V.G. Shavrov1, A.A. Glebov1,2, I.E. Dikshtein1, V.V. Koledov1,2,3, D.A. Kosolapov1,2,
 E.P. Krasnoperov2, T. Takagi4, A.A. Tulaykova1, A.A. Cherechukin1,2,3


1Institute of Radio Engineering and Electronics of RAS, Moscow, Russia

2Institute of Solid State Physics and Low Temperatures. RSC "Kurchatovsky Institute", Moscow, Russia

3Laboratory of Strong Magnetic Fields and Low Temeratures, Wraclaw, Poland

4Tohoku University, Senday, Japan


Received May 17, 2001


In the Ni2,15Mn0,81Fe0,04Ga alloy magnetic control of the one-way shape memory effect has been observed, that is restoration of the shape of a formerly deformed sample at constant temperature through the magnetic field induced martensite – austenite structural phase transition. The maximal deformation observed is 3 per cent. 


The Ni2MnGa alloy attracts great attention thanks to its unique peculiarity: its thermoelastic structural (martensitic) transformation evolves in the alloy in ferromagnetic state [1]. The alloy is being intensively explored the goal being production of new inteligent materials which change their shape and size under the action of external fields. Giant reversible strains (up to 6 per cent) caused by magnetic fields due to rearrangement of different martensite variants (see, for example, [2,3]) are revealed in this alloy. But to date magnetic restoration of the shape of a deformed ("trained") sample through the structural phase transition (SPT) at constant temperature, that is magnetic control of the shape memory effect (SME), it seems, has not been observed.

            Experimenting in this domain is hampered by relatively slight dependence of the structural transition temperature on the external magnetic field intensity for Ni2MnGa samples. In the paper [4] alloys were synthesised where Mn was partially substituted by Ni in Ni2+xMn1-xGa (x = 0 – 0.20), and it was shown that the SPT temperature TM increases and that of the magnetic transition (the Curie point) TC decreases with the increase in x; TM and TC getting closer and phase transitions coinciding into one first-order transition near x = 0.19. As it was expected the consequence of closer TM and TC was a bigger influence of the magnetic field on the structural transition temperature shift, which amounted to 1 K/T [5], the SPT temperature growth exceeding the temperature hysteresis in the field of the order of 10 T. This allowed of observing the magnetically caused reversible SPT at constant temperature [6]. The examination of the Ni2+xMn1-xGa alloys phase diagrams in the (T–x) coordinates showed that the phase transitions (x = 0.16 – 0.20) coincidence temperature was near room temperature which circumstance was a boost to experimenting. However the available samples turned out to be brittle and were often destroyed during the thermocycling even unloaded. It appeared that the addition of iron allows of enhancing the plasticity of the samples without substantial deterioration of their magnetic and thermoelastic properties. The present experiments were based on the Ni2,19Mn0,81Ga alloy with the coinciding martensitic and magnetic transitions; and the alloys with Fe partially substituted by Ni were synthesised.

            The purpose of the present research was to investigate experimentally the phase diagram in the magnetic field – temperature (H–T) coordinates of the newly synthesised Ni2,15Mn0,81Fe0,04Ga alloy and to demonstrate the SME by realising the magnetically induced martensite – austenite phase transition at constant temperature.

            The polycrystallic Ni2,15Mn0,81Fe0,04Ga samples were prepared by the arch-melting method in argon environment. Then 0.5 mm thick plates were cut out and polished in the high-temperature cubic (austenitic) state at a temperature higher than the finish temperature (TMA) of the martensite – austenite transition. As the polished sample temperature drops below the austenite – martensite transition temperature (TAM) a relief, related to the appearance of tetragonal low-temperature phase (martensite) and to its division into patches (variants, domains) with randomly oriented crystallographic axes, can easily be observed in the light reflected from the sample surface. The austenite – martensite phase transition is accompanied by a lattice compression along the c-axis by about 6 per cent and an extension along the a and b axes of 2 per cent. A microscope incorporating a video camera enables one to record reliably the start and finish temperatures of martensitic transition in distinct fractions of the polycrystallic sample and thus to acquire information about the inhomogeneity of the sample.

