You are on page 1of 23

Introduction to the Technique of Electron Spin Resonance (ESR) Spectroscopy

Physics Laboratory Course Dr. B. Simovi c April 20, 2004

Abstract Electron spin resonance (esr) spectroscopy deals with the interaction of electromagnetic radiation with the intrinsic magnetic moment of electrons arising from their spin. It has found applications in a wide range of different fields spanning from chemistry and biology to the novel and stimulating areas of quantum computation and single spin detection. This laboratory course provides a basic introduction to esr. It is intended to motivate students eager to learn about experimental aspects of spectroscopy.

Chapter 1 Introduction
1.1 What is ESR?

The phenomenon of electron spin resonance spectroscopy can be explained by considering the behavior of a free electron. According to quantum theory the electron has a spin which can be understood as an angular momentum leading to a magnetic moment. Consequently, the negative charge that the electron carries is also spinning and constitutes a circulating electric current. The circulating current induces a magnetic moment S which, if the electron is subjected to a steady magnetic eld H0 z, causes the electron to experience a torque tending to align the magnetic moment with the eld. The relation between the magnetic moment and the spin vector is s = gB S (1.1)

where B is the Bohr magneton1 and g is the Land factor.2 The energy e of the system depends upon the projection of the spin vector along H0 . Quantum theory stipulates that only two values are permitted for an electron Sz = /2, which means that the electron magnetic moment can only assume two projections onto the applied eld as shown on Fig.1.1 Consequently, 1 z = gB 2
1 2 e B = 2m = 9, 274101024 Joule.Tesla1 g = 2.0023 for a free electron



E + = gB/

E - = - gB/

e(a) (b)

Figure 1.1: (a) Schematic representation of a single electron spin in a steady magnetic eld H0 (b) Corresponding energy-level scheme.

and the ensemble of energy levels therefore reduce to the two values 1 E = gB H0 2 (1.3)

If electromagnetic radiation is applied at a frequency that corresponds to the separation between the permitted energies equal to E = E+ E = gB H0 = , energy is absorbed from the electromagnetic eld. This is the phenomenon of ESR.3 For electrons bound into an atom or a molecule, the phenomenon of ESR may not be observed at all, because electron spins pair o in atomic or molecular orbitals so that virtually no net spin magnetism is exhibited and the material is said to be diamagnetic. When an atom or a molecule has an odd number of electrons, however, complete pairing is clearly not possible and the material is said to be paramagnetic. In that case ESR can be observed. So far we have considered a single electron interacting with an external magnetic eld. In the present experiment, however, we deal with a macroscopic sample which means a statistical ensemble of magnetic moments. Therefore, we need to consider the relative populations of the energy levels N+ and N , which are given by the Boltzmann distribution:
The relation between the frequency of resonance and the amplitude H0 of the magnetic eld is : = 1.3996106 (g H0 )(Hz, Gauss)

Mn 2+ ion I=5/2 Iz
-5/2 -3/2 -1/2 1/2 3/2 5/2

Sz =1/2

5/2 3/2

Sz = -1/2

1/2 -1/2 -3/2 -5/2

Figure 1.2: Splitting of the ESR line in Mn2+ owing to hyperne interaction.

N+ E = exp( ) (1.4) N kB T where E = E+ E , kB Boltzmanns constant, and T the absolute temperature. Since it is the absorption, due to the slightly greater population of the lower level, that is observed, this dierence between the two populations should therefore be made as large as possible. At room temperature N+ N for a Zeeman splitting corresponding to a frequency of 10 GHz.


