bo h iabili y and oxida i e s ess
assays. Mo eo e , he impac o he alloy leacha es on he bioluminescence o Vib io ische i was also
in es iga ed. The ob ained da a showed ha only he di ec in e ac ion o he alloys pa icula es wi h
he applied o ganisms esul ed in ha m ul e ec s, ha ing all he alloys a compa able oxicological
po en ial o ha p esen ed by he e e ence ma e ial in he condi ions es ed. Al oge he , his s udy
p o ides new insigh s abou he sa e y o NdFeTi and NdFeSi alloys.
Ra e ea h (RE) elemen s ha e a key ole in many high echnologies, including he de elopmen o pe manen
magne s1, whe e hei excep ional physico-chemical cha ac e is ics and hei magne ic p ope ies make hem
excellen candida es o be used in he manu ac u ing o hese ma e ials. Among hem, neodymium (Nd) s ands
ou as one o he p e e ed elemen s in he pe manen magne ab ica ion, being he neodymium–i on–bo on
(NdFeB) he s onges comme cially a ailable pe manen magne . De eloped in he 1980s, and belonging o he
hi d gene a ion o pe manen magne s2, NdFeB magne s display ex ao dina y p ope ies3 o hei applica ion
in se e al ields, such as he au omo i e sec o 4,5, den is y6 o elec onic p oduc s7.
Due o se e al economic and poli ical ac o s8, as well as o he insu icien a ailabili y o RE elemen s, Nd
pe manen magne s a e comme cially expensi e. In addi ion, hei high demand, which is expec ed o inc ease
due o he dependence on pe manen magne s o he elec ical ehicle mo o s, will esul in a ise in hei p ice,
as obse ed in he las yea s9. This unce ain y in he ma ke o RE elemen s, combined wi h he en i onmen al
implica ions o hei ex ac ion10, such as he accumula ion o adioac i e ailings due o he adioiso opes
con ained in he RE elemen s deposi s11, ha e p omp ed e o s o de elop s a egies o educe he needs o
hese elemen s in he pe manen magne ield, becoming a p io i y wo ldwide. The de elopmen o RE-lean
pe manen magne s wi h a signi ican educ ion in he pe cen age o hese elemen s in hei composi ion is one
o he s a egies ha can be ollowed o accomplish his aim. Wi hin his app oach, RE(Fe,M)12 compounds,
whe e M ep esen s he ansi ion me als, a e conside ed in e es ing op ions o hei high magne o c ys alline
aniso opy, high sa u a ion magne iza ion and lowe RE con en 12. Hence, NdFeSi and NdFeTi, wo ypes o he
alloys belonging o his g oup13,14, show excellen magne ic p ope ies ha make hem good candida es o hei
implemen a ion in he ab ica ion o pe manen magne s wi h a educed RE pe cen age in hei composi ion.
Since me als ha e an impo an ole in a b oad ange o biological p ocesses, hese compounds a e essen ial
o li ing sys ems15. Consequen ly, homeos asis o me al ions is c i ical, and hei le els need o be kep wi hin
s ic limi 16, ou side o which hei lack o excess can lead o ad e se e ec s. The oxicological impac o me al
compounds is an issue ha has been widely s udied, exis ing nume ous esea ch wo ks whe e he gene al aspec s
and he mechanisms in ol ing hei oxici y we e e iewed17,18. In gene al, o ganisms a e exposed o me al
OPEN
1In e na ional Resea ch Cen e in C i ical Raw Ma e ials-ICCRAM, Uni e sidad de Bu gos, Plaza Misael Bañuelos
s/n, 09001 Bu gos, Spain. 2F aunho e Resea ch Ins i u ion o Ma e ials Recycling and Resou ce S a egies IWKS,
Ascha enbu ge S aße 121, 63457 Hanau, Ge many. 3CEA-LITEN, 38054 G enoble, F ance. *email: c umbo@
ubu.es; [email p o ec ed]
2
Vol:.(1234567890)
Scien i ic Repo s | (2021) 11:12633 | h ps://doi.o g/10.1038/s41598-021-91890-0
www.na u e.com/scien i ic epo s/
mix u es in he en i onmen bu , in mos cases, oxicological s udies a e pe o med conside ing he e ec s o
he me als independen ly. In his ega ds, some esea ch wo ks showed ha he e ec s o me al combina ions
can pose a highe oxicological impac han hese ma e ials indi idually19, cons i u ing a ele an ac o ha
should be aken in o accoun when he po en ial isks associa ed o me als a e being assessed. Among he di -
e en pe nicious e ec s ha hese compounds may cause, oxida i e s ess is a common ea u e o me al oxic-
i y, whe e he gene a ion o eac i e adicals can p oduce di e en ha m ul e ec s such as DNA damage, lipid
pe oxida ion o p o ein deple ion20.
