Phase Sensitive Faraday rotation

Aysha Aftab, Rabiya Salman and Sabieh Anwar

Tuesday, February, 09, 2010

Can light propagating through a medium be inuenced by the application of an external magnetic eld? You have observed optical activity in chiral molecules in your freshmen lab. The present experiment extends these concepts to magnetically induced birefringence through the historically important Faraday Eect, which reveals the rich interplay between optics and magnetism.

KEYWORDS Polarization Birefringence Faraday rotation Verdet constant Phase-Sensitive Detection Jones Calculus Laser Helmholtz coil Resonance in RLC series circuit.

APPROXIMATE PERFORMANCE TIME 2 weeks.

PRE-REQUISITE EXPERIMENT: Basic measurements with the Lock-in amplier.

Objectives

In this experiment, we will, 1. shed some light on the underlying mechanism of magnetically induced birefringence, 2. demonstrate the advantages of phase sensitive detection (PSD), 3. understand the mathematical formalism for polarized light and its manipulation, 4. build or use sources of uniform magnetic elds and measure the eld strengths using a commercial magnetometer, 1

5. calculate numerical integrals, 6. build resonant RLC series circuit and understand the resonance phenomenon, 7. calculate the Verdet constant of terbium gallium garnet (TGG) and of a diamagnetic liquid.

References
[1] Eugene Hetch and A. R. Ganesan, Optics, 4th edition, Pearson Education, Inc, India, 2008, pp. 318-320. [2] Daryl. W. Preston and Eric. R. Dietz, The Art of Experimental Physics, John Wiley and Sons, 1999, pp. 355-362. [3] Frank L. Pedrotti and Peter Bandettini, Faraday rotation in the undergraduate advanced laboratory, Am. J. Phys. 58, 542-544 (1990). [4] Lock-in amplier, user Manual, Stanford Research System, SR 510, http://www.thinksrs.com. [5] Aloke Jain, Jaynat Kumar, Fumin zhou and Lian Li A simple experiment for determining Verdet constants using alternating current magnetic elds Am. J. Phys. 67, 714-717 (1999). [6] David Jiles, Introduction to Magnetism and Magnetic Materials, 2nd edition, CHAPMAN and HALL/CRC, 1998, pp 20-25. [7] JAN SMIT, Magnetic Properties of Materials, Inter-University Electronic Series, Vol 13, McGRAW-HILL, New York, p 197. [8] Frank J. Loeer, A Faraday rotation experiment for the undergraduate physics laboratory, Am. J. Phys. 51, 661-663 (1983). [9] K. Turvey, Determination of Verdet constant from combined ac and dc measurents, Am. J. Phys. 64, 1561-1567 (1993). [10] Eric W. Weisstein, Jacobi-Anger Expansion from MathWorldA Wolfram Web Resource. http://mathworld.wolfram.com/Jacobi-AngerExpansion.html.

Theoretical introduction

Q 1. What is polarization of light? Write down the equation for linear and circular polarization. Also, show that linearly polarized light can be written as a sum of left and right circular light [1]? 2

2.1

Magneto optical eect in transmission geometry

Michael Faraday observed the relationship between electromagnetism and light in 1845. Faradays observation gave birth to the eld of magneto optics: the interaction of optical radiation with magnetic media or the interaction of light with an optically inactive medium placed inside a magnetic eld.

2.1.1

Birefringence

Some substances are optically anisotropic i.e., their optical properties are direction dependent. An atom can be viewed as a positive charge surrounded by an electron shell with some binding forces (the dipole oscillator model).

For an anisotropic substance, the z + binding forces on the electron are anisotropic implying that the spring constant will be dierent in dierent directions: an electron displaced from its equilibrium position along one y direction will oscillate with a dier- Figure 1: Negatively charged shell bound ent frequency than another direction. to positive nucleus by pairs of spring havSince the electric eld associated with ing dierent stiness light drives the electrons of medium at its frequency, these electrons reradiate. The resulting secondary wavelets recombine and light propagates through a medium. The speed of the wave through the medium, is therefore, determined by the dierence in natural resonating frequency of electrons and the frequency of the applied electric eld. With anisotropy, the whole process becomes direction-dependent. Since the refractive index, (n = c/v) is a function of speed, the anisotropy results in dierent refractive indices along dierent directions. This so-called birefringence manifests itself the in rotation of the plane of polarization [1].

2.1.2

Faraday rotation

Chiral compounds exhibit rotation of linearly polarized light due to natural birefringence, but this birefringence can also be induced in otherwise optically inactive materials either by applying stress, magnetic or electric eld. The Faraday eect is magnetically induced birefringence. Linearly polarized monochromatic light while transmitting through an optically inactive material, under the inuence of an axial magnetic eld, is rotated by an angle as shown in Figure 2. The angle of rotation is given by, 3

E Direction of propagation

B E d

Figure 2: Faraday rotation, The plane of polarization of light is rotated under the action of an axial magnetic eld.

