JEREMIAH A. CRONIN | DAVID F. GREENBERG VALENTINE L. TELEGDI Department of Physics, University of Chicago and The Enrico Fermi instiiuie for Nuciear Studies ADDISON-WESLEY PUBLISHING COMPANY READING, MASSACHUSETTS + PALO ALTO + LONDON » DON MILLS, ONTARIOADDISON-WESLEY SERIES IN ADVANCED PHYSICS COPYRIGHT © 1967 BY ADDISON-WESLEY PUBLISHING COMPANY, INC. ALL RIGHTS RESERVED. THIS BOOK, OR PARTS THEREOF, MAY NOT BE REPRODUCED IN ANY FORM WITHOUT THE WRITTEN PERMISSION OF THE PUBLISHER. PRINTED IN THE UNITED STATES OF AMERICA. PUBLISHED SIMULTANEOUSLY IN CANADA. LIBRARY OF CONGRESS CATALOG CARD NO. 67-21301.PREFACE 1947, the Dep: required every graduate student to pass a comprehensive examination in the fundamentals of classical and modem physics before allowing the student to begin his thesis research for the Ph.D. This examination, called the “Rasie” in the early days, and now known officially as the “Candidacy Soe ee We present here a selection of problems from these exams, together with solutions. It is our hope that the collection will be useful to students as a means of hoth measuring and improving their understanding af the subject, matter. This collection will serve this purpose only if it is properly used. Browsing through the solutions will not be of permanent value to the student; only after honest effort should he consult the solutions. We have tried to make each solution self-contained and coherent, and we have strived to avoid reproducing material found in standard textbooks. It is difficult to know how much knowledge may be assumed on the part of the reader; our general rule has been to assume the level of knowledge expected of a second-year graduate student at the University of Chicago. The solutions were prepared by two of the authors, Jeremiah A. Cronin and David F. Greenberg. The third author, Professor Valentine L. Telegdi, assisted in putting the solutions to Chapters 1 and 2 into final form. He also reviewed Chapters 6, 9, and 11, All three authors benefited from conversations with colleagues of the University of Chicago. In particular, they would like to thank Dr. Sol Krasner of the Physics Depart- ment for his numerous contributions. They are grateful to the Department for its support and encouragement, and to the secretarial staff of the Enrico Fermi Institute for their assistance in the preparation of the manuscript. Chicago, Illinois TA June 1967 . F sos nae LL‘The names of those faculty members of the University of Chicago Physics Department who have served on the Candidacy Examination Committees aver the years appear helow: EDWARD ADAMS, IT SAMUEL K. ALLISON HERBERT L. ANDERSON NORMAN E. BOOTH MORREL H. COHEN RICHARD H. DALITZ RUSSELL J. DONNELLY DAVID H. DOUGLASS ENRICO FERMI PETER G. 0. FREUND HELLMUT FRITZSCHE RICHARD L. GARWIN MURRAY GELL-MANN CLAYTON GIESE MARVIN L. GOLDBERGER ROGER H. HILDERRAND ULRICH E. KRUSE ANDREW W. LAWSON RICCARDO LEVI-SETTI WILLIAM TICHTRN LEONA MARSHALL BERNDT MATTHIAS MARIA G. MAYER PETER MEYER ROBERT 8. MULLIKEN DARRAGH BE. NAGLE YOICHIRO NAMBU PETER D. NOERDLINGER REINHARD OEHME EUGENE N. PARKER SHERWOOD PARKER JAMES C. PHILLIPS JOHN R. PLATT MICHAEL PRIESTLEY FREDERICK REIF J. J. SAKURAI MARCEL SCHEIN JOHN A. SIMPSON ROVAT. STARK EDWARD TELLER ROBERT W. THOMPSON DEREK A. TIDMAN GREGOR WENTZ. 8. COURTNEY WRIGHT CARL M. YORK WILLIAM H. ZACHARIASENProblems Mathematical Physics Mechanics Electromagnetism Electronies Optis =. 2. Quantum Mechanics Thermodynamics Statistical Physics Atomic Physics .” amor BB cear) 8 For the Experimentalist Solutions ‘Mathematicai Piysics Mechanics Electromagnetism Electronics nm ome Op oo Quanium Mechanics ‘Thermodynamics Statistical Physics Atomic Physies s State Nuclear Physics . HP Bbeowr)8 Appendix: Useful Information .PROBLEMSMATHEMATICAL PHYSICS 1, A community practices birth control in the following peculiar fashion. Each set. of parenta eontinnes having children nntil a son ia horn; then they stop. What is the ratio of boys to girls in this community if, in the absence of birth control, 51% of the babies born are male? 2. A die consists of a cube which has a different color on each of 6 faces. (a) How many distinguishably different kinds of dice can be made? (b) How many different ways are there to make a pair of dice? 3. Each face of a regular octahedron is to be given a different color. If eight different colors are available, how many distinguishable octahedra can be made? 4. In dealing 52 cards, consisting of 4 suits of 13 cards each, among two teams (each team containing two partners), what is the probability that a particular pair of partners obtains at least one complete suit between the 5. A certain process has the property that, regardless of what has transpired in an interval [0 to ¢], the probability that an event will take place in the interval [t, (f + A)] is X44. Assume that the probability of more than one n events have taken place, passing to the limit of h going to zero. Evaluate the average value of n and the average value of n® for the distribution func- tion. 6. There are about 6500 stars visible to the naked eye. Sometimes two stars appear very close together, though upon careful examination no phy- sical connection is found between them. Such a pair is called an optical double star. (a) Assuming the stars to be distributed at random on the celestial sphere, compute the expectation value of the number of optical double stars with ‘a wparation of no more than 1’ of arc. 12 MatTHEMaTIcaL PHysics—PROBLEMS (b) What is the probability that there are precisely two optical double stars? (c) Estimate roughly the probability of an optical triple star. 7, Find eigenvalues and normalized eigenvectors of the matrix 0 0 0 D 0 10 wl | oo 0 of © ono 8. Let Ay (i — 1,2,3) be the eigenvalues of the matrix / 2 -1 —a\ H= e 1 *} —3 2 3 ne Calculate the sums : : @) Bm and) Z 9. Calculate 1 = Tr [e'*e'>], where the components of @ are the three standard Pauli matrices 0; for spin 4. 10. Consider a symmetric second-rank tensor T with components T'x (é, k= 1,2,3). (b) Associate a surtace 1 = >; .,21'4X,X, (X; = Cartesian coordinates) with T. Give interpretations of the three invariants in terms of properties of the surface. lim ("" 7, —# tim [saci wth a>. sin? Sade.MATHEMATICAL PHYsICS—PROBLEMS 14. Calculate: @ n= ae vdr and (b) I, = J 15. Develop f(z) = cos (2*) in a Fourier integral. 16. Find f(t) by inverting the Laplace transform 8 Po mgayar, pra Jo”? 17. Calculate f ° dg |(+ cos $) (a) when @ > 1, (b) when a — ay + ke, ay, real, e > Oand 0 < a < lase 0, (c) when @ = —1. 18. Evaluate 1, = {"" de[oosh x and I, — f"” de/eosh' x. 19. Evaluate _ ("4g__0 +4008 1-Jraeg $+ 8ab cong? — (18h 20. The gamma function is defined by T(z) = Je frtetdt, — Re(z) > 0. fier cos tdt = T(x) cos 5, fer sintdt = T(x) sin@. ° 21. Show that sinh (47) align, Hwcace, 22, Evaluate by contour integration = dx me sige Show your contour and all poles and branch cuts in the complex plane. 34 ‘MaTHEMATICAL Puysics—PROBLEMS 23. Evaluate the series = (1) —Txt 2 nt” 720 by contour integral techniques. [Hint: Use the fact that the function 1)(sin 272) has poles along the real axis at z = 0, +1, +2,... Consider the contour above. 24, Consider the anaiytic function 2; Fe =o@n[1— 20 — pay] with p(z) = te —a)/z. Here a is real and positive. Choose the branch lines fram —oo to Q and from a ta oo. for plz) slong (a) Discuss the Riemann surface of F(z). (b) Show that there is one sheet where F(z) may be represented in the form Fla) = Fle) +(e) [7 ds — 0 and determine W(s). 25. Evaluate tim V7 ("ae where 7 is a positive integer. 26. Compute f(a, b) = (~ (dz/x) (e-** — 27. Find the sum of the following infinite series: SH=14 2x4 3c 4x7 +--+ for|z}<1. 28. A generating function F(z, t) of the Hermite polynomial H,,(z) is rea" = 3 Haye! (a) Express H(z) as a contour integral (h) Prove that, H(z) satisfies Hermite’s differential equation TH ,.dH _ Ga — GE + mnt =0. F(z,1) (c) Deduce the relation aH (x) = 2nH,.(2).MATHEMATICAL PHysICs—PROBLEMS = 5 29. A generating function for the Legendre. polynomials P,(z) is ——_!___se_ Fup, (= 2ar + F 2 Pe), Prove that 2Pi(z) = Pj_s(x) + LPi(z) where Pi(x) = dP,(x)/dz. Oe, 1) = = c086,|r] <1. 30. Given the Laurent series for e*-”) as 74,2" where A, = J,(1), obtain an expression for the Bessel function J,,() as an integral from —2 tom of & tei 81. The function $(z, y) is given on the plane z = 0. Find, for z > 0, a solution (2, y, 2) of Laplace's equation that reduces to $(c, y) on the plane 2=0. 32. Show that = [Teronag satisfies Bessel's equation of zeroth order and imaginary argument, that is Ko(x) = Jo(tz). Show that K(x) has the asymptotic torm De-* |S E for very large x; give the value of the constant D. 38. Calculate {r-dA over the surface of a torus 24, O: 2 = rain dy sing, cos d, a =rsin $y sin $, sin $, 2, = rin de cos chy 2 = 1 c08 $y. 35. Gaseous helium is flowing without turbulence at a velocity v down a pipe and into the atmosphere. Within a very short distance from the end of the pipe, the helinm ia rapidly dilnted to essentially zero concentration He Se ceeeaeeea xo x0 10 Sct up and solve the differential equation for the concentration of air in the pipe asa function of distance from the end of the pipe. Assume equilibrium conditions, neglect wall friction and end effects, assume no temperature difforonce, and assume that the coefficients of diffusion of O, and N, into Ho aro the aame and equal to D.6 MATHEMATICAL PHYsics—PROBLEMS 36. The equation describing the neutron density in a chain reacting pile is Vin + Kin =0. (a) With the boundary condition that the neutron density be finite and positive everywhere, and that it vanish outside the pile, find the radius of a spherical pile for a given value of K. (b) Now suppose that a thin layer of material of thickness ¢ is added to the surface, and that the neutron density in the layer is described by Vin — p'n = 0. Assume the boundary conditions at the interface are that n and grad n are continuous. Demanding that n vanish outside the pile and material layer, find for fixed values of K, pz and ¢, an expression for the radius of the internal region. Assuming K < yu, derive an approximate relation for the difference between the radii without and with the layer. 37. A point source of neutrons on the axis of a long square column of graph- ite 150 cm on a side emits 10° neutrons per sec. Calculate the flux of neutrons at a point on the axis 1 m from the source if the diffusion coefficient of the etn ae aiea ae aaMECHANICS 1. Derive the form of Stokes’ law by dimensional analysis. Assume that this foree is independent of the density p of the fluid. What happens when this assumption is dropped? 2. A gas bubble from a deep explosion under water oscillates with a period T ~ p%d’e* where p is the static pressure, d the water density, and’e the total energy of the explosion. Find a, 6, and c. 2 A satellite is ut into a cirenlar orhit. at. a distance R, ahove the center of the carth. A viscous force resulting from the thin upper atmosphere has a magnitude F, = Av*, where v is the velocity of the satellite. It is noted that this results in a rate of change in the radial distance r given by dridt = —C, where C is a positive constant, sufficiently small so that the loss of energy per orbit is small compared to the total kinetic energy. Obtain expressions for 4 and a. 4. A point mass m under no external forces is attached to a weightless cord fixed to a cylinder of radius #. Initially the cord 1s completely wound up so that the mass touches the cylinder. A radially-directed impulse is now given to the mass, which starta unwinding. (a) Find the equation of motion in terms of some suitable generalized coor- dinate, (b) find the general solution satisfying the initial condition, and (c) find the angular momentum of the mass about the cylinder axis, using the result of (b)8 MEcHANIcCS—PRoBLEMS 5. Consider a lawn sprayer consisting of a spherical cap (at. = 45°) provided with a large number of equal holes through which water is ejected with velocity vp. The lawn is not uniformly sprayed if these holes are evenly spaced. How must p(a), the number of holes per unit area, be chosen to achieve uniform spraying of a circular area? Assume the radius of the sprin- Kling cap is very much less than the radius of the area to be sprayed, and the surface of the cap is at the level of the lawn. 6. Find the differential equation for the contour of a constraining surface on which a point mass will oscillate with a period independent of the am- plitude. 7. Three masses (m;, ms, m;), forming the corners of an equilaierai iriangie, attract each other according to Newton’s Law. Determine the rotational motion which will leave the relative position of these masses unchanged. 8 A mass m moves in a circular orbit of radius r. under the influence of a central foree whose potential is —km/r". Show that the cireular orbit is stable under small oscillations (that is, the mass will oscillate about the circular orbit) ifn < 2. 9. ‘Iwo particles move about each other in circular orbits under the influence of gravitational forces, with a period 7. The motion is suddenly stopped at a given instant of time, and the particles are then released and allawed to fall into each other. Prove that they collide after a time 7/(4/2). responding quantities for the moon are 0.275r, and 0.604p,. A man standing on earth bends his knees, lowering his center of mass 50 em. Exerting his maximum strength he jumps straight up, raising his center of mass 60 em 11. A uniform thin rigid rod of weight W is supported horizontally by two vertical props at its ends. At ¢ = 0 one of these supports is kicked out. Find the force on the other support immediately thereafter. —MECHANICS—PROBLEMS 9 12, Three identical cylinders with parallel axes are in contact with each other on a rough plane, with two cylinders lying on the plane and the third resting on >) top of them, as in the figure. What is the minimum angle which the direction of the force acting between the cylinders and the plane makes with the vertical? 18. A yo-yo rests on a level surface. A gentle hori- zontai puii (see figure) is exerted on the cord so that = //--, the yo-yo rolls without slipping. Which way does it Roe roll and why? . 14, A horizontal circular disk of mass M is free to rotate about a vertical axis through @ point on iis rim. If a dog of mass m waike once around the rim, show that the disk turns through an angle given by the expression f 4m cos* o (31/2) + 4m cos* y" 15. A layer of dust is formed h feet thick (A small compared to the earth's radius) by the fall of meteors reaching the earth from all directions. Show, by considering angular momentum, that the change in the length of the day is approximately (5hd/RD) of a day, where R is the radius of the earth, and D and d the densities of earth and dust, respectively. Use a notation in which the initial quantities carry subscript zero, final quantities a sub- seript 1. The moment of inertia of a sphere about an axis through its center is (2/5) MR’, that of o thin-walled, hollow sphere of mass m and radius Ris (2/3) mR. 16. A simple gyrocompass consists of a gyroscope spinning about its axis with angular velocity @. The moment of inertia about this axis is C, that about transverse axis is A. The gyroscope suspension floats on a pool of mercury so that the only torque acting on the gyroscope is one constraining its axis to remain in a horizontal plane. If the gyro is placed at the earth’s equator, the angular velocity of the earth being ©, show that the axis of the gyro will oscillate about the north-south direction ; and for amall amp! that @ > Oi appruaimation. 17. The surface of a sphere is vibrating slowly in such a way that the prin- cipal moments of inertia are harmonic functions of time: amr (1 - eget), ip - BML | eco at), Tes = Ty = 510 ‘MEcHANICS—PROBLEMS where e <1. The sphere is simultaneously rotating with angular velocity Q(t). Show that the z-component of Q remains approximately constant. Show also that Q(t) precesses around z with a precession frequency w» — (3eQ,/2) cos wt provided 2, > w. 18. Three rigid spheres are connected by light, flexible rods with relative masses as shown below: O-0-2 = ny = Describe all the normal modes of the system and state whatever you can about the relative frequencies. 19. A rigid uniform bar of mass M and length Lis supported in equilibrium in a horizontal position by two massless springs attached one at each end. The springs have the same force constant k. The motion of the center of gravity ia constrained ta move parallel ta the vertical X axis. Find the normal modes and frequencies of vibration of the system, if the motion is constrained to the XZ-plane. 