Despi e he wide applica ion o RE-pe manen magne s in se e al ields, he in o ma ion abou hei sa e y
is sca ce. The implemen a ion o NdFeB magne s in den is y led o he eme gence o a ew s udies whe e he
cy o oxici y o pieces o his ma e ial and hei associa ed co osion p oduc s we e add essed21–23. Howe e , he
esul s obse ed we e in some cases con lic ing, which con i ms he need o de elop mo e oxicological esea ch
on Nd magne s. Rega ding he po en ial haza ds o he alloys applied in he ab ica ion o pe manen magne s
in powde o m, o ou knowledge, cu en ly he e is only one wo k pe o ming a p elimina y oxicological
e alua ion o hese ma e ials and hei associa ed leacha es24. Thus, due o his lack o s udies, he p o ision o
da a abou he possible isks o he me al powde s o RE-magne s, as well as o hose o new al e na i es o he
cu en comme cial magne s is c ucial, since i will p o ide ele an in o ma ion abou hei po en ial e ec s
be o e hei applica ion.
Conside ing ha NdFeTi and NdFeSi a e magne ic alloys in ended o subs i u ing cu en pe manen NdFeB
magne s, he aim o his wo k was o e alua e he po en ial oxici y o h ee new de eloped ypes o hese alloys
( wo ypes wi h he composi ion (Nd,Z ,Y)Fe10Si2 and one ype wi h he composi ion (Nd,P )1.15Fe11Ti) in powde
o m, and hei associa ed leacha es. Fo ha pu pose, bo h iabili y and oxida i e s ess assays we e pe o med
using he human cell line A549 and he yeas Saccha omyces ce e isiae. In addi ion, he sa e y o he alloys
leacha es was also assessed h ough he bioluminescen inhibi ion assay using he G am nega i e bac e ium
Vib io ische i. A comme cial Nd2Fe14B powde alloy was used in he assays o es ablish e e ence alues in he
s udied condi ions. Al oge he , he p esen ed da a p o ide a p elimina y e alua ion o he sa e y o NdFeSi and
NdFeTi alloys.
Resul s
Cha ac e iza ion o he alloys and hei associa ed leacha es. The su ace elemen al composi-
ion o he h ee RE-lean alloys and he e e ence sample was semi-quan i a i ely analysed h ough SEM–EDX.
Table1 shows he alues ob ained o each o he main elemen s comp ising he alloys (w %), wi h he excep ion
o B, which canno be de ec ed h ough EDX as i p esen s low pho on ene gy. As expec ed, he e e ence sample
showed he highes pe cen age o Nd in i s composi ion (≈16w %), while bo h NdFeSi samples showed he
lowes con en o his elemen (≈5w %). In he NdFeTi sample, he educ ion o Nd was no as e iden as in he
o he RE-lean alloys, p esen ing a 13.78w % o his elemen in i s composi ion. Howe e , i can be conside ed as
a ele an educ ion, since i ep esen s a dec ease o ≈13% wi h espec o he Nd con en in he e e ence. By he
same oken, ele an amoun s o o he elemen s ha we e added du ing he alloys p oduc ion as a eplacemen
o some o he Nd we e also quan i ied. Thus, P was de ec ed in he NdFeTi sample (4.27w % ± 0.39) while
signi ican pe cen ages o Z (9.29w % ± 2.82 and 8.89w % ± 0.80) we e ound in bo h NdFeSi S1 and NdFeSi
S2 samples espec i ely. Y, which was also employed o he ab ica ion o NdFeSi alloys, was no de ec ed in any
o he samples.
The mo phology and he size o he pa icles ha o m he alloy powde s we e also analysed and isualized by
SEM. Fo his pu pose, a small amoun om each sample was di ec ly used o he obse a ions. Figu e1 shows
he appea ance o he pa icles in NdFeB (1a, 1e), in bo h NdFeSi (1b and 1 ; 1c and 1g) and in NdFeTi samples
(1d, 1h). All o hem showed o be o med o a he e ogenei y o pa icles ha p esen a g ea a ie y o dimen-
sions ha ange om a ew o se e al hund ed mic ome e s. The bigges pa icles we e obse ed in he RE-lean
alloy powde s, whe e some o hem p esen ed sizes abo e 500µm, while in he comme cial NdFeB powde s he
pa icles we e sligh ly sho e han he o he s, wi hou exceeding he 500µM. Rega ding hei mo phology, a
Table 1. A e age weigh pe cen age (w %) o elemen s in he selec ed alloys de e mined by SEM–EDX.
NdFeB
Nd 15.97 ± 2.84
Fe 62.93 ± 3.43
B –
NdFeSi S1 NdFeSi S2
Nd 5.09 ± 0.51 4.97 ± 0.35
Fe 73.52 ± 5.69 73.10 ± 2.48
Si 6.51 ± 1.40 6.92 ± 0.48
NdFeTi
Nd 13.78 ± 0.68
Fe 69.96 ± 1.19
Ti 5.72 ± 0.23
3
Vol.:(0123456789)
Scien i ic Repo s | (2021) 11:12633 | h ps://doi.o g/10.1038/s41598-021-91890-0
www.na u e.com/scien i ic epo s/
a ie y o shapes was dis inguished, appea ing polygonal and ound pa icles. Mo eo e , in some o he images
(Fig.1e–h), pa icles in he nanoscale ange we e obse ed.
