- Email: [email protected]
On the nature of the phase transition in MnO
Solid State Communications, Vol. 18, pp. 701-703, 1976.
Pergamon Press.
Printed in Great Britain
ON THE NATURE OF THE PHASE TRANSITION IN MnO J. Konstantinovi6 and B. Babi6 Laboratory of Solid State Physics, Institute of Nuclear Science "Boris Kidri~"- Vin~a, Belgrade (Received6 October 1975 by P.G. de Gennes) The temperature dependence of the nuclear Bragg reflection (222) and of the magnetic Bragg reflection (111) of MnO was simultaneously measured by neutron diffraction on a powder sample. The intensity of the nuclear line exhibits a 40% drop at temperatures between 113 and 116 K. This dip was attributed to a decrease of the Bragg intensity due to structural fluctuations. If the effect of these fluctuations on the magnetic intensity is taken into account, the magnetization discontinuity is found to be greatly reduced or even to disappear, in disagreement with other authors. NUMEROUS experimental and theoretical studies have been carried out in order to explain AF-PM transition in MnO. 1 Discontinuity of magnetization in the vicinity of transition temperature was of particular interest in these studies. Recently published papers 2'3 explain the magnetization discontinuity and the disappearance of the lattice deformation by a first order magnetic phase transition. Besides neutron diffraction measurements of magnetization, useful studies of the dynamic properties of MnO have been done by the neutron inelastic scattering method. 4"5 Parallel existence of the first and second order transitions has also been suggested. 5 In this paper we present the results of the temperature dependence of magnetic Bragg reflection (111) in MnO powderfi Our method, similar to the one used in the study of NiO 7 was devised in order to enable us to follow simultaneously the temperature dependence of the magnetic Bragg reflection (111) and non-magnetic Bragg reflection (222). A neutron beam of defined wavelength series was obtained and measured with the neutron monochromator (monocrystal of pyrolitic graphite) and neutron time-of-flight spectrometer (Fermi chopper, time analyzer 512 x 8 ~sec). The tollowing wavelengths were used; ~n = 1.943 A for magnetic reflection (111); ~.iv = 0.971 A for nuclear, i.e. non-magnetic reflection (222); ~ I I I = 1.295 A for elastic incoherent scattering used as a monitor; hi = 3.886 A reduced to minimum by choosing great number of chopper turns in order to minimize the contribution of the neutron inelastic scattering. A battery of three He a detectors, with the effective cross section 6 x 10 cm, was situated at 20 -- 22o2 ' and
at 3 m distance from the sample. The reflected neutron beam (16 × 32 mm) was scattered by the MnO powder, placed in an aluminium container and closed in the liquid nitrogen cryostat. Heaters and aluminium screens were placed around the sample together with a temperature stabilizer "ARTRONIX", sensor and three thermocouples. This arrangement made it possible to stabilize and measure the temperature of the medium, the temperature of the sample and the temperature gradient in the sample. The temperature stabilization during the measurements was of the order of 0.05 K. The temperature of the sample was measured with precision of 0.2 K. The temperature gradient in the sample was less than 0.SK. The results are shown in Figs. 1 and 2. During the measurements the temperature was changed from lower to higher values, as shown by the arrow in the Fig. 2(b). The detector was fixed at the maximum of the Bragg's peak in the paramagnetic phase. The discontinuity of the neutron Bragg reflection intensity was observed at the temperature T1 = (115 -+ 1)K. Contribution of the magnetic elastic scattering disappears at 7'2 = (121 + 2)K. These results are compared in Fig. 3 with the measurement of Bloch et al. 2 In order to facilitate this comparison, the square root of the neutron intensity measured in this work was normalised to the results of Bloch et al. There is a very good agreement between the two sets of data although they were obtained by measurements on powder in one case and on monocrystat in the other case. However the interpretation of the data given by Bloch et al. will be criticized below. If one now considers the intensity IN of the nuclear Bragg reflection [Fig. 2 (a)], it is seen to exhibit a 40% dip of width 4 K around T1 = 115 K. This rather surprising effect was confirmed by a series of careful experiments. Although we are unable to provide a detailed explanation, it seems that the 40% dip can only be 701
ON THE NATURE OF THE PHASE TRANSITION IN MnO
702
'!I
I 18
•
16
?4
Vol. 18, No. 6 [o)
•
oo
• gO om
12 10
T=tl2.5 K
• .