= V Bd,

(1)

provided the magnetic eld remains uniform throughout the length d of sample. For non uniform magnetic eld, is given by, d = V B(z)dz. (2)
0

The proportionality constant V is a characteristic of the material, called the Verdet constant and is a function of the wavelength of light, temperature and refractive index of the material. It is the rotation per unit path length per unit applied magnetic eld. In other words, it quanties the induced birefringence. In this experiment you will measure this induced birefringence.

2.1.3

Larmor precession of the electron cloud in an applied magnetic eld

We now try to posit some foundational arguments describing the underlying mechanism of Faraday rotation. Consider an electron, moving in a circle of radius r in a plane whose normal makes an angle with an applied magnetic eld B. Since an electron is negatively charged its angular momentum L and magnetic moment e are opposite to each other. The magnetic eld exerts a torque on the magnetic dipole e , = e B = e B sin . Q 2. Referring to Figure 3 what is the direction of the torque on the magnetic dipole? According to Newtons second law, an angular impulse produces a change in angular momentum, dt = dL. 4

Thus, the attached vector L rotates in anticlockwise direction. The resulting precession traced out by tip of the vector L is shown in Figure 3. The angle of rotation through which angular momentums projection along the applied eld, L , moves in time dt is,

plane of electron rotation

Figure 3: Precession of angular momentum vector about the direction of the applied magnetic eld.

d = dL/L = dt/L sin and the precessional or the Larmor angular velocity becomes, d e B sin e B = = = . dt L sin L sin L The magnetic moment of circular current is given by L = e = iA = i(r2 ), where, e , 2 whereas the angular momentum of electron is given by, i = L = rp L = mvr = mr2 . Substituting Eqs (4), (5), (6) into (3), we get, L = ( e r2 )( )B 2 mr2 eB = , 2m 5 (7) (8) (5) (3)

(4)

(6)

showing that the Larmor frequency L is independent of the orientation of the current loop and the overall eect is the rotation of electronic structure about the direction of applied magnetic eld [3].

2.1.4

Semi-Classical description of induced birefringence

You must have realized from Q 1 that plane polarized light is a combination of left and right circular (l and r) polarized light. Now, if light of vacuum frequency f is traveling through a medium whose electrons are rotating at the Larmor frequency then the l and r components will rotate the electron clouds with frequencies f + fL and f fL . Therefore in the dispersive medium, (refractive index is frequency dependent,) the functional dependence of the respective refractive indices can be written as, nl = n(f fl ) and nr = n(f + fl ), If plane polarized light traverses a distance d then the optical path lengths for l and r light are nl d and nr d respectively, so the optical path dierence is (nr nl )d . The dierence of two refractive indices, the induced birefringence is, nr nl = n(f + fL ) n(f fL ) Using the Taylor Series, nr nl = (n(f ) + = 2fL From Equation 8, fL = Eq (10) becomes, nr nl = 2( eB dn )( ). 4m df L eB = , 2 4m dn . df dn dn fL ) (n(f ) fL ) df df (9) (10)

Since, phase change of a wave is k (= 2/) times the physical path traversed by the wave, the phase change for the two components is, l = ( r nl d )(2) nr d = ( )(2). 6 (11) (12)

E l E E r

E l l

Er r

(a)

(b)

Figure 4: Superposition of left and right circularly polarized light into linearly polarized light. (a) Entering the sample, both the l and r components are moving with same speed and (b) while passing through the sample, these components have travelled with dierent velocities.

When l and r waves enter the sample, the phase dierence is zero, but the phase dierence accumulates as light passes through the sample. The vector sum of the two electric elds on emerging from the sample is shown as E with a net rotation from its initial value. Since, E is an equal superposition of l and r components, we see from Figure (4) that, l = r + l r = . 2 Thus, the Faraday rotation angle is, = 1 2d ( )(nl nr) 2 d eB dn = ( )( ) 2m df e dn ( )Bd. = 2m df

(13)

Comparing Eq(1) and (13), the Verdet constant, V = e dn ( ) 2m df (14)

which is a function of wavelength and the dispersion [3]. The Faraday rotation is a direct result of nl = nr arising because of the frequency dependent refractive index. 7

2.1.5

Description of dispersion based on the Zeeman Eect

The physical reason behind the change in refractive index can also be explained through Zeeman splitting. The splitting of spectral lines of the atom when placed in magnetic eld is called the Zeeman eect, after the name of discoverer P. Zeeman. From Eq (3), e = LL L eB e = ( )( )= L B B 2m 2m (15)

Since, the left and right (l and r) components of light carry an angular momentum of + and respectively, the l component drives electron into left circular motion and the r component drives electrons into right circularly motion, resulting in dierent magnetic moments. Interaction of the magnetic moment e with the magnetic eld B slightly shifts the energy of atomic level by an amount E = (e B) = (e )B = ( e )B. m (16)

Thus, under the application of an axial magnetic eld, dispersion curves for left and right circularly polarized light are identical but displaced by the frequency dierence between the two Zeeman components, = E = e B m

which results in two dierent refractive indices nr and nl and therefore a dierent speed, at a given [2].

