CHAPTER 5 Diffraction of Light (Fraunhofer Diffraction)CHAPTER 5 Diffraction of Light (Fraunhofer Diffraction) In the middle of seventeenth century, Francesco Gri ‘when sunlight entered a darkened room through a ays ras Screen, the light spot on the opposite wall was large expected from ray of also observed th; a : than wou image was not sharp but was surrounded by coloured fre border op attributed this to the wave nature of fringes. gant ie, to xt waves spread as they travel and hence when. Pte characte | spreading-out as it passes through a: i of light. gum atid a “ is the edges of obec ich deflected the corpuscles eet meters ih ower once Thomas Young esablched ne ge of seconc wavelet renee oF vielen fom ios pi nt ber — aperture. Fresnel, thus applied Huygens oe wvelets in conjunction with the ore plain the intensity distribu 7 is obstructed by an aperture of observable interference fringes. eo if 3s phy ysically no difference betw Tre best we cur afew sources: SOP WO, often use: a diffraction is usually classified and analyzed eet paper me waves arrivin in can ure and scree atapemrded as plane be fein the diffraction es Known as Fraunhofer class of In laboratory, ‘uation is realized ing the source at Tee focal plane of a convex fens and also keeping the fereen at focal plane of gnother convex ens, aS 5.1a, On the the source of nt and/or screen are at fini aperture, the diffraction effect fed as Fresnel class of diffraction. In this case, one has to consider the curvature of both incoming and/or outgoing wavefronts; see Fig. by some simple system of aperture(s). I system ture(s). In Ch. 6, we shall analyze Fresnel class of diffraction effect qualitatively by introducing the concept ofhalt iod zones. i rasce sieges — ofa mare rigorous and general theory of eugieaeen of Fourier transforms is briefly described ing but the interference pattern between Troi a large (but finite) number of point n processes Pro quote Richard Feynman: “No one has fra gorence between interference dnd diffraction ne “io, roughly speaking, is to say that when ‘imerfering. then the result is usually ao faa large number of them. it seems that the in ight and the observation screen,Frounhofer Diffraction waves. In the following le slits called Fraunhofer diffrac their source; Fraunhofer diffraction ‘geometrical optics. 5.1 FRAUNHOFER DIFFRACTION FROM A Sing Let us fist consider Fraunhofer diffraction from a sing) rectangular aperture whose length is plane wavefront is incident the slit spread along ‘According to Huygen’s p the slit becomes a source of secondary of the slit-width represents a continuous array (line) of Huygen’s rom each of these point sources interfere to produce the (a) final n pattern on the element (of width dy) of the s| fer, in 1823, developed a theory to expl “faitraction based on the assumption that a waves inva he Peay pattems due toa single narrow si ciple, every part of the wave, , secondary ith a se Ned ini ions, we d escrity © Fr ey in fact LE Sup much I; fr ‘wavelets. Hence the ent width represents a point source. ‘Suppose one such element 4 at D resents te amplitude per unit width of the sit Element represen ie of secondary harmonic disturbances which is dE dE," 45.2) 1d by source element A are observed at point path between A and P is r, then the wave from A is given by B by (ra-+ 8) Fos the path distance from center 1 P and A= ysin0 is the extra path travelled by . (& sports? o5.3) ave replaced r ike source at ntral rere lies another element B (of wi th dy) ig of O. From geometrical symmetry, we note <) Orfrom B to P is (,~ A). Hence, wave disturbance a P vom element B is given BY =(f “Ub from each part of the to elements 4 and Bis p+ dE yy GD n of point P along distance y from the centre O of i width d)) is written as wl}cima ofthe single-sit diffraction pattern do not fal exactly where B= (B72) sin®. The resultant field at p natitaveen minima. The condition for maxima is obtained harmonic wave of amplitude (#, sinB/B), which, ny Gifterence B~ AS. where 5 = (8/2) sind is the