OP Set 2

25

OPTICS, SET 2 DIFFRACTION GRATINGS

In this set you will investigate Fraunhofer patterns produced by diraction gratings that is, by a regular array of many slits or reecting strips. The physics involved in these exercises is covered in Young & Freedman, 364,5. This set of exercises should take you two weeks to complete.

PRE-LAB
A regular array of slits, or grooves etched into a surface, is known as a transmission grating. Previously you investigated the Fraunhofer diraction patterns produced by light travelling through two slits. If we do the same calculation with more slits, the maximums of the diraction pattern remain in the same positions, but become progressively narrower. With hundreds of slits, these maximums become so narrow that we can measure their angular position with very high precision. We already know that, provided we are in the Fraunhofer regime, the position of these maximums depend linearly (approximately) on the wavelength of the light used. Hence a grating can be used as a spectrometer, an extremely sensitive device for measuring wavelengths and spectral intensities.

scrnDist (xd)

(slitWdth)

detector D (xd,yd)
yd

d (slitSepn) grating screen

(at large distance)

The grating equation Consider a plane electromagnetic wave of wavelength incident on a grating consisting of N slits, as shown. We imagine that the screen is so far away that the rays converging at D are essentially parallel, all making the same angle with the normal to the grating. Then theoretical analysis predicts that the intensity of light hitting the screen, as a function of , is given by IN () = IN (0) where tan = yd , scrnDist = 2 a sin and = 2 d sin . (2.2) sin( /2) /2
2

sin(N /2) N sin(/2)

(2.1)

You may care to show that, in the special case N = 2, these equations match exactly the corresponding equation in the textbook, Young & Freedman, Eq.(36-12).

26

OP Set 2

The term involving varies slowly with angle, and represents the diraction envelope pattern for a single slit. The term involving varies more rapidly, especially when N is large. The successive sharp peaks to either side of the centre are referred to as the rst, second, third, etc., orders of diraction. These peaks are quite narrow and each is separated from the next order by smaller subsidiary peaks. It can be shown from this expression that the maximum intensities in the diraction pattern, the bright fringes, occur at angles given by d sin = m (2.3)

where m is the order of the fringe. This is known as the grating equation (see Young & Freedman, Eq.(36-13)). Note especially that we have always assumed normal incidence, that is that the incident light is perpendicular to the plane of the grating. If not, Equation 2.3 will look a bit dierent. Resolving power In a spectroscopic laboratory, the usefulness of a grating is specied by its resolving power. This is a measure of the minimum wavelength dierence, , that can be distinguished by the grating. Its actual denition is the ratio of the average of two just distinguishable wavelengths divided by their dierence. See Young & Freedman, Eq.(36-14). R = /. (2.4)

Whether or not two wavelengths can be distinguished often involves a semi-subjective judgement, usually made on the following criterion. So long as the intensity peak for one of the wavelengths lies outside the rst minimum of the other, we agree to say that we can tell them apart. The two wavelengths are said to be resolved. If the two intensity peaks are closer together than this, then we agree to say that we cannot tell them apart. The two wavelengths are unresolved.

This recipe for making the decision is known as the Rayleigh criterion. The term involving in Eq.(2.1) determines how close together are the neighbouring maximums and minimums. You can use this, and Eq.(2.3), to nd what change in wavelength () will shift the intensity peak by this amount. Whence you can get this expression for the resolving power: R = / = mN (2.5)

where m is the (integer) order of the fringe in question. See Young & Freedman, Eq.(36-15).) Useful gratings have large values of N . Usually this is specied in terms of the number of grooves per millimetre and the grating width. Typical densities of grooves are 150 per mm up to 1200 per mm, and typical gratings are 50 mm or 100 mm wide.

OP Set 2 Advanced topic.

