I would use the experimental setup with the isotope, this gives us it’s ‘a’-value, which I’ll call a2. From the first experiment we can get Bj/J (I assume it’s the same for isotopes, but i am not sure about this). Now we have a measurement for a2, which can be expressed as a2=(mu.Bj)/(IJh^2). We can rearrange this to get mu out of this expression, since we know all other values.