= eB m

Figure 5: Refractive indices for left and right circularly polarized components of plane wave in the presence of magnetic eld. The dispersion curves for the two components are shifted by

2.2

Jones calculus

Jones calculus, invented by the American physicist R. Clark Jones, in 1941, is a useful formalism to understand the state of polarization of perfectly polarized light 8

as well as its transformation by various optical devices. For example, polarized light given by, E(z, t) = ox cos(kz t + x ) + oy cos(kz t + y ) iE jE is represented in the Jones formalism as, ( ) ( ) Ex (z, t) Eox eix t) = E(z, = ei(kzt) . Eoy eiy Ey (z, t) (17)

(18)

The two component column vector completely species the amplitude and phase of electric eld and hence its state of polarization. This is called the Jones vector. Most of the times, it is not necessary to know the exact phase but the phase dierence = y x between the x and y components. Moreover, ei(kzt) is always understood to be present. Accordingly, Jones vector can also be written as, ) ( Eox eix . E(z, t) = (19) Eoy ei Ignoring the term eix ( E(z, t) = Eox Eoy ei ) . (20)

For linearly polarized light = 0 or 180o , therefore the general form of Jones vector for linearly polarized light is, ( ) t) = Eox . E(z, (21) Eoy Jones vectors can be normalized such that the sum of the squares of their components is 1, i.e,
Eox Eox + Eoy Eoy = 1.

This normalized form discards the amplitude information needed for absorption calculations, but simplies analysis in many other cases. The normalized form of (21) at an angle w.r.t an arbitrary reference axis is, ( ) cos E(z, t) = , sin where, the angle is dened such that, cos = Eox 2 + Eoy Eoy 2 + Eoy

2 Eox

= Eox sin =
2 Eox

= Eoy . 9

Q 3. Write down the normalized Jones column vector for horizontally, vertically, left and right circularly polarized light? Suppose that the Jones vector for polarized incident beam Ei is represented by Et after transmission through an optical element then, the optical element can be represented as a 2 2 transformation matrix J, called the Jones matrix, given by Et = J Ei where J = ( ) j11 j12 . j21 j22 (22)

(23)

Equation 22 can be written as, )( ) ( ) ( Eix Etx j11 j12 . = j21 j22 Eiy Ety

(24)

If the beam passes through a series of optical elements represented by the matrices J1 , J2 , J3 , ..., Jn , then Et = Jn , ..., J3 , J2 , J1 Ei . (25)

The matrices do not commute, so they must be applied in proper order. Q 4. Show that the transformation matrix Jh for horizontal linear polarizer is ( Jh = ) 1 0 . 0 0 (26)

3
3.1

Experimental Technique
Why PSD in Faraday rotation?

You have already performed an introductory experiment of using the lock-in amplier, so without discussing the details of the technique and the instrumentation any further, we will only focus on why are we using phase sensitive detection (PSD) in this experiment. Consider a simple optical system used to measure the transmission of light through a medium. Let us suppose a small response obscured by overwhelming noise is to be measured. The output signal in this case will be, Vo = Vsig + Vnoise . (27)

The noise and signal amplitudes for such a system as a function of frequency are shown in Figure (7) [4]. The large peaks at 50 Hz and its multiples are due to electrical interference from the mains power lines. The noise power increases at lower frequencies (remember this is due to 1/f noise). Faraday rotation is extremely small in magnitude. If such a small signal buried in noise is to be 10

Display

Light source

Sample

Detector

Figure 6: A simple optical system.

Noise and signal amplitude

50

100

150

frequency (Hz)

Noise and signal amplitude

50

100

150

frequency (Hz)

Figure 7: (a) Noise and signal amplitude as a function of frequency. (b) Modulating the signal to a region of low noise.

measured, amplifying the signal will not improve the signal-to-noise ratio, the noise is amplied with the signal. A clever approach is to move the signal to a region of low noise, to higher frequency. For example, in the present experiment, we use an ac magnetic eld for inducing Faraday rotation instead of a dc eld produced by dc current or a permanent magnet. This technique gives two real advantages. The weak signal of interest buried in noise can be extracted successfully through PSD. Faraday rotation can be observed at smaller values of magnetic eld (e.g., 80 G rms). This circumvents the need for large, expensive, bulky, water-cooled electromagnets for producing large magnetic elds.