27

There is another kind of grating, consisting of a regular array of very thin reecting grooves a reection grating. Clearly such arrays will produce the same kind of diraction pattern as a transmission grating, if the reected light is allowed to fall on a screen.

diffraction angle d (stepLeng) detector D (xd,yd)

grating blaze angle b

screen
(at large distance)

However they oer one special advantage. In a transmission grating, most of the light goes straight through without being diracted sideways, and ends up in a bright band at the centre of the screen (called the zeroth order fringe). Very little energy reaches the screen where the diracted fringes are found. But in some reection gratings the reecting grooves are cut at an angle to the plane of the grating. This means that most of the incident light (assumed perpendicular to the plane of the grating, remember) does not reect straight back. Instead, following the laws of specular (mirror) reection angle of reection equals angle of incidence it goes at an angle dierent from the zeroth order diraction angle. Such gratings are said to be blazed. The trick is to construct the grating in such a way that the direction in which most of the light goes corresponds to the direction in which one of the fringes occurs (often the rst order fringe). If we restrict ourselves to the case where the light is incident normally to the plane of the grating, specular reection occurs at an angle = 2b , where b is the blaze angle the angle of the reecting grooves to the plane of the grating. Putting this angle into the grating equation (2.3) gives the blaze wavelength, b , for which the peak intensity will occur, m b = d sin 2b ( most often with m = 1 .) (2.6)

28

OP Set 2

IN LAB
The rst two exercises of the set employ the same Huygens Principle calculation we used in Set 1, to study diraction from a small number of slits. Our aim is to establish that the Fraunhofer patterns we observe are identical with those described by Eq.(2.1). Once this has been established, you will use this equation in the remaining exercises to study the properties of diraction gratings. To help with this, several pre-written MATLAB scripts are on the course web site: HuygensConstruction2 a slightly altered version of the function used in Set 1, which can handle dierent numbers of slits, within a small range (see page 43); op2driverA/B/C three GUI Data Input objects with facilities for inputting and changing various variables, to be used with dierent exercises in this set; Three functions which will calculate the Fraunhofer pattern under dierent circumstances ManySlitCalculation: for a small number of slits (see page 43); GratingCalculation: for a grating of up to 50 slits (see page 44); and BlazingCalculation: for a blazed reection grating (see page 44). Before starting these exercises, download these les to your local directory.

OP2.1 Double slit diraction Start from where you were at the end of Exercise OP1.6, examining the Fraunhofer pattern from two slits of nite (that is, not innitesimally small) width. You are asked to keep the screen distance constant (at the default value of 50 mm) and to observe what happens to the pattern as you change the width and separation of the slits and the wavelength of the light. (a) Download and run op2driverA and study the inuence of the width of the slits on the diraction pattern. Keep the slit separation xed at 0.010 mm and the wavelength at 600 nm. Change the slit width successively to these values: 0.002 mm, 0.005 mm, and 0.008 mm. Draw the resultant intensity patterns in these boxes. Make sure your drawing shows clearly which features of the pattern change and which stay the same. Slit width

0.002 mm

0.005 mm

0.008 mm

OP Set 2

29

(b) Study the inuence of the separation of the slits on the diraction pattern. Keep the slit width xed at 0.005 mm, and the wavelength at 600 nm. Change the slit separation to these values: 0.010 mm, 0.030 mm and 0.050 mm. Note: the graph is no longer smooth, so you will need to use more points in the calculation. Inside HuygensConstruction2 change the number of detectors N from 151 to 701. Draw the resultant intensity patterns in these boxes. Again, make sure your drawing shows clearly which features of the pattern change and which stay the same. Slit separation

0.010 mm

0.030 mm

0.050 mm

(c) Study the inuence of the wavelength on the diraction pattern. Keep the slit width xed at 0.005 mm, and the slit separation xed at 0.010 mm. Change the wavelength of the light to these values: 400 nm, 550 nm and 750 nm. Draw the resultant intensity patterns in these boxes. Again, make sure your drawing shows clearly which features of the pattern change and which stay the same. Wavelength

400 nm

550 nm

750 nm

30

OP Set 2 (d) Lastly we would like to compare this with the intensity patterns that you calculated previously for two point sources, and for a single slit. Set up the graph of the intensity pattern for the parameters: wavelength = 600 nm screen width = 30 mm slit width = 0.004 mm screen distance = 50 mm slit separation = 0.016 mm

When you were doing Exercise OP1.3, you saved a copy of the diraction pattern for two point sources a distance 0.016 mm apart, with the other parameters the same as you are using now. This le was named op1fig1.fig. Reopen this gure using Open from the File menu. Likewise, you were doing Exercise OP1.4(b), you saved a copy of the diraction pattern for a single nite slit, of width 0.004 mm, also with the other parameters the same as you are using now under the name op1fig2.fig. Reopen this gure, without closing the other two graphs. Arrange all three graphs on the screen so that the axes are all the same size. Study these graphs and answer this question: What is the relationship between the diraction pattern of two nite slits, and the other two patterns?

