Chapter7 InterferenceofLight

LectureNotesforModernOpticsbasedon Pedrotti&Pedrotti&Pedrotti Instructor:NayerEradat Spring2009

4/12/2009

InterferenceofLight

Interferenceoflight Interferencephenomenon dependsonsuperpositionoftwo ormoreindividualwavesunder strictconditions. Constructiveinterferenceleads toenhancementoftheresulting amplitudewithrespecttothatof theconstituents. Destructiveinterferenceleadsto diminutionoftheresulting amplitude p withrespect p tothatof theconstituents. Interferenceproducesan alternating gspatial p p patternof fringes.
4/12/2009 InterferenceofLight 2

Twobeaminterference:vectortreatment
Interference of the plane waves of same frequency arriving at point P: E1 = E01 cos ( ks1 t + 1 ) E2 = E02 cos ( ks2 t + 2 ) The combined disturbance at point P is: E = E1 + E 2 Es are rapidly verying functions in time with frequencies of ~10 1015 Hz and average to zero very quickly quickly. Irradiance or radiant power density is a measure of time-average of the square of the field (detector reading) I = Ee W / cm 2 E

2 t

I = 0 c EiE = 0 c E1 + E 2 i E1 + E2 I = 0 c E1 iE1 + E2 iE2 + 2E1 iE2 I = I1 + I 2 + I12

)(

1 in practice t averaging time s for the eye >>T(period 10-15s) 30

1) if E1 and E2 are orthogonal no interference happens. cos = 0 2) if E1 and E2 are parallel interference has maximum i effect ff t cos = 1 3) if E1 and E 2 have angle the interference term is: I12 = 0 cE01 iE02 cos where h = k ( s2 s1 ) + 2 1 is i the th phase h difference diff between b t the th waves. For a purely monochromatic wave is independent of time (a constant phase relationship between two waves).
4/12/2009 InterferenceofLight 3

Twobeaminterference
Interference of the plane waves of same frequency arriving at point P: I = I1 + I 2 + 2 I1 I 2 cos 1 1 2 2 ; I1 = 0 cE02 ; I12 = 0 cE01 iE02 cos = 2 I1 I 2 cos where h = k ( s2 s1 ) + 2 1 I1 = 0 cE01 2 2 For mutually incoherent fields and low quality laser sources the phases are random functions of time that very on time scales much larger than the E field oscillations but shorter than the detector averaging time. The interference term in this case will be: 2 I1 I 2 cos ( k ( s2 s1 ) + 2 ( t ) 1 ( t ) ) = 0 for all incoherent sources and low quality lasers and I = I1 + I 2 Light from independent sources even if they are same kind of lasers do not interfere. For mutually coherent beams ( light from same lase split and recombined ) the phase difference at detector

2 ( t ) 1 ( t ) = 0 if the light paths are equal and = constant and the interference term is:
2 I1 I 2 cos ( k ( s2 s1 ) + 2 ( t ) 1 ( t ) ) = 2 I1 I 2 cos ( k ( s2 s1 ) ) Even if travel distance difference is nonzero 2 ( t ) 1 ( t + t ) 0 so long as t << tc The coherence time of a source is inversely proportional to the range of the frequency components that make up the electric field tc = 0 = 1 .

Coherence length is the distance that electric field travels in coherence time: Lc = lt = c 0 = ctc For mutually coherent sources we assume the length traveled from the sources is much shorter than the coherence length, then the irradiance of the combined fields is: I = I1 + I 2 + 2 I1 I 2 cos .
4/12/2009 InterferenceofLight 4

Twobeaminterference
I = I1 + I 2 + 2 I1 I 2 cos

is the total phase difference between the beam from the point of separation separation. It can be due to
a) path length difference b) phase shift at reflecting beam splitters c) differing indecies indecies of refraction in the separate paths depanding on the sign of the cos the interference will be constructive (+) or destructive ( ). A pattern of alternating maxima and minima (fringes) will form at the plane of observation due to varying at different locations of the observation screen.
2 I max = I1 + I 2 + 2 I1 I 2 = ( E01 + E02 ) 2 = + = I I I 2 I I E E ( ) 1 2 01 02 min 1 2 If I1 = I 2 = I 0 I max = 4 I 0 and I min = 0

Fring visibility is a measure of fringe contrast and is defined as: Fringe visibility = I max I min I max + I min

To ensure maximum visibility in experimental settings, tti we make k sure the th amplitudes lit d are same in both arms when they arrive at the screen.
4/12/2009 InterferenceofLight 5

Conservationofenergy
Rewriting the irradiance for the interference pattern of two equal amplitude components, we get: I = I 0 + I 0 + 2 I 02 cos = 2 I 0 (1 + cos ) Using 1 + cos = 2 cos 2 I = 4 I 0 cos 2

we get

2 The energy is not conserved at each point of the interference pattern I 2 I 0 but it is conserved over one spatial period I av = 2 I 0 . So inteference causes redistribution of the energy of sources over the fringe pattern. The frine analysis performed performed here was for plane harminc waves but the results are general and hold for spherical and cylindrical waves. The only difference is that these waves do not have constant amplitudes and irradiances as they propagate.