Q 5. Can you think of a simple experiment that measures the noise spectrum of laser light detected by a photodetector? Q 6. What is Maluss law? How does a polarizer work? 11

3.2

Overview of the experiment

The plane of polarization of linearly polarized monochromatic light traversing through the sample S of length d placed under the inuence of an ac magnetic eld is rotated. Since the eld is oscillatory, the rotation angle is also oscillatory. Another polarizer set at an arbitrary angle relative to input polarizer subsequent to the sample is required to analyze the rotation. The analyzer converts the polarization modulation to an amplitude modulation by the way of Maluss Law. The emerging light beam carrying the information in the form of amplitude variations is incident upon a photodiode whose output appears in the form of current proportional to the light intensity. Let us suppose incident light polarized along the x-axis is propagating in the z direction. The electric eld in terms of Jones vector is, ( ) o = 1 Ao exp i(kz wt) E (28) 0 where Ao corresponds to the amplitude of the electric eld. Suppose, the analyzer is set at an angle w.r.t the polarizer. Jones transformation matrix for the analyzer

Figure 8: The relative angle between the polarizer and the analyzer is . is, Jrot () =

) cos sin . sin cos

(29)

Q 7. Write down the equation for electric eld after transmission through analyzer and derive the matrix (29)?(HINT: Use Figure 8.) After passing through the sample S of length d placed in magnetic eld, the plane of polarization of light is rotated by an angle , so the Jones vector after emerging from the sample is, ( cos sin ) (30)

and the corresponding electric eld is, ( ) = cos Ao exp i(kz wt). E sin 12

(31)

Figure 9: Rotation of plane of polarization of light, analyzer is oriented at relative to the polarizer.

Q 8. Since, the analyzer is set at an angle w.r.t the polarizer, show that electric eld of the light beam after emerging from the sample followed by the analyzer is, ( ) = cos( ) cos Ao exp i(kz wt). E (32) cos( ) sin The intensity of light measured by the photodetector is, I = kA2 [cos2 ( )]. 0 (33)

Q 9. Derive the expression (33). What are the dimensions of the constant k? In the subsequent discussion, we will normalize k = 1. (HINT: Use the concept of the Poynting vector.) Q 10. Write the Jones transformation matrix for the combination of the polarizer, sample and analyzer, placed in the same order.

3.2.1

Optimization of the analyzer angle

According to Eq. 33, the rotation of the plane of polarization manifests as a change in intensity at the photodiode. To get maximum change in intensity, the analyzer angle needs to be optimized. Dierentiating the intensity w.r.t , we get, dI = A2 2 cos( ) sin( ) 0 d = A2 sin 2( ). 0 Dierentiating again, ( ) d2 I = 2A2 cos 2( ) . 0 d2 Maximum change in intensity is obtained by maximizing 2A2 cos 2( ) = 0 0 13
dI d

(34) (35)

(36) or by setting
d2 I d2

= 0,

since, A0 = 0, we have, cos 2( ) = 0 ( ) = 45o . Since, the Faraday rotation is much smaller than , maximum I is obtained when the analyzer is set at 45o relative to polarizer. The measured intensity is, A2 0 [1 + cos2( )] 2 A2 0 = [1 + cos(2) cos(2) + sin(2) sin(2)] 2 (1 + sin(2) sin(2)) = A2 . 0 2 For = 45o and sin(2) (2), = I = A2 0 I (1 + 2). = 2 The eld is made oscillatory, with an oscillating frequency , B = B0 sin(t), and since the angle of rotation is directly dependent on the magnetic eld, = 0 sin(t), therefore, Eq. (37) can be written as, ( ) A2 1 + 20 sin(t) 0 I . = 2 3.2.2 Converting light intensities into photocurrents (38) (37)

(39)

The photodiode converts the photon intensities into current, thus i = idc + iac where idc = 20 and iac = 0 A2 sin(t). Modulated photocurrent due to Faraday 0 rotation iac is measured through lock-in amplier which displays the rms values, therefore the output of the lock-in amplier is, 0 A2 iac iac = = 0 . 2 2 Taking the ratio of the modulated current (shown by the lock-in amplier) to the steady current, we obtain, 0 A2 2 iac = 0 2 idc 2 A0 iac 0 = 2 idc 14 (40) (41)
A2

Figure 10: Signal measured by photodiode is made up of two parts, average light intensity, idc and modulated intensity at the frequency of ac magnetic eld, iac . The currents are proportional to the intensities.

and as far as the rms value of the Faraday rotation angle is concerned, iac 0 rms = = = . 2 idc 2 (42)

The dc component is measured by an oscilloscope in the absence of magnetic eld while the ac component is measured by the lock in amplier in the presence of the magnetic eld. For a uniform magnetic eld Verdet constant is determined from the experimental values of , B and d, = V Bd, whereas for non uniform magnetic eld, , is given by, = V
0 d

(43)

B(z)dz.

(44)

Q 11. What is the working principle of a photodetector? What does the photodetector measure? The electric eld or the intensity? Q 12. Can the photodiode measurement be aected by stray magnetic eld? (HINT: the Hall Eect)

3.2.3

Schematic of the experiment

Figure 11 shows the schematic diagram of the experimental setup for the observation of Faraday rotation.