ICP-MS analysis was pe o med on he alloy leacha es o quan i y he concen a ions o he di e en dissol ed
me al ions eleased om he powde s a 10g/L in wa e a e a 3-mon h incuba ion. The concen a ions o Nd, Fe
and Si we e analysed in NdFeSi S1 and NdFeSi S2 leacha es. The ob ained esul s showed ha he concen a ion
o Nd was highe in NdFeSi S1 han in NdFeSi S2 (988.14ppb and 651.38ppb espec i ely). Rega ding Fe and Si,
hese ions we e ound in highe concen a ions in he leacha es om NdFeSi S2, whe e he ob ained alues we e
9.06 and 234.51ppb o Fe and Si, whe eas in NdFeSi S1 he concen a ions obse ed we e 1.67 and 132.01ppb
espec i ely. In he case o he leacha es om NdFeTi, Nd and Fe we e also de ec ed (1602.59 and 1.43ppb
espec i ely) while he le els o Ti we e unde he limi o de ec ion (< 0.072ppb). Finally, he le els o Nd, Fe
and B we e also quan i ied in he e e ence sample, p esen ing alues o 0.68, 1.51 and 583.94ppb. Mo eo e ,
ions belonging o he o he elemen s added o educe he amoun o Nd equi ed o he alloy ab ica ion we e
also de ec ed. Thus, P was quan i ied in he NdFeTi leacha es (1153.60ppb). Fo i s pa , Y was ound in he
leacha es ob ained om NdFeSi samples (55.16 and 442.39ppb in NdFeSi S1 and NdFeSi S2 espec i ely). On
he o he hand, he concen a ions o Z , which was also used o he ab ica ion o bo h NdFeSi samples, we e
unde he de ec ion limi in bo h samples (< 0.005ppb).
Toxici y s udies using A549 cell line. The neu al ed up ake assay was applied o s udy he impac on
cell iabili y a e a di ec exposu e o he A549 cells o di e en concen a ions o he alloys. This assay is based
on he abili y o he neu al ed dye o be up aken by cells and s ain he lysosomes in iable cells, which can be
de ec ed and measu ed. Figu e2A displays he esul s ob ained a e exposing cells o 6.4, 32 and 160mg/L o
he di e en alloys. Resul s showed ha cell iabili y was no a ec ed a e 24h o incuba ion wi h any o he
samples. On he o he hand, A549 cells di ec ly exposed o he alloys suspensions showed o su e oxida i e
s ess when hey we e incuba ed o one hou a he highes concen a ions es ed. As shown in Fig.2B, he
le els o eac i e oxygen species (ROS) we e s a is ically signi ican in cells a e being exposed o 160mg/L o
all he alloys. Bo h NdFeB and NdFeSi S2 samples displayed simila le els o ROS a ha concen a ion, which
showed o be sligh ly inc eased compa ed o he con ol condi ion. Fo i s pa , his induc ion was much highe
in NdFeSi S1 and NdFeTi samples, whe e he le els o ROS we e mo e han 3 and 8 imes espec i ely han hose
p oduced by he o he s a he same condi ions (Fig.2B).
Rega ding he assays pe o med using he leacha es, cells we e exposed o dilu ions equi alen o concen a-
ions o 6.4, 32 and 160mg/L o each alloy. None o hem showed o ha e any signi ican oxic e ec on nei he
cell iabili y no oxida i e s ess in he condi ions es ed (Fig.3).
Toxici y s udies using Saccha omyces ce e isiae. Due o he mechanisms behind he esponse o
me al oxici y a e highly conse ed be ween humans and yeas s, S. ce e isiae has been ex ensi ely applied o
elucida e how me als and me alloids a ec bo h biological sys ems25,26. To de e mine he oxicological po en ial
o he syn hesized alloys, his ungal model was included in he assays. Two concen a ions (160 and 800mg/L)
and exposu e imes (2 and 24h) we e assessed (Fig.4). A he sho e exposu e ime, no signi ican di e ences
in he yeas iabili y we e obse ed in any o he s udied condi ions (Fig.4A). Howe e , a e a longe exposu e
pe iod (24h), he dec ease o yeas CFUs in some cases indica ed an impac o he alloy suspensions on he cell´s
iabili y (Fig.4B). In he p esence o 160mg/L, no nega i e e ec s on he iabili y o S. ce e isiae cells could be
de ec ed. Howe e , in he p esence o 800mg/L, a dec ease in he a e age cells iabili y was obse ed in cells
Figu e1. SEM images showing he mo phology o he magne powde s. NdFeB (a,e); NdFeSi S1 (b, ); NdFeSi
S2 (c,g) and NdFeTi (d,g). Images a, b, c and d: O iginal magni ica ion × 30 (Scale ba = 500μm); Images e, , g
and h: O iginal magni ica ion × 1000 (Scale ba = 10μm).
4
Vol:.(1234567890)
Scien i ic Repo s | (2021) 11:12633 | h ps://doi.o g/10.1038/s41598-021-91890-0
www.na u e.com/scien i ic epo s/
exposed o NdFeB, NdFeSi S1 and NdFeSi S2 samples, being only s a is ically signi ican in he la e sample and
in he case o he e e ence alloy.
The oxicological po en ial o he alloys leacha es was also e alua ed on S. ce e isiae, using he same exposu e
condi ions. Yeas cells we e exposed o a leacha e equi alen o a concen a ion o 800mg/L o each alloy o 24h
(Fig.5). Di e en ly o wha was obse ed in case o he di ec con ac expe imen s, none o he leacha e samples
p oduced a nega i e impac in he iabili y o his o ganism in he s udied condi ions (Fig.5).