~:a~
r0q
8 6
,%
(b)
" ", •
" ~
-
~
:
t~.t,.,~.,,~..,~,...,~-...',
o
~,.,,~.-~ ~ . .
2OO
I00
30
~ =015*t)K T2=(T21±2)K
3OO
o
n(chonets)
ee 20
;'..
&
, 2bo
3bo
T=121 K .
" 100
.~'.p• 200
\
n(chonel$)
• "
..
T=n55 K
% e~'-~,__o~_
-
Fig. 2. (a) Temperature dependence of the non-magnetic Bragg reflection (222)• (b) Temperature dependence of the magnetic Bragg reflection (111).
300 n{cho~els) ... Bloch et a l Phys Lett. 49A,354,(1974)
Fig. 1. Time-of-flight neutron spectra for magnetic Bragg reflection (111) - Xn; for nuclear Bragg reflection (222) - Xrv and elastic incoherent neutron scattering - Xm at three typical temperatures•
opo o u r m e c l s u r e m e n t s 6 x ox ox
attributed to local fluctuations in the crystalline structure. Strong structural fluctuations can actually be expected near a magnetic phase transition if there are strong magnetoelastic interactions, as is the case in MnO. The dip in the nuclear intensity suggests that the sub-lattice magnetization is not really proportional to the square root of the magnetic intensity IM. Actually the latter is essentially proportional to
ot
o
xo % o
o
IM ~ ~, (Mi)(Mj) exp [ l k ( R / - Rj)I. i,J This expression is of course proportional to the square of the sublattice magnetization I
o.5
i TITN
Fig• 3• Temperature dependence of the square root of neutron magnetic Bragg reflection (111); comparison with published data. magnetization turns out to be much less abrupt, suggesting a second order transition, although our data are not precise enough to rule out a weakly first order transition• We conclude that the transition may be a second order transition or very weak first order, in spite of the abrupt change in I m at T1. It should be pointed out that in both explanations (first order transition of the B e a n Rodbell type, and reduction of intensity due to lattice fluctutations) the interaction between spins and deformations plays a crucial role.
Vol. 18, No. 6
ON THE NATURE OF THE PHASE TRANSITION IN MnO
703
Acknowledgments - We are greatly indebted to J. Villain, Z. Mari~ and M. Popovi6 for very helpful discussions.
REFERENCES 1.
SHULLC.G.,STRAUSERW.A.&WOLLANE.O.,Phys. Rev. 83,333(1951);ROTHW.L.,Phys. Rev. 110, 1333 (1958); BIZETTE H., SQUIRE F.C. & TSAI B., Comptes Rendus 207,449 (1938); LINES M.E. & JONES E.D.,Phys. Rev. 139A, 1313 (1965); RODBELL D.S. & OWEN J.,J. Appl. Phys. 35, 1002 (1964); BLOCH D., CHARBIT P. & GEORGES R., Comptes Rendus 266,430 (1968); MOROSIN B., Phys. Rev. B1, 236 (1970); BARTEL L.C., Phys. Rev. B1, 1254 (1970); BLOCH D., FERON J.L., GEORGES R. & JACOBS I.S., J. Appl. Phys. 38, 1474 (1967).
2.
BLOCHD., MAURY R., VETTER C. & YELON W.B.,Phys. Lett. 49A, 354 (1974).
3.
BRAZOVSKIIS.A. & DEJALO~INSKII I.E., JETP 21,360 (1975).
4.
PEPY G.,J. Phys. Chem. Solids 35, 433 (1974).
5.
KROON. & BATA L., Neutron Inelastic Scattering H, Proceedings of a Symposium, Copenhagen, 111 (1968).
6.
Produced by A.D. Mackay, 198 Broadway, New York, NY 10038, U.S.A.
7.