The setup comprises these components: 15

Table 1: List of Equipment used in the experiment. Supplier Component laser 633 nm, 2 mW Thorlabs (HRR-020) Light source laser 405 nm, 40 mW B&W TEK (405-40E) Linear Polarizer extinction ratio=1000 : 1 Thorlabs (LPVIS050) Signal generator, 10Vpp GW-Instek (SFG-1013) Magnetic eld production Audio amplier, 150 W CERWIN VEGA Power supply, 10 A, 12 V Panasonic electronics Helmholtz coil, 120 G Homemade rms Detection element Photodiode Newport (818-SL) Stanford Research SysLock in amplier tem (SR-510) Oscilloscope GW-Instek Gaussmeter with axial LakeShore (410) Measuring instruments and transverse probes Kyoritsu (KEW SNAP Clamp meter 2017) LCR meter QuadTech, Inc. Optical breadboard 90 Thorlabs (PBI51506) 60 6 cm Optical rail, 60 cm long Thorlabs (RLA600/M) Rail carrier 2.5 cm Thorlabs (RC1) SS Post, 5 cm long Homemade Post holder, 7.6 cm long Thorlabs (PH3/M) Rotation mount, 2 deThorlabs (RSPO5/M) gree resolution Accessories Laser post, 20 cm long Thorlabs (P200/M) V shaped laser housing Thorlabs (C1502/M) Glass cell, 6 cm long Homemade Crescent shaped cell Homemade holder Teon crystal holder Homemade Laser safety glasses Thorlabs (LG4, LG10 ) OD=4, OD=7 M6 and M4 screws Thorlabs TGG crystal d=1 cm Castech Inc.

16

Helmholtz coil Photodetector analyzer B

Sample Laser Laser polarizer A

Optical rail C Laser power supply Audio amplifier Lock in amplifier Dc supply Signal generator

Figure 11: Schematic of experimental setup for Faraday rotation. (a) Light Source (b) Mechanism for producing and measuring an oscillating magnetic eld (c) Detection devices

3.2.4

Light source

Light from a lamp can be used after collimating it by a lens and passing through a color lter to make it monochromatic, however, since, LASER is a source of highly directional and monochromatic light and is easily available, it is convenient to use it as the light source. HeNelaser of wavelength 633 nm or an electrically pumped diode laser of wavelength 405 nm will be used in the experiment. Q 13. What is the basic principle of a laser? How does a HeNelaser work?

3.2.5

Mechanism for producing and detecting the magnetic eld

In principle, both ac and dc magnetic eld can be used in this experiment. Dc sources include permanent magnets or solenoids having steady current in their windings. Since, Faraday rotation is small in magnitude, of the order of microradians, so a large dc magnetic eld, of several kilo gauss will be required to achieve a sizeable rotation, which in turn requires large and bulky dc magnets or a large dc power supply to produce required eld [8]. However, using an ac magnetic eld, the rotation becomes oscillatory and can be tracked by PSD. For example, in this experiment, you will be provided with a Helmholtz coil capable of generating a eld of approximately 120 G rms. 17

The Helmholtz coil A pair of Helmholtz coils is used to produce a uniform magnetic eld over a large volume of space. It consists of two identical coils such that separation d of the coils is equal to their common radius a.

Figure 12: Magnetic eld at point P due to single circular coil carrying a current i.

Let us consider a single loop of conductor of radius a, carrying a current i. Using the Biot-Savart rule, magnetic induction at the point P , at a distance r is, dB = o i dl u, 4r2 (45)

where, u is the unit vector connecting the conducting element with the point at which the eld is to be determined, o is permeability of free space = 4 107 Hm1 . This geometry is shown in Figure 12. Substituting, r = into Eq (45), we get, dB = o (sin2 )i dl u. 2 4a (47) a sin (46)

The axial component of magnetic induction is dBaxial = dB sin . Therefore, dBaxial = o (sin3 )i dl u 2 4a (48) dl = 2a , we

Since, dl is perpendicular to u, and integrating round the coil 18

obtain the total axial eld, Baxial = o i (sin3 )2a 4a2 o i 3 = sin 2a o i a3 = 2a (a2 + z 2 )3/2 o i a2 = , 2 (a2 + z 2 )3/2

(49)

where z is the horizontal distance from the coils center to a point P in space. For N number of turns, Eq. (49) becomes, Baxial = N o i a2 2 (a2 + z 2 )3/2 o N i = . 2a

(50)

For the Helmholtz pair, if one coil is placed at z = 0 and the other at z = a, and if current ows through both the coils in same direction (referred to as superposition condition, Figure (13)), by symmetry the radial component of magnetic eld along the axis must be zero. Hence, the magnetic eld on the common axis of the coils becomes [6],

x
B =

Figure 13: Pair of Helmholtz coil with separation equal to the common radius and carrying the current in same direction.

z2 (a z)2 3/2 o N i [(1 + 2 )3/2 + (1 + ) ]. 2a a a2

(51)

Q 14. Show that at the point on the axis midway between the coils z = a/2, the eld is , 4 o N i B = ( )3/2 ( ). 5 a 19 (52)

Q 15. Using the binomial expansion (1 + z)n = 1 + nz + n(n1) + . . ., show that 2! Eq. (51) can also be written as, B = o N i (1 + c4 z 4 + c6 z 6 + ...), 2a (53)

where, c4 = 15/8a4 and c6 = 105/48a6 . Q 16. What do you conclude from equation (53) about the uniformity of the magnetic eld? How does the eld in the middle of the coils compare with the eld in the center of a single circular loop of the same radius?
DC supply Function generator

Audio amplifier

Helmholtz coil

Gaussmeter

Figure 14: Instruments for creating and detecting the oscillating magnetic eld.