20. A particle of mass _M hangs from one end of a uniform string of mass m and length L; the other end of the string is fixed. The particle is given a small lateral displacement 8 and released from rest. Set up the differential equations and boundary conditions to determine the motion of string and particle. Set up a transcendental equation that determines the natural frequencies, and solve the equation for the case m << M. 21. Set up a variational principle for the frequency @ of a membrane with surface tension 7’, of mass o per unit area, and with fixed edges; that is, find an integral over the area of the membrane, of which the extreme value is 22. If a watch is moved to a high altitude, does it run fast or slow? 28. A mass m is attached to a weightless string of length Z, cross section 8, and tensile strength T. The mass is suddenly released from a point near the fixed ond of the string. How small should the Young's modulus, Y, of tho string be, in order that it not break?‘Mxcuanic8s—PRosLems iL 24, A train of mass M, moving with velocity », is to be stopped with a coil-spring buffer of uncompressed length 7, and spring constant ky, which remains constant until the spring is fully compressed. At this point 1X l the spring constant & suddenly becomes very much greater than ky. Assuming a free choice of ke, What is the minimum value of J; if the absolute value of the maximum deceleration is not to exceed diux? 25. (a) A cylinder of radius R, length h and density p floats upright in a uid of density po. Tit is given » swell duwuward displacement uf amplitude 2, find the ciroular irequency @ of the resulting (undamped) harmonic mo- tion. (b) Show that for small oscillations, the motion of the fluid near the oscil- lating cylinder extends for a distance 8 ~ »/y/poa from the edge of the cylinder. The maximum gradient of veiociiy near ihe cylinder is thus AY [dr ~ aa. Neglecting the friction at the bottom of the eylinder, show that the maximum viscous retarding force on the jer i , cylinder i = Wy t 26. A liquid film of surface tension 7 is stretched ! | between two ciroular loops of radius a as shown. pep 28 Find the equation r(z). For what ratio (dja) is the : | cunfigurativn indivaied in the figure stable = Ay 27. A straight vertical strut, having length and T @ square cross section with side a, is firmly fixed to ~~ the gronnd Show that. the ma: out b wr YVi49?, shore ¥ is Young’s mod: without bonding is given by W = xtat¥/48!, where ¥ is Young’s modulus for the material of the strut. mum weight. it. ean carry on the free end 28. A rectangular beam with cross section (a x a) and length L has one ieal brick wall. Calenlate the deflection of ita free end due to ita own weight. The density is p and the Young’s modulus is Y. Assume small bending. and anchored in a ver 29. A thin uniform chimney is pivoted at its low end. Show that a section through the chimney at any point undergoes a flexion stress, and calculate the most probable point of rupture as the chimney falls. 20. The free surface of » liquid is one of constant pressure. If an imcom- pressible fluid is placed in @ cylindrical vessel and the whole rotated with constant angular velocity w, show that the free surface becomes a paraboloid of revolution.12 MEcHANIcs—PROBLEMS 81. An aircraft hangar of semicylindrical shape (with length L and radius R) is exposed to wind directly perpendicular to its axis at infinity with = —vAlr + Rtn) 008 8, L=10m;R=10m;v= 2 km/hr; air density = 1.2 kg/m*. 32, An air mass of 7’ = 280° is separated by a horizontal plane from an air mass at 7’ = 300°K, lying above it. Assume the presence of gravitational waves of wavelength X and small amplitude, causing a sinusoidal wave on the interface. Find the velocity of the wave as a function of the wavelength, assuming the interface is far from other horizontai interfaces. Treat the oscillations of the air masses as incompressible, 33. Two perpendicular semi-infinite walls, OA and OB in the diagram, t at the origin 0, and block the two-dimensional hydrodynami respective distances from the earth, What is the ratio of the tides induced by these two bodies at the equator? 35. Find the fundamental period of oscillation of an isolated mass of in- of the earth (63 under its own gravitational attraction. Assume the velocity flow is irrote- tional. 36. The coordinate systems S, and S; move along the z-axis of a reference coordinate frame S, with velocities v, and v, respectively, referred to S. The time measured in 8 for the hand of a clock in S, to go around onee, is t. What in the time interval f; measured in §, for the hand to go around?MEcHANICS—PROBLEMS 13 37. A rocket is shot out from the earth into interstellar space. Except for a short time in the beginning, the acceleration of the rocket, as measured by the passengers, is constant. The rocket has been aimed at a star a fixed distance from the carth, and moves on a straight line. According to clocks inside the rocket, how long will it take to get to the star? Denote the con- stant distance and acceleration by D and a’ respectively. 38. A particle of rest mass m moves on the z-axis of a Galilean frame of co rate of change of momentum) ma*z. Tv performs osciliations of amplitude a. Express the period of this relativistic oscillator in terms of a definite integral, and obtain an approximate value for this integral. 39. Antiprotana are captnred at rest in denterinm, giving rise ta the reac. tion p + D —+n + 7° (In this problem we ignore other possibilities.). Deter- mine the ° total energy. The rest masses are M(p) = M(p) = 938.2 MeV, M(D) = 1875.5 MeV, M(n) = 939.5 MeV, M(x*) = 135.0 MeV. 40. A positron (energy Z,, momentum p,) and an electron (energy E_, momentum p_) are produced in a pair-creation process. (a) What is the velocity of the frame in which the pair has zero momentum (barycentrie frame)? (b) Deduce the energy either particle has in this frame, and (c) give an expression for the magnitude of the relative velocity between the particles, i.e. the velocity of one particle as seen by an observer attached to the other. 41, A iast (extremely relativistic) eiectron enters a condenser at an angie @ as shown in the sketch. V is the voltage across the condenser and d is the distance between plates. Give an equation for the path of the electron in the condenser. eer 4 pe 42, The neutral 2° meson, of rest mass M, decays into two photons. The angular distribution of these +y-rays is isotropic in the rest system of the x°, If in the laboratory the 7° travels with velocity » in the z-direction, what in the probability P(6)dQ that a photon is emitted in the solid angle dQ about 4, when the meson decays in flight? Herc 6 is the angle as measured in the Inboratory with rexpect to the z-axis, and » may be comparable to the speed of light.14 MgcHANIcs—PROBLEMS 43. (a) If neutrons from a cosmic-ray interaction one light-year from the earth were to reach here with a probability of 1/e or greater, what must their minimum energy be? (b) If they then decay, what is the maximum angle to the fight path at which their decay electrons could be produced? (c) What is the maximum angle for the decay neutrinos? (d) At the angle calculated in (c), what is the maximum energy of the neutrino? 44. A precession of the perihelion of planetary trajectories has been derived relativity predicts such an effect because of tie dependence of inertial mass on velocity. Derive a formula for the special-relativistic precession for a planet of given angular momentum L, rest mass m, and energy H, moving in the gravitational potential of the sun. [Hint: Use polar coordinates u = 1/r i 8 the time explicitly.] 45. A helium-filled balloon floate inside a closed container filled with air at STP, in interstellar space. The container accelerates in a given direction, with acceleration equal to that due to gravity at the aurface of the earth. Which way does the balloon move relative to the acceleration?ELECTROMAGNETISM 1. The edges of a tube consist of equal resistors of resistance R, which are joined at the corners. Let a battery be connected to two opposite corners of a face of the cube. What is the effective resistance? 2, A rectangular wire mesh of infinite TT extent in a plane has 1 A of current fed =|} J } | J into it at a point A, as in the diagram, Je_fa and 1 A taken from it at point C. Find the current in the wire AC. %. Given two iron bars, identical in ap- $+#—+- pearance, one magnetized, the other not. Tell how to distinguish them without using external magnetic fields. (You are allowed to measure forces.) 4. A conducior is charged by repeated coniacis with a meiai piate which, after each contact, is recharged to a quantity of charge Q. If q is the charge of the conductor after the first operation, what is the ultimate charge on the conductor? 5. A variable capacitor is connected to a battery of emf E. The capacitor initially has a capacitance Cy and charge qo. The capacitance is caused to change with time so that the current J is constant. Calonlate the power applied hy the hattery, and compare it with the time rate-of-change of the onergy stored in the capacitor. Account for any difference. 6. When a capacitor is immersed in a medium having conductivity g, a resistance R is measured between the terminals. Show that, regardless of 1616 ELECTROMAGNETISM—PROBLEMS the geometry of the plates, RC = e/g, where ¢ is the dielectric constant of the medium and C is the capacitance in the medium. 7. An eccentric hole of radius a is bored parallel fo _ a (he to the axis of a right circular cylinder of radius 6 (b> a). The two axes are at a distance d apart. A current of I amperes flows in the cylinder. What is the magnetic ficld at the center of the hoie? a7 Eis =22=a2e 8. A conductor has the shape of an infinite conducting plane except for a hemispherical boss of radius a. A charge ¢ is placed above the center of the boss at a disiance p irom the piane. Caicuiaie the force on the charge. 9. Consider the thick hemispherical shell of inner radius a and outer radius h, symmetry (2-axis of diagram), show that a small compass needle placed at the origin will swing freely. 10. A thin uniform metal disk lies on an infinite conducting plane. A uniform gravi- tational field is oriented normal to the plane. Initially the disk and plane are uncharged: charge is slowly added. plate? 11. Calculate the capacity C of a spherical condenser of inner radius R, and outer radius Ry, which is filled with a dielectric varying as € =e +e, cos" 6, where @ is the polar angle. 12. A long straight wire carrying current J is placed a distance a above a semi-infinite magnetic medium of permeability y. Calculate the force per unit length acting on the wire; be sure to specify the direction of the force. 18. The electrical self-inductance of a circular loop of thin wire (such that the flux within the wire iteclf can bo neglected) is measured in two differentELECTROMAGNETISM—PROBLEMS = 17 surroundings: (A) The plane of the circle is placed in the X¥-plane and all space below this plane is filled with a medium of permeability » = 2. All space above the X Y-plane is evacuated; (B) » = 1 everywhere. What is the ratio of the self-inductance L in configuration (A) to that in configuration (B)? )—___,~__] banasensssag- | { L cy 14. There is a small inclusion of conductivity 0; inside a metal of conduc- tivity a). The inclusion perturbs an otherwise constant electric-field. Find the distance dependence of this perturbation far away from the inclusion. (Treat the problem only in the steady state.) 15. A long hollow cylindrical conductor of radius @ is divided into two parts by @ plane through the axis, and the parts are separated by a small interval. If the two parts are kept at potentials V, and V;, show that the potential at any point within the cylinder is given by (Vi + Vs) 7 Va) 4. (Ki — Tr a tno ve +2 SE Ge ) cos (2 — 18, where r is the distance trom the axis of the cylinder, as shown in the figure. 16. Determine the manner in which an initial charge density at any point inside a conductor will decrease with time. Approximately how long would pear? Comment on the validity of your solution. (Copper has a resistivity of 1.7 X 10-* ohm-em.) If the conductor is completely insulated, where does the charge go? 17. A small sphere of polarizability @ and radius a is placed at a great distance from a conducting sphere of radius 6, which is maintained at a potential V. Find an approximate expression for the force on the diclectrie aphere valid for a< r. 18. Derive the Clausius-Mosotti relation connecting the dielectric constant ¢ with the polarizability @ of a medium.18 ELECTROMAGNETISM—PROBLEMS: 19. Ina simple cubic lattice, the lattice constant is 2.00 A and the refractive index (say for sodium light) is n = 2.07. Suppose that the medium is sub- jected to a stress leading to a 2% clongation along one cube edge and a 1% contraction along the other two cube edges. Calculate the refractive index of the strained medium when the electric vector E is (a) parallel to and (b) normal to the principal strain axis. Consider the atomic polarizability @ to be a constant scalar quantity. {Hints: The local ficld acting on an atom in the medium described above can be found as foiiows. imagine a spherivai cavity about the atom, enclosing the six nearest neighbor atoms. Uutside the cavity, the medium may be regarded as continuous and isotropic. The local field at the center of the cavity may be expressed as te -E+PiSey = 1 where E is the applied field, E’ is the contribution from the polarized con- tinuum outside the spherical cavity, and EY is the contribution from the dipole induced in the ith atom within the cavity. In an anisotropic medium, = plied E.] RY! does not. vanish; moreover it. dependa upon the direction of the ap. 20. What is the critical angle for total external reflection for high-energy x-rays of wave length A, falling on a metal plate in which all N electrons per unit volume are essentially “free”? 21. The ionosphere can be considered as an ionized medium containing N essentially free clectrons per unit volume. Show that if a linearly polarized wave propagaies in the iuuusphere in a direciiun paraliel to thai of dhe small uniform magnetic field H produced by the earth, its plane of polarization will be rotated through an angle proportional to the distance traveled by the wave. Calculate the constant of proportionality. in a hollow metal pipe Sof rectangular cross section with perfectly conducting walls. What are the phase and group velocities? Show that there is a cutoff frequency below which no waves are propagated. 28. Inside a superconductor. instead of Ohm’s Law (J = cE). we assume London’s equations to be valid for the current density J: E == a coul(d)=—B, faa (in Gaussian units), and regard 2 as a constant. Otherwise, Maxwell's equa- tiona (with e =: 1, 4 = 1) and the corresponding boundary conditions are unohanavd.ELECTROMAGNETISM—PROBLEMS = 19. Consider an infinite superconducting slab of thickness 2d(—d d and z < —d), with E = D = 0 everywhere. If surface currents and charges are absent, compute H and J inside the slab. 24, Polarized light is passed parallel to the axis through a solid glass cyl- inder of length L which is rotating with an angular velocity Q around its axis, Find the rotation of the plane of polarization. (Assume constant indox of refraction n and permeability 1.) g 25. The slit lens in the figure above, containing an aperture long in compari- son with its width yo, separates a region in which the electric field is H, from a region in which the electric field is B;. A beam of charged particles focus- ing at a distance z, to the left of the aperture is refocused at a distance a, to the right of the aperture. If V, is the voltage through which the particles were accelerated before reaching the lens, show that d= 8). a my 2K, Use the approximations V, > Byx, and Bry, and 2, and 2, > yo 26. An ion moves in a helical path around the axis of a long solenoid wound so that the ion encounters # region in which the field intensity increases gradually from B, to B,. Under what circumstances will the ion be reflected? 27. The accompanying figure shows a section through the cylindrical plate (radius b) and filament (radius a) of a magnetron. The filament is grounded, the plate is at V volte positive, and » uniform mag- netic field H is directed along the axis of the cylinder. jectrons leave the filament with zero velocity and travel in curved paths toward the plate. Below what lovel of V will the current be suppressed by the field H7?20 ELECTROMAGNETISM—PROBLEMS: 28. The magnetic quadrupole shown in the figure hes the property of fo- cusing a beam of charged particles traveling nearly parallel to the z-axis. The focusing is in the x = 0 or y = 0 planes, depending on the sign of the charge of the particles. Find the simplest distribution of magnotic “poles” which focuses, in a similar ~~ LS way, uncharged particles with a mag- G S netic moment p polarized parallel (or antiparallel) to the z-axis. D 29. A well-collimated beam of protons moves through space in the form of a cylinder of radius R. The velocity of the protons is » and the number Per unit volume is p. Find the forces on a proton at a radius r from the beam ry of the heam 30. Set up the nonrelativistic equation of motion of an electrically charged particle about fixed magnetic monopole of strength 1. Find the constants of motion. 