Convince yourself, without necessarily computing anything new, that the total intensity distribution for two slits of nite width, is I2 () = I (0) sin( /2) /2
2

cos2 (/2)

(2.7)

Tutors initials

NOTES.

OP Set 2 OP2.2 Multiple slit diraction

31

At this point, if you havent done so already, open the m-le HuygensConstruction2 and observe what changes have been made in order to allow the number of slits to be changed. Make sure you understand how the calculation is done, and why it would be impractical to use this script for a very large number of slits. (a) Re-run op2driverA with the same parameters you used in Exercise OP2.1(d). Observe what happens to the diraction pattern as you change numSlits through 1, 2, . . . 5. Draw the resultant interference patterns for N = 2, 4, 6 in these boxes. Make sure your drawings show clearly which features of the patterns change, and which stay the same. No of slits

(b) Rerun the same computations, changing numSlits through, 1, 2, . . . 5. Measure and record in the following table, to three signicant gures only, the intensities at the middle of the screen. Calculate the ratio of the central intensity with N slits and the central intensity with a single slit (to two signicant gures) and enter the values here. N IN (0) IN (0)/I1 (0) (c) When we calculated these intensities, we modelled the light coming through each slit as coming from a xed number, M , of discrete point sources making up each slit. In terms of this model, it should be obvious why the central intensity increases as the number of slits increases. The same amount of energy ows through each slit, so with N slits, the total energy ux should be N times that for a single slit. But this is not what your results show or at least, should not be. Why do the ratios IN (0)/I1 (0) have the values you found for them? 1 2 3 4 5

Tutors initials

32 OP2.3 Mathematical formula for multiple slit diraction

OP Set 2

The point has already been made that it would be impractical to use the current script for a system with a large number of slits. In future we will use a MatLab function based on Eqn.2.1, rather than one based on Huygens Principle directly. But rst we need to make sure the two ways of doing the calculation give the same answers for a small number of slits. (a) Firstly, you need a MATLAB function which will calculate the right hand side of Eq.(2.1), for any positions along the screen. The m-le ManySlitCalculation contains, in skeleton form, such a function. The steps in the calculation are indicated, but initially each of the vectors you have to calculate are set equal to 1. The real statements are left for you to ll in, as follows. Firstly, write the three statements, using Eqs.(2.2), that calculate , and for a vector yd which has been passed to the function. Note: you will need to use the MATLAB function atan. Next write the statement that calculates (sin( /2)/( /2))2 , denoted by the name slitPatn. Do this in the most direct way, by adding the statement slitPatn = (sin(beta/2)./(beta/2)).^2; Test that this works properly, by setting up a vector in the command window for yd (redening scrnwdth if necessary) and calling the function in its unnished form. Then plot the result and you should see the usual nite slit diraction pattern: >> yd = [-scrnWdth/2 : scrnWdth/300 : +scrnWdth/2]; >> Irel = ManySlitCalculation(yd); >> plot(yd,Irel); There is a slight problem here. Although the output looks OK, you get an error message, because the in the denominator vanishes at the centre of the screen. There are several ways you can get round this, since you know that the value of sin / is 1 when = 0. However the simplest kludge is to make sure you never divide quite by zero. In MATLAB the smallest number you can add to 1.0, and make a dierence, is called eps. It is a number of order 1016 . Add this to your denominator, replacing the statement above by, slitPatn = (sin(beta/2+eps)./(beta/2+eps) ).^2; Run it again. You should get the same answer, but this time without the error message. (b) When you come to calculate the term involving in Eq.(2.1), which describes the contribution of the many sources, you have to use the same trick. Add this statement to the script, srcePatn = (sin(numSlits*(phi/2+eps))./(numSlits*sin(phi/2+eps)) ).^2; Test that this part works properly as previously, by adding a temporary statement slitPatn = 1 to override the calculation of slitPatn above. When you plot the result you should see the diraction pattern due to a number of point sources. Draw in this box the pattern you see when you do this test with the number of slits equal to 4.