            To study the T–H phase diagram the sample located in a transparent thermostat also containing a microscope, was placed in the field of a 0 – 15 T Bitter magnet. The sample surface resolved phase diagram was obtained by analysing the video that had been taken in the course of temperature variations across the martensitic transition point at various magnetic fields.

            The Ni2,15Mn0,81Fe0,04Ga sample T–H diagram is schematically shown in the Fig.1. The direct  martensitic transition temperature is TAM(H=0) @ 35 C whereas for the reverse transition TMA(H=0) @ 42 C. Both temperatures increase almost linearly as the magnetic field intensifies. The TM(H) dependence coefficient is about 0.7 K/T. The temperature hysteresis loop width for the martensitic phase transition is DTM = TMA - TAM = 7±2 K.

            In the fields of 10 T the transition point shift exceeds SPT hysteresis. At temperatures slightly aboveTMA(0) observation of the magnetic field induced reversible martensitic phase transition at constant temperature is possible [5]. To achieve this the sample temperature need be fixed within some magnetic field dependent interval, that can be called the magnetically controlled reversible phase transition interval, TMA(H=0)<T< TAM(H=10T), then by means of switching the field on and off the change in sample structural phase state at constant temperature is attained. In the Fig.1 this transition is represented by the "am" line. In the present case the controlled reversible phase transition interval was 1-2 K at a 10 T field intensity.

            The SME demonstration experiments were preceded by the sample "training", that is the loaded thermocycling procedure. In the present work the loading was that of bending. When cooling for the first time across the martensitic transition point the loaded sample undergoes an appreciable deformation and then, in the course of heating, regains its initial shape, performing work against the external force. After several cooling-heating cycles the sample deformation grows and the temperature hysteresis loop somewhat narrows and shifts to a higher temperature region. The phenomenon resembles one observed for the Ni2MnGa alloy [6]. Thus the samples "trained" for one-way SME in 1, 8 and 15 cycles were obtained to conduct experiments in the magnetic field. The recoverable deformation e that could be attained on the sample maximum curvature spot was 1-4 per cent.

            The physical explanation of the "reversible strain" can be inferred from the following.


Fig.1. T-H phase diagram of martensite transformation in Ni2,15Mn0,81Fe0,04Ga.


Under the action of sufficiently strong  external tension the temperature of martensite - austenite transition is increased. At cooling beyond the austenite - martensite transition point in nonuniformly stressed sample the process of martensite generation begins at the points of maximum stress. Appropriately oriented martensite variants are generated in the regions of contraction end extension, which lead to the giant enhancement of strain up to several percents and to a change in the sample shape. After heating beyond over the martensite - austenite transition point the initial shape of the sample is regained.

The experiment on the magnetic field induced SME is illustrated in the Fig.1 showing the T-H phase diagram of Ni2,15Mn0,81Fe0,04Ga together with the system thermodynamic state evolution trajectory ABCDE the course of the experiment. The initial state of the system at room temperature is below the point TMA(H=0), at magnetic field H = 0 and is marked by the point A. Initially the sample is "trained" and strain is maximum. At turning on a H=10 T magnetic field the system passes over to point B. Then with the field on the sample is heated up to T0, which satisfies TMA(H=0) < T0 < TMA(H=10 T). After that the field is switched off and the system undergoes the field induced SPT at constant temperature T0 (the point D). As a result a previously deformed ("trained") sample regains its shape (becomes straight) at constant temperature. This is the essence of the magnetic field induced one-way SME

The film presented is an excerpt of the video-recording taken in the course of this experiment in the time interval corresponding to CDE in Fig.1 at T=T0 . The recording begins at C, where H=10 T and the strain is maximum that is e ≈ 3 per cent. Then the field is made lower and the sample starts straightening (beyond the point D). Straightening goes on for another 15 seconds after the field is completely turned off. (The time scale is compressed by a factor 10.)  In the end the sample is totally straight (e=0) in the austenitic phase. The time delay of straightening relative to switching the field off is evidently connected with conveying the first order SPT latent heat to the sample.