81. A standard method for calibrating the orbits of charged particles of momentum p in static magnetic fields is to empirically determine the con- figuration in those fields assumed by a perfectly flexible wire carrying current I and under tension 7. Derive the physical basis of this method. [Hint: Derive the general differential equations for: (a) a particle orhit, dér/ds? = 2, and (b) the equilibrium configuration of a current carrying wire.] 32. Given an atom with spherically symmetrical charge distribution in an external magnetic field H, show that the field at the nucleus caused by the diamagnetic current is AH = —(eH/3me*) $(0) where (0) is the electrostatic potential produced at the nucleus by the atomic electrons. Estimate very roughly the magnitude of AH/H for an atom with atomic number Z = 50. 33. A spherically symmetric charge distribution of finite extent pulsates radially with some frequency @. How could one detect these radial pulsa- tions? Explain your answer. 34. A flywheel of radius X, with charge @ uniformly distributed along the rim, rotates with angular velocity w. What is the rate at which cnergy is radiated by the nyxtem?ELECTROMAGNETISM—PROBLEMS 21 35. Show that in the collision of nonrelativistic, spinless, identical particles, the emission of electric and magnetic dipole radiation does not occur, ae- cording to classical radiation theory. 36. A linearly polarized plane wave of electromagnetic radiation is incident upon an atom of polarizability @. According to classical electromagnetic theory, what is the electric field of the seattered wave at large distances? What is the total scattering cross section? 37. A thin copper ring rotates about an axis perpendicular to a uniform magnetic field Ho. (See figure.) Its initial frequency of rotation is @». Cal- culate the time it takes the frequency to decrease to 1/¢ of its original value a under ‘ire assumpiion thai ine energy goes into Joule heat. (Copper has conductivity ¢ = 5 x 10" ops, and HH, 4 density p=8.9 gmjem*. Hy=2006.) aris of rotation 38. A particle of mass m aud charge ¢ iv suspended on a siring uf iengils Z. ‘Ata distance d under the point of suspension there is an infinite plane con- ductor. Compute the frequency of the pendulum if the amplitude is suffi- ciently amall that Hook's Law is valid. Compnte how much energy the mese point will loee by ra plitude a. 39. Seven antennas, radiating as electric dipoles polarized along the z- direction are Placed along the lane. at coordinates z = 0. a excited in ie! {a) Calculate the angular distribution of the radiated power as a function of the polar angle 6 and the azimuthal angle $ (neglect constant multiplying factora) (b) Make » rough diagram of radiated power vs. angle ¢ in the zy-plane. (c) Consider the direction in which the radiated intensity is a maximum for (1) this array, and (2) for a single antenna by itself. How do these inten- sities compare? 40. Show that the energy loss per revolution of a singly charged particle due to radiation is ae to: L=5 ae 7 ind re = e2/mec. a a a whero R is the radius of the orbit « i pal :22 ELECTROMAGNETISM—PROBLEMS: 41. A light beam of intensity J, and frequency v» directed along the posi- tive z-axis is reflected normally by a perfect mirror moving along the positive zaxis with a velocity vp. What is the intensity I and frequency v of the reflected light in terms of J, and v9? 42, Two thin, parallel, infinitely long, nonconducting rods, a distance a apart, with identical constant charge density \ per unit length in their rest frame, move with a velocity e, not necessarily small compared to the of reference that is at rest, and in a frame of reference moving with the rods, and compare the results. 43. Consider an clectron moving in a time-dependent axially symmetric magnetic ficld with By = 0. Given the Lagrangian pens ee ao Laame{i- 5) 42a Bavxa) what conditions must be satisfied in order to have a circular orbit whose position and radius are constant in time? What is the angular frequency and energy of the electron in this orbit? Investigate the stability of the cirenlar orbits. Assume that the shape of the field near the orbit may be represented by B, = By(ro/r)", where B, is the instantancous value of the field at the equilibrium orbit r = 19; the z-axis is the axis of symmetry, n is positive, and B(r, z, t) = B(r, 2)T(t). Assume that the time variations of the external field are small in the time required for a single revolution. (a) Show that if m > 1, the orbit is unstable against radial oscillations. (b) Show that the sum of the squares of the radial and vertical oscillation frequencies is equal to the square of the equilibrium orbit frequency. 44. Find a covariant generalization for (a) the Lorentz force equation F = eE + v x B]; and (b) the equation of motion for a particle with spin 8, WS _ vg yp 2m where y in the gysummguetiv rativ,ELECTRONICS 1, A vacuum tube is used in @ conventional manner in combination with an LC circuit as an rf generator. If the distances between the electrodes are of the order of magnitude of 1 mm and the plate voltage is 200 V. what is the order of magnitude of the frequency above which the circuit will fail to operate as a generator? : 2. Show that the infinite chain of inductances L and capacitors C indi- cated in the figure below acts as a low-pass filter. Calculate the cutoff frequency @, in terms of w» = 1//TC. T tT tT IT JT 7. JT 3, A waveform produced by an electronic circuit is to be attenuated with 1 (uaaimum baud width). Using the high- impedance attenuation circuit shown below, draw in any additional com- ponents required to fulfill the above requirements, and determine the values of the required components.24 Exgcrronics—Prosiems 4. The triode in the tuned plate oscillator shown below has plate amplifi- cation factor and plate resistance R,. At what frequency is oscillation expected? Under what conditions will the circuit fail to oscillate? [Iint: Assume that the grid current is zero.] Cu 5. The diode circuit shown in part (a) of the accompanying figure is sub- jected to an input wave form at A shown in part (b) of the figure. (e) ) lly uncharged. plot the voltage at re perfe ‘Assuming that the capacitors are i Dow What is the limiting voltage at point B?OPTICS 1, A thin lens with index of refraction n and radii of curvature R, and R, is located at the interface between two media with indices of refraction n, and nz, as shown in the figure below. If S, and S, are the object and image distance respectively, and f, and f, the respective focal lengths, show that (f:/S:) + (f:/S2) = 1. ™ yt Ld 2. Light is normally incident on the interface between two media labeled 1 and 2, respectively. A wave of unit amplitude in medium 1, incident on the interface, is found to have reflected and transmitted amplitudes r and ¢ respectively. Similariy when the wave is incident in medium 2, the reflevied and transmitted amplitudes are found to be r’ and ¢’. Using the superposi- tion principle and time-reversal invariance, derive the Stokes’ relations e4i! =landr = —1, vy aA a plane surface film of uniform thickness d covering a semi- Assume that 4 = 1 for both media, and that the film and substrate have indices of refraction m, and m,. Find an expression for the wave reflected into the vacuum in terms of m,. m. and the vacuum wavelength X. Under 4. A method of color photography was invented by Lippmann in 1881. The photographic plate consists of an extremely fine-grain emulsion on a glass plate which is covered on the emulsion side with a layer of mercury to form a reflecting surface. It is exposed in a camera with the glass side toward the light. After development and with the layer of mercury in place, 2526 Oprics—PRoBLEMS the plate is now illuminated with white light and viewed by reflection. A faithfully colored picture of the image is seen. Explain what is happening. Light | Glass 5. Ina pin-hole camera the distance of the pin hole from the photographic plate is 10 cm. You want to take a picture of the sun in the visible spectrum (A = 5000 A). What diameter of the pin hole should you use in order to obtain the sharpest resolution? 6. A transmission grating 5 om wide is used to analyze, at normal incidence, the spectrum of sodium. Determine the minimum number of lines needed to resoive the sodium D doubiet at A anc A respectiveiy, in the first order. In these conditions, what will be the angular separation of the two members of the doublet? 7. A microwave detector is located at the shore of a lake 0.5 m above the wavelength rises slowly above the horizon, the detector indicates successive maxima and minima of signal intensity. At what angle @ above the horizon is the radio star when the first maximum is received? 8. An opaque screen has two very narrow parallel slita a distance d apart. The screen is illuminated by a long, straight, incandescent metal ribbon of small width w and lucaled « distance Z in front of the diagram screen. A colored-glass filter placed in front of the screen transmits light of wave- length d. The transmitted light is made to interfere on an observing screen placed a large distance behind the opaque screen. As the distance d between Ribbon Opaque I ELF 7 RB. ObservationOprics—PROBLEMS 27 the slits is increased, it is found that for a distance d = dy, the interference fringes on the observing screen disappear. What is the width w of the ribbon? 9. A diffraction grating is made up of N slits each half as long as the previous one. The siits are separated by a distance d. What is the anguiar distrivution of the intensity for light of wavelength \? 10. Consider a reflection grating whose grooves, though equidistant, have nd to some small value much less than 1. 11. A black screen with a circular opening of radius a is located in the ay- plane, with center at the origin. It is irradiated by a plane wave ap = exp (ike), k= 2x/r. Determine the approximate zeros of intensity on the positive z-axis for za. 12. Light is passed through an array of perfect. polarizers. The polariza- tion planes are approximately lined up in a fixed direction, but there is a random error between two neighboring planes with a Gaussian distribution B exp (—a6!) where @ is the relative angular deviation. Find the average attenuation coefficient of the system per polarizer for a beam of light after passing the first polarizer. Assume a >> 1. 18. A plane monochromatic wave is incident upon an imperfect. linear diffraction grating consisting of NV parallel to the grating. The equilibrium positions of the apertures correspond to a perfect linear grating with grating spacing d. The time needed to photograph the diffraction pattern is very long com- pared to the periods of oscillation involved. The probability distribution the root-mean-square displacement being the same for all apertures. Show that the intensity distribution (i.e. intensity as a function of angle between direction of incidence and direction of observation) in the Fraunhofer diffraction pattern of such an imperfect grating can be expressed I= $l + N(Q~ d)ie. 1, is the intensity distribution for the corresponding perfect grating formed by the apertures when they are in their equilibrium positions; i, is the in- tensity distribution in the diffraction pattern for single aperture. Express ¢ in terms of the rma displacement.28 © Oprics—PRoBLEMS jrlicep, 14. Light from a source of frequency f is led through the system shown. Tf the npper eandnit. carries a liquid having index of refraction m moving with velocity u, and the lower conduit contains the same liquid at rest, what is the minimum value of u that will cause destructive interference at P’? fed] foamy fp ts 4 Pal A iow Ci G C y —_— U ?, 15. In order to observe the sun in monochromatic —_ Optic axis light, the French astronomer B. Lyot invented the eegete [45° birefringent filter, consisting of a series of bire- & twice as thick as the previous one. Polarizing films Plane of palorization (P) are mounted between the crystals and at each Fig. 2 end, (See diagram 1.) All the orystals are mounted with their optical axes parallel and at wn of the light. The polarizatio: also alll parallel and at 45° to the direction of the optic axes. (Sce diagram 2.) Ouly certain bands of light are able to penetrate Ube filler. For » filter containing s elements, calculate the transmission as a function of wavelength 2. Also find the width AA of the bands that can pass the filter, and the wavelength separation between such bands. ‘ht angles to the direction ofQUANTUM MECHANICS 1. Let Band C be two anticommuting quantum-mechanical operators, i.e. {B, C}, = BC + CB =0. Let wp be an eigenstate of both B and C. What can be said about the corre- sponding eigenvalues? For B = baryon number andC = charge conjugation, the relations {B,C}, = 0 and C? = 1 hold. What does your result imply in this ease? 2. Three matrices M,, M,, M,, each with 256 rows and columns, are known to obey the commutation rules [M;, M,] = iM, (with cyclic permutations of 2, y, 2). The eigenvalues of one matrix, say M,, are +2, each once; +3 each 8 times; +1, each 28 times; +4, each 56 times; 0,70 times. State the 256 eigenvalues of the matrix M? = M2 + M3 + M2. 3. Find the eigenvalues of the matrix = a. LE. eT. ik = 1.2.3, D=(r x pl). 4. If one considers systems capable of emitting particles of half-integral spin, one encounters operators U obeying the commutation relations (1) [0,3] = @ (FF = fue + Fu) +, where J is the angular momentum of the emitting system. Find selection Q) and 2), § J,and J? ge (eigenvalues m and j(j + 1), respectively). In other words, what matrix elements (m’j’| U |mj) can be nonzero? [Hint: Let X, = j(j+1).] 5. Prove that all the wave functions belonging to the maximum eigenvalue of the square of the total spin operator of a system of N clectrons are sym- metric in the apin coordinates of the individual electrons. 2030 QuaNTum MzcuAaNice—PRoBLEMS 6. Prove the Thomas-Reiche-Kuhn sum rule meet, — 2, 7. Show that the source-free Maxwell equations may be put into the “Dirac form’ through the introduction of the Kramers vector F = E + iH, F* = E — iH, nd the definition of suitable matrices 8. Use this representation of the above, p is s momentum operator. 8. Use the Bohr-Sommerfeld quantization rule to calculate the allowed energy levels of a ball which is bouncing elastically in a vertical direction. 9. A three-dimensional isotropic harmonic oscillator has the energy eigen- values eo(n + $) where n = 0,1,2...What is tho degroc of degencracy of the quantum state n? ‘ 10. ‘Three mass points of mass m are confined to move on a circle of radius r. Their mutual distances are fixed and equal so that they form an equilateral triangle. The three mass pointa obey Bose statistics and have no spin. Discuss the rotational cnorgy levels of the system. Bi. Derive the dipule-dipule maguviic interaciion energy: of » pruion aud an antiproton at a fixed distance a, in eigenstates of total spin, in terms of the proton magnetic moment jig. Two magnetic dipoles have the interaction energy vai r {s-ng — alacniern 12, Solve and classify the eigenvalues of the Hamiltonian Hi = Alo? + of) + Boo, where @,, @; are the Pauli spin matrices for the particles (1) and (2) respec- tively. (The Pauli principle is not considered.) 18. A system consists of two distinguishable particles, each with intrinsic spin }. The spin-spin interaction of the particles is Jo, +, where J is ® oon- stant. An external magnetic field H is applied. The magnetic momenta ofQuanruM MECHANICS—PROBLEMS 31 the two particles are ae, and Ag,. Find the exact energy eigenvalues of this eystem. 14. An electron is contained inside a sphere of radius R. What is the pressure P exerted on the surface of the sphere, if the electron is in (a) the iowest S state? (b) the lowest P state? 15. A particle of mass m moves in a potential V(r) = —V». when r a. Find the least valne of V, auch th 16. Consider the one-dimensional Schrédinger equation with vin = [eer for x > 0, eo forz <0. Find the energy eigenvalues. netic flux through the electron orbit is quantized. Neglect electron spin. Show also how knowledge of the energy levels found here nonrelativistically may be used to determine the relativistic corrections to the energies. 18. A quantum-mechanical system in the absence of perturbations can exist in either of two states 1 or 2 with energies E, or Ey. Suppose that it is acted upon by a time-independent perturbation 10 Vu vel, ooh Wa 07 where V,, = V8. If at time t = 0, the avatem is in state 1. determine the am. plitudes for finding the system in either state at any later time. Use the particle in the potential V=oo fore <0, V=u for z > 0. Take 2-9 as the trial function 20. An clectron with charge ¢ and mass m is confined to move on a circle of radius r. It is perturbed by uniform electric field F parallel to one of the diameters of the circle. Find the perturbation of the energy levels up to torms of the order of F*. Notice in particular the anomalous behavior of tho firut excited state.32 QUANTUM MECHANICS—PROBLEMS 21. In the Stark effect on the ground state of an atom, if the applied electric field F is amall, the energy of the state is perturbed by an amount propor- tional to the square of the applied ficld F. Thus AZ = —}a¢ F* where a is the polarizability of the atom. Obtain an expression for the polarizability of the hydrogen atom in its ground state by using perturbation theory. Evaluate the expression in an approximate way which gives an upper and lower limit to the polarizability, thus showing that 4a’ < a < (16/3)a’, where @ is the Bohr radius. 