(c) You are now ready to calculate the complete function, and can verify that the diraction pattern from a number of nite slits is accurately described by Eq.(2.1), as follows:

OP Set 2

33 Take out the statement you added in OP2.3(b), which sets slitPatn equal to 1, so that it once again calculates slitPatn using . Now compute the diraction pattern again, using the Huygens Principle method of calculation, without using the driver (because you want to do your own plotting). Plot the result so that you can draw another graph on top of it. >> [yd Itot] = HuygensConstruction2; >> plot(yd, Itot); >> hold on; Then calculate the complete theoretical diraction pattern, given by Eq.(2.1), using the relative intensity distribution computed by the function ManySlitCalculation (with the correct slit width and spacing), multiplied by the central intensity that you recorded in exercise OP2.2(b). Plot this pattern on top of the graph you just computed. >> Irel = ManySlitCalculation(yd); >> plot(yd, Irel*..., red); (You ll in the multiplier.) Repeat with the number of slits equal to 2, 3, . . . 6. Estimate the discrepancy in the heights of the rst order fringes in all cases. Discrepancy:

Tutors initials

NOTES.

34 OP2.4 Transmission diraction grating

OP Set 2

We now consider a grating when used as a simple spectrometer. Schematically the experimental arrangement is something like this. (Compare with Young & Freedman, Fig.36-20.)

o e t e c t r

u r c e

o c i m a t

o r

You will notice that the detectors are arranged around a semicircle, with the diraction angle varying between /2. Although real diraction gratings might consist of 105 slits, we will conne ourselves to numbers less than 100. We will not attempt to use Huygens Principle to calculate the intensity pattern at the detectors. Instead we will use Eq.(2.1). (a) Open the m-le GratingCalculation and examine its contents. It should match exactly the statements in ManySlitCalculation with these small changes: The function requires the vector theta as input rather than the distances yd we have used till now. The variables scrnDist and scrnWdth are no longer needed. The wavelength lambda is passed to the function, rather than being a global variable. Youll see why in a moment. The intensity at the middle of the screen is chosen so that, while arbitrary, it reects the dependence on the number of slits in the grating and the width of each. Now run the new script op2driverB. The GUI input panel it opens up is similar to the previous one, except that the variables scrnDist and scrnWdth are no longer present, and there is facility to input two dierent wavelengths. Currently both these are set at the same gure. Check that everything runs without errors. (b) Start o with these parameters: wavelength1 = 600 nm wavelength2 = 600 nm slit width = 0.0003 mm slit separation = 0.0015 mm number of slits = 20

Observe the pattern as you change the number of slits to 10, 20, 30, . . . 100. Draw the resultant intensity patterns for values of 10, 50 and 100 in these boxes. Make sure your drawing shows clearly which features of the pattern change and which stay the same. No of slits

10

50

100

OP Set 2

35

(c) With the same default values, and 50 slits, change the slit separation to these values: 0.001 mm, 0.003 mm and 0.005 mm. Draw the resultant interference patterns in these boxes. Again, make sure your drawing shows clearly which features of the pattern change and which stay the same. Slit separation

0.001 mm

0.003 mm

0.005 mm

(d) With the same default values, and 50 slits, change the slit width to these values: 0.0002 mm, 0.0006 mm and 0.0010 mm. Draw the resultant interference patterns in these boxes. As usual, make sure your drawing shows clearly which features of the pattern change and which stay the same. Slit width

0.0002 mm

0.0006 mm

0.0010 mm

(e) Lastly, with the same default values, and 50 slits, change both wavelengths together to these values: 400 nm, 550 nm and 750 nm. Draw the resultant interference patterns in these boxes. Again, make sure your drawing shows clearly which features of the pattern change and which stay the same.

