HAL Id: jpa-00217829
https://hal.archives-ouvertes.fr/jpa-00217829
Submitted on 1 Jan 1978
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of
sci-entific research documents, whether they are
pub-lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
NEUTRON SCATTERING STUDY ON THE
SPIN-FLOP SYSTEM CsMnBr3.2D2O
J. Basten, E. Frikkee, W. de Jonge
To cite this version:
JOURNAL DE PHYSIQUE
Colloque C6, supplément au n° 8, Tome 39, août 1978, page C6-819
NEUTRON SCATTERING STUDY ON THE SPIN-FLOP SYSTEM Cs.MnBr3.2D20
J . A . J . Basten, E. Frikkee and W.J.M. De Jonge
Netherlands Energy Research Foundation (ECN), Westerduinweg Z, 17SS LE Petten (NH) The Netherlands
Department of Physios, Eindhoven University of Technology, Eindhoven, The Netherlands
Résumé.- La variation de l'aimantation alternée avec la température et le champ magnétique dans le système "spin-fiop" CsMnBr3.2D2Û a été étudiée par diffraction neutronique. Le diagramme de phase magnétique a été obtenu. Les valeurs des exposants critiques & || et B_j_ sont respectivement 0.321(6) et 0.326(7).
Abstract.- The dependence of the staggered magnetization on temperature and magnetic field in the
spin-flop system CsMnBr3.2D20 was studied by means of neutron scattering. The magnetic phase diagram
is reported. The critical exponents B|| and B[_are found to be g|| = 0.321(6) and gj_= 0.326(7)
Neutron scattering experiments on weakly a n i -sotropic antiferromagnets exhibiting a spin-flop transition, can provide several tests of the theory of bicritical behaviour, which w a s developed recent-l y / 1 , 2 / . Particurecent-larrecent-ly, the order parameters both in the antiferromagnetic (AF) and the spin-flop (SF) p h a s e , i.e. the components of the staggered m a g n e t i -zation, M ' ' along and M 4/ perpendicular to the easy axis, can b e observed b y neutron scattering. CsMnBr3 2D2O was chosen because of the convenient location of the spin-flop bicritical point (H, -26 k O e , T, = 5.3 K ) , and because o f its orthorombic structure
(space group P c c a ) . Its pseudoonedimensional H e i -senberg properties are an additional interesting feature. Below T 6 . 3 K the AFordering can b e d e s -cribed/3/ by the magnetic spacegroup Pc'c'a' (see figure 1 a ) .
^ i ^S N^ 0 T T 6
O-•Gic— > » 4 — ' 3v i V r v -V 9
9-\ ] ^J l—I—I—I—U,
(a) (b)Fig. 1 : (a) the array of magnetic moments in CsMnBrs.2D20. (b) T h e a c -plane of the reciprocal lattice. Magnetic and nuclear reflections have b e e n indicated by filled and^open circles, respectively. <(> is the angle between K and the intermediate ( c ) -axis.
The b-axis and the c-axis a r e the easy and interme-diate direction, respectively / 4 / . In this
experi-ment the scattering vector K was confined to the a c -plane. In this plane magnetic and nuclear scat-tering occurs at different reciprocal lattice points (see figure lb) both in the AF-and in the SF-phase.
The experiments were performed on a diffrac-tometer at the HFR-reactor at Petten. The disk-like
crystal of 15x15x4 mm3 was fixed in a
superconduc-ting magnet (field homogeneity 1%0) with H parallel
to the easy (b)-axis. After mounting the crystal no final adjustments could be made. From the required canting angle of the cryomagnetic system (+ sample) out of the vertical position, the mismatch between
->• o
H and the b-axis was estimated as 0.3 ± 0.3 . Near T, the temperature stability was better than 1 mK with a maximum drift of 5 mK over a day.
The variation with H and T in the peak inten-sities" of three magnetic Bragg reflections, viz.
