Eligiendo el n´ umero de cubetas y la funci´ on hash

logging.

Temperatures below 30 K were measured with a 1000 ft carbon

resistor while above 30 K a Cu/constantan thermocouple was employed. The carbon resistor was found to be slightly field dependent below « 10 K. For measurements at other than the lowest attainable temperature (4.28 K) errors of up to 0.2 K can result from this dependence. In general, the errors in temperature were found to be less than 0.1 K.

Magnetization measurements at six field strengths between 5 and

46 kG were made over the temperature range 4.2 to 220 K, for [Mn^^sar] (CF^SO^)^. In order to prevent sample alignment at high magnetic fields and low temperatures, samples were ground with

t For Fe(II) and Co(II) cages, 5 and 10 kG main fields were employed in order to prevent sample alignment.

I

I

Vaseline and a uniform mull was prepared. Measurements were made

with both increasing and decreasing magnetic fields, and there was no evidence of sample orientation.

Once the data were corrected for both Vaseline and the sample

holder, the mass of the sample was determined by comparing the gram susceptibility of the neat powder at 10 kG with that of the mull,

over a temperature range of 30 K.

Before measuring susceptibilities over a wide temperature range,

the room temperature magnetic moment of all compounds were measured

on a second Faraday balance consisting of a Newport 4 inch

electromagnet and a Cahn RG electrobalance. Field strengths of 5 to 6.5 kG, calibrated against Hg[Co(CNS)^] were used to ensure no field dependent impurities were present in the sample. The results obtained with the room temperature and Oxford Faraday balances

generally agreed to better than 0.05 B.M.

Some room temperature magnetic moments were measured by the Faraday method on a Newport Instruments Gouy balance with a 3/2" type C electromagnet and a Faraday kit conversion as supplied. The diamagnetic corrections for the metal-free ligands were measured

using their hydrated bromide salts, i .e. sar.4 V2 HBr.3H20 and cq ( N l ^ ^ sar.5HBr.3H2O. Tabulated values of Pascal's constants were used to determine the susceptibilities of the water and HBr molecules

for each compound and resulted in x^(sar) = -300 x 10 cm mol and —6 3 — 1

Xj((NH0)0sar) = -172 x 10 cm mol . Unless otherwise stated all d

L L

measurements are in the unrationalized e.g.s. electromagnetic units which has been the standard practice of magnetochemists for many years. S.I. units are rationalized, and so the conversion from e.g.s. e.m.u. to S.I. magnetic units involves factors which include 4 tt , since the permeability of free space is 4 tt x 10“ 7 kg ms“2A~2

in S.I., but is unity, and dimensionless, in the e.g.s. system. For convenience, the well-known e.g.s. system is retained here.^

In the forthcoming discussion of the magnetic behaviour of the

cage complexes, an octahedral ligand field is assumed to be the initial perturbation of the free ion term. Other perturbations that remove orbital or spin degeneracy, i . e . spin-orbit coupling, zero-

field splitting, Jahn-Teller distortions etc. are considered as subsequent perturbations where relevant.

2.3.3 Computer Assisted Calculations

Susceptibility calculations were performed with a Fortran

program COM/MAG^* on the Monash University Burrough's 6700 computer.

The indicator of "best fit" was obtained by minimization of the R factor, defined by Equation 10,

R = ~ X ( c a l c ) 1 ■ x 100 ...(10) x(obs)

Some data were fitted using a Fortran program CAMMAG, developed

ft*?

at Cambridge University. ^ The input data for this program were the

molecular dimensions of the cage .required for the angular overlap model (AOM) parameters. The output of the program yields spectral transition energies and assignments, magnetic susceptibilities and

their orientations and molecular g-tensors. Further details of this program are provided in reference 62.

t In the e.g.s. system the Bohr magneton is 0.92731 x 10“ 20 erg gauss” * and in S.I. it is 0.92731 x 10- 2 3 A m molecule- *. Using these values, the magnetic moment, |i , is 2.828(%mT)/2 Bohr magnetons in S.I. units. These two expressions give the same numerical value for the magnetic moment, and are valid in either system of units.

o

r* X

g

UI w

s

<

8

CC fe

o

UJ tt

f

o

z

m H Z O

S

3

00 100 200 TEMPERATURE CK3 300 Figure 2.2

Variation of the magnetic moment and the reciprocal susceptibility with temperature for [V sar-2H](PF6)2.5H20 .

Figure 2.3

Variation of the magnetic moment and the reciprocal susceptibility with temperature for [V(NH3)2sar-2H]Cl4.5H20 .

2.4 RESULTS AND DISCUSSION

2.4.1 The 3d * Configuration

Attempts to isolate the Ti(III) cage have so far been unsuccessful. However, success has been achieved in the isolation of what are believed to be the first hexaamine complexes of V(IV). The

[V sar-2H](PF6)2.5H20 and [V(NH3 )2sar-2H]Cl4 .5H20 complexes, (sar-2H indicates deprotonation at coordinated N-atoms), exist as doubly deprotonated salts.

The magnetic moments measured at room temperature for these

complexes are = 1.80 B.M. for [V sar-2H]2+ (9 = 3.2 K ) , and

U = 1.93 B.M. for [V(NH3)2sar-2H]4+ (0 = 3.1 K ) , consistent with the expected 3d* configuration. Both complexes show little variation of their moments when cooled to liquid helium temperature (Figures

2.2 and 2.3).

The action of a cubic ligand field on the free ion D term lifts

its degeneracy and leads to a T2g term lyin8 lowest. The six-fold degeneracy of this ground term may be lifted by spin-orbit coupling, as discussed by K o t a n i ^ (cf . Figure 2.4); it may also be removed by a ligand field of symmetry lower than cubic. It has been shown that ligand fields of low symmetry have comparatively little effect on the effective magnetic moment unless they raise the orbital degeneracy of

2

the T2g term ^ more than ten times the spin-orbit coupling

57 9

constant. However, for small splittings of the T 2g term, the moment should vary markedly with temperature.

A study of the magnetic properties of six V(IV) c o m p l e x e s ^

illustrates well both mechanisms for lifting the degeneracy of the

2

T2g term, and provides a useful model for discussion of the

In document Estructuras de Datos y Algoritmos (página 123-126)