Chaub a Cryogenic Laboratory, Faculty of Physics, Vietnam National University, 334 Nguyen Trai, Thanh Xuan, Hanoi, Viet Nam b Center for Materials Science, Faculty of Physics, Vietnam Na
Trang 1Journal of Magnetism and Magnetic Materials 282 (2004) 44–48
Nanostructure and magnetization reversal process in
N.H Duca,*, D.T Huong Gianga, N Chaub
a
Cryogenic Laboratory, Faculty of Physics, Vietnam National University, 334 Nguyen Trai, Thanh Xuan, Hanoi, Viet Nam
b
Center for Materials Science, Faculty of Physics, Vietnam National University, 334 Nguyen Trai, Thanh Xuan, Hanoi, Viet Nam
Available online 28 April 2004
Abstract
Studies of the naturally formed nanostructure and magnetization reversal process were performed for the sputtered Tb(Fe0.55Co0.45)1.5/Yx(Fe0.7Co0.3)1x multilayers (0pxp0.2) with a TbFeCo layer thickness tTbFeCo=12 nm and YFeCo layer thickness tYFeCo=10 nm The structural investigations showed that nanocrystals are naturally formed and coexist within the amorphous matrix in Y0.1(FeCo)0.9layers In this state, low magnetic coercivity and large parallel magnetostrictive susceptibility are observed The results are discussed in terms of the crystalline discontinuity of the soft YFeCo layers
r2004 Elsevier B.V All rights reserved
PACS: 75.60.Jk; 75.70.Cn; 81.07.Bc
Keywords: Spring-exchange multilayers; Nanocrystalline structure; Magnetization reversal; Giant magnetostriction
1 Introduction
The exchange-spring concept [1] opened an
alternative route towards new high-performance
hard magnetic materials By associating a coercive
hard magnetic phase with a large magnetization
soft phase, it was expected that new high-energy
product materials could be prepared
Exchange-spring behavior was found in various systems
However, as far as our knowledge, no material has
been found with properties clearly superior to
those of usual hard magnetic materials
Mean-while, this concept has successfully been applied to the so-called giant magnetostrictive spring-ex-change multilayers, where high magnetostrictive layers and soft magnetic layers alternate [2]
magnetostriction as large as 890 106 and a huge parallel magnetostrictive susceptibility (wlJ=dlJ/dH) of 8 102T1 in an applied field
of about a few 10 mT for magnetostrictive/soft magnetic TbFeCo/FeCo multilayers In this case, magnetization reversal is thought to be nucleated within the soft layer in a low applied field and propagates from the soft layers into the magnetos-trictive layers [4] In the TbFeCo/FeCo multi-layers, the soft FeCo-layer is continuous (Fig 1a), thus the nucleation of reversal occurs at some
*Corresponding author Tel.: +84-8-7680978; fax:
+84-4-8340724.
E-mail address: ducnh@vnu.edu.vn (N.H Duc).
0304-8853/$ - see front matter r 2004 Elsevier B.V All rights reserved.
doi:10.1016/j.jmmm.2004.04.010
Trang 2defect points on the sample surface and interfaces.
In this context, one expects that the reversal can be
nucleated in an easier way in discontinuous soft
phase, i.e in layers in which the FeCo nanograins
are embedded within a non-magnetic matrix (Fig
1b) Practically, an excellent magnetic softness has
recently been reported for Tb(Fe0.55Co0.45)1.5/
Y0.2(Fe0.8Co0.2)0.8 (denotes as Terfecohan/
Y0.2(Fe0.8Co0.2)0.8multilayers, in which the
nanos-tructure of YFeCo layers was formed from an
amorphous phase by heat treatments[5–6]
As far as our knowledge, the transformation of
the amorphous state in the RCoFe (R = rare
earths and/or light transition metals) layer is
shown to be dependent on the R-concentration
[7] At a critical R-concentration, the
nanostruc-tured RFeCo layer is expected to be naturally
formed in as-deposited multilayers By this way,
an optimization of the magnetostriction and
magnetostrictive softness can be reached right
after depositing or/and annealing at low
tempera-tures
In this paper, a direct approach to the natural
nanostructure and large parallel magnetostrictive
susceptibility will be applied for the Terfecohan/
Yx(Fe0.7Co0.3)1x multilayers The results are
discussed in terms of the structural discontinuity
caused by the formation of the nanostructure in
the soft magnetic layers
2 Experimental
{Terfecohan/Yx(Fe0.7Co0.3)1x}n multilayers
with x=0, 0.1, 0.2, n=50 and the individual layer
thicknesses are tTbFeCo=12 nm and tFeCo=10 nm were fabricated by RF-magnetron sputtering The typical power during sputtering was 200 W and the
Ar pressure was 102mbar Composite targets were used which consisted of segments of different elements (here Tb, Y, Fe, Co) The substrates were glass microscope cover slips with a nominal thickness of 150 mm Both target and sample holders were water-cooled Samples were annealed