The oxicological po en ial o he alloys suspensions in S. ce e isiae was also de e mined in es iga ing hei
abili y o induce oxida i e s ess. As in he iabili y expe imen s, concen a ions o 160 and 800mg/L we e used
Figu e2. E ec s o di ec exposi ion o A549 cells o di e en concen a ions o he alloys suspensions. (A)
Viabili y o A549 cells (Neu al Red assay). Resul s a e exp essed as % o con ol (un ea ed cells). (B) Oxida i e
s ess (ROS le els) in A549 cells. Resul s a e exp essed as he ela i e luo escence alue o he con ol (un ea ed
cells) which was assigned a alue o 1. Da a ep esen he mean o 3 biological eplica es (± s anda d de ia ion,
SD). Di e ences we e es ablished using a One-way ANOVA ollowed by Dunne pos hoc es o compa e e e y
mean wi h he con ol, and conside ed signi ican a P ≤ 0.05. **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.
Figu e3. Exposi ion o A549 cells o di e en concen a ions o he alloy leacha es. (A) Viabili y o A549 cells
(Neu al Red assay). Resul s a e exp essed as % o con ol (un ea ed cells). (B) Oxida i e s ess (ROS le els)
in A549 cells. Resul s a e exp essed as he ela i e luo escence alue o he con ol (un ea ed cells) which was
assigned a alue o 1. Da a ep esen he mean o 3 biological eplica es (± s anda d de ia ion, SD). Di e ences
we e es ablished using a One-way ANOVA ollowed by Dunne pos hoc es o compa e e e y mean wi h he
con ol, and conside ed signi ican a P ≤ 0.05. **P ≤ 0.01.
5
Vol.:(0123456789)
Scien i ic Repo s | (2021) 11:12633 | h ps://doi.o g/10.1038/s41598-021-91890-0
www.na u e.com/scien i ic epo s/
Figu e4. Colony o ming uni s (CFUs) o S. ce e isiae cells exposed o di e en alloys suspensions a wo
concen a ions (160 and 800mg/L) and wo exposu e imes: 2h (A) and 24h (B). Resul s a e exp essed as he
pe cen age (%) o CFUs de e mined o each exposu e condi ion using as e e ence alue he non-exposed cells
condi ion, which was assigned a alue o 100%. Da a ep esen he mean o 3 biological eplica es (± s anda d
de ia ion, SD). Di e ences we e es ablished using a One-way ANOVA ollowed by Dunne pos hoc es o
compa e e e y mean wi h he con ol, and conside ed signi ican a P ≤ 0.05. *P ≤ 0.05, **P ≤ 0.01.
Figu e5. CFUs o S. ce e isiae cells exposed o di e en alloys leacha es du ing 24h. Resul s a e exp essed
as he pe cen age (%) o CFUs de e mined o each exposu e condi ion using as e e ence alue he non-
exposed cells condi ion, which was assigned a alue o 100%. Da a ep esen he mean o 3 biological eplica es
(± s anda d de ia ion, SD). Di e ences we e es ablished using a One-way ANOVA ollowed by Dunne pos
hoc es o compa e e e y mean wi h he con ol condi ion, and conside ed signi ican a P ≤ 0.05.

6
Vol:.(1234567890)
Scien i ic Repo s | (2021) 11:12633 | h ps://doi.o g/10.1038/s41598-021-91890-0
www.na u e.com/scien i ic epo s/
o expose yeas cells o 2h (Fig.6). In his case, only a s a is ically signi ican inc ease in he ROS le els could
be obse ed in he yeas cells exposed o 800mg/L o NdFeSi S2 sample. The obse ed esul s, conside ing he
high concen a ions used in he exposu e expe imen , indica e a low abili y o he di e en alloys o induce
oxida i e s ess in S. ce e isiae.
Toxici y de e mina ion o he alloys leacha es applying he Vib io ische i bioluminescence
inhibi ion assay. Due o hei op imal sensi i i y and eliabili y, es s applying V. ische i as model o ganism
a e equen ly used o he de ec ion and eco oxicological sc eening o oxic subs ances in a a ie y o subs a es,
including soils and wa e 27, as well as o he s udy o he sa e y o di e en elemen s such as REs and me als28.
To de e mine whe he he ion le els de ec ed in he alloy leacha es we e abo e he sa e y limi s, V. ische i was
exposed o hem a di e en concen a ions wi h he aim o s udy hei e ec on he na u al bioluminescence
o his mic oo ganism. As shown in Fig.7 and Table2, none o he leacha es led o an inhibi ion o he ligh
in ensi y a he concen a ions es ed. Figu e7 displays he e olu ion o he bioluminescence o V. ische i du ing
a 30min-incuba ion pe iod in he p esence o leacha es equi alen o concen a ions o 160 and 800mg/L o he
alloys. A signi ican d op in he ini ial bioluminescence peak alue was obse ed in all he samples incuba ed
wi h he leacha es. Howe e , om his poin on, he in ensi y o he ligh ba ely dec eased, p esen ing all sam-
ples simila bioluminescence dec ease pa e ns along ime. The bioluminescence inhibi ion pe cen ages a e 5,
15 and 30min exposu e a e showed in Table3. Nega i e alues we e ob ained o all he samples a all he ime
poin s es ed, which indica es ha he d op in he ligh in ensi y du ing he incuba ion was lowe han he na u-
al a enua ion o he mic oo ganism e en in he ea ly s ages o he incuba ion.