(1 0 3 ) , (1 0 1) and (3 0 1 ) , were recorded in the
AF-phase, where I(t) = C(ic) (Ms't)2» and in the
SF-phase where I ( K ) = C ( K ) (M-Lsin<j>)2. Here<j)is the angle between K and the intermediate (c)-axis (see figure lb). The left hand side of figure 2 shows some results of H-scans near the spin-flop transi-tion at different T. The data have been corrected for non-magnetic contributions. The proportionali-ty factors C ( K ) have been eliminated to achieve that all three curves coincide in the Af-phase. The same set of constants C ( K ) could be used for all tempe-ratures , which indicates that extinction can be ne-glected. The SF-intensities are now found to be pro-portional to the corresponding sin2<J>, showing the
c-axis to be the intermediate axis, indeed. From these curves, ( M " )2 and (M-L)2 can be obtained
se-parately, as shown in the right hand side of
figure 2.
Fig. 2 : (Left) The field dependence of the peak- intensities of three magnetic Bragg reflections for different T ; Triangles ( 1 0 3), squares ( I 0 1) and circles (3 0 1). Only a small fraction of the data in a scan is sh0.m. (Right) The resulting field de- pendences of
(~2:)
and (MA)', obtained from the three curves on the left, are indicated by open and closed circles respectively.+-
Although the observed rotation of Mst appears to be quite rapid, it is not a first-order spin-flop transiiion. In order to observe this first-order
-+
character, the mismatch J, between H and the b-axis is not allowed to exceed J,
-
0.08' at T = 0, as can be calculated151
from the anisotropy field (770 Oe) and the exchange field (273 kOe)/4/. For T>Q the situation is even worse as $,(TI = Qc(O).+
(Tb-T)/Tb. For a mismatch $>>J, the rotation of Mst from an Af-like to a SF-like ordering is correctly described by the molecular-field expression 161 tan(29+2$) - = s i n ( 2 $ ) / ~ o s ( 2 J , ) - ( H / H S F ) ~ ( 1 )
-+
Here 8 is the angle between M and the easy axis st
and HSF is the spin-flop field. The lower insert of figure 3 shows a comparison between the observed and calculated variation of
(MI'
)
'
and(M&)~
withst
H. The calculated curves are obtained from the ob-
served M2 st
.
2+ (MA) (dash-dotted line) andtheir theoretical ratio cos29/sin28, where 8(H) is calculated form eq. (1) with J,=0.5O. The agreement is striking. HSF(T) taken as the point of intersection of these curves is shown in figure 3.
Fig. 3 : The magnetic phase diagram of CsMnBr3.2D20. The upper insert shows details near the endpoint of the spin-flop line. The lower insert shows a compa- rison between the observed and calculated (solid lines) behaviour of ('):M; and
(M&)~
with H for an arbitrarily chosen T. The value of HsF(T) is taken as the point of intersection of both curves. (See further the main text).The location of the second-order
AF-P
boun- daryT
:
(
H
)
and SF-P boundaryT$(H)
are obtained from best fits of the magnetic intensities to( ~ ~ 1 ~ ' ~ ) ~ -
B(E)~'~~'
1
.
Here E = (T/T~-II
andE = (H/Hc-1
I
in temperature and field scans, respec- tively. From these fits with B,f3
and T or Hc va- riable, we also obtained a statistical average for the critical exponents, yielding f3 = 0.321(6) andI
I
f3
= 0.326(7). In a spin-flop system with orthorom-1
bic anisotropy (as we have in the present case) the theoretical predictions are
f3
=B
=0.325. No changeI I
I
in
f3
along both phase boundaries was faund.References
/ 1 / Kosterlitz,J.M., Nelson,D.R. and Fisher,M.E., Phys. Rev.
B13(1976) 412 and references therein
-
/2/ Domany,E. and Fisher,M.E., Phys. Rev. (1977) 3510 and references
131 Swiiste, C.H.W., De Jonge,W.J.M. and Van Meyel,J.A.G.W., Physica
76
(1974) 21141
Botterman,A.C., Ph. D. Thesis (Eindhoven, 1976) (unpubli- shed)/ 5 / Rohrer,H. and Thomas,H., J. Appl. Phys.