ends Upon ‘aves from the center O and either of the end potretlifterenge ihe ‘The intensity at P is proportional to square of the aes the stig 2 AMPlitude. . oy Me: B or BF where J, includes £3 and the constant of proporti oS, .condary : Portion ") 0, = 0, so that sinB/B = |. Hence, 1, represents recay NO tha, vjetermined BY ° Mt ntensiy ae wie curves y= Band jum of intensity J, = os Jovan on the same Braph S19 Fe 0 occurs when SinB = op 3 printersetion orn for so, | Peles ive the value of Br B= k(6/2)sin® = max tanB= B (see Fig. 5.3). nat intersections m=+ 1,49 =i 7 dt “ Ie we denote the Psvordinate of point P by Yo, then appre, 09 te e 143m, 246m Fig. 53 sneer wees fal Iength ofthe son) ae mel | 21 itenslyDistbstion axis, such that ‘ances ¥;, on both sides ae ae’ intensity of fist order maxima at B = 1.431 is given by Eq. 5.7: Del Fan 1B = 1438) = 1 (Sg), = 0049 /, - Yat oh . where J, = KB-=0) is the peak intensity of principal (zero order) maximum. 459) weereig that peak intensity of first-secondary maxima is only 4.9% of Tatof the principal maximum. Similarly, we find that peak intensity of aretd order secondary-maxima (at = 2.46n) is only about 1.7% of J,- thin i The diffraction pattern is symmetrical about the p Maxima. The principal maximum occurs at Y,~ 0; the wit principal maximum may be regarded as the distance bere snot first-ord nce between the “Thus, most ofthe light intensity is confin cipal maximum i. two | tah essentially represents the image of thf'slit. However, the edges of ” weishage are not sharp; a series of secondary maxima and minima blur = 2fh ay the image. This is due to diffraction which inevitably occurs because of I are Se, or aon At = Ze @n) terference of secondary waves among themselves. Thee 49 denotes the angular width (or spread) of the principal 5.1.2 Angular Spread } reduced. te that principal maximum will widen as width & of the We have argued that principal maximum of the diffraction pattern essentially constitutes the image of the slit. Though angular width of the ‘The peak intensity of principal m: image A0 = 21V6 is independent of the distance Z between the apertare J, Up = 1, at = 0).and src, the Hiner width would increase With 25 shy ing, es in space. The is a consequence of diffrac nk true even if there is justa device which progia Ml ofaimension. The wave nature of ight completely exehues of rectilinear propagation of light. rey * The above expression (S.11) for angular divergence is yah slaperure. However, similar spread occurs fr other ype Mh too, viz. circular apertures, etc. ans Example 5.1 A parallel beam of light of wavelength 5000 A is incidentonay slit of width 0.4 mm. Fraunhofer diffract patter is obser screen with the help of convex lens of focal length 30 cm a ut bi slit ai (i) Whatis the width of the central maximum? Calculate the distance between the tw first order on both sides of the central maximum. Solution seconday man (A) The width of central maximum is defined as the di ‘wo first order minima; itis given by Eq, 5.10: lance bt = 2fh _ 203% 5000 10-0 I LO =O mm ‘The first order secondary maxima occur at @ such that K(b/2) sin® = 1.43. B= Ladin © 13D ° : sri. Solving for O, We get e bad 43 50002107 «178.107 in =" Od x 10" scimation sin0 = 0 for 0 to be very small. Hence, can use the APPT two first order secondary maxima is Wh width DENS a 1.78 «10? rad nce between them on the sereen «1.78 x 10-3 * 30 cm 20x fn? 1.06 mm. given by tern due to a fong narrow slit is observed at a ‘. Determine the width of the slit which wou maximum of 30°, 90°, and 180° we find p= 2%5690x6 5 ag =30° = 1/6 rad, 5 = 2.17 pm 0-909 =n/2 rad, = B= 0.72 wm Ao = 180° =m rad, b= 0.36 pm ‘Thus, as lit becomes narrower, there is more angular spread of the beam. in width & < 0.36 im, the spreading occurs over the entire field of 5.1.3 Diffraction from a Rectangular Aperture 1g sections we considered diffraction from a slit having a b much smaller than its length. The fringe pattern therefore is uniform along the length of the slit (X-axis) while