The hyperne interaction

Hyperne interaction is the interaction between the magnetic moment of an electron with the magnetic moment of the nucleus in its vicinity. Nuclei individually associated with the electron spin system often have a magnetic moment I which also has dierent allowed orientations (2I + 1) in H0 . The magnetic eld associated with the nuclear moment then can add to or subtract from the applied eld experienced by the electron spin system associated with it. In the bulk sample some electrons will therefore be subject to an 4




Figure 1.3: DDPH-Molecule(C6 H5 )2 N NC6 H2 (NO2 )3

increased eld and some to a reduce eld. Consequently, the original electron resonance line is split into (2I + 1) components. For example, when the electronic spin of a transition metal or a free radical4 interacts with its own nuclear spin the hyperne interaction is described by the Hamiltonian term Hhf s = AI.S (1.5)

with A the coupling constant. The hyperne coupling constant varies with the nuclear species, and it is a measure of the strength of the interaction between the nuclear and electronic spins. Fig.1.2 illustrates well the phenomenon: the hyperne interaction between the electronic spins5 and the nuclear spin I=5/2 in the Mn2+ ion splits the resonance line of the 3d electrons into six sub-levels. In molecules, the unpaired electron circulates between several atoms and the resulting hyperne structure is the result of a Hamiltonian term of the form Hhf s = Ai mi (1.6) where the projection mi of the ith nuclear spin on the magnetic eld direction may take on the following 2Ii + 1 values: Ii , Ii 1, Ii 2, ...., 1 Ii , Ii . For example, the hyperne interaction with the two equally coupled nitrogen nuclei (I=1) in DPPH molecule (see Fig.1.3) leads to a splitting of the resonance into ve components of respective intensity 1:2:3:2:1.
A free radical is an atom, molecule or ion containing one unpaired electron. By contrast, a transition ion can have several unpaired electrons. 5 The electronic conguration of the free ion Mn2+ is 3d5 .


The dipole-dipole interaction

For a large concentration of electronic spins, the electronic magnetic moments also interact appreciably with each other, and this can alter considerably the ESR spectra. The interaction is mediated by the dipolar eld associated with the magnetic moment S H(r) = 0 1 3(S .r)r (S + ) 3 4 r r2 (1.7)

Combining equation1.2 and 1.3, we see that the energy of dipole-dipole interaction between two adjacent electrons distant of r lays between Edd and Edd with 0 2 2 1 Edd = g B 3 (1.8) 8 r The dipolar interaction induces therefore a broadening of the resonance line, which increases with the concentration of dipole moments.

Chapter 2 ESR spectrometer

We need four essential components to build an ESR spectrometer: A monochromatic microwave source A waveguide for guiding the microwave power to the sample A cavity designed to ensure a proper coupling between the sample and the incoming wave. A detector for microwave power to detect the response of the sample to microwave irradiation. A schematic drawing of the ESR spectrometer is shown in Fig.2.1


Microwave parts

The dierent parts used in this experiment are listed below: 1. A gun oscillator is a monochromatic source of microwave. The fundamental frequency is here 10GHz. Tuning of the frequency is achieved by slowly turning the screw on the top of the metallic case of the oscillator. The frequency can be read out with the frequency counter located next to the source. 2. A calibrated attenuator is use to control the level of microwave power from the source. The scale is logarithmic: 7









Figure 2.1: ESR spectrometer

P0 P with P0 the output of the microwave source. xdB = 10Log 3. A T-hybrid is the 4-ports device sketched in Fig.2.2. A wave entering from the source at input 3 splits equally into two waves travelling to 1 and 2. The port 4 being orthogonal, no transmission from port 3 to port 4 is allowed. Also, no reection occurs at port 3 and 4 owing to the presence of the source and the detector.1 Waves are therefore reected only from ports 1 and 2. Let be the dierence between these two reected waves. If = 0 then the two recombine in port 3. If = , the two recombine in port 4. In this experiment, the relative phase and amplitude of inputs 1 and 2 can be controlled with an attenuator and a moveable short located on the right arm of the hybrid.
Reections occur because of impedance mismatch along the path of the wave. However. both the detector and the source are designed to keep the impedance equal to its free-space value at their entrance plane.

To Detector


To cavity

From source

Figure 2.2: Magic T.