In our experiment, the Helmholtz coil is constructed from 18 gauge copper wire (diameter 1.2 mm). Each multilayer coil consists of 18 turns in 18 layers, the coils outer and inner diameters are 10.2 cm and 6.5 cm respectively. The length of each coil is 2.7 cm and radius is 4.5 cm. Inductance of the coils, determined using the LCR meter, is found to be 7 mH with a resistance of 1.5 for each coil, so the total inductance of the Helmholtz pair is 15 mH and the total resistance is 3 . The Helmholtz coil pair constitutes a series RLC circuit. At resonating frequency r , the inductive reactance XL is equal to the capacitive reactance XL and total impedance is purely resistive. The resonating frequency is, 1 r = LC or fr = 1 . 2 (LC) 20

Q 17. Calculate the resonating frequency when a capacitor of 0.97 F is connected in series with the coil? Why is the Helmholtz coil made resonating?

(a)

(b)

Figure 15: Orientation of Gaussmeters (a) axial and (b) transverse probe for eld measurement.

The Experiment
1. Assemble the setup according to Figure (11). Turn on the audio amplier and the function generator. Amplify an approximately 1 V, 100 Hz sinusoidal signal through audio amplier. Apply this amplied output to the Helmholtz coil. 2. Increase the frequency of the ac signal applied to the coil. Tabulate the frequency against current passing through the Helmholtz coil (Table 2) and plot the frequency response. The current is measured with the help of a clamp meter or an ammeter. Frequency (Hz) 100 . . . 2500 Current (A rms)

Table 2: Mapping the frequency response of the Helmholtz coil.

3. Set the function generator at the resonating frequency. Increase the current by increasing the gain of the amplier (Table 3) or by increasing the voltage of the signal generator. Measure the magnetic eld using the Gaussmeter in ac mode (LakeShore, Model 410) equipped with transverse probe at the midpoint between the two coil. Plot a graph between current and magnetic eld. Do you observe a linear relationship as predicted by equation (52)? 21

Current (A rms) 1 . . . 1.6

Magnetic eld (G rms)

Table 3: Linear relationship between the current (A) and the magnetic eld (Gauss).

Next you will determine the Verdet constant for Terbium Gallium Garnet(TGG) crystal and a diamagnetic liquid such as carbon disulphide CS2 . Never touch the lateral surfaces of the TGG crystal. 4. Turn on the provided laser. The HeNetakes about 30 mins to warm up and reach a stable value. Never look directly at the laser light. Always wear safety goggles when operating the laser. For the time being, keep the laser power switched on but close the shutter of the laser head. 5. The TGG crystal is 1 cm long but CS2 is lled in a 6 cm long glass cell. We cannot expect the magnetic eld in between the coils of Helmholtz pair to be uniform over this large a distance. So you need to map the magnetic eld prole and perform numerical integration of the magnetic eld as suggested in Equation 44. The next few steps will help you calibrate the magnetic eld. 6. Fix a scale with the edge of the sample holder. Place the glass cell over the crescent shaped sample holder. Open the laserhead shutter. Adjust the height of the laser and sample holder to pass the beam through the center of the coils. You might need to adjust the heights of optical components by translating the stainless steel posts. Close the shutter again. The height of the sample holder will now be kept xed. Mark the end points of the sample on the scale. 7. Remove the sample. Fix the axial probe on another holder on either side of the Helmholtz coil and turn on the Gauss meter. Select the 200 G range and ac mode. Switch on the audio amplier and tune the function generator to the resonating frequency, 1.22 kHz. Current is now passing through the Helmholtz coil. Select some value of current and measure the corresponding magnetic eld at the midpoint between the Helmholtz coils (A eld of 90 G rms is a reasonably good value). Move the probe away from the center of coils on both sides. Check that magnetic eld is not reaching the polarizers and photodiode. If required, adjust the distances by moving the rail carriers along the length of the optical rail. Map the magnetic eld prole by moving the probe along the length of the sample with a step size of 0.5 cm, for dierent values of current (similar to the observation tabulated in (4)). Remove the probe and turn o the magnetic eld. Tabulate the variation in magnetic eld along the length of the sample as in Table 4. 22

8. Estimate the numerically integrated magnetic eld over the length of the sample d,
k = k Bo (z) z k .

(54)

Table 4: Numerical integration of the magnetic eld. i (A rms) z (cm) . . . .1 . . . .2 .3 .4 .5 .6 .7 .8 .9 1 Bo (G rms) 1.1 1.2 1.3 1.4

(G cm)