intensity variation occurs along Yanis. A rectangular aperture is a slit of width 6 and comparable length /. In this case, diffraction spreading occurs along X-a asitoccurs along Y-axis, A det integration over both Vand Y di distribution is product of di similar manner led study (see Sec, $2.3) which involves wensions shows the resultant intensity tion along X and ¥ directions: ab =A 2}sn0, 4 aperture, it where 2. for the slit in d !) : ular symmetry about optic axis, we and a=h{z)sin®, 1e screen (kept at far distance from the Note that B is same as obtained before consider i ae Ed ee og boa ‘analogous manner, & corresponds to length LAE? tng fraction pattern is Known as the iry pattern. postion of coordinates (sy) on the sereen: Angles, ana xe fsinO,, y= sind, Sa where fs focal length of the lens. The wo-dimens tives zero intensity at points (x,y) where either sional pay ny ‘in, = =f ‘| airy bsind, =m y= peacm: or Isin@, ==> x= ER where m, 1 are non-zero integers. on Screen Fig. 5.6 inciple, find the fraction pattern 16) ar aperture in yas we did for However, when the | ircular, the source 4 7 dimensi fi ‘Soros, henge elementary source point cad Aperture isdefined by two coordinates (x, jpoe(r.0). Correspondingly, we ry ve to perform double : | From Eq, 5.10, we note that sp cee proportional to dimer ‘puede eton ofthe lnger erivation would be attempted in See. frac ; s given by an equation similar to Eq. 5.7. ¥ te ony iferenee stead of sine function (sinB), we now get another funetion Jy(B). is known as first-order Bessel function.but with deere 0, we have is ity occur at those points where zeros of (p) are given by (excluding P= 0): ue B = 3.832, 5.136, 7.016, ... ‘Thus, the first minimum occurs when 2B _ 3832 _ 1220 iD" "xD ~~ D ind = 0; hence, the angular spread ofthe cental may img Orn sind = 30=20=1224 a 60) where a is radius of the aperture, (Note that th ; . spread is symm ¢ axis). The central spot, also called the iy dive eps the effective image of the aperture. Detailed calculations show ttre 84% of the intensity is confined within central maximum, Example 5.3 Suppose 2 laser beam of wavelength 6328 A is used to illuminge te surface of moon. The diameter of the b staf Th eam atthe source outs Sa reading occurs only due to diffraction, how lar a wou iid be produced on the surface of moon? The isan ‘surface is 370,000 km; any scatter due to atmospher 135 ted, the angular spread only diffraction jven by Eq. 5.13 h 6328 * tor") 1225 12 25x10 - «107 5 ane by we find c <; therefore, cose. term terms 6080 a sn sinB term. The product ofthese two terms means varies more 89/0 scilating interference pattem on which a slow'y tat we ton pater i superimposed as shown in Fig. 5. Th oscil bution therefore represents a diffraction modulated i e pattern. inert eof intensity occur at all those points where interference term vanishes, ie., where. ting core =0-34sin0= (3h (6.16) Besides above, intensity occurs also at the points where diffraction term vanishes, i.¢., where 1 = 0, 1, 2y om sinB=O=>bsind=mk = m=1,2,3, 5.17) [Note that = 0 corresponds to diffraction maximum as (sinB/B) = | When diffraction minima coincide with maxima of interference pattern, those bright fringes of interference pattern are missing from the final pattern, ‘The interference maxima occur when i . (5.18)138 Hence, at the diffraction minima coincide with or Thus, in case slit separation dis an integral mul above cont orn = 42, #4, position © where both Eqs. 5.17 a tiple of si exactly satisfied, For instance, age on with - are the orders of missing interference 2b then nape Example 5.4 ee A doubles central diffraction peak. What is the slit s aes Miia centr Width soy Solution The double slit diffraction pattern is symmeti ; is symmetrical about 0 = both central fraction maximum and central Conecsasp e020 maximum coincide. eee Suppose first order diffraction from Eq. 5.17, where b is sli According to Ei ©, such that minimum occurs at angle 0: we find (m= 1): ‘ence, 4. 