4/12/2009

InterferenceofLight

Youngs(1802)doubleslitexperiment
To demonstrate the interference pattern and fringes Young generated two coherent point sources by splitting the wavefront of a wave generated by a point source to two in-phase point sources. Today we uese a laser as the first source. We can assume that the amplitudes of the splitted waves are equal. Our goal is to develope an expression for the irradiance at an arbitrary point, P on a screen at a distance L from the sources. a is the separation of the sources. The irradiance at P is: I = 4 I 0 cos 2 ( / 2 ) where is the phase difference of the waves arriving at P In this geometry phase difference is caused by the optical path difference = s2 s1 a sin phase difference associated The p with is

= k ( s2 s1 ) = ( 2 / )
I = 4 I 0 cos 2 a sin I = 4 I 0 cos 2
4/12/2009 InterferenceofLight 7

Youngs(1802) doubleslitexperiment a sin

I = 4 I 0 cos 2

constructive interference a sin = = m , m = 0, 1,... destructive interference a sin = = ( m + 1/ 2 ) , m = 0, 1,... For y << L sin = tan y ay then I 4 I 0 cos 2 L L

ay m L bright fringe m y = , m = 0, 1,... = ,ma m x L a m + 1/ 2 ) L ( ay 1 = m + ym ,min = , dark fringe L 2 a m = 0, 0 1 1,... Fringe separtion (peak to peak):
a The minima are situated midway between the maxima. We can design experiments with this technique to measure the wavelength or width of the very tiny slits.
4/12/2009 InterferenceofLight 8

ymaxima = yminima = ym +1 ym =

is constant for all fringes.

Formationofthefringeswithanalysisofthecrestsandvalleysofthe sphericalwavesfromtwosources
minima One valley and one crest maxima Crests of both sources

4/12/2009

InterferenceofLight

Thebrightfringesurfacesfortwocoherentpointsource generatedby g yrotating gthep patternaroundthexaxis

4/12/2009

InterferenceofLight

10

DoubleslitexperimentswithvirtualsourcesI Lloyds y mirror

Theinterferencepatternisgeneratedbysuperpositionofthelightfromtheactual sourceSandthevirtualsourceSthatisimageoftheSonthemirrorMM

4/12/2009

InterferenceofLight

11

DoubleslitexperimentswithvirtualsourcesII Fresnels Fresnel smirror

Theinterferencepatternisgeneratedbysuperpositionofthelightfromthetwo virtualsourceS1 andS2thatareimagesoftheSonthemirrorM1 andM2

4/12/2009

InterferenceofLight

12

DoubleslitexperimentswithvirtualsourcesII Fresnels Fresnel sbiprisms

Theinterferencepatternisgeneratedbysuperpositionofthelightfromthetwo virtualsourceS1 andS2thatareformedbyrefractioninthetwohalvesofthe biprism. p

4/12/2009

InterferenceofLight

13

Interferenceindielectricfilms
Dielectric here means index of refraction is real and positive. No considerable absorption hapens at the wavelength of interest interest. To create two coherent sources from one we can use two techniques: 1) Amplitude division. For example by partial reflection at two different interfaces. 2) Wavefrot division. Fro example placing two slits in front of the waverfront. A transparent film bounded by parallel planes divides a beam of light into two parts.

( amplitude division ) some of the light is reflected from the first interface and some from
the second interface. Two reflected portions from first and second planes can be brought together by a lens to interfer at a point on a screen.

4/12/2009

InterferenceofLight

14

Thinfilminterference
The path difference is: = n f ( AB + BC ) n0 ( AD ) = n f ( AE + FC ) n0 ( AD ) + n f ( EB + BF )
= 0 it can be shown

Snell's law: n0 sin i = n f sin t AE = AG sin t = ( AC / 2 ) sin t 2 AE = AC sin t sin t n = AD 0 2 AE = AD AD = AC sin i sin i nf n0 AD = 2 AEn f = n f ( AE + FC ) n f ( AE + FC ) n0 ( AD ) =0 = n f ( EB + BF ) = 2n f EB = 2n f t cos t this can be related to the angle of inceidence through the Snell's law. For normal incidence = 2n f t. The phase difference corresponding to the path difference is: = k = 2 / . The net p phase difference has to take into account the p possible p phase change g by y the reflection. P optical path difference, r equivalent path difference arrising from the phase change. Constructive interference: P + r = m where m = 0,1, 2,... D t ti interference: Destructive i t f P + r = ( m + 1/ 2 ) where h m = 0,1, 0 1 2,... 2 What is the value of r for dirrerent types of interface ?
4/12/2009 InterferenceofLight 15