22, Two ideniicai particies of spin } obey Fermi siatistics. They are confined in a cubical box whose sides are 10~* cm in length. There exists an attractive potential between pairs of particles of strength 10-* eV, acting whenever the (Take the mass of the individual particles to be the mass of the electron.) 23, Consider an atom with one 2p electron placed in an electric field with orthorhombic symmetry. The potential V of the field is V = Aa* + By? — (A + B)2', Show that the expectation value of L. is zero. Neglect: electron spin. You may assume that V is small compared to the strength of the atomic potential. 24, Two identical plane rotators with coordinates 6,, 6, are coupled accord- ing to the Hamiltonian H = A(pi, + Pi,) — B cos (6, — 62), where A and B are positive constants, (Note: 6, + 27 is equivalent to 61.) From the Schrédinger equation determine the energy eigenvalues and eigen- functions when the following conditions hold: (a) Tn the ease R< AR, disensaing only terms linear in R. Wateh ont. for dogencracies. (b) In the case B > Ah’, by reducing the problem to an oscillator problem (small oscillations). 25. A particle of mass m moves in a two-dimensional potential well Determine the expectation values of the coordinate operators 2, y for the ground state when a small perturbation V — Fix + Fry (Fi, Fs are con- stanta) is applied. Consider terms only to first order in F, and F;. You need not compute the final matrix clements. It is sufficient to express them in integral form and state explicitly which ones are not zero.QUANTUM McHANIcs—PROBLEMS 33 26. Consider two identical linear oscillators with spring constant k. The interaction potential is given by H’ — cx,r, where 2, and z, are the oscillator variables. (a) Find the exact energy levels. (b) Assume c < kand compute the lowest pair of excited states in first-order perturbation theory. (Give energy levels in first order and eigenfunctions in zeroth order.) 27. A two-dimeusivual useiliaiur has the Hamilivnian Hadi t pp that bone + 9", where h land § < 1. Give the wave functions for the three lowest energy §=0: of thes: te the Srst-order pert storder port 840. 28. Consider an electron in a uniform magnetic field in the positive z-direc- tion. The result of a measurement has shown that the electron spin is along the positive x-direction at.t = 0. For t > 0 compnte quantum-mechanically the probability for finding the electron in the state (a) S: = 4, (b) S: = —}, and (c) 8, 29. Tritium (the isotope H*) undergoes spontaneous beta decay, emitting an electron of maximum energy about 17 keV. The nucleus remaining is He’. Calculate the probability that an electron of this ion is left in a quantum state of principal quantum number 2. Neglect nuclear recoil, and assume that the tritium atom was initially in its ground state. 30. An atom with J = 4, m, = § is in a uniform magnetic fieid. Suddeniy the field is rotated by ¢ = 60°. Find the probability that the atom is in the sublevels m, = +4 or m, = —4, relative to the new field, immediately after the change in field. 81. What is the physical basis for the selection rule that 2 transition from one state of angular momentum zero to another state of angular momentum zero by emitting a photon, is forbidden? Is there any way in which a zero- zero transition could be made by the emission of light? What is the physical haaia for the fact that a radiative transition requiring a large apin change proceeds slowly? 32. Show that the photoelectric absorption cross section of an atom, for the ejection of an electron from the K-shell, goes like Z* for photon energies large compared with the binding energy of a K electron. Use simple per- turbation theory, and neglect recoil.34 QUANTUM MECHANICs—PROBLEMS 38. A spherical square well has depth V, and radius a. A particle of positive energy E and mass m is caught in a state of angular momentum L ¥ 0. Estimate the lifetime + of the particle, ignoring the angular momentum inside the well. 34, A quantum-mechanical system is initially in a state of angular momen- tum L, = 0. It decays by electric dipole emission of light, to a lower state of angular momentum Z, ~ 1, This state in turn decays by an electric dipole momentum Z, = 6. Both quania are observed with suitabie detectors. Calculate the probability, W(¢), that the directions of propagation of the two quanta form an angle ¢. Does the result depend on whether the system is an atom or nucleus? [Hint: Use second-order perturbation theory.] 35. Consider the scattering of a particle by a simple cubic structure with lattice spacing d. The interaction with the lattice points is —2naht v= ae — 1). Treat the scattering in Born approximation. Show from your result that the condition for nonvanishing scattering is that the Bragg law be satisfied. 36. The expression Bigs : Ve — Ve) gives the probability that one particle per second will pass through a unit area normal to the direction of J. A beam of particles with uniform velocity v enters a region where some of them are absorbed. This absorption may be represented by the introduction of a constant complex potential V,— iV, into the wave equation. Show that the cross section per atom for absorption is@ = 2V,JANv where N is the number of absorbing atoms per unit volume. 37. A particie is scattered by a compietely absorptive (“biack") sphere of radius @ which is large compared to the de Broglie wavelength X/2x = 1/k. How do the 7, and 6, in the scattering amplitude £8) = 3S (21+ 1Y(y.et — 1)P.losn8) 3 behave as a function of 1? Evaluate the clastic scattcring cross section, the reaction cross section, and the total cross section. 38. Calculate the differential and total cross sections for scattering of a spinless particlo of mass m incident on an infinitely massive, infinitely hard aphoro of radius a. Treat tho caso for which the particle moves sufficiontlyQuantum MEcHANICS—PROBLEMS 36 slowly that the D-wave phase shift is negligible. Give the answer in the form of a polynomial in (ka) and retain only terms in the crose sections which are of lower order than a? (ka)'. You may use the following recursion relation, valid for both regular and irregular solutions: If F, satisfies Pia) + Zr He) + Fes — 25H] =o, then 39. Consider the scattering of a beam of spinless particles by a hard sphere of radius a: V0 forra. (a) Find the total cross section where a A. Show that i in the forward direction the various P,(cos 0) = Jo(n 8) for large n, small 6 Ze asl} = Sale) $0) = HF 2m + Ile — 1)Py (cos 0) potential. Suppose a etream of particles ie incident from the left. Find the Telative intensities of retlected and transmitted beams.THERMODYNAMICS 1. Derive the following relations (the Mazwell equations): (aT\ _ _(aP\ (aT\ _ (av\ (aP\ _ (28\ a \av)s~ “NaS)v? — \OP)s ~ \8S)n? = OT ~ NV)” “ and F a avy _ ; Gr).= (5), Td8 =C,aT + T($7) av; (2) and Gr), = 7(5r), ~?- 2. Calculate the Joule-Thomson coefficient (@U/@V), for a van der Waals gas, where 3. Tn certain phase transitions it is noted that there are no discontinuous changes in enthalpy H or volume V, but rather in their first derivatives with respect to temperature. Derive the two thermodynamic relations which supplant the Clapeyron equation. Let the subscripts 1 and 2 refer to the two phases. 4. For water at 27°C, (1/V)(@V/27)p = 0.00013 deg-'. Find the change of temperature in a large body of water carried by a curront to a depth of 5. Consider a pressure-volume diagram for a given mass of a substance, on which there is a family of adiabatic curves. Prove that no two of these adiabatic curves can interacct. 8. One mole of H,0 in cooled from 25°C to 0°C and frozen. All the heat taken up by the refrigerating machine, operating at maximum theoretical an‘THERMODYNAMICS—PROBLEMS 37 efficiency (no entropy created), is delivered to a second mole of H,0 at 25°C, heating it to 100°C. (a) How many moles of H,O are converted to steam at 100°C? The heat of vaporization 0’ at 100°C is 9730 cal/mole. The heat of fusion 2 at 0°C is 1438 cal/mole. How much work must be done by the re- frigerator? 7. Two identical perfect gases with the same pressure P and the same number of particles NW, but with different tomperatures 7, and 7',, are con- fined in two vessels, of volume ¥, and Vz, which are then connected. Find the change in entropy after the system has reached equilibrium. 8. A gas-tight, frictionless piston of small thermal conductivity slides in a * thermally insulated cylinder. Both compartments A and B contain equal ly the temperature of the gas sis 7, in volume A, and 37, in B. The system is to be considered as in mechanical equilibrium at all times, and eventually it will be in thermal equilibrium as well. (b) What will be the total change in the entropy of the entire system per mole of the gas in it between t = 0 and t = oo? (c) How much useful work could have been done by the system (via a suitable transfer mechanism) por mole of gas if the transfer of heat from one compartment to the other had been accomplished reversibly? 9. If one mole of a monatomic gas undergoes a free expansion, what is the change in temperature in terms of the initial and final volumes and the change in entropy What is the approximate change in the enthelpy? 10. One gram of water at 20°C is forced through an insulated porous plug under a pressure of 10‘ atm into a lab where the pressure is 1 atm. Find the state of the water emerging from the plug. Assume that the density at 10¢ atm is the same as that at 1 atm. The latent heat of evaporation is 540 cal/gm. 11. A vessel contains helium gas at 10°K (above critical point). The vessel is thermally isolated and the gas is allowed to escape slowly through a capil- lary tube until the pressure within the vessel is 1 atm and the temperature is 4.2°K (normal boiling point of helium). If the gas is perfect, find the initial pressure P, for the vessel to be entirely filled with liquid at the end of the process, Data: Latent heat of He at 4.2°K = 20 cal/mole; C, of helium gas = Beal/mole deg.38 THERMODYNAMICS—PROBLEMS 12, A long thin metal bar vibrates in its fundamental mode of longitudinal oscillation. In what range of frequency does the oscillation tend to be iso- thermal? Take Young’s modulus ¥ as 10" dyn/cm’, the density p as 10 gm/co, the heat conductivity & as 1 cal/om-deg-sec, and the specific heat C as 0.1 cal/gm-deg. 13. Two vessels, each of volume V, are connected with a tube of length Land small cross section A (LA < V). Initially, one vessel contains a mix- ture of carhon monoride ata partial preamre P, and nitrogen at. a partial P Py while the other voacel Pr contains nitrogen at a pressure P. Th coefficient of diffusion of CO into N, or [—* Ny into CO is D. Calculate the partial (° v v presanre of CO in the firat, vensel a a J \ function of time. — 7 14. An iron sphere of radius 20 cm is heated to a temperature of 100°C throughout; ite surface is then kept at constant temperature 0°C. Find the temperature of the center 10 min after the cooling has begun. The ratio of thermal conductivity to apecific heat per unit volume is K/Cp = 0.185 in ogs unite. 15. A sphere of radius Ris immersed in an infinite liquid ot temperature 7 tained at that temperature thereafter. The conductivity of the liquid is K, its specific heat C, and its density p. (a) Express the temperature at any point outside the sphere, and at any ime t > 0 ag a defini (b) Find eaplicitly the limiting temperature distribution ast — w. 16. Find the lemperature depeudeuve of the eleetrumaguetiv energy density £ within a cavity with perfectly reflecting walls, using thermodynamic arguments. 17. A spherical satellite of radius r, painted black, travels around the sun in a circular orbit at a distance D from its center (r < D). The sun, a sphere of radius R, radiates as a blackbody at a temperature 7, — 6000°K and subiends an are Za = 32’ as seen from the satelite. What is the equilibrium temperature 7’ of the satellite? 18. The Helmholts free energy A of a ferromagnet can be assumed to depend on the magnetization for no external field as follows ‘ a for temperature 7’ near the Curie temperature 7',; a and f are positive and‘THERMODYNAMICS—PROBLEMS 39 approximately temperature-independent. Ay is the free energy of the un- magnetized state, and may be regarded as approximately temperature independent. Derive the temperature dependence of the average magnetiza- tion M(T) which follows from Eq. (1) if fluctuations are neglected. What is the influence of fluctuations on M(T')? Obtain also the magnetic suscepti- bility above the Curie temperature. 19. The velocity of sound v in a paramagnetic gas is altered by a magnetic m per. mole Hi = Hi/T, io iowest order in yy. The specific heat (for H = 0) may be regarded as temperature independent. 20. According to Meissner, B = 0 in a superconductor. In the normal state, M in negligible. At a fixed temperature T < T., an the external field H in lowered below the critical field H.(7) = Hy [1 — (/T)*], the normal state undergoes a phase transition to the super- conducting state. Show that this viewpoint is correct by finding the difference in Gibbs free energies of the normal and supercon. ducting metal at temperature T < T,, i.e. G,(T, H)— GT, H). At H =0, compute the latent heat of transition ZL from the normal to the superconducting state, and the s discontinuity in the specific heat. Is the phase transition first or second order? The Gibbs function in @ magnetic field is given by @=U—TS— HM. 21. From thermodynamic principles, determine the vapor pressure F, for a very small droplet of liquid of radius r, in terms of the vapor pressure P.. of a large body of the same liquid having a negligible surface-to-volume ratio. [Hint: For finely divided matter the ratio of area to volume inereases greatly, and surface phenomena become dominant.) 22. A rocket weighing 1000 kg is shot into space. Under the assumption that all stellar bodies have an average mass of 10°° kg and move about with a random velocity averaging 10 kg/sec. what is the average velocity the rocket will tes very long time? (Negl ability of the rocket falling into a star.) 28. Consider hydrogen gas at temperature 7 (around liquid nitrogen tem- perature) and pressure P (about 1 cm Hg), with a concentration of ortho- hydrogen equal to z. Derive an expression for (a) the molar specific heat of this gas, dnd (b) the thermal conductivity K of this gas.40 ‘THERMODYNAMICS— PROBLEMS Besides 7, P, x and fundamental constants, your expressions should contain only three parameters characterizing the H, molecule, the molecular mass M, the internuclear separation R, and the collision cross section 7. ‘You may assume that at liquid nitrogen temperature, the number of molec- ules in rotational states (J > 2) is negligibly small. 24, Compute the ratio of thermal conductivity of helium gas at P = 0.1 atm and 300°K to that at P = 0.5 atm and 300°K. Compute also the ratio of 25. Obtain the mean vector velocity of gas molecules escaping through a small hole in an enclosure kept at temperature 7’. Let WV be the number of particles per unit volume of the gas. Collecting shest | I firbfene ——__+ >, F___ Hole 26, A small circular opening of radius a, which is small compared to the mean free path of mercury. is cut in the wall of a very thin-walled rectangular tank containing mercury vapor at a temperature 7’ and very low pressure P. At a distance h above the hole and parallel to the wall of the tank, a collect- ing sheet of metal is placed and cooled so that when any of the mercury atoms strike it, the vapor condenses at once. Derive an expression for the distribu- tion of mereury in gm/cm? on the collecting sheet at time ¢ in terms of polar angle @ between the normal to the hole and the point of collection. See the figure above.STATISTICAL PHYSICS 1. A young man, who lives at location 1° A of the city street plan shown in the figure, walks daily to the home of his >~— fiancee, who lives m blocks cast and n blocks north of A, at location B. Because "=4 he is always anxious to see his fiancee, his 4 route always approaches B, i.c., he never doubies back. In how many different ways | | | | can he go trom A to BY Begs 2, According to the Stefan-Boltzmann law, the energy of Hohlraum (black- hody) radiation depends on the temperature ax TF with K the ordi tegral). Derive the exponent K in the energy- temperature relation for a photon gas enclosed in an NV-dimensional Hohlraum. 