36 Wavelength

OP Set 2

400 nm

550 nm

750 nm

Tutors initials

OP2.5 Diraction grating spectrometer Thus far we have explored diraction patterns for a single wavelength. (a) Test problem: A grating was used in the third order with light from a helium discharge tube. The red line at 667.8 nm appeared at an angle 40.0 . Run op2driverB using the following starting parameters: wavelength1 = 667.8 nm wavelength2 = 667.8 nm slit width = 0.0002 mm slit separation = 0.0030 mm number of slits = 40

and, by adjusting the slit separation in small steps, nd how many slits per mm the grating must have had. Compare your answer (to four signicant gures) with the theoretical value that you would get by using Eq.(2.3).
measured value theoretical value

Number of slits per mm:

One of the most common uses of a grating spectrometer, however, is to investigate the light emitted by certain atomic or molecular species, which we know consists of several discrete, very pure frequencies that is, line spectrums. The next two exercises involve exploring the (Fraunhofer) diraction pattern when there are two frequencies present. (b) Run op2driverB using the following parameters: wavelength1 = 700 nm wavelength2 = 500 nm slit width = 0.0002 mm slit separation = 0.0030 mm number of slits = 40

Observe the diraction patterns you see. Identify the rst, second, third, etc. order fringes associated with each wavelength. You should be able to appreciate a possible source of diculty with diraction gratings. The rst order green fringe occurs at a smaller angle than the rst order red fringe. Similarly the second order green comes before the second order red, and the third order green before the third order red. But the fourth order green also comes before the third

OP Set 2

37 order red. The two diraction patterns are said to overlap. With a spectrum consisting of many wavelengths, this kind of overlapping can lead to confusion. Note: The word overlap here means that two whole spectrums sit on top of one another. Do not confuse this meaning with how it is used later in Exercise OP2.6 when you deal with two lines in one spectrum overlapping one another.

(c) Test problem: By keeping the rst wavelength xed at 700 nm, and varying the second wavelength, nd the largest dierence in wavelength (to three signicant gures) for which there is no overlap in the rst, second and third orders. (This means the third order lines must be just on the point of overlapping.) Wavelength dierence:

Tutors initials

NOTES.

38 OP2.6 Resolving power

OP Set 2

(a) We now wish to investigate the resolving power of a grating. We rst do a preliminary exploration of which parameters determine the ability of a grating to distinguish two close frequencies. Start o with these values: wavelength1 wavelength2 slit width slit separation number of slits = = = = = 550 nm 575 nm 0.0002 mm 0.0020 mm 20

If you look at the second or third order fringes, it should be perfectly obvious that there are two discrete frequencies present, even though they have very similar colours. However from the rst order fringe it is not so obvious. If you zoom, you will see that the total intensity has a slight dip in the middle, but from the simulated spectrum it is hard to tell whether there are two narrow lines present, or a single wide line. The lines are unresolved in rst order. However, if you change the number of slits to 40, the two lines are clearly resolved in all orders. As suggested by Eq.(2.5) the resolving power of the grating is increased by increasing the number of slits. (b) The Rayleigh criterion for judging when the two lines are just on the borderline of being resolved is that the maximum of one line occurs at the same angle as the rst minimum of the other. Refer back to the discussion on page 26 of these notes. Change the number of slits to 23 and verify that the Rayleigh criterion is now just satised. (Note: you need to be careful when zooming, so that the maximum does not disappear out the top of the window.) In most real world situations, the Rayleigh criterion cannot be invoked in this form. Since all you can measure is the total intensity from both lines together, you cant say where the rst minimum of either one occurs. For one dimensional sources, such as we are dealing with here, it can be shown theoretically that when the maximum of one line occurs at the same angle as the rst 100% minimum of the other (that is, when the 79% Rayleigh criterion is just satsied) the total intensity halfway between the two is 79% of the maximum intensity of either(to two signicant gures). Therefore for all one dimensional sources the criterion is taken to be the two lines are considered just resolved when the ratio of the minimum intensity between the two lines to the maximum intensity is equal to 79%. For the conguration you have set up, measure this ratio, together with an estimate of its uncertainty. Does it match the theoretical value? Rayleigh ratio, minimum/maximum:

(c) Test problem: Using a grating with 50 slits, and a slit separation of 0.0030 mm, nd the smallest dierence of two wavelengths around 550 nm that can be resolved in third order. Estimate the accuracy to which you can make this measurement. Hence calculate the measured resolving power, R, using Eq.(2.4), and compare it with the theoretical value, calculated from Eq.(2.5).