5.18, mth order interference maxima occur at ange sing, = mh, fringes c ineide with first diffraction minimum (on both is 1.5 mm ration is 1-5 mm. . i separ of interference maxima Wi 1) if al is an integer. in first diffraction peak ample find the ratio of widths of central diffraction peak to ference fringe. ak is the distance between two first ima lie symmetrically at angle 6 to diffraction ma. These yok sind'=O'= 5 sus, the angular width of central diffraction peak is ‘The y= 2A 20-5 sinilrly, width of contra nterference maximum isthe stance between ma given by Eq. 5.16. These minima also lie jes to 8 = O axis, where rder mi two first or {gmmerically at angle 0" (say) on both sind"=>> (n= 0) Hence, angular width of central interference maximum is » 205 near width on observation screen is proportional to angular he required ratio is 2d 2x15 _ 6 03 Example 5.6 How many interference fringes are contained in the central diffraction peak if} = 0.2 mm and d = 0.9 mm? Find the sp: fnges if they are viewed at a distance 1 m fic is $500 A. Also determine the relative inten: 1g between interference he slits. Wavelength of y of first three fringe:@ Gi) ) Solution ‘he first diffraction minimum occurs at angle g x © elven, sino'= by positions of interference maxima are given by 0; =a te ng sin, ="7-= Tay =qgsino Hence, for = 4, 0, <6" and n= 5, 0,> ; Th nterference maxima lie within central faction YD ogy are 2 x 4 + 1 = 9 bright fringes within the cen Thu Central peak, 8, ‘Angular position of interference maxima is gi ver dsin8 =m BY Ba. 5 a Hence cos0d® = “dn For consecutive maxima, d= |; corresponding spacing between bright fringes: gives ‘hear, woe bowk dcos0 Ata distance L from the slits, spacing between frn oe 8 bean, dy= Ldo= 4 For L= 1m, 4= $500 A, and d= 0.9 mm, we get 55010710 _ a= A597 = 061 mm The intensity of diffraction pattern is given by Eg, 5.5, J(®) = 41, ( i Central (n= 0 4) = Oand § = 0. Hence, we get eile 10=0)=41, ist order interference maxima oceu i isto ee maxima occurs when d sind, = 2. wl B= Fbsin0 = % jght-h 141 nd order interference maxima occurs when dsin®, Cores a = 2m, and f= (2/4.5). Hence, we get sine /48)* - mo==4 (232Fag°) = 199% Ss FRAUNHOFER DIFFRACTION a ‘our analysis to Praunhofer diffrction pattern produced we extent parallel slits, each of width 6, the distance between two mo 5 being @. its tue that a plane Hight rm nates all the Nslits. ly ill : ny waves from al sai paw a by a converging lens a are then fore “he resultant amplitude of jautern on the screen is. the toh of waves from each of the rposit"pall calculate it by choosing we ts: Mative method discussed in last 33. See. jet wave amplitude, at point P, due vie9 soright coming out from sTit tis Eq. 3-6: sia) at B sshere 7) =P isthe optical path between midpoint S, of slit | and P. a sseribed before. “ amplitude due to 2,at P, is siren in0 is the optical path difference between beams emergi ikewise, we can write the nemetric series, we get we ‘summing the geo" Ep= (z sl Berm a sinf) (singh ‘2)) onan B singe 73) ) Sin “The above expression, in fact, FEPrESENIS a plane esupg, i nt ={( p, sinB)(sin Ne’ Wa Ep ={(# GE SAS) where 0 = KA/2 = (d/2)'sinO and 7, =(%) —(V— 1) “6, 7, therefore, becomes )A/2). Thy a \ ay iy wo tB) ( ‘Thus, the intensity distribution is again a prodi Bry st term (sinfv/B)? is due to diffraction produced tye tom: Indi slit, and the second term (sinNo/sinay arises due to int V equally spaced point sources (viz. S,, Sy, .., S). erference pail ‘Note that Eq. 5.22 reduces to single slit diffraction pat 2 tor N= I, and to two-slit pattern (Eq. 5.15) for N= 2, Pattern (Eg 5 wns of Maxima and Minima described by (sinf¥/BY? acts as an envelop on the fringe patary Palen by interference term (sin No/sina)?. Let us first look at eee route pattern. 1c Merferenee We find that as c.