3. If the sun behaves as a blackbody at 6000°K with diameter 10° km, what: is its total microwave-emitted power per megacycle bandwidth at Sem? 4. Three particles at the corners of an equilateral triangle cach carry a quantum-mechanical spin }, their mutual spin Hamiltonian being given by =X (0,-0.4 0-0, + 01-01). List the energy levels of this spin system, giving their total spin values and degeneracies. Dednee the partition function. Z. 5. Consider gas contained in volume V at temperature 7. The gas is composed of N distinguishable particles of zero rest mass, #0 that energy £ and momentum p of the particle are related by £ = pe. The number of single-particle energy states in the range p to (p + dp) is 4xVp*dp[h*. Find the equation of state and the internal energy of the gas and compare with an ordinary gas. 4142 STATISTICAL PHYsics—PROBLEMS 6. 10" weakly interacting spinless particles, each with the mass of the electron, are identical in appearance but obey classical statistics. They are confined in a cubical box which is 10-* om on an edge. Each particle under- goes a potential interaction with the box which is of two sorts. One is at- tractive and leads to a bound state well localized near the center of the box and having an energy of —1 eV. The other interaction is a strong re- pulsion which prevents the particle from escaping through the walls of the box. Find at what temperature the pressure in the box is one atm. 7. A surface has o temperature of 800°C. Atoms of sodi which strike the surface are found to be 99% ionized when they evaporate. Atoms of chlorine are found to be ionized negatively by one part in 10° when they evaporate from the same surface. What is the electron affinity of chlorine? Th 8. Helium atoms can be adsorbed on the surface of a metal, an amount of work ¢ then being necessary to remove a helium atom from the metal surface to infinity. The helium atoms are completely free to move, without mutual interaction, on the two-dimensional metal surface. If such a metal surface is in contact with helium gas at a pressure P, and the whole system is in equilibrium at temperature 7’, what is the mean number of atoms adsorbed per unit area of the metal surface? Express your answer in terms of quan- tities given in the problem, and fundamental constants. 9. An LC circuit is used as a thermometer by measuring the noise voltage across an inductor and capacitor in parallel. Find the relation between the rms noise voltage and the absolute temperature 7. mA nucleus can therefore be in any “Of three quantum states labeled by the quantum number m, where m= 0, +1. Because of electric interactions with internal fields in the solid, a nucleus in the state m = 1 or in the state m = —1 has the same energy ¢ > 0, while its energy in the state m = 0 is zero. Derive an expression for the entropy of the N nuclei as a function of the temperature 7, and an expression for the heat capacity in the limit e/kT <1. ii. Consider a system of three-dimensionai rotvators (with two degrees of freedom and no translational motion) in thermal equilibrium according to Boltzmann statistics; take account of the quantization of energy. Calculate the free energy, entropy, energy and heat capacity (per rotator) in the case of high temperatures, making use of Euler's approximation formula: BMF) free) AIO seo)SratisTicaL PHYsios—PROBLEMS 43 12. The average energy of a system in thermal equilibrium is (E>. Prove that the mean square deviation of the energy from (E), <(E — (H))*) is given by kB — KB) = ETC,, where C, is the heat capacity of the entire system at constant volume. Use this result to show that the energy of a macroscopic system may ordinarily be considered constant when the system is in thermal equilibrium An ascom Pi Fs Only nearest neighbors interact. When the spins of the neighbors are both up or both down, their interaction is J. When one is up and one is down, the interaction energy is —J. (In quantum-mechanical language, the energy Jn! a} hetween a‘neighhoring pair i and j.) What. ia the partition fnnetion Z of the assembly at temperature 7? 14. A highly simplified theory for the temperature dependence of the molar specific heat ¢ associated with the transition from paramagnetism to fer- romagnetism of ions with spin ¢ = 4, gives a curve of the type shown in the accompanying diagram; i.e., ¢ = Cmax (2T/Ts — 1) for T.|2 < T < Ts, and ¢ — 0 otherwise. What is cnax in terms of fundamental constantet © 1/2 Ts; 15. Show how a consideration of the appropriate partition functions leads to Fermi-Dirac and Bose-Kinstein statistics. Find also the distribution resulting from “parastatistics,” in which no more than two particles may be characterized by @ given set of quantum numbers. 16. How would Debye’s theory of the specific heat of solids be altered if the phonons (sound quanta) obeyed Fermi-Dirac (instead of Bose-Einstein) this asmmption, the temperatnra dependence of the rod the wr and wand vory Debye temperature, (Constant factors need not be evaluated.) 17. Assume that the inside of a metal represents a square potential well for electrons. Starting from Pauli’s principle, derive the velocity distribution of the electrons as a function of temperature. Derive the formula for the current in a vacuum diode with planar electrodes for negative plate voltage.44 SratisTicaL Poysics—PROBLEMS: 18. Show that a neutrino star, if sufficiently dense, may be treated as a degenerate gas of relativistic fermions. Derive a condition relating the mass and radius for the equilibrium of such o star. 19. Derive an expression for the magnetic susceptibility of a dijute solution of permanent dipoles, each of magnetic moment MM, if the dipole can assume (a) any position, and (b) only two positions, with respect to a weak magnetic field. és spin $ and magnetic moment p, and obeying Fermi-Dirac statistics, the lining up of the spins produces » magnetic moment/volume. Set up general expressions for the magnetic moment/volume at arbitrary T' and H. Then for low enough temperature, determine the magnetic susceptibility of the gas in the limit of zero magnetic field, correct to torms of order T?. Note the integral ae Zea de s(Z) + vo] 21. Consider an oil drop of radius 0.0001 om, in a gas of viscosity 180 mi- cropoise and temperature 27°C. What is the rms displacement of this drop after 10 sec? Neglect gravitational effects. 22. Ina hot plasma, all the atoms may be regarded as completely ionized. Although the ions have long-range forees due to Coulomb interactions, macroscopically the plasma is electrically neutral. This suggests that the Coulomb interactions are screened, sot so become short-range. EstimateATOMIC PHYSICS 1, Give a labeled energy-level diagram of the hydrogen atom for principal proton to be a point charge. Give a similar energy-level diagram for those levels of the helium atom which correspond to the excitation of orte electron only out of the ground state. Point out and explain the similarities and two diagrams 2. Find the three lowest terms of the carbon atom. Neglect spin-orbit cou- pling. Write wave functions for the three terms. 3. Find the three lowest terms of the nitrogen atom. 4. Give the electronic configurations for zirconium and for hafnium, and explain why the chemical separation of these two elements is difficult. The atomic numbers are 40 and 72, respectively. 5. Assume ihe iowesi-ierm vaiues in sud Re Be 41,448; Bp * Pras 24,484; 3d *Dyaga ‘12,274; 4p OR, 6,858; 45° "Si, 15,705; | 4p *Pyasa 1,180; 4d "Diasa 6,897; 5s Sin 8,246; Sp "Pras 6,407; Indicate what. transitions would he ohserved when excited hy (a) illumination with light at 4123 A, (b) bombardment by 3.3 eV electrons, if the sodium is initially in 38. wi (vapresed in em 6. Obtain an approximate expression for the energy shift of the ground state of the hydrogen atom due to the finite size of the proton, assuming that tho proton is a uniformly charged ephore of radius R = 10°"? em. 4546 Atomic Puysics—PROBLEMS 7. An electron is placed in a potential Lt auet +) +82, states? Sketch their relative positions. Calculate the linear Zeeman effet for B parallel to 2 and for B parallel to x. 8 Ina hydrogen-like atom, the 28 and 2P levels are separated by a small energy difference A, due to a small effect which has a negligible influence on the wave functions of these states. The atom is placed in an electric field E. Neglecting the influence of more distant levels, obtain a general expres- sion for the energy shifts of the n = 2 levels as a function of the Geld strength E. (Note: Neglect electron spin in this problem. Do not evaluate explicitly any nonzero integrals which may occur in your discussion. ] 9. The spectral line of mercury (Hg) at 1849 A splits under the influence of a magnetic ficld of 1000 G, into threo components separated by 0.0016 A 10. The electron and positron have the same (absolute) magnetic moment, but opposite g-factors. Show that the “ground state” of the ete” atom (positronidm), «'S., ‘8: doublet, eannot have o linear Zeeman effec if this terme of the total magnetic: moment operator is tre 11. Solve exactly for the energy eigenvalues when a magnetic ficld B is applied to a doublet P level (say of sodium), ignoring hyperfine structure. The Hamiltonian is H +L + plo(L + 28)-B, 3 where ¢ is the fine-structure splitting when B = 0. 12, Caleulate the hyperfine Assume the electron is in an S state, and give the answer in electron volts. hydrogen atom 13. The ground state of the hydrogen atom is split into two hyperfine states separated by an interval AZ = 1.42 x 10° cps. What is the hyperfine split- ting in the deuterium atom? The respective magnetic moments are Py = 2.8jty and jig = 0.86j1q, Where jn is the nuclear magneton. 4. A D,q term in the optical spectrum of ,,K** has a hyperfine structure with four componente. (a) What is the spin of the nucleust (b) What interval ratios in the hyperfine quadruplet are expected?Atomic PHysics—PROBLEMS: 47 15. An atom with no permanent magnetic moment is said to be diamagne- tic. If one neglects the spin of the electron, what is the induced diamagnetic moment for » hydrogen atom in its ground state when a weak magnetic field B is applied? 16. (a) Find the magnitude of the Doppler broadening for an argon glow- tube of temperature 300°K. Assume a wavelength of 0.5 ys for the radiation. (b) At what presure should collision broadening and Doppler brondening the argon atoms as solid spheres of radius i A. 17. Find the alteration of intensity (ratio of intensities of succeeding lines in an electronic band spectrum) of the following molecules, which are as- ; se ai @ Or), (b) (H*): (c) (He*), (a) (He), Assume kT so large that one may ignore the variation of intensity due to the Boltzmann factor. 18. The potential of a diatomic molecule with nuclei of masses M, and Mz, can be approximated by V(r) = —2 v(z -s rational energy levels for small oscillations by a power-series development of the effective potential. 19. The far-infrared spectrum of HBr consists of a series of lines spaced 17 cm~! apart. Find the internuclear distance of HBr. 20. The dissociation energy of the H, molecule is 4.46 eV, while that of the D, molecule is 4.54 eV. Find the zero-point energy of the H, molecule. 21. The muon has ciectric charge equai to that of the electron and a mass roughly two hundred times that of the electron. It is known that a #-mesic hydrogen atom may exist. This atom is known to combine with a proton to form a “mulecule” H?, consisting of p*pt~. Assuming that the muon for the internucicar equilibrium distance r, the zero-point energy of the molecule, and the binding energy of the mulecule. The following data are given: internuclear distance in Hi is 1 A, the zero-point energy is 0.14 eV, and the binding energy is 2.7 eV.10 SOLID STATE 1. Many metals can occur with both the body-centered cubic and the face- centered cubic structure, and it is observed that the transition from one structure to the other involves only an insignificant volume change. Assum- ing no volume change, find the ratio Dp/Dj, where Dy and Dy, are the closest distances of approach of metal atoms in the two structures. 2. Tolman obtained a value for e/m for electrons by the mechanical accel- eration of a sample of metal. How can this be done? Assume the electrons free. 3. Given the following data concerning lithium chloride: (a) lattice energy = 192 keal/mole = A; (b) heat of formation — 97 kcal/mole — B; (c) ionization potential ot lithium atom = 5.29 V ; (d) heat of sublimation of lithium = 38 keal/mole = D; (e) heat of dissociation of chlorine molecule = 58 kcal/mole = E. Compute the electron affinity # of the chlorine atom. 4. A sample of germanium shows no Hall effect. If the mobility of electrons in germanium is 3500 cm*/V-sec and that of the holes is 1400 em*/V-seo, what fraction of the current in the sample is carried by electrons? 5. The rovui-temperature modifi cred cubic {boo}, and this transforms at 910°C into @ face-centered cubic (fec) modification; the heat of transformation L is 253 cal/mole. The addition of a small amount of carbon lowers the transition temperature. Estimate the change in the by (The modification.) 6. A crystalline body in a state of thermally excited elastic vibrations may be treated as a system of N distinguishable independent quantum-harmonic oacillatora of the same angular frequency « (Kinstein'a model). Give the expression for the distribution law of the system. Compute the averageSot StaTe—PRoBLEMS 49 energy of the system at high and low temperatures; find the molar specific heat C in those limiting cases, and discuss the validity of the model in those cases. 7. The lattice specitic heat of a certain form of carbon has a temperature dependence T*, instead of the more common 7? dependence for solids. What: can you infer about the structure of this particular phase of carbon? °K. From optical dais, below the conduction band. Calculate the conductivity at T, = 9. A parallel beam of 25 eV electrons is incident on a thin polycrystalline sheet. of a metal which is known to have a cubic lattice with lattice constant. 5 A. When the diffraction pattern formed by the electron passing through the sheet is photographed, it is found that the smallest ring has an angular diameter of 120°. What is the depth of the potential well for the metal? 10. A crystalline insulator is placed in an electric field E. Show that an electron in the crystal will oscillate according to e(x — x,)-E — (ke + 3) — elles), where € aud & are ihe energy and momenium of ihe eieviron. Extimaie a typical amplitude for the oscillation, and a typical period. 11. Positrons annihilate electrons in condensed media with the cross section where » is the relative velocity hetween the annihilating par- annihilation photons are emitted isotropically in the center-of-mass frame of the electron and positron. () If only the outermost electrons are annihilated, show that the two annihilation photons are emitted in opposite directions within a cone of half-angle, 8 radians, where 8 ~ 1/137. (b) Predict the distribution W(8) for a metal, knowing the number of con- duction electrons per unit volume. In this case, will there be a temperature cffect? 12. Consider the propagation of a sound wave with frequency @ and wave number k, in a metal of N atoms per unit volume in the frec-clectron model. Derive an expression for the velocity of sound in the metal. Estimate nu- merically for aluminum, which has atomie number 27 and Z = 3 valence clectrons. [Hint: Consider the lattice as a “heavy” plasma with screening by the clectrons.]1 NUCLEAR PHYSICS 1. Pions and muons each of 140 MeV/c momentum pass through a trans- parent material. Find the range of the index of refraction of this material over which the p’s alone give Cerenkov light. Assume m,c! =140 MeV; myc? — 106 MeV. 2, The existence of a massive meson was suggested by Yukawa to explain the nuclear forces. From the uncertainty principle, derive a relation between the range of the force and the mass of the meson. Estimate the mass of the meson. 3. A high-cnorgy beam of antiprotons enters a liquid-hydrogen bubble chamber of iength i. Let , be the cross section for eiastic scattering and ssume both are independent of energy. Derive an expression P,(I) for the probability that an incident 6 undergoes double olasti ing, i.e. is elastically scattered twice and leaves the chamber 4. The lifetime of an unstable nucleus is determined by the interval be- ‘tween two events representing its creation and decay. ‘The mean life of such nuclei may be determined in the following way: Each pulse from detection of creation eventa ia put into a coincidence circuit after passing through a known time delay. Each pulse from decay events ie put into the same coincidence circuit with no delay. The coincidence rate is observed for two different delays, t, and t;. Assume that, in a certain measurement, the decay rate 2 is approximately known, and that 1/ is much larger than the resolv- ing time of the coincidence circuit. Assume no background or accidental coincidence problems. How would you find ® from the observed coincidence rates C, and C; corresponding to delay times ¢, and ¢,? If you have a total time T for the experiments, how would you divide your time between observing C, and C41 (Assume you cannot observe tho two ratos simultancously.) What delays t, and f, would you uset nONucteas Puysics—PRosLems 51 5. Experiments (deuteron “stripping”) show that the ground state of 0" is formed from that of O'* only by acceptance of a neutron of orbital angular momentum 1 = 2. The first excited state is formed by the accoptanco of neutron with angular momentum 1 = 0. What can you conclude about spin and parity of (a) the ground state, (b) the first excited state of O'"! 