There are multiple sources of diamagnetic correction depending on how you prepare and mount your sample. For typical SQUID samples you need to correct for the gelatin sample holder (some groups use Teflon buckets or quartz tubes), the straw any other packing material as well as the sample diamagnetism. For very small samples the diamagnetism of the straw and capsule is important. but for larger samples the sample diamagnetism dominates.
Unfortunately the parameters for the straw and capsule tend to change slightly due to inhomogeneity in the straw and capsule composition and weight, but talk to an experienced co-worker in your lab and they may be able to advise on the typical diamagnetic correction for straws and capsules in your lab.
The sample diamagnetism can typically be *estimated* from Pascals; constants for the constituent elements/ions. A recent set of tabulated data can be found in J Chem Ed:
http://www.e-fizika.si/magnetism/DiamagneticCorrectionsandPascalsConstants.pdf
One alternative approach to adjust for sample diamagnetism works on the following basis; the diamagnetism is temperature independent whereas the paramagnetism is temperature dependent (Curie-Weiss like). The sample susceptibility is given by
chi(obs) = chi(p) + chi(d)
So chi(p) = chi(obs) - chi(d)
If the sample follows Curie-Weiss behaviour then a plot of 1/chi(p) vs T should give a straight line, i.e. 1/[chi(obs)-chi(d)] should give a straight line.
Thus you can adjust chi(d) to give the best fit to a straight line in the high temperature region. Take care that the value of chi(d) you derive is of a similar magnitude to what you expect based on Pascals constants.
In terms of practicality if you are using Excel and you think your sample is Curie-Weiss like then estimate by eye the high temperature region in which C-W behaviour holds. Make an initial estimate of the diamagnetic correction based on Pascals constants (see above).
Plot 1/[chi(obs)-chi(d)] vs T in this region, add a linear trend line to your graph and ensure it shows the R^2 parameter which is the goodness of fit (1.000 for a perfect fit), then adjust the diamagnetic correction by small increments until you get an R^2 value close to 1.000.
Ensure your chi(d) value is not too dissimilar (same order of magnitude) to that initially guessed.
You should fit linear line in the high temperature region (higher than Tc) only. There should be sufficient data points in that region.
The Curie-Weiss law describes the temperature dependency of the magnetic susceptibility of paramagnets. The inverse susceptibility of a paramagnet should show linear increase with temperature. If this is not the case, i.e. for materials showing cooperative effects (ferromagnets, antiferromagnets ...) you can try fitting the linear part of the dataset as Mr. Sher pointed out.
If that does not lead to feasible results or you do not observe a linear dependency you can first try to apply a diamagnetic correction of your data via diamagnetic increments based on the compound you measured.
If your data still shows nonlinear behaviour you can try to use a modified Curie-Weiss law were a temperature independent correction is introduced that allows you to fit nonlinear curves.
The Curie-Weiss law *means* that there is a linear dependence between 1/chi and T... so fitting a non-linear curve is nonsensical and should not be considered.
There are a variety of reasons why compounds do not obey Curie-Weiss behaviour, though they normally deviate at low temperature so, as pointed out above, you can normally fit the high temperature region. For molecular species deviation from C-W is normally due to a depopulation of excited states. For example, Co(II) in octahedral symmetry has a 4T ground term (S = 3/2, L(eff) = 1). Spin-orbit coupling gives rise to J(eff) states in the range J(eff) = +5/2, +3/2 and +1.2 (|L+S|...|L-S|) separated by lambda.J(eff) where lambda is the spin-orbit coupling constant. Depopulation of the excited J(eff) states at low temperatures (< 70K or so) leads to deviation from Curie-Weiss law. Additional information on your system would be helpful to help provide more insight.
If there are no other excited magnetic states, the inverse susceptibility shoud follow exact linear dependence of temperature. Of course, diamagnetic contributions have to be properly subtracted. Linearity of the inverse susceptibility means that there is a functional change on approaching the critical susceptibilitry. This is a typical crossover in the sense of RG theory.
Thank u very much for your answer. Sir, how to subtract the diamagnetic contribution from the curve.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science