4. The detector is a crystal rectier (diode) which consists of a semiconducting material. The incident microwave power causes the current to ow. The current I increases with the microwave power P and the sensitivity of the detection strongly depends on the slope dI/dP which is specic to each diode. 5. The waveguide is a rectangular opened-ended metallic tube delimiting a dielectric media in which electromagnetic waves propagate according to Maxwell equations. Boundary conditions have to be fullled by the electrical and magnetic components of the wave on the metallic walls. Consequently the propagation is restricted to a set of modes occurring at well-dened frequencies which are the characteristic values of the wave equation. There is a cut-o wavelength above which no propagation is allowed and which corresponds for a rectangular waveguide of width a, to c = a/2. 6. The cavity is a closed metallic box with an iris to allow the microwave to couple in and out (see Fig.2.1) Any cavity possesses resonant frequencies at which the energy stored reaches large values. These frequencies are related to the dimensions of the cavity. The quality factor Q of a cavity measures the frequency width of the resonance or equivalently 9

To moveable short


Quality factor

Total reflection



Figure 2.3: The mode of the cavity used in our ESR spectrometer can be monitored with a network-analyzer which measures the reected power versus frequency. The vertical scale is logarithmic. The sharp dip shows the absorption of part of the incident power by the cavity. The amplitude of the dip quanties the amount of absorbed power by the cavity. Note that both the amplitude of the dip and the frequency at which the mode occurs change when a sample is introduced in the cavity.

its selectivity. It is dened like Q = 0 Energy stored Energy loss (2.1)

Q-values in general are of the order of magnitude of the volume-tosurface ratio of the resonator, divided by the skin depth in the conductor at the frequency of resonance. Figure 2.3 shows the detailed characteristics of the rectangular cavity used in the present experiment.



Electronic equipments

1. An oscilloscope with its X and Y channels. 2. A plotter also with its X and Y channels. 3. The pre-amplifier amplies in two successive steps: a rst dc amplication and an ac amplication through a 115 Hz-bandpass lter. Beware of not saturating the dc stage of amplication. 4. The GUN power supply delivers 15V. 5. A Hall-probe measures the static magnetic eld. As indicated by its name, the principle of operation is based on the Hall-eect. The probe should therefore be positioned vertically relative to the magnetic eld in order to get the maximum sensitivity. Be careful when manipulating the probe. It is indeed fragile and expensive. 6. The gaussmeter converts the voltage measured from the Hall-probe into a value of magnetic eld (the maximum deviation corresponds to 1V). 7. The two sets of coils I and II generate the static magnetic eld H0 and the small modulating eld Hm respectively. 8. The magnet power supply supplies the current to the pair of coils I which produces the static eld. It is controlled by the ramp-generator. The current is extremely stable in order to avoid spurious noise that could interfere with the measurement. 9. The low frequency oscillator creates a sinusoidal current at the frequency of 115 Hz in the second pair of coil (II). This produces the eld Hm and provides the external reference signal for the lock-in amplier 10. The ramp-generator produces a ramp of magnetic eld by varying continuously the current in the pair of coils I. The voltage output of the ramp can be connected to the X-channel of the plotter if the variation 11

of static elds are too small to be accurately read from the gaussmeter. For larger amplitude of change, the analogue output of the gaussmeter can be converted to a digital signal and sent to the X-channel of the plotter. 11. The lock-in amplifier amplies signals at frequencies close to the frequency of a reference signal. More details concerning the principle of operation of a lock-in are given in the following section. Here we emphasize that the reference signal should come from the low frequency oscillator. Be careful to connect the 115Hz signal to the proper input of the lock-in amplier.


Chapter 3 Detection scheme

3.1 ESR signal

In a resonance experiment, the phenomenon of absorption of electromagnetic energy by the sample is generally translated into a variation of the complex impedance Z = R+jX of an oscillating circuit. When the sample experiences a time-varying eld, the absorptive component of the susceptibility to this eld changes the resistivity R while the dispersive components changes the reactance X.1 For a complex susceptibility sketch the variation through resonance of and and also R and X as a function of frequency. The sensitivity of a given detection scheme to absorption or dispersion can thus be dened as the change of the measured quantity - a voltage, a current or a power - arriving at the detector for given values of R and X. In the ESR spectrometer designed in Fig.2.1, we detect variations in the power reected by the cavity. In other words, we work with a reectioncavity spectrometer. Hence we need to consider the relationship between the impedance dened as the entrance plane of the cavity Zc and the reection coecient : Zc Z0 = (3.1) Zc + Z0
We recall here that the absorption (dispersion )is related to the imaginary (real) part of the complex susceptibility and changes in phase quadrature (in phase) with the timevarying eld.


