cMODULE 159-2(ESR)
J.V. Acrivos, SJSU

ESR.2:

I. Given:

(a) System of particles possessing angular momentum in units of .

(b)Spin Angular Momentum == S;

orbital angular momentum == L;

nuclear spin angular momentum == Ii,

(c) The objective of the lecture is to determine the shape of the esr absorption line versus externally applied magnetic field Bz.

II.

The background knowledge for the lecture is the classical Larmor precession frequency solved in freshman physics is then:

wL = ge Bz0, (1.6')

where Bz0 is the field at the center of the absorption line and the finite width due to the lifetime of states given by the Uncertainty Principle:

, (1.7)

leading to a half width at half height (HWHH) of:

, (1.8)

where T2 is called the spin state lifetime; it is determined by the surrounding molecules as well as the chemical reactions that the species carrying an unpaired spin undergoes.
The intensity of the transition is determined by the area under the absorption curve, it depends on the magnitude of the moment square and on the difference in population between the states which in turn depends on the energy separation hv and the absolute temperature T:

N+/N- = exp(-hv/kBT), (1.9)

where the signal intensity is proportional to the fractional excess population:

(N+ - N- )/Ntotal = (if T >> hv/kB) ~ hv/kBT, (1.10)

The magnetization M = k Mz + j My + i Mx, in an external flux field

B = kBz + iBx

obeys the kinetics:

dMz/dt = -(Mz-M0)/T1; dMx/dt = -Mx/T2; dMy/dt = -My/T2, (1.11)

where T1, T2 are called the spin - lattice and spin - spin relaxation times respectively. The solution obtains the Bloch Equations:

, (1.12)

, (1.12')

which gives the susceptibility:

At resonance the imaginary part of the susceptibility (called the absorption) is at a maximum and:

|M|/M0 , (1.13)

The spin system absorbs power from the radiation field:

P = = dW/dt = <Hx dMx/dt + Hy dMy/dt> = w X" H12. (1.14)

Tha absorption of power raises the spin temperature and equilibrium is reached when dW/dt is equal to the rate at which the spin system transfers energy to the lattice containing the spin system.

At exactly resonance w=wL and for a = , the magnitude of component X"H1 of the magnetization in the equatorial plane (relative to z) is:

X" H1/M0 = - a H1/(1 + (a H1)2 T1/T2), (1.15)

which passes through a maximum when |a H1| = (T2/T1)1/2. The absorbed power at resonance is:

P = dW/dt = M0H/T1 ((M0-Mz)/M0), (1.16)

where the quantity s == (M0-Mz)/M0 is called the saturation parameter

III.

ESR detection measures the derivative of the power P absorption with respect to the external field. For a Lorentzian line shape we obtain:

L' = dPL/dBz = d(X"H1)/dBz =Ntotal 2S S(S+1) hv/kBT a *

(a)2 B1 (Bz-Bz0)/[1 + (a (Bz-Bz0))2 +(a)2T1 /T2 B12]2 . (2.1)

It is convenient to define a== and x = (Bz-Bz0) then for T1 = T2 the normalized line shape is:

L' ==L'(B1, x)/ {Ntotal 2 S S(S+1) hv/kBT a} =

(a B1) (a x)/ [1 + (a x)2 + (a B1)2]2 , (2.1')

and the fields of maximum/minimum slope then determine a, i.e.,

dL"/dx = 0 as (a B1,0) 0, for (a xms) = 1/31/2 (1 + (a B1)2)1/2.

Here we note that the normalized relation relation (2.1') evaluated at xms depends on (a B1).

L'(B1, xms) == 16/33/2 a B1 /[1 + (a B1)2]3/2

In order to compare L' for different values of T2, note that the line shape depends very strongly on the relaxation times, e.g., when T1=T2 doubles the last term in the denominator reduces the intensity of L' in the wings (Figure 1.1) but increases in absolute value more than twofold near xms. When the product (a B1) increases, L'ms== L'(xms) gives significant values only in the linear region of Figure 1.2.

Figure 1. 1: Typical Lorentzian ESR absorption derivative.

Figure 1. 2: Saturation of typical Lorentzian ESR absorption derivative.

The results in Figure 1.2 indicate that B1 has to be reduced until the linear region in Figure 1.2 produces the best results. L' in Figure 1.3 shows the kind of problems that could be encountered when a B1 increases above the linear region.

Figure 1. 3: Reduction of ESR absorption derivative by B1 when T1 = T2 =10-6 s. Note that xms(saturated) ~ B1/31/2 allows for the callibration of B1 directly.

A Gaussian ESR absorption line shape extends to the wings:

G' = dPG/dBz= Constant'/T * a/(T2<> 1/2)

a2/(T22<> ) B1(Bz-Bz0) exp(-a2(Bz-Bz0)2/(2T22<>)), (2.2)

dn(G')/dBz = [1 - a2/(T22<> )(Bz-Bz0))2]/(Bz-Bz0),

The value of T2 is determined from the line widths. <> is the second moment of the esr absorption line. The values for the spin-lattice relaxation times are determined from saturation experiments. The results depend on the line shapes. In the Curie-Weiss temperature regime the absorption can be measured relative to a standard which does not saturate easily, i.e.,

a B12 T1,reference/T2,reference <<1.

An absorption ratio of sample to reference is defined as (a B1,0)2T1/T2 0 for both the reference and the sample for Lorentzian line shapes:

1/r0 = = (L'ms xms2) reference/(L'ms xms2) sample)=

(Tsample /Treference)* (Ntotal, reference/ Ntotal, sample)*

((Sreference(Sreference+1)/ (Ssample (Ssample+1)), (2.3)

Ntotal, sample is the only unknown to be determined from r0 in (2.3) in the absence of saturation. As the amplitude B1,0 increases above a limiting value, the ratio r0 decreases because the sample may saturate, i.e.,

r = =r0(a B1) /r0(a B1,0)/(B1/B1,0) =

(1+(a B1,0) 2T1/T2)3/2/(1+(a B1)2T1/T2)3/2, (2.3')

so that as (a B1,0)2T1/T2 approaches zero, the value of T1T2 is obtained with good accuracy for:

1/r = (1+(a B1)2T1 /T2)3/2

= 10 to 1.33

with errors r/r 5% and B1/B1,0 to better than 0.1% obtain:

(T1 /T2)1/2 = (1/r2/3-1)1/2/(aB1)

with an absolute error of

(T1 T2)1/2(r/3r)/(1 - r2/3)

(T1 T2)1/2depends on the chemical dynamics.

IV.

Gorter related T1 to the heat capacity and heat capacity and heat transfer coefficient at constant field H, CH andH:

T1 = CH/H. (2.4)

The sample temperature is TS and the lattice's is T0 at field H and TST0 at the rate:

dTS/dt = (T0-TS)/T1, (2.5)

when |T0-TS|/T0<<1.

Phase transitions at a critical temperature, say Tc, may be identified by the changes in both C and and consequently T1:

T1(Tc)/T1 CH/CH -H/H.

T1 NMR measurements by Slichter et al., use it for nuclei in hyperfine contact with the Bose condensate. For cuprates:

CH(Tc)/CH

and it appears that:

H(T<Tc)/ H CH(Tc)/CH

So that by esr T1 relaxation measurements one can find Tc of phase transitions.

Return to Acrivos' Front Page