at different temperatures TA=200C, 300C,
350C, 400Cand 450Cfor 1 h in a vacuum of
5 105mbar
The crystal structure of the sample was investi-gated by X-ray diffraction using the D5005 Siemens with a copper anticathode The magneti-zation was measured with a vibrating magnetiza-tion magnetometer (VSM) in a magnetic field upto 1.4 T at room temperature The magnetostriction was measured by using an optical deflectometer (resolution of 5 106rad), in which the bending
of the substrate due to the magnetostriction in the film was determined
3 Experimental results and discussion
In the x=0 sample, the large X-ray diffraction intensity at 2y=45 is characteristic of the (1 1 0) reflection of BCC-Fe (Fig 2) No other diffraction peaks are observed indicating that the TbFeCo layer is amorphous The intensity of the BCC-Fe reflection is strongly reduced in the x=0.1 sample This is attributed to the formation of BCC-Fe nanocrystals in the YFeCo layers The cross-sectional HRTEM image shown in Fig 3reveals the coexistence of nanograins (with an average grain size of about 10 nm) and of an amorphous phase in the Y0.1(Fe,Co)0.9 layers This transfor-mation to the nanostructure was associated with the reduction of the thermodynamic driving force for the crystallization caused by substitution [8] Finally, the (1 1 0) reflection almost disappears at x=0.2 reflecting the fact that the whole layer is now amorphous Similar phenomenon was ob-served for the Terfecohan/YxFe1xmultilayers[4] Low-temperature annealing (at TAp350C ) is usually performed to relieve the stress induced during the sputtering process At present, as
Fig 1 Illustration of the magnetostrictive exchange-spring
multilayers with a structural continuous (a) and discontinuous
(b) soft magnetic layer.
Trang 3illustrated in Fig 4, the microstructure of the
350C-annealed samples with x=0 and 0.1 is
almost the same as that of the corresponding
as-deposited ones However, the modification of the
amorphous state to form BCC-Fe nanostructured
phase is observed in the Y0.2(Fe,Co)0.8layers
The magnetic hysteresis loops measured as a
function of the magnetic fields applied in the film
plane are presented in Fig 5 For all samples, the
observed curves are characteristic of in-plane
magnetization By, increasing x from 0 to 0.1, the (magnetic) coercivity (MHC) decreases from 4.8
to 3.1 mT then it increases again to 6.2 mT for x=0.2 In these samples, the YFeCo composition
is ferromagnetic at room temperature[7] It is that the FeCo (and/or YFeCo)/TbFeCo coupling im-poses in-plane magnetization The smallest coer-civity value found in the sample with x=0.1 may
be attributed to the specific nanostructure of this sample as observed by X-ray diffraction results at room temperature In the TbFeCo/FeCo multi-layer, the soft FeCo-layer is continuous We thus expect that the nucleation of reversal occurs at some defect points on the sample surface In
Fig 2 X-ray diffraction patterns of as-deposited Terfecohan/
Y x (Fe,Co) 1x multilayers.
Fig 3 Cross-sectional HR-TEM image of the as-deposited
Terfecohan/Y 0.1 (Fe,Co) 0.9 multilayer.
Fig 4 X-ray diffraction patterns of 350 -annealed Terfecohan/
Y x (Fe,Co) 1x multilayers.
Fig 5 Magnetic hysteresis loops of as-deposited Terfecohan/
Y (Fe,Co) multilayers.
Trang 4TbFeCo/Y0.1(FeCo)0.9, the FeCo grains are
nano-crystallized and embedded within an amorphous
matrix Each FeCo nanocrystals are decoupled
from each other Soft phase reversal can then be
nucleated at any of the nanocrystals, on a defect
position Statistically, it is expected that nucleation
will be easier than in FeCo-pure system This
explains qualitatively the observed difference in
coercive field values between the samples with
x=0 and 0.1 For x=0.2, the whole Y0.2(FeCo)0.8
layer becomes continuous in the amorphous state,
then the magnetic coercivity is enhanced again
Magnetic softness improvement due to stress
releasing effects is clearly provided by the
reduc-tion of the magnetic coercivity with the same
factor of 2 in the 350C-annealed samples with
x=0 and 0.1 This coercivity decreasing factor
increases up to 4 in the x=0.2 sample Moreover,
it is worthwhile to note that the coercivity in the
350C-annealed samples with x=0.1 and 0.2 is
almost comparable (i.e MHC equals to 1.7 and
1.6 mT) In this context, it is possible to argue that,
besides the stress releasing effects, the
nanostruc-ture formed in Y0.2(FeCo)0.8 layers due to heat
treatment must be the reason for the low magnetic
coercivity mechanism