Absorption aa Dispersion ba

Figure 3.1: This construction shows how the absorption and dispersion components of the signal are connected to the variations of the reection coecient . with Z0 the characteristic impedance of the wave. At resonance, = 0 for a cavity on tune (X=0) and matched with the waveguide (Zc = Z0 ). Far o from resonance the cavity is equivalent to a short, nothing goes through the iris, and consequently 1. Changes in the reection coecient are given by the vector = R + X (3.2) R X and the ESR signal (S) is therefore dened as the scalar product S = 0 . (3.3)

This means that the extent to which the absorption and dispersion components aect the modulus || depends on the phase of changes in expressed relative to . This can be readily seen in Fig.3.1. Based on Fig.3.1 and what has been said above explain how to detect the dispersive component (ba) of the signal.


Field modulation

When the magnetic eld is scanned through the region of resonance, the spin system in the resonant cavity absorbs a small amount of energy from 14

the microwave magnetic eld H1 , and produces a slight change in the resonant frequency of the microwave cavity. The DC detection of ESR is severely limited, however, by the drift of the amplier and the 1/f noise. For these reasons, most ESR spectrometers incorporate magnetic eld modulation which transfers the relevant signal from DC to AC. The principle is the following. When the magnetic eld is modulated at the angular frequency m , an alternating eld 1 Hm sin m t (3.4) 2 superimposed on the constant magnetic eld H0 + H . H is the local eld induced by the surrounding of the considered electron. It is therefore H which determines the broadening of the ESR line. The constant magnetic eld is normally swept over the range H0 from 1 (H0 2 H0 ) to (H0 + 1 H0 ) in a time t0 where H0 is the magnetic eld 2 strength at the center of the scan. At any time t during the scan, the instantaneous magnetic eld strength H is given by H = H0 + H + Hmod = H0 + H0 ( where t 1 1 ) + Hm sin m t t0 2 2 (3.5)

t 1 ) (3.6) t0 2 Consequently, the signal at the input of the detector is sinusoidal with the frequency m and with an amplitude proportional to the derivative s/H. Note that in order to consider the eld (H0 +H ) as constant, the scan should be slow enough so that there are many cycles of the modulation frequency during the passage between the peak-to-peak (or half amplitude) points of the resonance line. H = H0 ( Sketch the eect of eld modulation on the ESR line.


Principe of a phase-sensitive detection

A lock-in detector compares the ESR signal from the detector with a reference signal and only passes the components of the former that have the proper frequency and phase. If the reference voltage comes from the same oscillator 15

that produces the eld modulation voltage, the ESR signal passes through while noise is suppressed. Thus a lock-in detector only accepts signals that lock to the reference signal. Hence the name of phase-sensitive detector. The operation of a lock-in detector is simple. A reference oscillator produces a reference signal vr vr = Vr cos(t + ) (3.7) where is a phase angle. At resonance the sample absorbs microwave energy and produces the ESR signal voltage es es = Es cos(t + s ) (3.8)

where s is another phase angle that may be close to . The multiplier produces an output that is the product es vr of the ESR signal and modulated voltages 1 es vr = Es Vr cos(t + s ) cos(t + ) = Es Vr [cos(2t + + s ) + cos( s )] 2 (3.9) The low pass lter removes the rst term to produce the dc output voltage Vout 1 Vout = Es Vr cos( s ) (3.10) 2 and we see immediately that the sensitivity is maximal if is set equal to s , which gives 1 Vout = 2 Es Vr (3.11) When the electron spin resonance signal is measured by the lock-in amplier it still contains a considerable amount of noise. A large part of this noise may be removed by passing the signal through a low-pass lter. The lter has associated with it a time constant or response time 0 , which is a measure of the cuto frequency of the lter. Another way to say this is that the lter fails to pass frequencies that are much greater than the inverse of the time constant 1/0 ; it attenuates, distorts, and retards those incoming waveforms that have frequencies in the vicinity of 0 , and it transmits undisturbed those frequencies considerably below 1/0 . The waveform that is impressed on the ESR response lter may be considered as the derivative of the absorption (or dispersion) line, and for comparison with the above 16

criteria, its eective frequency may be taken as the inverse of the time that it takes to scan through the resonance from one peak to the next. In other words, if the time that it takes to scan through the magnetic eld range is very short compared to the time constant 0 , then no signal will appear on the recorder; if this time equals the time constant, then a distorted signal will result; while if one waits many time constants to complete the scan, then the recorder will faithfully reproduce the true lineshape.