Discussion
The apid g ow h o indus ies whe e me allic powde s a e applied, oge he wi h he isks ha hey may ep esen
o heal h and he en i onmen 29, make he a ailabili y o da a conce ning he po en ial oxici y o hese ma e ials
a p io i y issue. The applica ion o addi i e manu ac u ing echnologies in he pe manen magne ield in ol es
he use o me al powde s, o example, o he p oduc ion o 3D-p in ed magne s30. Hence, i is impo an o
know he speci ic haza ds associa ed o hei manipula ion and o he was es gene a ed a e hei use, conside -
ing no only he po en ial impac o he small me al pa icula es, bu also ha o hei co osion p oduc s, which
can be a sou ce o con amina ion h ough he elease o me allic ions o he en i onmen , whe e hey may pose
subs an ial nega i e consequences31. In his s udy, he oxicological po en ial o h ee RE lean alloys wi h he
same c ys al s uc u e ( wo ypes wi h he composi ion (Nd0.2Z 0.7Y0.1)Fe10Si2 and one ype wi h he composi ion
(Nd0.75P 0.25)1.15Fe11Ti), de eloped as al e na i es o cu en NdFeB pe manen magne s, was e alua ed apply-
ing di e en in i o assays and using model o ganisms ep esen a i e o human and en i onmen al exposu es.
Fi s ly, he powde s and leacha es unde s udy we e cha ac e ized h ough di e en echniques. Thus, he
su ace elemen al composi ion o he alloys was semi-quan i a i ely analysed h ough SEM–EDX, con i ming
ha he e e ence NdFeB sample p esen ed he highes pe cen age o Nd in i s composi ion, while bo h NdFeSi
samples had he lowes con en o his elemen . In he case o he NdFeTi sample, and in spi e o he ac ha he
educ ion o his elemen was no as signi ican as in he o he s, i also ep esen s a ele an dec ease in he inal
Nd con en . In addi ion, due o he ac ha mo phology and size o he pa icles a e ac o s ha could in luence
he oxici y o he powde alloys, hei appea ance was analysed and obse ed by SEM. I was con i med ha all
he samples used in his wo k consis ed in a he e ogeneous popula ion o pa icles p esen ing i egula mo pholo-
gies (mainly polygonal shapes) and a a ie y o sizes in he mic o and nanome ic scales, wi h a mino di e ence
in he bigges pa icles size, which we e ound o be sligh ly la ge in he RE-lean alloy powde s, p obably due
Figu e6. ROS induc ion analysis o S. ce e isiae cells exposed o di e en alloys du ing 2h a wo di e en
concen a ions (160 and 800mg/L). Resul s a e exp essed as a bi a y luo escence alues. Da a ep esen
he mean o 2 biological eplica es (± s anda d de ia ion, SD). Di e ences we e es ablished using a One-way
ANOVA ollowed by Dunne pos hoc es o compa e e e y mean wi h he con ol condi ion, and conside ed
signi ican a P ≤ 0.05. *P ≤ 0.05.
7
Vol.:(0123456789)
Scien i ic Repo s | (2021) 11:12633 | h ps://doi.o g/10.1038/s41598-021-91890-0
www.na u e.com/scien i ic epo s/
Figu e7. E olu ion o V. ische i bioluminescence in ensi y du ing he 30min-exposu e o alloys leacha es.
Each poin ep esen he mean o 3 biological eplica es (± s anda d de ia ion, SD).
Table 2. Bioluminescence inhibi ion o V. ische i cells exposed o di e en concen a ions o alloy leacha es a
5, 15 and 30min (pe cen age o con ol).
Sample
% o bioluminescence inhibi ion
5min 15min 30min
NdFeB 160mg/L − 8.35 ± 9.77 − 22.08 ± 9.63 − 29.91 ± 4.07
NdFeB 800mg/L − 1.47 ± 3.63 − 15.35 ± 4.30 − 27.52 ± 11.68
NdFeSi S1 160mg/L − 0.77 ± 3.94 − 12.23 ± 4.04 − 37.59 ± 7.17
NdFeSi S1 800mg/L − 4.88 ± 4.37 − 11.07 ± 6.24 − 24.85 ± 6.39
NdFeSi S2 160mg/L − 1.01 ± 3.97 − 8.98 ± 3.67 − 37.45 ± 2.91
NdFeSi S2 800mg/L − 0.12 ± 3.06 − 8.67 ± 5.05 − 32.51 ± 4.17
NdFeTi 160mg/L − 0.82 ± 2.16 − 11.73 ± 6.48 − 31.33 ± 4.98
NdFeTi 800mg/L − 0.14 ± 2.64 − 6.55 ± 2.76 − 32.35 ± 5.79
Table 3. Samples, alloy composi ion and c ys al s uc u e o he ma e ials analysed in his s udy.