—yn, the interference terms leads to a maxi vaxima: an (SinNa) _ |. ( Neosva)? tine) “an S) (52) These are called the principal maxima of the resultant intensity paten, That is, principal maxima occur when xdsind i a= MNO wim, ordsinO=nk = 0, £1, 42. (52 The above condition imy iat waves from adja inp ion implies thi 's from adjacent slits are and hence waves from all the slits mutually augment each other. and hence waves i other. Ta ie i, ——= ree) Principat od principal maxima. The amplitud nsity is proportional to a) (Principal Maxima) ...(5.25) or intensity, of different principal maxima is not .¢ factor (sinB/B)? which is different at different fof central maxima (i.e. principal maxima of zere rat (sinB/B) = 1. 8 P= NM interference pattern oceur when sinNa. = 0, but sine a= Pr a =f (5.26) = 0, No 2Ny soo, We Get = and hence since = 0; ng for principal maxima. Thus, between two xima (say p = 0 and p = N), we obtain (V~ 1) (N— 1), In between two minima, intensity must = 2s rgace, we have (N'~ 2) secondary maxima between cipal maxima. | Mr gecondary maxima is much smaller compared to that (Moreover, their intensities fall offas we move away structive to note that there are no secondary ne for N= 3, two for N= 4, and so on. As N’ we have S S Wen hence that contained within all reas Ss (LIN®). Consequently, for large N; secondary maxima are gecbat not visible and are of PO practical 5.4.2.Missing ‘lar principal max‘ Order of Principal Maxima yumm of order» may be absent ifat its position \rtby Eq. 3-24, we find that diffraction term vanishes, Ponds, ifboth the following conditions are simultaneously satisfied. dsind = 7 bsinO = and then /= Oat the position where nth order pr occurred. Thus, the missing order 1 of pr condition (see Eq. 5.19): n=(Spr mo ipal maxima should have given by the5.4.3 Width of Principal Maxima we seen above that princip: Forp= sninima correspond 10 p = a ons of these adjacent minima by (9, » N Since 40, is small, we put (53) sin(®, £48,) sind, + 40, cos@,, ‘40, being equal to nearly half 1 ‘a, is taken as the width of pri dsin(Q,£80,) =r = he which makes use of the diffraction wi formed consists of multiple, equidistant slits on an otherwi seen wise However, one takes a transparent sheet on whieh say th a diamond point operated with i the het econ 3d mechanism, The grooves act as opaque ea, 1e number of slits created on a good quality grating may be as ner ,000 per em, and the size of the gratis RS it wave of wavelength 2, falls normall ns of the principal maxima formed on the paler dsin,, =m 4s wl ed grating equation, is used to determine the izangle 8, of nth of order maxi lement ‘constant. If the source of reg according 10 2.. However, zero-order central wavelengths. Higher order max ea for ly separated by the grating, Thus by of diffraction for various colours, we ean determine js polychromatic, ree is POE) different wavelengths col dsinQ= mi = "22 i ihe spectral ange (ue. ange of wavelengths) such that 1 ind OF petween maxima of orders n and (n+ I) ‘Suppose 0 Over rresponding shortest and longest ‘wavelengths. Note that th core at which nth order maxima i formed is due to largest dsin®, = "2 sier hand, the minimum angle at which (w+ 1th order maxima jg formed is due 10 shortest wavelength 2): ® dsinOye = (1+ DA 1p of nth and (n+ Ith order maxima if, 28,5 ing case for which Aachen “The value (1, ,) is called free spectral range ofa gratin; upon order 72 ‘and hence is smaller at higher orders. 5.5.2 Dispersion by a Grating ies angular separation of different wavelengths; by ig the grating equation 5.29, we define the angular dispersion of grating as, or ®, 0 cA dos, 26.31) Combining Eqs. 5.29 and 5.31, we find Dyn _ tan, a deosd, (5.32) aaah man ilent OF pra However fferen ete innger angular separation at higher angles of diffi Foal IF, is smatl (so that cos, = 1), we find that 7 ay! 1 “ (a/an3 in thie case. the spectrum iS called normal spt" "ig Petry separation a between two wavelengths is Proportions hee ‘between wavelengths. toding mel 5.5.3 Resolving Power of Grating % Resolving power ofa grating refers to its ability to prod (of maxima) of two closely spaced wavelengths, Ice dig isaemtp Hence, the two wavelengths can be resolved if | a +O) = he nQ+ AR) = rh or Redon ‘Thus, the resolving power of grating dey