OP Set 2
wavelength dierence R (measured) R (theoretical)

39

(d) Test problem: Imagine that you need a grating which is powerful enough to separate two spectral lines of wavelengths 600 nm and 615 nm that is, to be sure that there are in fact two distinct lines present. Using a slit separation of 0.0030 mm, nd the total number of slits necessary to do this in rst order. Repeat for second and third orders. As a check, use Eq.(2.5) to calculate the expected number of slits for each order. measured value Number of slits, rst order Number of slits, second order Number of slits, third order expected value

Tutors initials

NOTES.

40 OP2.7 Advanced question: Blazed reection grating

OP Set 2

We now consider a spectrometer using a blazed reection grating. Schematically the experimental arrangement is something like this.

cope teles

detector

grating

collimator

light source

The incident beam of light comes in from the right, is reected and diracted by the grating and some of the light enters the telescope tube situated at a diraction angle , as shown. Now we know, from what we have seen so far, that when light diracts, the amount of light energy that spreads sideways is much less that the amount which goes straight ahead. In this case, straight ahead corresponds to the direction in which a single ray reects from the mirrored surface, that is = 2b . Therefore, as was mentioned on page 27, the maximum amount of light will be detected when the diraction angle happens to equal twice the blaze angle; and only very small amounts of light energy will enter a telescope set at any other direction. This is the property which we will investigate in this last exercise.

(a) It should be clear that, in terms of the calculations we have been doing so far, the kind of diraction pattern we expect from this kind of grating will be that from a transmission grating in which the slit width is essentially equal to the slit separation. For this exercise we will replace these two variables by a single variable, the step length. Open the m-le BlazingCalculation and examine its contents. It should match exactly the statements in GratingCalculation with these changes: The variables slitWdth and slitSepn have been removed from the global statement and replaced by the variable stepLeng. The two MATLAB statements that calculate the single slit pattern and the many point source pattern, have been rewritten with this change of variable names. A new variable, blazAngl, has been added to the global statement, for the blaze angle b . Its value will be entered elsewhere (in degrees, not radians!). The statement which calculates the quantity , has been rewritten so that the maximum of the single slit pattern occurs, not at = 0 but rather at = 2b . The nal statement returns, not the total, combined intensity distribution, but the two component parts separately. You will see why shortly. (b) You are supplied with another pre-written GUI object, named op2driverC. Run this and check that it behaves in a similar fashion to previous ones. It has been constructed to call the function BlazingCalculation halfway through its operation.

OP Set 2

41 You should see the familiar diraction graphs, with this change. The graph which displays the diraction pattern (for a single wavelength) as a line graph does not show the total intensity distribution. Instead it displays the single slit pattern and the multiple source pattern separately. Only in the lower graph, the intensity simulation, can you see the combined eect of both these eects. Make sure you understand what the displays are showing.

(c) Run op2driverC using the following parameters: wavelength = 400 nm step length = 0.002 mm blaze angle = 0 degrees number of slits = 50

Change the blaze angle in steps of 10 and show in these boxes what happens to the diagram with the two diraction patterns separately (the upper graph). Make sure your drawings show clearly which features of the patterns change, and which stay the same. Blaze angle

10

20

Then change the wavelength to 750 nm, and repeat these observations. Show what happens to the two diraction patterns diagram in these boxes. Blaze angle

10

20

(d) Test problem: A grating has 750 lines (slits) per mm. What blaze angle should the grating have if it is to be used (with normal incidence) to select the rst order fringe for a wavelength of 650 nm? (Hint: Dont try to change the actual number of slits. So long as that number is large enough (about 50), the answer to this question will not be aected.)