Sample name Alloy composi ion C ys al s uc u e
NdFeB Nd2Fe14B P42/mnm, space g oup 136
NdFeSi S1 (Nd0.2Z 0.7Y0.1)Fe10Si2I4/mmm, space g oup 139 (1:12)
NdFeSi S2 (Nd0.2Z 0.7Y0.1)Fe10Si2I4/mmm, space g oup 139 (1:12)
NdFeTi (Nd0.75P 0.25)1.15Fe11Ti I4/mmm space g oup 139 (1:12)
8
Vol:.(1234567890)
Scien i ic Repo s | (2021) 11:12633 | h ps://doi.o g/10.1038/s41598-021-91890-0
www.na u e.com/scien i ic epo s/
o he di e en me hodologies applied o hei gene a ion (je milling was used o ob ain he NdFeB powde s,
while RE-lean alloy powde s we e p oduced a e smashing magne pieces wi h a mo a ). Rega ding he leacha es
analysis, di e en concen a ions o Fe, Nd o Si, as well as P o Y, which we e applied du ing he ab ica ion o
some o he alloys, we e de ec ed by ICP-MS in he samples ob ained di ec ly om he alloys used in he expe i-
men s. Unexpec edly, he concen a ion o Nd in NdFeB leacha es was e y low compa ed o ha quan i ied in
he RE-lean alloys, which p esen lowe pe cen age o Nd in hei composi ion. Since he mo phology o he
pa icula es is simila in all he samples and none o he alloys we e coa ed, hese a ia ions, as well as hose
obse ed be ween bo h NdFeSi samples, could be a ibu ed o he di e en suscep ibili y o me al leaching o
he elemen s applied du ing he alloys manu ac u ing.
Due o he widesp ead dis ibu ion o me als, hei impac in he biological sys ems is a opic o c i ical in e es
and many esea ch wo ks ha e add essed his issue. In he p esen wo k, o e alua e he oxicological po en ial
o he di ec in e ac ion o he RE-lean alloy powde s, he human cell line A549 and he yeas S. ce e isiae we e
used. Mo eo e , he e ec o he alloy leacha es was s udied in bo h o ganisms and in he luminescen bac e ium
V. ische i. Ou esul s showed ha only he di ec in e ac ion o high concen a ions o he pa icula es we e
able o cause ha m ul e ec s. Mo e speci ically, s a is ically signi ican le els o oxida i e s ess we e obse ed
in human cell lines a e being exposed o 160mg/L o all he alloy powde s, while he iabili y o S. ce e isiae
was comp omised when exposed o NdFeB and NdFeSi powde s a 800mg/L. The numbe o a ailable s udies
assessing he oxici y o RE-pe manen magne s is sca ce, wi h mos o hem mainly ocused on he analysis o Sm
and Nd alloys, wo o he mos common ma e ials applied in den is y. In he speci ic case o Nd magne s, se e al
esea ch wo ks ha e been pe o med analysing he sa e y o magne pieces in he mm scale using mammalian
cell lines, and eaching in some cases di e en conclusions ega ding hei po en ial cy o oxici y21–23. Conce n-
ing he biological impac o pe manen magne pa icula es, Pé igo e al. showed ha MRC-5 cells exposed o
NdFeB nanopa icles coa ed wi h oleic acid a concen a ions up o 100µg/mL p esen ed a iabili y ≥ 80%32.
Mo eo e , in a ecen s udy whe e we desc ibed he po en ial oxici y o MnAl based magne s powde s, he
sa e y o a comme cial NdFeB magne powde , used as e e ence, was e alua ed oo in he A549 cell line and S.
ce e isiae, showing simila e ec s o hose obse ed on he NdFeB magne s udied in he p esen wo k24. The
sa e y o me al mic o and nano-sized pa icles in euka yo ic cellula models has been add essed by di e en
s udies. Fo ins ance, A ms ead e al. epo ed ha bo h nano- and mic o-CoC Mo can induce oxici y, showing
ha he e ec s on cell iabili y and oxida i e s ess we e speci ic o dose, exposu e ime and cell ype33. Fo i s
pa , Li e al. desc ibed he oxici y o di e en elemen al me al powde s in os eoblas -like SaOS2 cells34. These
au ho s showed ha in some cases he elemen al me als exhibi ed di e en deg ees o cy o oxici y depending
on i he ma e ials we e in bulk o powde o ms, he la e p esen ing highe oxici y. These au ho s associa ed
hese di e ences wi h he highe abili y o powde ed me als o elease ions o he medium, esul ing, he e o e,
mo e cy o oxic. By he same oken, Palombella e al. ca ied ou a compa a i e mul idisciplina y s udy based
on he size e ec o ze o- alen i on, cobal , and nickel mic opa icles and nanopa icles using human adipose
s em cells (hASCs) as model o ganism, and e alua ing di e en pa ame e s such as cy o oxici y, mo phology,
cellula up ake, and gene exp ession35. These au ho s obse ed ha while nanopa icles we e mainly in e nal-
ized by endocy osis and could pe sis in he esicles wi hou any appa en cell damage, mic opa icles we e no
in e nalized. Thus, hese au ho s sugges ed ha he nega i e e ec s obse ed in hASCs could be explained by
he elease o ions in he cul u e medium, o o he educed oxygen and nu ien exchange e iciency ha he
p esence o mic opa icles a ound he cells could cause.