Reflectedwaveamplitude
Amplitude of the reflected electromagnetic wave at an interface n n Er = a b Ei na + nb Three cases are recognized: 1) If n a < nb ( external reflection ) then E r and E i have different signs so the reflected wave has half-cycle phase difernce with respect to the incident wave. 2) If n a > nb ( internal reflection ) then E r and E i have the same signs so the reflected wave and incident wave are in-phase. 3) If n a = nb then Er = 0 no reflection happens.

Er

na

nb

Ei

AmplitudeofreflectedEMwavefrom na nb aninterface f Er = Ei
na + nb

Conditionforconstructiveanddestructive i interference f by b athin hi film fil

No relative phase shift at interfaces : Constructive: r + p = 0 + 2n f t cos t = m 2n f t cos t = m 1 Destructive: r + p = 0 + 2n f t cos t = m + 2 m = 0, 1, 2, 3,... Half - cycle relative phase shift at interface : Constructiv i e: r + p =

1 + 2n f t cos t = m 2n f t cos t = m + 2 2

1 for normal incidence 2n f t = m + 2 1 + 2n f t cos t = m + 2n f t cos t = m 2 2 for normal incidence 2n f t = m Destructive: r + p = m = 0, 1, 2, 3,...

Antireflectingcoating
For antireflecting coatings we want the reflected wave to vanish. 1 if r + p = m + then the reflected wave is out of phase and desctructive interference will happen. 2 Usually n s > n f > na so both reflectons are external and no phase shift happens at the interfeces r = 0. To minimize the film thickness and loss we take m = 1 and for one layer antireflection coating this is the so-called quarter wavelength layer. 2 4 This works only for one wavelength. In order to make it for a brader spectrum designers use many layers. An antireflection coating is perfect if amplitudes of the waves coming from both surfaces are equal. To achieve that we need a special relationship between the indexes of the layers. The reflection coefficient 1 n2 / n1 of a surface is: r = the r at two interface have to be equal. 1 + n2 / n1 At air-film interface ra f = 1 + n f / n0 n f ns = = n2 r r ns a f f s f = ns n0 for n 0 = 1 n f = 1 ns / n f n0 n f air For maximum At film-substrate interface rf s = extinction 1 + ns / n f Example: for the yellow-green yellow green ( 550nm ) that is eye eye's s most sensitive range n f = 1.22 1 22 the clasest material to this is index is MgF2 with n = 1.38.
4/12/2009 InterferenceofLight 19

p = 2t =

t=

1 n f / n0

Dielectricmirrors,filters,lasermirrors We can also design complete reflectors with alternating high-low /4 layers that
create constructive interference for the reflected light ( here we have phase shift ) . Advantage of these mirrors is that they don't heat up (no absorption) and are perfect reflectors for a specific wavelength. wavelength Good for making laser resonators. resonators We can also design low-pass, high-pass, band-pass filters with very high precision. Optical thin film design is a branch of optics that deals with these applications.

4/12/2009

InterferenceofLight

20

Localizedandnonlocalizedfinger
Nonlocalizedfringes g formeverywhere y andwecanp placeascreenontheirp pathtosee them.Forexamplefringesfromadoubleslitareobservableateverydistancefromthe sources. Localizedfringes formataspecificlocationinspaceandthescreenhastobeplacedthere toview i them. h For F example l fringes fi from f athin hi film fil arelocalized l li dati infinity fi i and dweneed d a converginglenstoprojectthemonascreen.Thefringesinthepicture(ontherightside) arefringesofequalinclinationofHeidingerfringesformedbyparallelraysnotpossible pointsources. withp

4/12/2009

InterferenceofLight

21

Generatingnonlocalizedrealfringeswithdielectricfilmsusingpoint sources

4/12/2009

InterferenceofLight

22

Interferenceatawedge
Pathdifferencebetweentworaysis almostequalto2t Wave1haszerophaseshiftdueto reflectionfromtheglassairinterface. Wave2has1800 phaseshiftdueto reflectionfromairglassinterface. Thatiswhytheleftcornerappears dark.Lightpathdifferenceiszerobut phasedifferenceduetoreflectionis 1800orhalfacycle.
na nb Er = Ei na + nb

1 2

Fringesofequalthickness
For a wedge with verying thickness the path difference = 2n f t cost varies even if the angle of incidence is constant. For a fixed direction of incident light bright or dark fringes will be associated with a particulat thickness (fringes of equal thickness). Fringes can be viewd with the arrangement in the figure and are called Fizeu Fringes. At normal incidence cos = 1 and the light g path p difference is p = 2n f t bright m 2n f t + r = 1 + m dark 2
4/12/2009 InterferenceofLight 24