6. Which of the following particles may undergo two-pion decay? Give your reasoning for each of the three cases. a o ” Here, J and P are the intrinsic spin and parity, J is the isospin. Assume atrict isospin and parity’ conservation. 7. Particle A decays by strong or electromagnetic interactions to particle Band particle C. If A has spin }, prove that the decay products must come out isotropically even if 4 is polarized. iu of the 17 is aero aud the parity of the deuteron is even, show how the existence of the reaction x (at reat) +d—+n+n determines the parity of the >. 8. Knowing that the 9. The elementary particles A, p, n, *, x*, x~ have the following (internal) quantum numbers (I = isospin, I, = third component of isospin, 8 = strangeness). Muitipiet Particie i is A A -1 0 0 P +4 N }__—_] o £ a + +1 x x 0 1 0 [esate [eee The weak nonleptonic decays . A—-Nix qd) obey the selection ruies [Sntetet tate — Srinat stare | | Zintear state — Zrrmat aeata | =52 NUCLEAR PHysics—PRoBLEMS (the so-called “AI = } rule”), and of course charge conservation. Calculate the branching ratio 4 — Rate(A — px"), ~~ Rate(A—> nz) The strong interactions conserve I, I,, S, and the electric charge. In par- ticular, the prqceas x+N—7K+A ) proceeds via strong interactions. The ratio of the cross sections is (e+ p—K°+A)_ @+n—>K +A) Show how this determines the isospin of the K°. ve and angular momentum coupled by an interaction —2a8-L, where a is a positive constant. Use this model to predict the spins and parity of the following nuclei (a) He (b) stir (ce) 3B (@) WW 11. A radioactive element is believed to have the following decay scheme: a 0.24 Mev =e 2 + fe vy 1.3MeV = 2.1 MeV 7 Suggest a series of experiments to measure the energies of the radiations and to establish the correctness of this scheme. 12. The radiations emitted by a certain radioactive species were studied in a B-ray spectrometer. The B-spectrum was resolved into two components of 0.61 MeV and 1.436 MoV maximum energics. The higher.energy com.Nuciear Puysics—PROBLEMS 53 ponent was about four times as abundant as the low-energy one. When the rays accompanying the A-spectrum were allowed to strike a thin silver foil placed in the source position of the spectrometer, the following five photoelectron energies were measured: Photoelectron _E (MeV) Intensity A 0.216 strong B 0.237 weak a o.R01 weak D 9.299 vory woak Ez 1.042 very weak The K and L binding energies in silver are 25 and 4 KeV, respectively. Draw a plausible decay scheme for the radioactive species under investiga- tion. 13. Show that the conversion of a high-energy photon into ani electron- positron pair can occur only in the presence of matter. 14. Pions are prod lear ov; ph graphic emulsions. It is observed that up to a kinetic energy of appro: mately five MeV only negative mesons emerge from silver nuclei which are in the emulsion. Why are positive pions not observed below this energy? 15. ,,Si?? decays to its “mirror” nucleus A*’ by positron emission. The maximum positron energy is 3.48 MeV. Assuming that the nuclear radii may be given in the form r,4", where A is the mass number, estimate 1, from the above data. 16. N‘'' decays to an excited state of U'' with §~ emission of maximum energy 3.72 MeV. This excited O""-state decays by neutron emission. F"” decays by positron emission of maximum energy 1.72 MeV; no radiation ocours following this decay. The mass differences are: n — H! = 0.78 MeV, (a) N'" — 0" = 8.80 MeV, (2) 0! + H! — Fi = 0.50 MeV. (3) O* has excited levels at 6.05, 6.13, 6.9, 7.1 MeV and higher. (a) Using only the above data, calculate the energy of the emitted neutrons in the laboratory system. (b) Show all energies involved in a level diagram for the nuclei with 4’ = 17. (c) Which qualitative features of the level diagram follow from charge independence? Add the further levels which can be predicted from charge indopendence.54 Nuciear Paysics—PRoBLEMS 17. The threshold for the reaction N"(n, 2n)N"* is 10.6 MeV. Suppose the nitrogen in the air is exposed to the radiation from a radioactive series con- taining the following alpha emitters: x 5.0 MeV ag not measured Z 10.0 MeV Would you expect any IN* to be formed by the reaction N* (at, om)N**/ 18. A certain nucleus may be considered to be a sphere of charge Z and radius a, which will undergo nuclear reaction when hit by either a neutron or proton. Compute the ratio of the reaction cross sections (a,/,) according 19. A charged particle is slowed down in a photographic emulsion from a velocity of 10° cm/sec to thermal velocities. Does the grain density increase or decrease if the particle is (a) an electron? (h) 2 Z—=12 20. Calculate classically and nonrelativistically the differential scattering cross sectiqn at small angles, for fast magnetic monopoles of strength g scattering from fixed nuclei of charge Ze. Estimate the energy loss dk /dz of a monopole traversing a nonmagnetic sample of these nuclei, having N nuclei per unit. volume. 21, Derive a formula for the low-energy differential scattering cross section from » poieuiial which is capable of producing oue weakly buuud sine with the scattered particle. 22, A neutron is bound to a force center by an attractive force of range ro = 107" em. The ground state of the system has a binding energy of 1 KeV. center? Assume the potential function is well behaved. ergy for this force 23. A beam of 100-KeV neutrons is attenuated to 50% of its initial intensity in passing through 10 gm/cm! of carbon. What can you say about the s-wave phase shift for the scattering of neutrons from carbon nuclei? 24. It is required to make a quick estimate (in the absence of tables of cross section) of the fraction of 14-MeV neutrons which will be clastically scattered (over all angles) from a sheet of lead 2.0m thick (of density 11 gm/cm*) placed in 14-MeV neutron beam. Make your best guess, listing your assumptions.NucLeak Puysics—PROBLEMS 55, 25. A thick target of Mn‘* is bombarded during a time ¢ with a deuteron beam (current i) in order to produce Mn'*, which decays with a half-life Typ. Calculate the number of active nuclei present at the end of the irradia- tion, assuming a range R of the deuterons and an average cross section over the range of the deuterons. Numerical example: ¢=48x 10°°A Typ = 2.6 he ¢=52hr R= 110 mg/em* o = 10 cm’. 26. Show, by explicit calculation of the scattering cross section of ortho- and para-hydrogen for thermal neutrons, how the relative sign of the triplet and singlet scattering length may be determined. 27, Suppose that the differential cross section da/d{2 in the center-of-mass frame for the reaction p + p— x* +D is A + B cos* @ at energy E. What is the do/dQ for the inverse reaction x* + D —+ p + pat the same center- of-mass energy? The experiment is done with unpolarized beams. 28. Neutrons can be scattered in the Couiomb fieid of a nucieus because of their magnetic moment. Write down the Hamiltonian of the interaction, and calculate the spin-averaged differential cross section in the Born ap- proximation. (Treat nonrelativistically.) 29. Assume that an iron atom has an average magnetio-dipole moment per unit volume given by #g(r), where py is the magnetic moment of the iron atom. Thermal neutrons of momentum k,, polarized along the direction of un, with ky perpendicular to H4;, are scattered in iron. Note that the scat- tering has two sources: that. due to the nuclear force hetween neutrons and iron nuclei, and that due to magnetic forces between neutrons and iron alums. Compare these contributions to the scattering. Also calculate the total spin and nonspin flip cross sections. Treat the scattering as though due to a single iron atom. [The vector potential A(r) due to a magnetic dipole Wis given by A = (u x r)/r*,] 30. The dominant decay mode of the neutral sigma hyperon (spin }) is radiative, ie. 3°—> A° + photon. Although electric-dipole radiation is forbidden because Z* has no charge, the decay may proceed via magnetic- dipole radiation, through an effective interaction H Ty0°V X AL ____geh (He 1 Maje Here 7s, denotes the operator that converts % into A, and gefi o/(Ms + Ma)e may be interpreted as a transition magnetic moment, which interacts with the magnetic field B = V x A of the radiation field. Taking the amplitude56 Nuctzar Puysics—PRosLems for emitting a photon to be the plane wave* A= e( ye exp(i-k-x — it), a 20V, esiimaie the 2° iifeiime, for y = i. The masses of ihe pariivies are M(%)=1190MeV = and = M(A) = 1115 MeV. 31. An element of low atomic number Z decays with positron and neutrino emission; the maximum kinetic energy of the positron is W = 00 KeV, and the probability per second of the decay process is I’,. Using Fermi’s A-decay (for an allowed transition), calculate the probability T', for K-capture by the same nucleus. 82, A system which may be taneously two nonidentical particles with relativistic energies. Assuming the partition of energy among the two particles to be governed entirely by phase-space considerations (energy-independent matrix element), obtain particles. Discuss the relation of the result to the process of A-decay. 33. (a) Assume that the muon decays into an electron and two distinet neutrinos with a transition rate = 2% 2 Vp dpdn Pa TOY TR do where g is @ coupling constant, V a normalization volume, w the energy of the three emitted particies, and p the momentum of the electron. By cai- culating the number of neutrino states per unit energy interval, dn/dw, obtain an expression for the momentum spectrum of the electron. (Assume electron energy ~ pe.) b) Calculate g* from the mean life of the muon, T) = 22% 10% sec (m, = 207 m,). 34, Assume the existence of a particle of electronic charge ¢, spin 4, and rest mass M, which disintegrates spontaneously into an electron (mass m) and a photon. The mean lifetime of the particie is T' (in its rest system). Conversely, these particles can be created by irradiated electrons with light of appropriate frequency. What is the frequency a» required if the electrons are (practically) at rest? And what is the probability of the process per * In those units, e/hc = 44/137. In units whoro et/he = 1/137, wo have A = c(2zh/wV)'"e oxp (i kex — iat),NuciEaR PHYSICS—PROBLEMS 57 second (per electron) if the incident photons have an energy U(c) dw per unit volume and per frequency interval dw? What is the answer to the second question if the heavy particle has spin } instead of $2 45. in a chain-reacting device consisting of a homogeneous mixture of uranium and carbon, neutrons emitted by the uranium move about with an average velocity » and mean free path 4 = 10cm. They will, on the average, be captured after N = 100 collisions. For each neutron captured, k— 1.04 a he o. Give t! © average. Give th equation for the time rate of increase in the neutron density. if the device is built in the form of a cube, find the critical size. 36. Neutrons are slowed down in hydrogen. (a) After n collisiona what: is their average energy and the diatrihntion function p,(Z) for the number of neutrons per unit energy? Assume that the scattering of a neutron by a proton is spherically symmetric in the center- of-mass frame of reference. (b) If neutrons of energy E, are produced at the rate of q/em* sec™' and the total scattering and absorption erosa sections are known functions of the energy, find the steady-state neutron flux (product of neutron density and velocity) as a function of energy and the known cross sections. 37. Slow positive and negative i-mesons (1, .-) have, in a vacuum, a mean life 7, for decaying into an electron and two neutrinos. Negative u’s can also be captured into atomic orbits and very rapidly fall into the K-shell, where they are close enough to the nucleus to be absorbed by it. (a) Making the plausible assumption that the probability of this nuclear absorption is proportional to the fraction of time the -meson spends inside the nucleus, how does the probability depend on the atomic number of the material in which the meson is stopped? (b) If the lifetime in the K-shell of hydrogen is rq, what are the moan lives r,, T. of positive and negative muons in zine (Z = 30)? tT» = 2.10 x 10-* sec tm = 2.075 x 10-* sec Note: Neglect time required for transition from outer shells to K shell. 38. A cyclotron produces a beam of deuterons of 200 MeV energy. When incident on a Be target, a narrow neutron beam is produced by the process of stripping (i.e. the proton is removed by a nuclear collision, allowing the noutron to continue its original direction of motion). Calculate approximately the distribution in angular spread of this neutron beam due to the internal motion in the deuteron. Use an approximate form for the deuteron wave function (limit of zero range) and take the binding energy of the deuteron to be 2.18 MeV.58 Nucizak Puysics—PRosiems 39. Study the vibrational excitations of a nucleus of atomic weight A and charge Z, in the liquid-drop model. Include the effects of Coulomb repulsion and surface tension. Develop a criterion for instability against nuclear fis- sion. To find the surface tension, use the fact that it contributes a term M, = U,A** to the semi-empirical mass formula with U) = 14 MeV. | Hint: Use the methods of Problem (2-35).]12 FOR THE EXPERIMENTALIST We include here a set of problems concerning experimental physics. Most of these problems have no unique answer, with solutions of varying degrees of sophistication possible. Furthermore, because of technical developments. the appropriate response to some questions may depend strongly on time. We have, therefore, not attempted solutions to these problems. . 1. Design an efficient furnace and accurate temperature-measuring equip- ment tor studying the electrical properties of solid materials in the tem- perature range 1800-2300°C. The volume of the furnace is to be 10 cm’. Include methods for end-effect corrections for electrical connections to the sample in question. 2. De. otomet: the range 1500-2000 A. The de: 2, Design « spectrophotometer for use in the range 1500-2000 A. The design is to include a source or sources of light, dispersive agent, and method or methods of detection. 3. Devise a set of experiments that will measure the mechanical properties of so-called “Bouncing Putty” or “Krazy Klay”. a silicon compound which bounces when it strikes a hard surface suddenly but flows under its own weight. 4. Design a strong focusing synchrotron to produce 50-BeV protons, paying special attention to size of magnet, gradients, ratio of straight sectors to curved sectors, energy storage in the magnetic field, radio-frequency system, and orbital stability. the following isotopes: Rb**, Na*‘, or Cs’**. 6. Design a magnet to produce a field whose homogeneity and stability are +.01% over a range of 1000-20,000 G. The region where B is to be constant is 6 in. in diameter and 1 in. high. Include measuring equipment to deter- mino B to +.05%. 5960 For THE EXPERIMENTALIST—PROBLEMS 7. Given the problem of measuring minute normal displacements of a surface, describe three of the most sensitive methods you can devise, and give an estimate of the minimum displacement which each can detect. &. Discuss the reiative merits of a quartz prism and a vacuum refiection- grating spectrograph for the following applications: (a) The observation of the fine structure due to the Lamb-Retherford effect. (b) To study the UV bands of CO between 1300 and 1550 A. (c) To study the Doppier effect in the aurora due to particles coming from the sun. (d) Measuring spectral intensities by the rotating-sector disk method. 9. Suppose that you suspected the existence of gamma-rays of about 0.5 MeV energy in the radiation from the sun. Assume a fus of about 5 per em? per sec at the earth’s orbit. Assume that the absorption coefficient of air for gamma-rays is 0.1 cm*/gm. Devise an experiment to prove that such radiation existe, that it ia nat corpuscular, and that it really comes from 10. Describe any set of experiments from which the fundamental constants e. m, ¢, h can be determined, and estimate the accuracy obtainable with these experiments. 11, What techniques would you use for measuring (a) Magnetic fields in the ranges: (b) Temperatures in the ranges: (1) 10 to 107 G, (1) 107 to 1°K, (2) 1G to10kG, (2) 1° to4°K, (3) 100 to 250 kG (1 kG = 108 @) (8) 1000? to 500° K In case you can think of several methods suitable for a given range, state the experimental circumstances on which you would base your choice of any one method. 