You will now measure idc , the dc component of the detected signal, in the absence of magnetic eld, with the sample in place. Perform optical alignment if required. 9. Connect the photodiode to the oscilloscope, select the dc input mode. Remove the background reading from ambient light either by placing a black tube or by making the background level at the datum. Open the laserhead shutter. 10. Rotate the analyzer angle and nd out the maximum and minimum intensity. Set the analyzer at angle of 45o approximately w.r.t polarizer. Note down the value of voltage at the oscilloscope and divide it by 1 M, the input impedance of oscilloscope, to get idc . Next, you will determine iac , the (rms value of the) ac component of the photocurrent ( intensity), using the lock-in amplier. 11. Activate the magnetic eld. Provide the reference signal from the oscillator to the lock-in amplier. Select the reference frequency mode, f Hz in the reference channel of lock-in amplier. Select the current input mode I, from the input section. Make the oset equal to zero. Connect the photodiode to the input BNC connector. Make sure that no error indication (unlock or overload) occurs. Turn on the band pass and line frequency lter. The eect of each key press on the lock-in amplier is indicated by a nearby LED. Select a suitable sensitivity (usually 200 nA) and time constant (3 ms) for the pre lter. Check that no oset is introduced by the lock-in amplier at the selected sensitivity scale. 12. Adjust the phase located in reference section to make the output equal to zero. Then introduce a phase shift of 90o , bringing the reference and input signals in-phase. 23

13. Rotate the analyzer angle in steps of 10o and tabulate iac , for any xed value of the magnetic eld. You will observe that the maximum rotation occurs when analyzer is at an angle of approximately 45o relative to polarizer. Q 18. sent? What does the reading on the lock-in amplier physically repre-

14. Fix the analyzer at 45o relative to the polarizer. Increase the magnetic eld, from an initial value of 10 Gauss, in steps of 5 or 10 Gauss by increasing the current. The transverse probe of Gaussmeter can be xed to observe the magnetic eld in the center of Helmholtz coil. 15. Tabulate the values for iac for each value of current (and hence the corresponding magnetic eld) passing through the coil. Q 19. sample? Use the above results to calculate the Verdet constant of your

Q 20. Clearly quantify your uncertainties. What are the major sources of error? Q 21. Can you measure idc with the help of the lock-in amplier?

16. Remove the HeNelaser from setup, place AlGaN diode laser. Turn the diode laser on, its warm up time is 15 minutes and its output is linearly polarized, therefore, remove the polarizer A, use analyzer B only and repeat all the steps.

4.1
4.1.1

Applications of Faraday rotation

An optical isolator acts as a photon valve, passing radiation in one direction and blocking in the other. An isolator is shown in Figure 16. Polarizer A is used to make the beam horizontally polarized which is then passed through a 45 Faraday rotator C, followed by another linear polarizer(analyzer, B) at 45 relative to A. If any of the light is reected or backscattered from analyzer, it undergoes an extra rotation of 45 by C and thus is blocked by A. In a LASER, if any of the emitted light returns into the active medium through an unwanted reection, it can generate instabilities in the emission. Optical isolators are used to prevent the unwanted reection in lasers [7].

4.1.2

Domain Observation

Light will have dierent characteristics after reection or transmission by regions having dierent orientations of magnetic moments. Let a sample be made up of three domains, the magnetization of each domain is shown in Figure 17. Plane polarized light of wavelength , passing through domain A, is rotated through some angle 1 , while interacting with C, is rotated 1 . If the analyzer is at 1 , A may 24

0 45 A C
o

45 B

Figure 16: Optical isolator, backscattered radiation undergoes an additional 45 rotation by C , thus is blocked by polarizer A.

be dark, C bright and D of intermediate shade. For analyzer at 90 , D will be dark, A and C will be equally bright. For analyzer set at 1 , A will be bright, C dark and D of intermediate shade, i.e, polarization direction may be turned one way or the other, depending on magnetization, thus resulting in dierent intensities, it is therefore possible to image magnetic domains [7].

Figure 17: The domain imaging through magneto optic rotation.

4.1.3

Circulator

Optical circulators are used in ber optics, to separate light traveling in opposite directions. Figure 18 shows one such circulator. It is made up of two Foster Seely Prisms and a 45 rotator placed between the prisms. In these prisms the rejected polarized light is internally reected, so that it exits perpendicular to the axis of prism. Horizontally polarized light entering along a passes straight through the prism, is rotated to 45 by the rotator using Faraday rotation and emerges from second prism at b. However, any light reected back to the circulator entering at b with polarization azimuth 45 undergoes a 45 rotation through the rod, thus 25

polarized at 90 and exits from port c. Similarly, light entering at c emerges at d and entering at d exits at a [7]. A circulator has at least three ports. The

45 o rotator a oc 90 b d 135o

Figure 18: 45 rotator placed between two Foster-Seely prisms constitutes a circulator light beam if entering from one port after passing through circulator exits from the second. Another light beam entering from the second port or light reected from second port exits from third port and so on.

