31 May 2016

magnetism by Evans method

In-Class Activity

Submitted by Adam R. Johnson, Harvey Mudd College

After I teach my students about magnetism and magnetic properties in coordination compounds, I spend a day showing how the data is collected and analyzed. I teach them about the Gouy balance, the Evans method of determining magnetism by NMR, and SQUID magnetometry. I also show them real data that I collected as an undergraduate or graduate student, and have them interpret and analyze it.

The only experiment that we can do locally is the Evans method, so I spend more time on this technique. We use the method during the metal acac laboratory.

This learning object includes a short powerpoint presentation outlining the basics of the method, some real data collected by students at Harvey Mudd College, and some links to primary literature in the field to help guide your teaching of the material.

A related learning object goes through the history of the field.

Learning Goals: 

students will understand the basics governing the shift of the NMR signal in a magnetic environment
students, given a data set, will be able to determine the number of unpaired electrons in a coordination complex

Equipment needs: 

none. If you want to collect your own data, then you will need an NMR spectrometer, a paramagnetic compound, and appropriate glassware.

Implementation Notes: 

I do this as a short lecture followed by problems done in small groups in class. If the students don’t finish the problems, they do them as homework.

The powerpoint has some notes to help with the presentation

There are 5 student datasets and a brief explanation

there is a faculty-only file which is an excel spreadsheet which does the calculations. The spreadsheet has the three-term, two-term and one-term versions of the equation, showing that the 2nd and 3rd terms can safely be ignored in most cases; the largest sources of error are the measurment of the mass and volume of the solution!

Time Required: 
one 50 minute class period
Evaluation Methods: 

I usually ask a magnetism question on a homework and then again on my midterm or final exam. I give the students the equations they need. Importantly, for me, magnetic properties are used to help the students predict something about structure. For example, if a Ni(II) 4-coordinate 16-electron complex has unpaired electrons, that suggests that it is tetrahedral rather than square planar.

Evaluation Results: 

This is a fairly straightforward calculation, so the students generally do not have problems with the math (unless they forget to multiply the field strength of the NMR magnet by 1,000,000.

Creative Commons License: 
Creative Commons Licence


Hi Adam, did you use the internal glass capillary to obtain the NMR spectra? Do you also need to use a very long relaxation time or just a single scan spectrum to get the spectra?


I ran them with a glass capillary of CDCl3, and it seems as if the spectra are too broad and amorphous to see anything of note. I can't see both the solvent peaks I was expecting to see.

Yes, we do it with the internal glass capillary and it usually works very well. The signals for the acac ligands are usually broadened into the baseline but we can see residual CHCl3 and TMS. Send me an email if you want and I can try to help troubleshoot.


I ended up adding iPrOH as a secondary standard. It was too hard to identify the CHCl3 peaks because they were so small. This ended up working just fine though I'll probably add tBuOH next time. Thanks!

I do this experiment with Cr(acac)3, Fe(acac)3, and Mn(acac)3 and I use a glass capillary in a 5 mm NMR tube. I have an insert and capillary set from New Era and this has served me very well.

I usually dissolve the acac complex (between 5 and 10 mg) in 0.6 mL of a CDCl3/CHCl3 mixture (1:1). This fills the interstitial space once the capillary and insert are placed in the tube. I also fill the capliiary with the same mixture of CDCl3/CHCl3.

I then run 8 scans as I acquire my spectra. I have no trouble seeing any of the CHCl3 peaks in the spectrum and have never had to add another standard (iPrOH or t-BuOH).  I would more than happily share my spectra with you, although they are pretty silimar to the ones that Adam posted.

Oh, and I forgot to mention that I use CDCl3 that does not have TMS in it. I have found that the TMS peak is so large that it swamps out the signals that I want to see. 

Thanks Anthony. I didn't realize that the residual solvent peak in my NMR solvent would not be intense enough to see in the spectra Upon adding iPrOH at 10% by mass, I was able to see this peak and make the necessary measurements. It was broad and had no coupling, so I initially had trouble identifying it conclusively in some situations.

Your 50/50 CHCl3 mixture is also a good option to ensure a strong proton signal that is obvious to identify. I'll probably go with that next year.

The VIPEr community supports respectful and voluntary sharing. Click here for a description of our default Creative Commons license.