Me al ions’ oxici y in S. ce e isiae has been epo ed o be simila o ha obse ed in animal cells36. Mechanis-
ic oxici y analyses o di e en me allic elemen s in yeas indica e a as and gene al ac i a ion o me al-speci ic
oxida i e de ence and p o ein deg ada ion p ocesses25. This as esponse allows a apid adap a ion o acu e me al
exposu e ia e ec i e es ablishmen o me al de ences a p o eomic le el, esul ing in a ela i ely as educ ion
o he de oxi ica ion pa hways ac i i y, which could explain why he obse ed ROS le els in he me al alloys
exposu e condi ions we e simila o hose shown by he non-exposed yeas cells.
Since all he alloys leacha es used in he p esen s udy showed o be sa e o he di e en o ganisms, i is
mo e easible o a ibu e he oxida i e s ess obse ed in A549 cells, as well as he dec ease in he iabili y o S.
ce e isiae, o he di ec in e ac ion o he mic o and nano-sized pa icles o each sample wi h he selec ed cellula
models. Thus, hese obse ed e ec s a e p obably consequence o he combined ac ion o he mic opa icles ou -
side o he o ganisms, oge he wi h he e ec s o he nanopa icles, ha could be in e nalized and cause cellula
s ess by di e en mechanisms, as i was al eady desc ibed o di e en me al oxide nanopa icles in A549 cells37.
Bo h essen ial (Cu, Fe, Mn, Zn…) and no essen ial (Al, As, Cd, Pb, Hg…) me al ions can cause oxici y.
Essen ial me al oxici y is mainly due o hei excessi e accumula ion in he o ganism, while he ha m ul e ec s
induced by no essen ial me als a e associa ed o bo h hei accumula ion and o he dis u bance o essen ial me -
als’ unc ion15. Szi ák e al. epo ed ha en i onmen al low concen a ions o some me als such as Fe, Pb o Cd
a e enough o p oduce an inc ease in he in acellula le els o ROS in cells o he eshwa e alga Chlamydomonas
einha d ii38. In his ega d, i is also impo an o ema k ha , in he en i onmen , o ganisms a e gene ally
exposed o mix u es o me als so, in spi e o being unde oxic le els, hei combina ion can esul in ad e se
consequences17. This e ec is usually unde es ima ed since many esea ch wo ks a e ocused in he s udy o he
o ganisms´ esponses o hese ions indi idually. In ou assays, he o ganisms we e exposed o alloy leacha es
equi alen o he concen a ions used in he di ec con ac expe imen s, by which hey con ain mix u es o ions.
This cons i u es one o he s eng hs o he p esen wo k, since, as explained abo e, i is ep esen a i e o mo e
ealis ic condi ions. None o he leacha es showed o be oxic a he di e en concen a ions es ed, con i ming
ha he me al ions mix u e p esen ed in he samples do no pose a signi ican challenge o any o he o gan-
isms applied. These esul s a e conco dan wi h hose ob ained in ou ecen wo k, ocused on he e alua ion o
he oxici y o MnAl based pe manen magne s, whe e he leacha es o NdFeB comme cial powde s we e also
e alua ed in he same o ganisms and condi ions24.
9
Vol.:(0123456789)
Scien i ic Repo s | (2021) 11:12633 | h ps://doi.o g/10.1038/s41598-021-91890-0
www.na u e.com/scien i ic epo s/
Conclusions
The sa e y o h ee Nd-lean alloys wi h he same c ys al s uc u e and designed o be applied in he manu ac u -
ing o pe manen magne s was e alua ed in his wo k. The selec ed alloys included wo NdFeSi and one NdFeTi
alloys, and a comme cial NdFeB sample, which was used as e e ence.
The po en ial oxici y o he alloys in powde o m and hei associa ed leacha es was analysed using h ee
model o ganisms which can be conside ed ep esen a i e o human (A549 cells) and en i onmen al exposu es (S.
ce e isiae and V. ische i) and applying di e en in i o assays. In spi e o he ac ha some wo ks ha e sugges ed
ha me al pa icles oxici y is a ibu able o hei abili y o elease me al ions o he su ounding en i onmen ,
ou esul s showed ha only he di ec in e ac ion o he alloy pa icula es wi h he o ganisms ha e esul ed
in pe nicious e ec s, being he obse ed e ec s compa able o hose caused by he comme cial powde s. Thus,
oxida i e s ess was s a is ically signi ican in A549 cells exposed o 160mg/L o he alloys, while in S. ce e isiae,
yeas cells exposed o bo h NdFeSi samples showed a educ ion in hei iabili y a 800mg/L, being s a is ically
signi ican in he case o NdFeSi S2.
In summa y, he p esen wo k cons i u es a p elimina y analysis o he biological impac o h ee speci ic
RE-lean alloys, indica ing ha he pa icles o hese speci ic ma e ials could pose a oxicological isk due o hei
di ec in e ac ion wi h he o ganisms, while he eleased ions seems no o be in ol ed in hei oxici y. Howe e ,
mo e esea ch is needed o cha ac e ize in dep h he e ec s o hese ma e ials, as well as hei beha iou in mo e
ealis ic scena ios.