12, Discuss a method for determining Avogadro's number with better than 5% accuracy. 18, The “cl life is to ie an evacuated container in a pile and measure the rate of accumulation of hydrogen. List three factors that make this a difficult ex- periment to perform. 14. Discuss nuclear reactions which aro observed when lithium is bom- barded by protons and when it in bombarded by neutrons, How are those at? Sctisaarhass:For THE EXPERIMENTALIST—PROBLEMS 61 15. Discuss methods suitable for measuring high, medium, and low resis- tance. In what range of resistances is each method the best? What is the approximate accuracy of each method? 16. Discuss the theory of operation of the high-energy machines which are now being built throughout the United States. What do you consider the main difficulties in constructing each of these machines? To what kind of experiment is each best adapted! Suggest brief programs of research for Eiaborate in detail one such program. What in your opinion will be the probable effect of these machines on our knowledge in physics? (This problem was given in 1948.) 17. How do yon measnre the wavelength of electromagnetic radiation in all different frequency ranges? Describe briefly the type of equipment required in each region. 18. Design an audio oscillator and power supply to operate at a nominal frequency of 50 ke which does not contain any inductances. The oscillator ia to provide 3.0 W of power to a 200-ohm load. 19. The nuclear magnetic resonance method for measuring nuclear magnetic moments makes it desirable to measure the magnetic field between the poles of @ magnet to as high a degree of absolute accuracy as possible. Propose a means for carrying out such a determination for a field of the order of 10,000 G and discuss the limits of accuracy. Take the diameter of the poles as 6 in. and the gap as 1 in. it is desired to have water moving as fast as possible in a pipe of at least 1 in, inside diameter and of length 5 ft. ‘The pipe is to be provided with end windows for observation. Design the pipe and the subsidiary equipment and 21. Design an experiment for measuring the lifetime of an artificial bete- active substance produced by slow neutron bombardment, assuming a lifetime of the order of 4; sec, and allowing for the possibility that fairly intensive longer lives are also excited. Estimate the accuracy. 22. The bathysphere is a hull which can be lowered to great depths below the surface of the sea and can safely accommodate an observer for the pur- pose of exploration of the ocean depths. Design a bathysphere for a depth of 1000 m, giving particular attention to mechanical details such as the observation windows, illumination facilities, devices for renewing the air, ote. Give a# complete a design an possible.62 For THe EXPERIMENTALIST—PROBLEMS 28, Design a seismograph for detecting three-dimensional seismic motion of maximum amphtude 1 mm and maximum period 30 see 24. Starting with a 3-phase, 440 (5 pereent)-volt line, specify a power a 40-in cyclotron Assume an ambient temperature of 30°C + 10°C, a maxi- mum magnetic-ficld intensity of 18,000G and a mmimum field intensity of 16,0006 Specify eoft aren for the magnet’ The controller should keep on constant ag a function of time to within 01 pareant 25. In a synchrotron electrons are accelerated to an cnorgy of 3 x 10 eV Devise an apparatus for measurmg the shape of the y-ray spectrum originat- ing from an internal thin target of tungsten struck by the electron beam 26. Construct a spectrometer for measuring the diffraction, refraction and reflection of z-rays of a wavelength of 40 A 27. Heavy nucle: have been observed at the top of the atmosphere by means of nuclear phoio-emulsous Tim unpuriani thai the properties of these particles be determined experimentally usmg methods other than photo- graphic emulsions It is further important to design the experiments in such ‘a way as to indicate the origin of these particles Develop an experimental method which can successfully attack this problem and devise the methods vf analysis iv be used to iterpiet the data In particular, describe un detaid all aspects of the experiment and estimate intensities expected 28, Design a practical experiment to detect the neutral mesons which are ampposed to emerge from a heryllmm target. which is homharded with 1 uA of 450 MY Caloulate th od of 450 Mc¥ protons Calculate th produc! of the neutral mesons which your method would be able lo detect Show how you would distinguish these mesons from neutrons and gamma-rays which would be present 29. What 1s the deepest hole which can be dug into the earth? Explain the limitations and design roughly the practical equipment necessary to dig and maintain the hole 30. A high-impedance pulse generator produces periodic pulses of the shape shown in the diagram on p 63 The pulse rate 1s 10!/sec and the pulse height 1301 V Work out the basic design of an oscilloscope to observe these pulses so that the properties of the pulee described may be measured to +5 percent The deflection sensitivity of the cathode ray tube 18 20 V/in Basie circut diagrams and schematic drawings should be used to designate electronic components The propertios of each circuit shown should be caloulated or narafully eatimatad.For THE EXPERIMENTALIST—PROBLEMS 63 $1. The neutron resonances in the range 1-100 cV for various nuclei arc to be measured. A nuclear reactor is available as a source of neutrons. The neutrons may be brought through the reactor shielding by means of a cylin- drical pipe measuring a cm in diameter, so that neutrons ranging from thermal to over 10* eV, are present at the end of the pipe. The resonances are to be determined by reflecting a collimated beam of neutrons from a crysial spevirumeier. The reflected uvuirums are tv be deiwvied using Proportional counter filled with BF, gas. (a) Draw a plan view of the arrangement of the experimental apparatus showing all components, giving approximate dimensions for the apparatus, sample to be measured, etc. (b) Develop the theory of the crystal spectrometer, discussing resolution, limitations on beam solid angles, beam flux requirements, effective penctra- tion of neutrons into the crystal, and the true measurement of resonance width. 32. Design apparatus for measurement of the Lamb-Retherford shift in the 4868-A, linc of HoII (He*) to an accuracy of 10 percent. The Lamb-Rether- ford shift for the n = 2 state of hydrogen is approximately 1000 me. 33. Design apparatus for determining the velocity of light to an accuracy of 1 part in 10° by measuring the velocity of propagation of microwave radiation across a cavity. Show how microwave absorption of ammonia can be used for frequency standardization. (a) Give the overall design and show what measurements are to be made. (b) Give details and precautions necessary to obtain the required accuracy. 34. A few rough measurements have been made of the decay of the neutron giving a half-life of 9-18 min. What difficulties may arise in the Chalk River experiment, whose description follows? Show how a coincidence measurement would improve the experimental evidence for the decay process. Show how this experiment is related to neutrino emission.64 For THE EXPERIMENTALIST—PROBLEMS THE RADIOACTIVE DECAY OF THE NEUTRON J. H. Robson, Chalk River Laboratory. The positive particle from the radioactive decay of the neutron has been identified as a proton from @ measurement of charge-to-mass. A collimated beam of neutrona emerging from the Chalk River pile passes between two electrodes in an evacuated tank. One electrode is hold at o positive potential, up to 20 keV, while the other electrode is grounded and forms the entrance aperture to @ thin-lens magnetic speciromeier, the axis of which is perpendicular io the beam of neutrons. The positive decay particies can be focused on the first electrode of an electron multiplier. The background counting rate is 60 cpm. A peak of 80 cpm is observed above background when the magnetic field is adjusted for protons of energy expected from the electrostatic field. When a thin boron shutter is placed in the neutron beam, the proton peak disappears. Preliminary estimates of the collecting and focusing efficiency and the neutron flux indicate a minimum half-life of 9 min and a maximum of 18 min for the neutron. determining their mass—with a precision of 5 percent—by deflecting them in an electric and magnetic field. Discuss sources of error. The mass is assumed. to be around 200 electron masses. 36. A penetrating shower of 10" eV total energy consists of 5 charged mesons, 6 neutral mesons and 2 protons. Describe in detail apparatus by which (a) cach individual particle can be identified, and (b) its energy can be determined. 37. A new phuivconduciive umierial, PUTe has a loug-waveleugiht direst for light absorption near 6 j1. Describe the expected photo-response signal and the noise as a function of wavelength and temperature. Compare the highest: expected signal-to-noise ratio at room temperature with that. of What experiments would be necessary to determine whether the photo- conductivity is intrinsic or due to an impurity; whether it is a bulk effect, surface effect, or contact effect; to determine the sign of the charge carri- ers, number/em*, mean free path. and other important parameters? What be as a semiconductor? 38. A beam of 2-MeV neutrons of flux N cm-* sec! is incident normally on a 100 pg cm * film of glycerol tristearate (C5:H,1O¢) (density 0.862 g/em*) producing hydrogen recoila which enter a gas. Develop the theory for and design a detector and recording system utilizing the ionization from theFor THE EXPERIMENTALIST—PROBLEMS 65 recoils to measure the incident neutron flux W and the maximum energy Exux- Call the bias energy of the detector system E, and the incident neutron energy E,. Describe the properties of the detector selected for the measurements. Calculate the expected response of the detector for recoil protons. Draw 1 cross-section view of the detector, indicating all components with suitable specifications such as dimensions, materials, ete. Draw block diagrams showing all functions to be porformed by electronic circuits. Tabuiaie the estimated specifications for each of these circuits, What is the efficiency # of the film? What is the energy of proton recoiling at an angle of zero degrees with respect to the incident neutron in the detec- tor? What is the pulse size distribution in the detector? What is the observed counting yield of the detector? Show how any spurivus background proion re ments. av above measure jo way be devermined 39. Consider the reaction Au!” + n?—> Au'® > Hg!" Hy! The half-life for the decay of the excited state of mer: life, giving limitations and sources of error in the measurement. 40. Given the reaction I'*' + n°— 1'*®—+1? The latter decay, for which the daughter nuclei are unknown, occurs with a half-life 25 min. Outline experimental procedures by which you could obtain the complete decay scheme of the I'#® produced, including limitations and orrors of the method you propose. 41. Discuss critically at least three pieces of evidence for a finite age ot the visible universe. 42. Discuss the possibility of producing a magnetic field of 10¢ G in the laboratory. Explain the limitations, indicate the major design considerations, and describe the unin features aud compoucnis uf apparatus iv real field or to reach as near to this field as you think feasible. 48. Design a cryostat for making measurements of specific heat at liquid helium temperatures. 44. Design apparatus for the measurement of viscosity as a function of temperature and pressure in nitrogen gas at about 5000 atm pressure. 45. Oxygen is paramagnetic, a property which is unique among the common gases in the atmosphere. Design an apparatus using this property for ana- lyzing the oxygen consumption in human respiration.66 For THE ExPERIMENTALIST—PROBLEMS 46. The velocity of light is known, at present, only to within a few parts in one million. Devise an experiment to measure the velocity of light to one part in 10°, Discuss the equipment needed and the necessary corrections to be made. Try to make the experiment feasible. 47. Suppose that the neutron has a permanent electric-dipole moment. Design an experiment to detect such a moment if it is larger than that pro- duced by two elementary charges (+ and —) separated by 10°! cm. Why t? 48. Dyson, in an interview after returning from the U.S.S.R. (June, 1956) said, “Since World War II there have been six fundamental experiments done in physics; one in Italy, one in England, and four in the United States.” If you were to select six such experiments with the boundary condi- tions as specified, which six would you select? Tell the significance of each experiment. 49. What is the experimental evidence for the assertion that the atomic number of » chemical clement is equal iv iis uuciear charge? 50. State, and discuss in a few sentences, one good method for detecting beams of each of the following entities: (a) Electrons of 100 eV energy (h) Positrons of 10° eV energy (b) Photons of 10-1 eV energy @) Free radicale (c) Photons of 10-* eV energy (i) Alkali atoms (d) Photons of 10** eV energy (&) Phonons (c) Metastable atoms (1) Neutrons of 10° eV energy (f) Thermal nentrons (ma) A particles {g) Antineutrinos {n) Spin waves 51. Consider the partial circuit diagram to the right: Estimate the total rms noise generated in the grid circuit of this high-impedance DC amplifier at 300°K. Assume usual values for the characteristics of a suitable tube. What. simplifying assumptions have you made? — 10! ohms ‘What precautions are necessary in the use of such a circuit? List other possible sources of noise in the first stage. 52. Design a coincidence-anticoincidence circuit with the following pro- perties: (a) There 18 one output channel. (b) There are four input channels I, IT, III, IV.For THE EXPERIMENTALIST—PROBLEMS: 67 (c) An output signal results when there is input to either channel I or channel II, or both; but no output pulse results if there is input to either channel III or channel IV, or both. 88. The rotor of a Beams-type uitracentrifuge is ing at 150,000 rpm. The speed is expected to be constant to within 0.1% over any 5-min interval. Describe how you would design, construct, and use an apparatus to check the speed regulation to this precision tromagnet powered by a single flashlight cell. After making some estimates of what is needed to perform this stunt, comment on whether this is a string- ent test of such a cell. Note that the circumstances allow the presence of a aupplementary piece of soft. iron concealed in the roof of the car. 55. What type of detector would you use for each of the following purposes? (a) Detection of protons and A-rays with equal efficiency and pulse height. (b) Efficient detection of 20 MeV A-rays with zero efficiency for 50-MeV protons. (c) Efficient detection of 50-McV protons with zero efficiency for 20-MeV A-rays. (d) Efficient detection of 5-MeV y-rays with zero efficiency for detecting 5-MeV a-particles. (0c) Efficient detection of fission fragments with zero efficiency for detecting @-, B-, or y-rays. 56. Design a permanent magnet for use in cosmic-ray research. The magnet is to produce a field of 10,000 G over @ 1 in. gap. The pole pieces are to be circular with a diameter of 10 in. The basic form of the magnet is to be as shown in the following diagram: 15,000} T YY | 24,000 | | | 20, WU 10,000 1400 | 2 aoe =| z= |e=2| | [=z | «0 | oo Re a 4 4 Seton Alnico V Note change of scale at zero. (e)68 For THE EXPERIMENTALIST—PROBLEMS The particular shape and dimension of each piece, with the exception of the pole face itself, are completely variable. The hysteresis loops for Alnieo and soft iron are given in the figure. The design should include: (a) Length and cross section of Alnico piece, (b) Cross section of soft iron yoke, (c) Apparatus and procedure for magnetization. Discuss the feasibility of constructing each item. Note: Cost of Alnico = $4.00/Ib: cost of soft iron = $0.05/Ib; cost of copper = $0.50/Ib.SOLUTIONSMATHEMATICAL PHYSICS 1. This particular method of birth control does not affect the ratio of the sexes. At each birth, the probability of having a male child is 51%, and each 1 is an independent event, the ratio for the comumunity is hence 52. 48. 2. (a) Choose a specific color, say color A, and use it for the upper surface of the die. Choose any color B for the lower surface. There are five choices for B. The cube may now be rotated so that some color is facing us. Now the three remaining colors over these remaining three sides in (3!) ways. Therefore there are a total of 30 possibilities. We can also obtain the result by considering the rotation symmetries of the die. If there were no rotation degeneracies. we would have (6!) possibil- ities. However, there are six sides that can be chosen facing up, and once such a choice is made, any one of four sides may be taken as facing forward. This exhausts all degeneracies and gives 6!