Chapter 4 Experiment
The following questions are of course suggestions and should not substitute the initiative of the experimentalist. The cabling of the spectrometer is described in Fig.4.1. 1. Read the manual and by doing so try to answer each of the exercises marked with the symbol 2. Measure the magnetic eld as a function of the current in the magnet for large and small, slow and fast eld changes. Observe the hysteresis. 3. Measure the amplitude of the modulation of the eld as a function of the modulating current. 4. Measure the ESR signal on a powder sample of DPPH in absorption and dispersion. Estimate the linewidth and the g-factor. 5. Measure the ESR signal of DPPH in solution for dierent concentration. Estimate the dipole-dipole coupling constant. From that estimate the range of concentration where the dipolar interaction becomes stronger than the hyperne interaction and compare with experiments. 6. Repeat 6 and 7 for Mn2+ in solution. 7. Observe and discuss the eect of varying the amplitude of the modulating eld on dierent ESR lines.










DC or AC



y Scope x





Figure 4.1: Schematic representation of the cabling of the spectrometer. Note that both DC and AC detection schemes are represented. The dashed lines show the path of the AC signal.




All the books listed below are in principle available at the library of the physics department at ETH. 1. H. A. Atwater, Introduction to microwave theory, McGraw-Hill (1962) 2. C. P. Poole.Jr, Electron Spin Resonance, Interscience Publishers (1967) 3. A. Abragam and B. Bleaney, Electron Paramagnetic Resonance of Transition Ions, Clarendon Press . Oxford (1970) 4. D. J. E. Ingram, Radio and Microwave Spectroscopy, Butterworth & Compagny (1976) 5. C. P. Poole.Jr. and H. A. Farach, Theory of Magnetic Resonance, 2nd edition, Interscience Publishers (1987) 6. C. P. Slichter, Principle of Magnetic Resonance, 3rd edition, Springer Verlag (1996)


1 Introduction 1.1 What is ESR? . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 The hyperne interaction . . . . . . . . . . . . . . . . . . . . . 1.3 The dipole-dipole interaction . . . . . . . . . . . . . . . . . . . 2 2 4 6

2 ESR spectrometer 7 2.1 Microwave parts . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.2 Electronic equipments . . . . . . . . . . . . . . . . . . . . . . 11 3 Detection scheme 13 3.1 ESR signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3.2 Field modulation . . . . . . . . . . . . . . . . . . . . . . . . . 14 3.3 Principe of a phase-sensitive detection . . . . . . . . . . . . . 15 4 Experiment 18 4.1 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20


List of Figures
1.1 1.2 1.3 2.1 2.2 2.3 (a) Schematic representation of a single electron spin in a steady magnetic eld H0 (b) Corresponding energy-level scheme. Splitting of the ESR line in Mn2+ owing to hyperne interaction. DDPH-Molecule(C6 H5 )2 N NC6 H2 (NO2 )3 . . . . . . . . . . 3 4 5

ESR spectrometer . . . . . . . . . . . . . . . . . . . . . . . . . 8 Magic T. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 The mode of the cavity used in our ESR spectrometer can be monitored with a network-analyzer which measures the reected power versus frequency. The vertical scale is logarithmic. The sharp dip shows the absorption of part of the incident power by the cavity. The amplitude of the dip quanties the amount of absorbed power by the cavity. Note that both the amplitude of the dip and the frequency at which the mode occurs change when a sample is introduced in the cavity. . . . 10 This construction shows how the absorption and dispersion components of the signal are connected to the variations of the reection coecient . . . . . . . . . . . . . . . . . . . . 14 Schematic representation of the cabling of the spectrometer. Note that both DC and AC detection schemes are represented. The dashed lines show the path of the AC signal. . . . . . . . 19