4/12/2009

InterferenceofLight

25

Newtonrings

4/12/2009

InterferenceofLight

26

Filmthicknessmeasurements

4/12/2009

InterferenceofLight

27

4/12/2009

InterferenceofLight

28

Stokesrelations
In reality y many y internal reflections happen pp in the films. The Stokes relations offer a formalism to track the rflection and transmssion coefficients for the electric fields arriving at an interface. Let E i = amplitude of the incident light, E t = amplitude of the tranmitted light, E r = amplitude of the reflected light We define the reflection coefficient r as: r = E Er and the transmission coefficient t as: t = t Ei Ei

E r = rEi At the interface Ei is devided th to parts E t = tEi According to the principle of ray reversibility both cases in fig a and b are valid. valid When they happen together (figure c) we mark the reverse senario by primed parameters. But b and c are physically E = r 2 + tt ' E tt ' = 1 r 2 i i equivalent. So 0 = ( r ' t + tr ) Ei r = r ' for angles of incidence related through the Snell's law. There is a phase difference between the rays incident from each side.

Stokes relations between the amplitude coefficients

4/12/2009

InterferenceofLight

29

Multiplebeaminterference

4/12/2009

InterferenceofLight

30

MultiplebeaminterferenceI
We now consider interference between a narrow beam of amplitude E 0 and angle of incidence

i and the multiple reflections from inside of a film of thickness t , index n f surrounded by air.
We want to find superposition of the reflected waves from the top of the plate. The h phase h difference diff between b the h successive i reflected fl d beams b is: i = k = k 2n f t cos t Incident beam: E 0 eit E1 = ( rE0 ) eit E2 = ( tt ' r ' E0 ) ei(t ) E3 = tt ' r '3 E0 ei(t 2 ) E4
5 i t 3 ) 0

( ) = ( tt ' r ' E ) e (

...EN = tt ' r '( 2 N +1) E0 e ( ER = EN = rE0 e

N =1

i t ( N 1) )

for N=2,3,...
i = E0 e r + tt ' e r ( 2 N 4) ei( N 2) N =2 it

this form does not hold for E1 that never passes through the film. film
it

+ tt ' E0 r
N =2

( 2 N 3)

i t ( N 1)

N =2

x N 2 = 1 + x + x 2 + ... =

1 where x = r '2 e i and x < 1 the series converges 1 x

InterferenceofLight 31

4/12/2009

MultiplebeaminterferenceII
tt ' r ' e i ER = E0 e r + using i the th Stokes St k relationships l ti hi 2 i 1 r ' e
it i 2 1 r re ER = E0 eit r 1 r 2 e i

I R = ER I R = ER

eit 1 e i =E r 1 r 2 e i
2 2 0 2 0 2

r 1 e i = E0 eit 1 r 2 e i e (1 e ) )
it i

1 r 2 ei

using 2cos =e + e Ii

E I and R = R 2 Ii E0

2r 2 (1 cos ) (1 cos ) = E 2r IR = 4 2 1 + r 4 2r 2 cos 1 r 2 r cos + 2r 2 (1 cos ) = Ii I R = Ii 1 + r 4 2r 2 cos

Using the conservation of energy equation I i = I R + IT we find IT = ET

2

1 + r 4 2r 2 cos 2r 2 (1 cos ) Ii 4 2 Ii = 1 + r 2r cos

2 1 r2 IT = 1 + r 4 2r 2 cos

I i

4/12/2009

InterferenceofLight

32

MultiplebeaminterferenceIII
2 2 1 r 2r (1 cos ) IR = 1 + r 4 2r 2 cos I i and IT = 1 + r 4 2r 2 cos I i A minimum for reflected irradiance occurs when (1 cos ) = 0 cos = 1 = 2 m 2

= m = 2n f t cos t condition for minimum reflectance

2 1 r2 cos =1 I IT = I i This also provides the condition for maximum transmission IT = i 4 2 1 + r 2r cos All the secondary and higher order reflections are in-phase with each other and they all have phase difference with the first reflection. So they all cancel out with the first reflection.

tt ' r ' E0 E2 = = 1 r 2 for interface of air and glass n1 = 1, n2 = 1.5, r 2 = 0.04 and 96% of the light is E1 rE0 cancelled in the first reflection so ignoring the higher order reflections is perfectly justified. When cos = 1 the reflection maximum occurs occurs. 1 1

= m + = m + = 2n f t cos t condition for maximum reflectance 2 2

4r 2 and in that case I R = 1+ r2 4/12/2009

I 2 i

1 r2 and IT = Ii 2 + 1 r
InterferenceofLight 33