Magnetostriction l (=lJ-l>) data are
deter-mined For all samples, the magnetostriction
obtained is comparable to the value deduced from
the data of the single-layer Terfecohan samples,
e.g lTbFeCoB103 [9–12] using the following
expression[2–3,6]:
/lS ¼lYFeCotYFeCoþ lTbFeCotTbFeCo
tYFeCoþ tTbFeCo :
Low-field parallel magnetostriction lJ data are
presented in Fig 6(a) for the as-deposited and
350C-annealed films with x=0.1 Like in
mag-netic hysteresis loops, there is a so-called
(magne-tostrictive) coercive field (lHC), where l=0 in the
magnetostrictive hysteresis loops Experimentally,
thelHCis observed to be equal to theMHCvalue
obtained from the magnetization measurements
In addition, it is in good agreement with that
already reported in Ref [12] that the
magnetos-trictive response to applied fields is always
strongest in the magnetizing fields just above the
coercivity Because the performance of
microsys-tems is determined by the parallel magnetostrictive response to an applied field, the observed behavior
is an important factor to consider the working point for the magnetostrictive films in microsys-tems In this case, the value of the parallel magnetostrictive susceptibility wlJ is usually dis-cussed The low-field dependence of the parallel magnetostrictive susceptibility is shown in
Fig 6(b) for the corresponding films It can be seen from this figure that the as-deposited x=0.1 sample with the natural nanostructure exhibits already a wlJvalue as large as 3.2 102T1 This value is almost 5 times higher than that of as-deposited x=0 sample and is comparable with that of 350C-annealed x=0 one The magnetos-trictive softness, in particular, is strongly improved after annealing at 350C: wlJ reaches a maximal value of 14.5 102T1 at the magnetic field
Fig 6 Low-field magnetostriction (a) and parallel magnetos-trictive susceptibility (b) data of the as-deposited and 350 C-annealed x=0.1 multilayer.
Trang 51.9 mT This wlJ value is almost 15 times higher
than that obtained in Terfenol-D[13]and 2 times
higher than that obtained in multilayers by
Quandt et al [2,3] Low-field lJ and wlJ data are
presented inFig 7for the x=0.2 films annealed at
400C(i.e for the thermally induced
nanostruc-tured film) It turns out that the maximal wlJvalue
equals 15.6 102T1 This is in good accordance
with the coercivity data that the excellent magnetic
as well as magnetostrictive softness can be
obtained either in naturally or in thermally formed
nanostructured multilayers
4 Concluding remarks
For conclusion, it is worthwhile to mention that,
in general, a large l value is not obtained in
conjunction with large wlJ This is because magnetostriction is intimately associated with magnetocrystalline anisotropy and it is well known that the coercivity tends to be high in large anisotropy systems The obtained spectacular result illustrates the significance of the approach, which we have developed in view of optimizing both magnetostriction and magnetostrictive sus-ceptibility in the spring-exchange magnet type multilayers with structurally discontinuous soft layers
Acknowledgements This work was granted by the State Program for Natural Scientific Researches of Vietnam The authors acknowledge discussions with D Givord
References
[1] R Coehoorn, D.B de Mooij, J.P.W.B Duchteau, K.H.J Buschow, J Phys (Paris) Colloq 49 (1988) C8;
R Coehoorn, D.B de Mooij, D de Waard, J Magn Magn Mater 88 (1989) 101.
[2] E Quandt, A Ludwig, J Betz, K Mackay, D Givord,
J App Phys 81 (1997) 5420.
[3] A Ludwig, E Quandt, J Appl Phys 87 (2000) 4691 [4] N.H Duc, D.T Huong Giang, V.N Thuc, L.V Vu,
N Chau, Appl Phys Lett., submitted.
[5] N.H Duc, J Magn Magn Mater 242–245 (2002) 1411.
[6] N.H Duc, P.E Brommer, in: K.H.J Buschow (Ed.), Handbook of Magnetic Materials, Vol 14, Elsevier Science, North-Holland, Amsterdam, 2002, p 89 [7] P Hansen, in: K.H.J Buschow (Ed.), Handbook of Magnetic Materials, Vol 6, Elsevier Science, North-Holland, Amsterdam, 1991, p 289.
[8] U Herr, H Geisler, H Ippach, K Samwer, Phys Rev.
B 59 (1999) 13719.
[9] N.H Duc, K Mackay, J Betz, D Givord, J Appl Phys.
79 (1996) 973.
[10] T.M Danh, N.H Duc, H.N Thanh, J Teillet, J Appl Phys 87 (2000) 7208.
[11] N.H Duc, D.T Huong Giang, V.N Thuc, N.T Minh Hong, N Chau, Physica B 327 (2003) 328.
[12] N.H Duc, T.M Danh, N.A Tuan, J Teillet, Appl Phys Lett 76 (2001) 3848.
[13] E Tr !emolet de Lacheisserise, D Gignoux, M Schlenker, Magnetism, Vol 2, Kluwer Academic Publisher, Dor-drecht, 2002, p 227.
Fig 7 Low-field magnetostriction (a) and parallel
magnetos-trictive susceptibility (b) data of the 400 C- annealed x=0.2
multilayer.