Solid State Communications, Vol. 20, pp. 863—868, 1976. Pergamon Press. Printed in Great Britain TRANSITION FROM FERROMAGNETISM TO MICTOMAGNETISM IN Fe—Al ALLOYS* R.D. Shull, H. Okamoto and P.A. Beck Department of Metallurgy and Materials Research Laboratory, University of illinois, Urbana, IL 61801, U.S.A. (Received 9 August 1976 by A.G. Chynoweth) Alternating low-field susceptibility and steady-field magnetization measure- ments show the occurrence of mictomagnetism in Fe—Al alloys with either the [Fe 3A1I-or the [FeA1J-typeatomic order over almost the en- tire composition range of these two phases. Fe70Al~ [Fe3A1]becomes ferromagnetic on cooling below T~ = 400 K, it becomes superparamagnetic on further cooling below the “reversed Curie temperature” Ta ~ 170 K and it becomes mictomagnetic below the freezing temperature of Tf = 92K. Alloys Fe,~ 5A129 ~ to Fe73A127 with [Fe3Al] order transform direct- ly from the ferromagnetic to the mictomagnetic state, without inter- mediate paramagnetism. RECENT STUDIES 13 indicate that Fe—Al solid solu- under argon atmosphere, homogenized by annealing at tions with the [FeAt] ordered crystal structure (Csa- 1200°C for four days and quenched in iced brine. All type) extend between the limits of approximately 24 alloys were then annealed at 600—700°C for periods of and 51 at.%Al. The [Fe 3AI]-type ordered crystal struc- one to five days in order to achieve [FeA1J-type order- ture extends from somewhat below 25 to slightly above ing. The Fe~Al32 [FeAt] specimen was then given four 32 at.% Al. Arrott and Sato 4 noted a very unusual days of annealing at 540°C, followed by a ten-day anneal feature in a 30.4% Al alloy in the [FeAl] ordered state, at 500°C. The Fe 55A145 and Fe52A148 specimens were which was ferromagnetic at room temperature. On cool- subjected to a more elaborate heat treatment: cooling ing this alloy, the spontaneous magnetization disappeared from 700 to 540°C at 20°C/day and then to 470°C at below 180K. Arrott and Sato suggested the likelyhood 10°C/day. This heat treatment was previously found 3 of a transition on cooling from ferromagnetism to anti- to improve considerably the [FeAl] order in Fe 49A151. ferromagnetism as a result of indirect exchange inter- [Fe3Al]-type ordering in alloys with 25—32at.%Al was action. 5 However, Pickart and Nathans6 found no generated by additional heat treatment (following that evidence for long range antiferromagnetism, by means designed to produce [FeA1]-type order). The additional of neutron diffraction. 01ni7 confirmed this for the heat treatment consisted of seven days of annealing at ordered equiatomic alloy, FeAJ which, according to the 460°C,followed by a 23—30 -day anneal at 440°C. indirect exchange model5 should have only antiferro- Steady-field magnetic measurements were made by magnetic interactions. It is now quite clear from the the Faraday method, using a vacuum electrobalance. analysis of magnetic data8 that any moment that occurs Alternating low-field susceptibility measurements were in FeAJ is associated with defects in the atomic ordering made at 200 Hz with a differential transformer and a and that, in the annealed condition, the alloy is para- lock-in amplifier; the primary current was held constant magnetic. Kouvel has shown9 that the alloy Fe~Al~ in all measurements at a value producing an alternating [FeAI] has a displaced hysteresis loop at 1.8 K, after field amplitude of 4 Oe. The quadrature component of being cooled to that temperature in a magnetic field. the output of the lock-in amplifier was corrected for the The present work was undertaimn in order to clarif~y the output with the specimen removed from the transformer nature of the transition from ferromagnetism to micto- and it was standardized for specimen mass and for the magnetism on cooling of alloys near 30 at.% Al and to amplification used. The standardized signal gave a repro- determine the temperature and composition limits of ducible relative index, S, of the susceptibility. Reman- the paramagnetic to mictomagnetic transition in the ence measurements were made by cooling the 5 mm dia. Fe—Al system. spherical specimens from room temperature to the tern- Alloys with 25, 27,28.5,29.5,30,30.4,31,32,45, perature of the measurement in a field of 12.6 kOe, and 48 at.% Al were melted in an induction furnace shutting off the field and 10 mm. later quickly rotating the specimen 180° around an axis perpendicular to the * direction of the applied field between two coils connec- Work supported by a grant from the U.S. ERDA, . . Contract No. E(l1-1)-1 198. ted in senes. The current passing through the coils 863