its V ruled on the grating. Further, ts (534) pends linearly on the m resolving power is higher ae Example 5.7 (2) A transmission grating has 5000 Tines/em, Hi orders can be seen when itis illuminated by (@) Show that second and third order spectra of whi overlap ina diffraction grating, (©) How many lines must a grating have if ther ‘order spectrum observed? ‘many spend d= to 2 Correspondingly, the limits o ‘aisbles wand vare obtained from Eq. S2.28: E aye for s'=0 ywe have vj = a0 ‘eéifraction a straight edge is given by iE‘The field amplitude at a Point P (0, therefore, becomes (4, = 0): z Mler-C0) +45¢6)-soy), Ey) = HE (C0) ~ C42) 4 ¢5(00) = S(-20)] je ]= fete Son the edge of geometical shady 10,y)=40 Gy cae his is the sa 7 me intensh Tesult as we ‘obstructed wavettoy Bot in Sec, 6.10, Ea. 6.21. 1 ig the arbitrary p nt, n (x,y), We have At any other y=, So that field amplitude is EQ,y) =- Ee aitl(os ~ Clu)) +1(05~ seu) Hence, intensity J Tear) = ~2f (05cm)? (05-50 ++(82.32) merical sags Yalu of Cu) and Su) frm numerical values of Fresnel integrals oscil changes). oselineas “There is an alternative way to estimate intensity ve reading the values of Fresnel integrals on agrapphical representation called Corns spiral. ‘$2.9 METHOD OF CORNU'’S SPIRAL If the values of Fresnel integral C(t) and S(t) are plotted aga variable + on the same graph, the resulting graph is called Cornu sp phasor diagram as shown in Fig. 2.5. Note that values of C(«) are read along X-axis while that of S(z) are read along Y-axis. The m along the curve denote the values of variable +. We interpi 1d S(x) correspond to real and imaginary CC) a ral represents the value of the integra. yan Wt on the SP! cla + SO) + jn Fig. S2.5 correspond to values x = co and J Feo) = 405); origin O denotes the value + = Fig. 82.5 ihe feld amplitudes in Fresnel diffraction theory contain terms ow, the i ras (Con) - CC) + (S¢42) - Se) field intensity is given by square modulus of amplitude; therefore, ‘The fie (Clq) C(n))? +(S(2)- SCE)? ‘Note that the above quantity is simply the length-square o! ints %, and t, on a Cornu spi 1h dsemined by reading distances on Cornu spi ‘Toillustrate, let us consider the diffraction intensity due to a straight edge as given by Eq. $2.32: Hea) =42{(ce@)—Cemyy +(S1)-S04))"]ont = a - i Opitley Wo kino that CQ) = 8(m) © 0.5 Is polit Eon the spiral Orig = « Otigin corresponds tou, © 0 ors © 0, Hence Intensity at polnt x= 6 fg : : 10,9) =2-COY » AD (C08)? 1059?) « 4 Nov, ax we move Into the Hlumination zone, #4 becomes nega) ¥> 0, An x Inerenses, 1 (or t) decreases and hence observation id moves nlong the spiral Into third quadrant, The upper end ofreprent ' line or chord EP remains fixed at £ while lower ond moves wn i spiral, Chord £P continues to increase ns more and more of lov ha _ first Fresnel zone is exposed (sce discussion in Sec, 6,9), thet maximum of Intensity occurs when we reach point ’, where chord Ain is,maximum, At this point © -1.22 on spiral which cortapénde distance of about x = 1.2 ( nd Tk) within the illumination zone ing 7 screen, The intensity at distance x is therefore, given by the iy (neh where P\E is length of the chord on spiral, Its value can be read on thy graph (Fig, $2.5): HE = 1.65. Hence, we get fy, * 1361p, However, it is more accurately found by looking at the values of Fresne| integrals (Table $2,1): C(-1.2) = -0.7154, S(-1.2) = -0.6234 ‘Therefore, we have In = (Roy =4{os+ 017154) + (05+ 062347] &137ly As we further increase distance x on screen, we reach point ?, on the spiral which corresponds to a minimum of intensity, since EP, is a minimum. Point P, has w, = —1.87. Reading the graph, one finds that Jy, = 0.78 Ig. The length of chord EP continues to oscillate between maxima and minima till we get an almost uniform illumination with vanishing fringe contrast. Conversely, as point P moves into the shadow, the representative line EP on the Cornu spiral retracts The upper end of the phasor remains fixed at the p: “while the lowerendir I fiof;the spiral pw region