            So, in the experiment we have directly demonstrated the effect of magnetic field induced shape memory effect at fixed temperature in the absence of external tensions. The main distinction between the magnetic field induced SME and the effect of giant magnetic field induced strains due to martensitic variants rearrangement in Ni-Mn-Ga single crystals can be stated as follows. The latter effect can be attributed to the sample shape change in low-temperature phase, and it is not related directly to the process of austenite - martensite transition. This effect is due to the fact that different martensitic variants have differently oriented tetragonal c-axes and, accordingly, different easy magnetization axes. Under the action of an external magnetic field variants acquire different magnetic energies. As a results martensitic domains boundaries are forced to move. The energetically advantageous variants tend to grow and disadvantageous ones tend to shrink.

Alternatively, the magnetic field induced SME can be ascribed to the influence of external fields on the martensite - austenite boundaries in a sample, which is in the intermediate state in the region of martensitic transition hysteresis.  The saturation magnetization of the martensite is higher than that of austenite. As a result in sufficiently high external magnetic fields, exceeding the field of magnetocrystalline anisotropy of the martensite the structural transition temperature rises and the martensite - austenite boundaries  move so that the martensite volume is enlarged.

 We can outline some peculiarities of the magnetic field induced SME as compared with the effect of giant strains due to magnetic field induced martensite domains rearrangement in Ni-Mn-Ga single crystals:

1. The magnetic field induced SME is a broader effect because a sample can be "trained" for any kind of strains for example, contraction, extension, torsion, etc. Polycrystals can be used as samples as follows.

2. The experiments on giant strains due to the magnetic field controlled martensite domains rearrangement some times produce ambiguous results because the martensite domains structure can be often treated as a thermodynamically unstable system, being very sensitive to internal defects of the structure of the material. As a result even in the case of thorough reproduction of external conditions the sample can end up in a different shape each time.

3. The magnetic field SME is observed in the fields which are 1-2 orders higher, than those leading to strains due to the magnetic field controlled martensite domains rearrangement.

4. Given a time interval the sample with magnetic field induced SME can do more a bigger work, than the one with martensite domains rearrangement and can change its shape against  greater external forces.

5. The martensite domain rearrangement is a faster effect  than the magnetic field induced SME. The characteristic times of the magnetic field induced SME are determined by characteristic times of the martensite transition latent heat absorption.

            The authors are grateful to A.N.Vasil'ev, V.D.Buchel'nikov, and V.I.Nizankovski, for discussions and to V.V.Hovailo for the help in preparation of the samples.

            The work is supported by RFBR, grants 99-02-18247, 01-02-06053.



1. P.J. Webster, K.R.A.Ziebeck, S.L. Town, M.S. Peak. Magnetic order and phase transformation in Ni2MnGa. Philos. Mag. B. 49, No. 3, 295 (1984).

2. R.C. O'Handley, S.J. Murray, M. Marioni, H. Nembach, S.M. Allen. Phenomenology of giant magnetic-field-induced strain in ferromagnetic shape-memory materials. J. Appl. Phys. 87, No. 9, 4712 (2000).

3. R.Tickle, R.D.James. Magnetic and magnetomechanical properties of Ni2MnGa. JMMM 195, 627 (1999).

4. A.N.Vasil’ev , A.D.Bozko, I.E. Dikshtein, V.V.Khovailo, V.D. Buchel'nikov, M. Matsumoto, S. Suzuki, V.G. Shavrov, T. Takagi, J. Tani. Structural and magnetic phase transitions in shape memory alloys Ni2+XMnXGa. Phys. Rev. B59(2), 1113 (1999).

5. A.A. Cherechukin, I.E. Dikshtein, D.I. Ermakov, V.V. Koledov, L.V. Koledov, T.Takagi, A.A. Tulaikova, V.G. Shavrov. Reversible structural phase transition in Ni-Mn-Ga alloys in a magnetic field. JETP Lett., 72, No. 7, 373 (2000).

6. V.V. Kokorin, V.A. Chernenko, Martensitic transformation in ferromagnetic Heusler alloy. Phys. Metals and Metallography, 68, 1157 (1989).


c3.gif (955 bytes)



c4.gif (956 bytes)