(OPTIONAL) Measurement of the the Verdet constant using higher harmonic components

The light rotated by the Faraday medium incident on the photodetector from analyzer, contains fundamental as well as components at higher frequencies. The rms values u1 and u2 , at f and 2f , respectively of these current components are measured, where f is the frequency of ac signal passing through Helmholtz coil. The ratios u1 /Uo and u2 /Uo can also be used to determine the Verdet constant, where, Uo is the steady output from the photodiode under zero magnetic eld and analyzer set for maximum transmittance [9]. The power transmitted through a Faraday rotator is, I = A2 o [1 + cos 2( )] 2 A2 o = [1 + cos 2( o cos(t))] 2 A2 o = [1 + cos 2 cos(2o cos t) + sin 2 sin(2o cos t)]. 2
m=1

(55)

Using the Jacobi-Anger expansion, we obtain[10], cos(2o cos t) = Jo (2o ) + 2 sin(2o cos t) = 2 where the Bessel function is, J (x) =
q=0 m=1

(1)m J2m (2o ) cos(2mt)

(1)m J2m+1 (2o ) cos((2m + 1)t)

x (1)q ( )2q+ q!(q + + 1) 2 26

and is the factorial function, given by,1 (n) = (n 1)! Therefore, Eq (56) becomes, [ ( ) A2 o m I = 1 + cos(2) Jo (2o ) + 2 (1) J2m (2o ) cos(2mt) 2 m=1 ( )] m + sin(2) 2 (1) J2m+1 (2o ) cos(2m + 1)t .
m=0

(56)

Let the amplitude of coecient of the terms containing t and 2t be represented by s1 and s2 respectively. Then, s1 = = = = = A2 o 2(1)0 J1 (2o )| sin(2)| 2 [ ]2q+1 (1)q 2o 2 Ao | sin(2)| q!(q + 1 + 1) 2 q=0 [ ] 0 1 3 (1)2 5 2 (1) Ao o + + + ... | sin(2)| 0!(2) 1!(3) o 2!(4) o [ ] 1 3 1 5 2 1 Ao o + + ... | sin(2)| 1! 1!2! o 2!3! o ] [ 1 4 1 2 2 Ao o 1 + + ... | sin(2)| 2 o 12 o A2 o 2J2 (2o )| cos(2)| 2 [ ]2q+2 2o (1)q 2 Ao | cos(2)| q!(q + 2 + 1) 2 q=0 [ ] 1 2 1 4 (1)2 6 2 Ao + + + ... | cos(2)| (3) o (4) o 2!(5) o [ ] 1 4 1 6 2 1 2 Ao + ... | cos(2)| 2! o 3! o 2!4! o [ ] A2 2 1 2 1 4 o 1 o + o ... | cos(2)| 2 o 3 24

(57)

(58)

s2 = = = = = Since,

(59)

o = V Bo d Substituting equation (60) in (58) ] [ 1 1 2 2 4 (V Bo d) + (V Bo d) ... | sin(2)| s1 = Ao V Bo d 1 + 2 12 [ ] 1 1 2 4 = Uo V Bo d 1 + (V Bo d) + (V Bo d) ... | sin(2)|. 2 12

(60)

(61)

is the conventional symbol to generalize the factorial function. Since, we are using for numerically integrated magnetic eld, therefore, we have used to denote the general form of factorial function.

27

where, Uo is the steady power on photodetector when polarizers are set for maximum transmittance (in the absence of applied magnetic eld). Substituting equation (60) into (59), we obtain, [ ] A2 1 1 o 2 2 4 s2 = (V Bo d) 1 (V Bo d) + (V Bo d) + ... | cos(2)| 2 3 24 [ ] 1 Uo 1 2 2 4 (V Bo d) 1 (V Bo d) + (V Bo d) + ... | cos(2)|. = 2 3 24

(62)

The f and 2f components are determined through lock-in amplier which displays rms values, so from equation (61), the rms value of the rst harmonic component of output current (ignoring higher order terms) is, u1 Uo V Bo d | sin(2)| 2 = Uo V Bd| sin(2)|

(63)

where B = Bo / 2, B represents the rms value of the eld measured by the Gaussmeter. Similarly, from (62) the rms value of the second harmonic component of output current is, u2 Uo (V Bo d)2 | cos(2)| 2 2 Uo = (V Bd)2 | cos(2)|. 2 (64) (65)

Both equations (63) and (65) can be used to determine Verdet constant. In short, we have three three dierent means of measuring the Farday rotation, Method 1. The gradient of the plot of u1 or iac against B for = 45o results in the Verdet constant. This is, in fact, the method you have used in previous section. Since, Uo = 2idc and u1 = iac , Equation (63) is actually Eq. (43) in disguise. Method 2. Determine the gradient of the least squares-t line to a plot of u1 /B against | sin 2| for xed Uo . Equate the gradient to V BdUo and nd the Verdet constant [9]. Method 3. Determine the gradient of a plot of u2 against B2 when = 90o . equate this to to V 2 d2 Uo / 2 and nd the Verdet constant. Q 22. Find the Verdet constant for TGG at 405 nm using methods 2 and 3.

28