Me hods
Alloys and sample p epa a ion. The ma e ials used in his s udy a e speci ied in Table3.
NdFeSi S1 ((Nd0.2Z 0.7Y0.1)Fe10Si2 s ip-cas lakes) we e p oduced by he s ip-cas ing echnique. 15kg o
pu e elemen s (pu i y highe 99.5%) we e melded induc i ely in an alumina c ucible unde ine a mosphe e.
The mel was pou ed a 1350°C ia a undish on o a wa e -cooled coppe wheel. The su ace wheel speed o he
quenching wheel was se o 1.8m/s o a oid o ma ion o α-Fe. The lakes we e c ushed a e he quenching o
lakes wi h a diame e o app oxima ely 20mm and a hickness o 320µm.
NdFeSi S2 ((Nd0.2Z 0.7Y0.1)Fe10Si2 mel -spun ibbons) we e syn hesized by he mel -spinning echnique. 50g
o he s ip-cas lakes we e used as inpu ma e ial. Du ing mel -spinning, he lakes we e mel ed induc i ely in
a qua z glass c ucible in a gon a mosphe e. A 1355°C he mel was cas ed ough a ec angula nozzle (0.6 × 7
mm2) by an A o e p essu e o 350mba and subsequen ly quenched on o a wa e -cooled coppe wheel. The
su ace wheel speed was se o 10m/s esul ing in lakes wi h a ine mic os uc u e and smalle g ains compa ed
o he s ip-cas lakes o NdFeSi S1.
NdFeTi ((Nd0.75P 0.25)1.15Fe11Ti ibbons) was p oduced like he NdFeSi S1 sample, bu using di e en equip-
men . Thus, he s ip-cas ing echnique was applied, i.e., induc ion mel ing ollowed by cas ing on a wa e -cooled
o a i e coppe wheel wi h an s ip-cas ing u nace om ULVAC. The linea speed o he wheel was se o 1.4m/s.
As e e ence sample, comme cial Nd2Fe14B alloy powde s p o ided by ARELEC Magne s and Magne ic
Sys ems we e used. This sample was p oduced by s ip cas ing and je milling, and i was named in his wo k
as NdFeB.
Wi h he aim o ob ain a uni o m powde , NdFeSi S1, NdFeSi S2, and NdFeTi samples we e smashed in a
mo a . The ob ained powde was used o p epa e he s ock solu ions o he expe imen s. In he case o he
e e ence sample, he powde p o ided by ARELEC was used di ec ly o p epa e he s ock solu ions.
S ocks o he powde alloys esuspended in wa e a 10g/L we e p epa ed o pe o m he di ec con ac
expe imen s. In o de o ob ain he alloy leacha es, powde s we e le in wa e a 10g/L du ing 3mon hs a 4°C.
A e his incuba ion, he samples we e cen i uged and he supe na an s wi h he leacha es we e eco e ed and
subsequen ly il e ed h ough 0.22 polye he sul one memb anes. Be o e pe o ming di ec con ac es s, he alloy
powde s we e ea ed o ob ain a homogenized solu ion as desc ibed below: samples we e o exed a ull speed
o 1min and hen submi ed o ul asonica ion o 20min a low powe in ensi y. Finally, be o e p epa ing he
di e en concen a ions used in he expe imen s, an addi ional o ex s ep was pe o med.
Analysis o he su ace composi ion and mo phology o he alloys. The su ace composi ion o he
alloys used in his wo k was semi-quan i a i ely analysed by SEM–EDX (Scanning Elec on Mic oscopy–Ene gy
Dispe si e X- ay spec oscopy) in he Uni e si y o Bu gos using a JEOL JSM-6460LV mic oscope equipped
wi h a X-MaxN ene gy dispe si e de ec o . To pe o m he elemen iden i ica ion and quan i ica ion, a leas 3
di e en a eas we e selec ed and analysed in each ma e ial.
The mo phology and he size o he pa icles ha a e pa o he di e en alloys powde s we e analysed by
Scanning Elec on Mic oscopy. A small quan i y o each sample was di ec ly examined using JEOL JSM-6460LV
a he Mic oscopy Facili y o he Uni e si y o Bu gos.
ICP‑MS analysis o he alloys leacha es. Wa e samples con aining he leacha es om alloys aqueous
suspensions (10g/L), ob ained a e incuba e powde s a 4°C o h ee mon hs, whe e analysed by induc i ely
coupled plasma mass spec ome y (ICP-MS) using an Agilen 8900 ICP-QQQ ins umen . Fo da a acquisi ion,
5 eplica es we e used.
Model o ganisms and cul u e condi ions. A549 lung cance cell line was pu chased om Sigma. I
was cul u ed in comme cial Dulbecco’s Modi ied Eagle’s Medium (DMEM) supplemen ed wi h 10% ( / ) e al
bo ine se um (FBS) and 100 U/mL penicillin and 100mg/L s ep omycin, and incuba ed unde s anda d condi-
ions in 37°C humidi ied 5% CO2 a mosphe e.