NEGOVETI~I. & KONSTANTINOVI~ J., Solid State Commun. 13,249 (1973).
Pergamon Press.
Printed in Great Britain
ON THE NATURE OF THE PHASE TRANSITION IN MnO J. Konstantinovi6 and B. Babi6 Laboratory of Solid State Physics, Institute of Nuclear Science "Boris Kidri~"- Vin~a, Belgrade (Received6 October 1975 by P.G. de Gennes) The temperature dependence of the nuclear Bragg reflection (222) and of the magnetic Bragg reflection (111) of MnO was simultaneously measured by neutron diffraction on a powder sample. The intensity of the nuclear line exhibits a 40% drop at temperatures between 113 and 116 K. This dip was attributed to a decrease of the Bragg intensity due to structural fluctuations. If the effect of these fluctuations on the magnetic intensity is taken into account, the magnetization discontinuity is found to be greatly reduced or even to disappear, in disagreement with other authors. NUMEROUS experimental and theoretical studies have been carried out in order to explain AF-PM transition in MnO. 1 Discontinuity of magnetization in the vicinity of transition temperature was of particular interest in these studies. Recently published papers 2'3 explain the magnetization discontinuity and the disappearance of the lattice deformation by a first order magnetic phase transition. Besides neutron diffraction measurements of magnetization, useful studies of the dynamic properties of MnO have been done by the neutron inelastic scattering method. 4"5 Parallel existence of the first and second order transitions has also been suggested. 5 In this paper we present the results of the temperature dependence of magnetic Bragg reflection (111) in MnO powderfi Our method, similar to the one used in the study of NiO 7 was devised in order to enable us to follow simultaneously the temperature dependence of the magnetic Bragg reflection (111) and non-magnetic Bragg reflection (222). A neutron beam of defined wavelength series was obtained and measured with the neutron monochromator (monocrystal of pyrolitic graphite) and neutron time-of-flight spectrometer (Fermi chopper, time analyzer 512 x 8 ~sec). The tollowing wavelengths were used; ~n = 1.943 A for magnetic reflection (111); ~.iv = 0.971 A for nuclear, i.e. non-magnetic reflection (222); ~ I I I = 1.295 A for elastic incoherent scattering used as a monitor; hi = 3.886 A reduced to minimum by choosing great number of chopper turns in order to minimize the contribution of the neutron inelastic scattering. A battery of three He a detectors, with the effective cross section 6 x 10 cm, was situated at 20 -- 22o2 ' and
at 3 m distance from the sample. The reflected neutron beam (16 × 32 mm) was scattered by the MnO powder, placed in an aluminium container and closed in the liquid nitrogen cryostat. Heaters and aluminium screens were placed around the sample together with a temperature stabilizer "ARTRONIX", sensor and three thermocouples. This arrangement made it possible to stabilize and measure the temperature of the medium, the temperature of the sample and the temperature gradient in the sample. The temperature stabilization during the measurements was of the order of 0.05 K. The temperature of the sample was measured with precision of 0.2 K. The temperature gradient in the sample was less than 0.SK. The results are shown in Figs. 1 and 2. During the measurements the temperature was changed from lower to higher values, as shown by the arrow in the Fig. 2(b). The detector was fixed at the maximum of the Bragg's peak in the paramagnetic phase. The discontinuity of the neutron Bragg reflection intensity was observed at the temperature T1 = (115 -+ 1)K. Contribution of the magnetic elastic scattering disappears at 7'2 = (121 + 2)K. These results are compared in Fig. 3 with the measurement of Bloch et al. 2 In order to facilitate this comparison, the square root of the neutron intensity measured in this work was normalised to the results of Bloch et al. There is a very good agreement between the two sets of data although they were obtained by measurements on powder in one case and on monocrystat in the other case. However the interpretation of the data given by Bloch et al. will be criticized below. If one now considers the intensity IN of the nuclear Bragg reflection [Fig. 2 (a)], it is seen to exhibit a 40% dip of width 4 K around T1 = 115 K. This rather surprising effect was confirmed by a series of careful experiments. Although we are unable to provide a detailed explanation, it seems that the 40% dip can only be 701
ON THE NATURE OF THE PHASE TRANSITION IN MnO
702
'!I
I 18
•
16
?4
Vol. 18, No. 6 [o)
•
oo
• gO om
12 10
T=tl2.5 K
• .