42

OP Set 2 Compare your result with the theoretical value that would be predicted by Eq.(2.6).
measured value theoretical value

blaze angle:

(e) Investigate the eect of changing the step length (or the number of lines per mm) by doing again the computation you made in part (d), and then repeating it with gratings with 500 lines per mm, and 250 lines per mm. Show what happens to the two diraction patterns diagram in these boxes. Lines per mm

750

500

250

(f) As described here, a blazed grating seems to be a very limited instrument. Any actual device seems able to observe at one wavelength only, or at least wavelengths within a certain small range. As an exercise, estimate the range of wavelengths over which the grating described in part (d) could be used to give useful measurements. The criterion you will use depends on how you intend to use the instrument. Enter your estimate of the range in this box, and state the criterion you used to make your judgement.

There is, however, one further variable that can be changed, which we have not included as a parameter in our calculation. Look again at the diagram on page 27, and describe how you could use such an instrument to observe a complete range of dierent wavelengths.

Tutors initials

OP Set 2

43

Sample m-les and functions

Exercise 1
function [yd, Itot] = HuygensConstruction2; % HUYGENSCONSTRUCTION2 Calculates the intensity distribution % for a small number of slits. Modified for Exercise 2. % All physical variables are global and set elsewhere global lambda scrnDist scrnWdth slitWdth slitSepn numSlits; % Numbers of sources and detectors can be changed easily M = 26; % Number of sources N = 501; % Number of detectors. K = numSlits; % Number of slits % xs,ys are coordinates of the source (row vectors) xs = zeros(1,K*M); ys = []; % There are K slits. Each slit represented by a (1xM) vector ysSlit = [-slitWdth/2: slitWdth/(M-1): +slitWdth/2]; % The vector representing all slits together has K*M elements. % Concatenate K vectors each of size (1xM) for count=1:K, ys = [ys, slitSepn*count + ysSlit], end; % Adjust centers of slits symmetrically about origin ys = ys - (K+1)/2*slitSepn; The rest of the script is identical with HUYGENSCONSTRUCTION. See Exercise Set OP1.

Exercise 2
function Irel = ManySlitCalculation(yd) % MANYSLITCALCULATION Function to calculate the theoretical % diffraction pattern from a number (N) of slits. global lambda scrnDist scrnWdth slitWdth slitSepn numSlits; % First theta beta phi calculate theta, beta and phi = 1; = 1; = 1;

% Next calculate the diffraction pattern from a single slit slitPatn = 1; % Then calculate the diffraction pattern from N point sources srcePatn = 1; % Last multiply the two to produce the relative intensity Irel = slitPatn .* srcePatn ; return;

44

OP Set 2

Exercise 3
function Itot = GratingCalculation(lambda,theta); % GRATINGCALCULATION Calculates the intensity distribution % for a grating of up to 50 slits. global slitWdth slitSepn numSlits; % First calculate phi and beta. Note: vector theta is passed. beta = 2 * pi * slitWdth * sin(theta) / lambda; phi = 2 * pi * slitSepn * sin(theta) / lambda; % Next calculate the diffraction pattern from a single slit slitPatn = (sin(beta/2+eps)./(beta/2+eps)).^2; % Then calculate the diffraction pattern from N point sources srcePatn = (sin(numSlits*(phi/2+eps))./(numSlits*sin(phi/2+eps)).^2; % Choose an appropriate value for I(0) proportional to numSlits and slitWdth I0 = (numSlits*slitWdth)^2; Itot = I0 * slitPatn .* srcePatn ; return;

Exercise 4
function Intens = BlazingCalculation(lambda,theta); % BLAZINGCALCULATION Calculates the intensity distribution % for a blazed reflection grating of up to 50 slits. global blazAngl stepLeng numSlits; % First calculate phi and beta. Note: vector theta is passed. beta = 2 * pi * stepLeng * sin(theta-pi*blazAngl/90) / lambda; phi = 2 * pi * stepLeng * sin(theta) / lambda; % Next calculate the diffraction pattern from a single slit slitPatn = (sin(beta/2+eps)./(beta/2+eps)).^2; % Then calculate the diffraction pattern from N point sources srcePatn = (sin(numSlits*(phi/2+eps))./(numSlits*sin(phi/2+eps))).^2; % Return the two intensity patterns separately Intens = [slitPatn;srcePatn]; return;