Magnetic susceptibility of antiferromagnetic nickel diglycine dihydrate

The magnetic susceptibility X of nicke] diglycine dihydrate has becn measured between 0.6 and 120 K. The high-temperature data were taken in powder and follow a Curie law behavior with µ '" = 3.189 µ 8 and an antiferromagnetic Curie tem~rature T<: = 2.47 K. At low temperatures the x vs temperature curves for powder and oriented single crystals deviate from the Curie law; they show a rna1'imum ,tat about 2.21 K and an abrupt decrease at lower temperatures. The observed behavior of ,t for nickel diglycine dihydrate indicates dominating antiferromagnetic interactions between Ni ions. A phase transition to an antiferromagnetic phase occurs below 2.2 K.

The chernica l environment o f t he met a l i on s in metal derivatives of aminoac ids is simi lar to that found in some proteins(l) and their study could be useful in learning rnore about the el ectronic structure and magnetic coupling between metal ions in these rnacromolecules.
A leas t squares f i t o f Eq. (1) to t he data in Fig .l gives C:l.269 emu K/mole and a posi tive Tc:2.47 K In this paper we report measurernents of the magnet i c susceptibility X in nicke !digl ycine dihydrate 0 J in a wide temper ature range.
The data were t aken primarily in c rvstalli ne powders , but preliminary data in oriented s ingle crystals ar e also presented .The suscept ibili t y data show a trans i tion o f nickel d iglycine dihyd r ate (NiDD) to an an tiferromagnetic phase below 2 . 2 K; at h i.ghe r temperatures single ion interactions are the most important.A detailed description o f the rnagnetic properties of NiDD is not yet available and we use an effec tive molecular field mode l to discuss the experiment al data.

EXPERIMENTAL DETAILS
Nickel di glycine dihydra t e was synthesized as outlined by Stosick(2) and detailed by Sen et al (3).The material was puri fied by successive recrystallizations.Single crystals weighting about 10 mg were obtained by slow evapora tion (typically two months) at room ternperature from a water solution.Powde r samples were obtained f rom the singl e c r ystals.
A Faraday magnetometer was used to measur e X between 0.6 K and 120 K. tn chis tem9erature range the d i amagnet ic corrections are telatively srnall.The magnetic suscept ibility of the sample holder was subtracted from the data.
In single crystals, which cleave along t he bc plane , we were able to identify t he b axis and measure X in the b, c and bxc directions .

RESULTS AND CONCLUSIONS
-1 The inverse molar susceptibility X vs. temper ature T data for powdered NiDD are shown in + refl e cting the e ffect of t he spinorbic interaction on the c rystal field levels of the ion(4) .The magnetic susceptibility )( vs. T data for powdered NiDD below 20 K is shown in Pig. 2. The data in this range were taken with a field of 1400 Gauss and were found to deviate from the Curie law behavior.The susceptibility shows a rounded maximwn <x=0.172en;ui/mole} ar~und T-2.2 K, and an abrupt decrease of X 'Wl.th decreas1ng T below 1.5 K, as expected for dominating antiferromagnetic interactions.
Pigure 2: Experimental values for the magnetic susceptibility X vs. temperature T for powdered NiDD.
We show in Fig. 3  the bc plane, each connected to the f our nearest Ni atoms by strong hydrogen bonds.The magnetic interaction between nearest Ni atoms in the same plane is the strongest because the smaller interatomic distance and the existence of the strong hydrogen bond.The Hamiltonian describing the magnetic properties of NiDD in the presence of an external magnetic field is d{=r.°lf(i,>.).'D<>.> .s(i,>.) i=l ,N >.

+ r.
s(i,>.) .J(i,>.,i' , >.').s(i',>.')(2) i,i'=l,N >.,>.' In Eq. ( 2), S(i,A} is the spig of the Ni ion in the A site of the ith unit cell, D{A) is the fine structure tensor for Ni ions in the >.site and J(i,>.;i' , ).') is the magnetic interaction between S(i,>.) and S(i',>.' ).We call >.=A and A~B the two sites for Ni ions in the unit cell of NiDD.Specific heat measurements (6) on NiDD indicate a phase transition at T 0 =0.88K and a complex Schott ky contribution which allows to obtain D/k=-14.SK and IELkl=l.l3K for the principal values_of the tensors D(A).The principal directions of D{).) are not known and the crystal data give no reason to assume that they are the same for both Ni atoms in the unit cell.
A simple molecular field model has been used to analyze the effect of the last term of Eq. ( 2). the crystal data on NiDD discussed above suggest a two sublattice model, where each sublattice is identified wich one of the Ni atoms in the unit cell .Even if the identificat ion of the magnetic sublattices is different from that given by Berger and Friedberg (7) to explain their data in nickel nitrate dihydrate, the algebra involved in our calculations is very similar to theirs.Then the rnagnetic field acting on the Ni ions in the A and B sublattices is reolaced in Eq. ( 2) by and where ii is the external field, MA and ~ are the A and B sublattice magnetizations and n 1 and n 2 the molecular field parameters(7}.The calculation given by Berger and Friedberg has been ext ended in order to introduce the effect of the non axial spin Hamiltonian parameter E which is non-zero for NiDD (6).In our calculation we have assumed D(A) diagonal and equal for both Ni ions in the unit cell of llliDO.
Using the value of Tc given above , and the values of T Fig.l.i ncl.ica ting that the antiferromagne t ic interactions ' s number and s~1 for Ni 2 + ions , we obtai n ehe values g=2.253 and µeff=3.186 µB for the gyromagneti c factor and the effec t ive magnetic moment.These values are expected for high spi n oc tahed ral compounds of Ni 2 Figure 1: Figure 3: 0 , D/k and lE/kl obtained from the NiDD using this model are in poor agreement with the experimental data particularly below 5K.The reasons for this disagreement are, probably, the fact that the effective field model neglects any effect originating in short range order and .alsobecauseweare not considering that the tensors D().) are not di-Magnetism & Magnetlc Materials-1981agonal and are rotated for the two Ni ions in the unit cell of NiDD.It seems that a_theory t.Jlat consider the different orie.ntations of D(A) and D(B) , and a less simplistic magnetic structure of NiDD is needed to explain the data.Paramagnetic resonance measurements of the Ni2+ ions and neutron diffraction studies of the ordered phase would be useful for this purpose.