~:a~
r0q
8 6
,%
(b)
" ", •
" ~
-
~
:
t~.t,.,~.,,~..,~,...,~-...',
o
~,.,,~.-~ ~ . .
2OO
I00
30
~ =015*t)K T2=(T21±2)K
3OO
o
n(chonets)
ee 20
;'..
&
, 2bo
3bo
T=121 K .
" 100
.~'.p• 200
\
n(chonel$)
• "
..
T=n55 K
% e~'-~,__o~_
-
Fig. 2. (a) Temperature dependence of the non-magnetic Bragg reflection (222)• (b) Temperature dependence of the magnetic Bragg reflection (111).
300 n{cho~els) ... Bloch et a l Phys Lett. 49A,354,(1974)
Fig. 1. Time-of-flight neutron spectra for magnetic Bragg reflection (111) - Xn; for nuclear Bragg reflection (222) - Xrv and elastic incoherent neutron scattering - Xm at three typical temperatures•
opo o u r m e c l s u r e m e n t s 6 x ox ox
attributed to local fluctuations in the crystalline structure. Strong structural fluctuations can actually be expected near a magnetic phase transition if there are strong magnetoelastic interactions, as is the case in MnO. The dip in the nuclear intensity suggests that the sub-lattice magnetization is not really proportional to the square root of the magnetic intensity IM. Actually the latter is essentially proportional to
ot
o
xo % o
o
IM ~ ~, (Mi)(Mj) exp [ l k ( R / - Rj)I. i,J This expression is of course proportional to the square of the sublattice magnetization I
o.5
i TITN
Fig• 3• Temperature dependence of the square root of neutron magnetic Bragg reflection (111); comparison with published data. magnetization turns out to be much less abrupt, suggesting a second order transition, although our data are not precise enough to rule out a weakly first order transition• We conclude that the transition may be a second order transition or very weak first order, in spite of the abrupt change in I m at T1. It should be pointed out that in both explanations (first order transition of the B e a n Rodbell type, and reduction of intensity due to lattice fluctutations) the interaction between spins and deformations plays a crucial role.
Vol. 18, No. 6
ON THE NATURE OF THE PHASE TRANSITION IN MnO
703
Acknowledgments - We are greatly indebted to J. Villain, Z. Mari~ and M. Popovi6 for very helpful discussions.
REFERENCES 1.
SHULLC.G.,STRAUSERW.A.&WOLLANE.O.,Phys. Rev. 83,333(1951);ROTHW.L.,Phys. Rev. 110, 1333 (1958); BIZETTE H., SQUIRE F.C. & TSAI B., Comptes Rendus 207,449 (1938); LINES M.E. & JONES E.D.,Phys. Rev. 139A, 1313 (1965); RODBELL D.S. & OWEN J.,J. Appl. Phys. 35, 1002 (1964); BLOCH D., CHARBIT P. & GEORGES R., Comptes Rendus 266,430 (1968); MOROSIN B., Phys. Rev. B1, 236 (1970); BARTEL L.C., Phys. Rev. B1, 1254 (1970); BLOCH D., FERON J.L., GEORGES R. & JACOBS I.S., J. Appl. Phys. 38, 1474 (1967).
2.
BLOCHD., MAURY R., VETTER C. & YELON W.B.,Phys. Lett. 49A, 354 (1974).
3.
BRAZOVSKIIS.A. & DEJALO~INSKII I.E., JETP 21,360 (1975).
4.
PEPY G.,J. Phys. Chem. Solids 35, 433 (1974).
5.
KROON. & BATA L., Neutron Inelastic Scattering H, Proceedings of a Symposium, Copenhagen, 111 (1968).
6.
Produced by A.D. Mackay, 198 Broadway, New York, NY 10038, U.S.A.
7.
NEGOVETI~I. & KONSTANTINOVI~ J., Solid State Commun. 13,249 (1973).