/(6 x 4) = 30 ways. In general, for @ figure with N faces and painted with N different colors the number of possibilities is N!/R, where R is the number of discrete rotations which leave the figure invariant. (L) Here une must ake account of Ube facl Uhal the individual dive are indistinguishable. Thus if each die may be colored in N different ways [from part (a), NW = 30] the number of distinguishable pairs of dice is N(W + 1)/2 = 465. 3. We treat the regular octahedron as we did the die in the previous problem. The symmetry viewpoint is particularly useful. If we take one of the six corners pointing up, we get. 6 possibilities; there is still a four-fold rotational degeneracy, giving 24 possibili- ties. Alternatively, if one of the eight sides is fixed in space, the octahedron will still entertain a three-fold rotational degeneracy. Hither way, there are 24 dege- nerate positions. We obtain therefore, 81/24 = 71/3 possibilities. n12 MATHEMATICAL Puysics—SOLUTIONS Another way of counting requires more care. We draw the octahedron flattened; the dashed lines join when the figure is folded into 3D. Color any surface with some color (shaded in the figure), and face that side up. Now color all surfaces but A, B, C. Because all other surfaces are uniquely specified with reference to these, the number of possibilities is 7-6-5-4. There are now only 2 independent choices for A, B, C. For if a given color is chosen for any of these, a rotation will bring these surfaces into one anoth. cr. Thus, the first of the three to be filled gives no oxtra possibilities, while the remaining iwo ieave Z possibilities. So the answer here is 7-6-5-4-2 = 71/3. 4, The total number of distinct ways the 52 cards may be dealt is (52!). The total number of ways a particular team may be dealt two particular complete suits is a = (26!) x (261). ‘I'he number of ways they may be dealt one particular suit regardless of whether they have another complete suit ia b = (39!) x (261)/(13!). Thus the number of ways they may be dealt only one particular complete suit is (b — Su). The term Su is the number uf ways the viher three sui may be obtamed in addition and must be subtracted out. If we are not m- terested in which suit they may have, the probability of being dealt one and only one complete suit is Py = 4(b — 2a)/52!. The probability of obtaining any two complete suits is P, — 6a/62!. (The factor & arises from the 6 different pairs of suite possible.) The probability of obtaining at least one complete suit and possibly two is P, + P, = (4b — 6a)/52!. Note that it is a very good approximation to neglect the probability, a, of obtaining two suits in the latter probability. Is of I taken place, while in the remainder an event has not, is (Ak)*(1 — nA)"=", where V = t/h is the total number of intervals. The number of ways we can specify n intervals out of N is N!/n!(N — n)! and therefore the pro- bability of n events is Pr= aa air — aye", Upon letting N go to infinity, and using Stirling’s formula for large N, Ni ~ /Sat N%e-*. se) — RIM)" — e-™. we obtain it (2 — iN)" — e™, we obtain gth h an even Ni~ VGaN N%e-™, together with ¢ Quy ae P, Note that 3 P, = 1. Also ny ~ eH OM aye ogMATHEMATICAL PHYSICS—SoLUTIONS 73 and Gt) = eM SE MOM eons 2 er = Oy 4 a M > 1 this distribution becomes Gaussian in x, where n=M+a and |al 4, the fraction of the sphere which it may occupy without combining with @,B,+,8, or the other stars, is 1 — (n — 3)p. Note that, if (N — 3)p > 1, the above probability is zero. This is because in this case the stars are so dense that it is impossible to keep them separated s0 as to form only two double stars. In the limit Wp <1, one may obtain an interesting expression for XS (1 — pil — 2p) --- [1 — (N — 3)p}. Form log X — DJP log (1 ~ jp). Then, sinee jp <1 for all j < N. one4 MarammaticaL Pysios—SoLvtions may expand log (1 — jp) = —jp. Hence vo thay one ubiains X =e"? wheu terme uf the order Np! ace negligible (for this problem N*p* ~ 1.2 x 10-4). The number of independent ways two pairs may be formed from the N stars is wai Thus the probability of seeing any two double stars and no more is the Poisson distribution goo cst Ee 2 eee ce ace DP a — (32) X — yee? — 0.008. (c) Consider three definite stars labeled a, 8, y. The probability that stars A, lie within 1’ of arc of at is p*. Ifa triplet is defined only when it is isolated from the other stara, i.e. ia R not: part of some larger multiplet, then the prob. Jf th, pa py. If, however, one does not care whether the triplet happens to be part of a larger multiplet, then the probability is simply p*. Sinoe in this problem Np <1, both these probabilities are p* for all practical purposes. ‘The number of independent triplets one may form from the N stars is wi. 313i ~ 3T Thus the probability of observing some triplet is N*p'/3! = 2 x 10-*. 7. (8) M = M* and M* = M~', i.e. M is both Hermitian and unitary. Hence eigenvalues j are +1 only. Further, Tr M = ¥; Mu = 0, 80 that Z ps =0. Hence the four possible p's are +1, +1, —1, —1. Write, for any cigonvector x, hy x) =17), forw— 41,2, —%,and2, — Ty Wal hence 1 0 wre dl*) awe aa gl*| Vtio whi \i/ \o)MATHEMATICAL PHysics—SOLUTIONS 75 are two convenient choices. Similarly, for 4 = —1, the corresponding choices are ial ie) x= 7 ") and Te a - : f ' : | ‘ 0, -1 (b) The pedestrian approach is to compute w i — 1) 8. As H and H? are diagonalized simultaneously, (b) is the sum of the eigen- values of H*. The traces of H and H? can be computed without diagonaliza- tion, as traces are invariants, Thus (0) 3 = Te Hiding) = 3 Hu = 0 = det | m — plj mith and wy foe diag) = SH, =o 42. (by) DM = Tr Wing) = Z Wylly = 2 My = 42. The last step is true bocause the given H is symmetric. 9 We use the properties (a) of = 1, and (b) Tr (o:) = 284. From the first of these, expanding the exponential ef" = cos. a + (22) sina, where a =|s|, we have aleacted — cos a con h — wane?) sin asin b +12 ack b ~S cos b sin a + i: sin ® cos a. a Because of (bh), we drop terms linear in @. Further, Tr[(o-a)(o-b)) = Tr[o.7a,b,] = 28,0, = 2a-b. Therefore, T = 2 cos a cos b — (2a-bjab) sin a sin b. 10. (a) From the property of determinants, det (AB) = det (A)-det (B) where A and R are arbitrary square matrices, it follows that the determinant of a sccond-rank tensor is invariant under coordinate transformations. If A’ =UAU" where det (U) = det (U7) = +1, then det (4’) = det (A). U” is the transposed matrix of U and the coordinate transformation is Xj = Z,UyX,. The cubic form in 2, det (7 — 2), is invariant in view of the above discussion. What is more, each coefficient of the various powers76 MATHEMATICAL PHYsics—SOLUTIONS of 2 is invariant because these powers are linearly independent. The coef- ficients are I, = det (7) LA (Palas t Paka t Pala — Tula — Pala — PesPoed T,= Tr (T) = Py + Te + Ta. (b) The geometrical significance of J, is best determined in the coordinate system where 7’ is diagonal. Denoting these diagonal elements by we see that the equation for the surface 1=2 TeX defines an ellipsoid with principal axes a, b, and ¢. The volume of this ellipsoid is 4x 4x 1 we O = Zab — | _____ | - a3 fers Honea I, — (4r/20)* A plane which lies in the X,, X; plane defines an ellipse through the equa- tion VS TiXi+ TeX$ +27 2XiX, (Xs = 0). This ellipse has an area A, where x*/4j = T1,T2 — Th. Thus nan(he hed) where A, and A, have been defined correspondingly. The geometrical signi- ficance is that if an ellipsoid is cut by an orthogonal trihedron (an open surface formed by the intersection of three planes) whose apex coincides with the conter of the ellipsoid, the cum of the reciprocal aquares of the arcas cut out is independent of the orientation of the trihedron. Similarly the invariance of J, is equivalent to the statement that for an orthogonal triad whose origin coincides with the center of the ellipse, the sum of the reciprocal squares of the lengths of the intercepts with the surface is independent of the orientation of the triad. ity at @ puint zo is au nib-vider puis, may be expanded in a Laurent series: Ze — a", & where a = Pp deelle — 5)'ShenMATHEMATICAL PHysIcs—SOLUTIONS 7 When integrating f(z) about a closed contour which encloses this point and no other singularity, only the term in 1/(z — 2,) survives and the integral has the value ari a (n — Ider WE The residue is defined to be (a) The residue is ame i 1fay_2 (b) The residue is alo(aaa)|... 12. The integrand has poles as indicated in the diagram, and the path of integration may be deformed as shown, because no singularities are crossed in making such a deformation. As ¢—+0 tho poles appear at k= ta. In closing the contour ai infinity in the | upper half-plane, only the pole at k = +a is enclosed and the integral becomes . ~1df io. _ 3x, On-gr ae ee ap lena ~ Ba 18. Since the integrand is analytic at z = 0 we may deform the path of integration to that shown: raf sin’zy, eee Jase On this contour expand sinz=(2)(e*—e) and sin’z ( x (e* — Be + Be — 4), Those terms which have a positive imaginary exponent are evaluated by closing the contour above with a semicircle of infinite radius, while those with nogative imaginary exponent are evaluated along the contour closed below. Those terms evaluated along the contour closed below give zero contribution because they enclose no singularities. ‘'hose evaluated above have a residue at z = 0 and J becomes _1 ett — gett), _ 2xi_ Gi)? _ 3x, Ta apl (AS )e = eS =F 14, (a) Consider the integral of z/(e? — 1) over the contour shown below. The contour is closed at oo, and since it encloses no singularities of z/(e? — 1)18 MATHEMATICAL PHYsIos—SOLUTIONS we obtain “ydy _o oe —T “ade ("2 . arta t TAR = Slave. (Note: The fact that the real part of [7 ydy/(e¥ — 1) is a simple integral serves aa motivation for chonsing this cantonr.) One ahonld now prove the simple relationship = atdx _ ny (" ata = jesse Sz eset trom which we finally conclude that jy 2 dzj(e* — 1) = x*/6. { clos ot Tiofinity oL_»——__. (b) Using the same contour and identical arguments together with the results of I,, one obtains J, = x*/15. It. ia worthwhile to remark that the integral I, is‘needed to calculate the energy density of blackbody radiation. 15. The Fourier inversion of the equation Sie) = ["" Fajetsak yields { 7 7 Py = [ee seyee, Writing f(z) = (e** + e-'*")/2, we have for (k): roth rk Fy) = where J = j ** ew dy. The integral may be written in the form yar TD Jw through the substitution y = (1 + )s//2. The original path of integration is deformed to the one shown in the accompanying diagram. The circular ares give zero contribution in the limit of large . Finally we have F(k) = Ve (00 # + si £).MaTHEMATICAL PHysics—SOLUTIONS 2 16. The function F(p) = a*/(p* + a?) is analytic except for poles at p = tia. This allows us to write P(e + ik) = le eet F(t) dt, with o> 0. Then by Fourier inversion with respect to k, we have emit) = & C eM F(o + ikjdk, for t>0. The integral may be done by contour integral methods after closing the contour with a large semicircle in the upper half-plane. The function F(c + ik) has poles at k = +a + ig. We then have 2 Dri ciation har FB tererers — grees, ews or f(t) = a sin (at). 17. Let z =e; then raf" _#b_ ag foatusp t where the integral is to be taken over the unit circle, |z| = 1, in the complex z-plane. The roots of the denominator are at z = —a@ + Ja? — I. (a) For @> 1 the root (—a + /a@ — 1) is inside the unit circle while the other lies exterior to the unit circle. Thus the only contribution to the integral comes from the pole at (—a@ + »/@ — 1). The residue of the in- tegrand at thie pole is (—i(a* — 1)-!). Finally then I = 2xla? — 1)", (b) For |a| <1 we write the roots as z, = —a + iT — a". Now for a = dy + ie, where 0 < @. <1 and e > 0. one has, as e — 0. 24% — (dy + ie) $iVT— Oh + and to first order in e, —(1 — af)? and I = —2ni (1 — af) 7. (c) For @ = —1 we may write —1 + cos = —2sin* (f/2). Thus , which diverges.80 MATHEMATICAL Puysics—SoLvrions 18. Making the substitution u = sinh z, the integrals become pe du pe du n=fry Fu and n=f a + ue ‘This form is much easier to handle by the methods of complex integration. Both integrals may be closed in the upper half-plane with the integrand of I, having a pole at u = +i of residue 1/2i, while the integrand of I, has a double pole at u — {i and thus a residue id=fz=e0 === 1 ====e dul@t an = Thus we finally obtain =x and I= ary 19. Let z= circle: ‘*; then the integral becomes a contour integral on the unit pa §% _aet +l) + 2be__ Yi & + Blayle + afhy(2ab) There are poles at z = 0, 2, = —a/b and 2, = —b/a. (a) Take |a| > [b|, Inside the contour there are poles at z = 0, 2, = —bja: yp 2af__1_alb at +1) + 2b(—b/a) , ] ** DabL~ (ja) (—bja + ab) ste (b) Take |b| > [a]. Inside the contour there are poles at 2 = 0, 2, = —a/b: 2xf__1__a(a’/b? + 1) + 26(—a/b) , ,] zabl (a/b) (oja — ajo) J 2x 2x, Fa t= > 20. Integrate z*-'e-* around the contour shown. fertetde because no singularities are enclosed. But. See = 0 as the radius of indentation {| —+0 a8 R—+00. On T,, we have [* t-le~! dt: "OTs a Jo sit Fyre . Teste tdy = emt [yr te dy We). [ieesleeeslecees t mMATHEMATICAL Puysics—SOLUTIONS 81 Multiply by exp (—ziz/2) and take real and imaginary parts to get the desired results. 21. We choose the rectangle with sides y = 0, y = 1, « = +R, with in- dentations as shown below: ry M% Pe ml m = Fea ar +h ~ ff - 15 ede, nef" edz, 1 ) a sinh xz sinh 2” sinh x2 = aun) ean = Lth= | atorrm dy + Same Dee SS: eatzet) eaten Lth=f" Rglagas if . . sinh x@ +4) « sinh x(—# +a) [xe — Near 2 = 0, e/sinh xz = 1/72, and near z =i, e/sinh xz = — therefore ee xe 1 Now sinh x(i + z) = —sinh (xz) = —sinh x(i — 2); 80 F eieens dz (* enter ae t+h=) gnhez — J, ainhxz end ~ (en — o-22) melee — et) a A+ Gares a+ (ee te — m0 as R—> co and e—+ 0. Therefore £™ (ea — 9-92) _ f@ sinh ax dr Jo “Qsinh we ©” ~ Jo “sinh wx 1 11 a =e eee 22, Thoro are poles of first order at z = +4; the square root branch cut is taken ftom 0 to co on the real axis. The contour shown encloses the pole82 MaTHEMATICAL PHYSIC8B—SOLUTIONS at z = 1, and never crosses the cut. One has 2dz _ x(1+4) FH1~ JE rs ao gy [7 etd =a+as, @+h [ | Pole ot from which av [7 de x J Pote ot ~Y Jo + 72" I We have used the fact that /@ = (1+ a)//2. 28. The function ein (ex) has poles at z (n is an integer), the ln. LBiet sin (ey round « counterclockwise contour enclosing Suieaae only at the point z= n(n % 0) yields 1 qy aimmame- Se The sum over n yields (= _ aE ~ KS ram 2 sin (ex) ’ where C is the contour shown in the probiem, This contour may be closed by adding semicircular contours above and below the real axis. These ares add nothing to the integral. Hence S(cy"_ lg _d& Eat 7a Te eine where C’ is the clockwise contour enclosing only the singularity at z = 0. ‘The function 1/(2* sin 72) has a residue 14z*/720 at the point z = 0. ‘There- fore tne txt (aay _ txt, on 720 2%, (a) Possible multivaluedness of F(z) arises from two sources: (1) The multivaluedness in the definition of the logarithm, ie. we log @ = log || + 66 + 2nip, where @ = | |e” with —x < 6 < x and pis an integer. (2) Having made 8 definite choice for the branch of log «, there is still two-valuedness in F(z) due to the two-valuedness of p(2).MATHEMATICAL PHYsIcs—SOLUTIONS 83 We define p,(z) = We — a)jz, where lim p,(z + te) = for z <0 and z >a; or Fa then on the second sheet of p(z) we take py(z) = —py(z). For a fixed value of p, F(2) is a two-sheeted function with Fy) = V2 = hog w_(2) on the first sheet. and Fale) = — 4/2 log o.(2) on the other. Here we have defined (ay One should verify that. { @.(2)@(2) = 1, } w4(e) aud wie) <0 fore >a, is} eo and o(2)>0 for z <0. Since the logarithmic branch cut is superimposed on that of p(z), the only branch cuts of F(z) are for z <0 and a a . = 2 Mog « (a + io) + log m (2 — io)). Now for x >, w-(z) is negative and one must calculate the sign of the imaginary part of @.(z + io) to decide whether the imaginary part of the logarithm is (ix + 2ip) or (—ix + 2zip). A simple calculation proves that the imaginary parts of both w(x — io) and @.(z + io) are positive.84 MarHematicaL Puysics—SoLuTions Thus one finds D,{x) = (2xi + 4xip) for x >a. Fy. (b) The function Ge) _— SO ™~ is analytic everywhere in the finite plane Se except at the branch cuts of F(z). Hence for a point z not on the branch cut we have mM——_—_€$UL_E™ 1 gO) de! a=, foe. where the contour is shown at the right above. Since the function G(z) goes to zero as jzj—> oo, the circuiar portion of the contour may be expanded to 2—> co and gives zero contribution in this limit. Because of this, the contour reduces to id Fe) — Fe) _ Alf _Disyde iE tts 2% -= (6 aNe— 7) Ou the one sheet fur which Dye) = 0 (uamely ihe principal bras we have 1 of lug @} FQ) = Fle) + @— 4) [Oa 25. We make the substitution y= »/nz, and the integral then becomes . (* dy tm [ras a+ vin" i 1 =e becomes [levaMATHEMATICAL PHysics—SOLUTIONS 85 A more difficult procedure is to evaluate {7 dz/(1 + 2%)" first by contour integral methods, and then take the limit. 26. We note f(a, b) = f(bja) = —f(a/b). Let y = bja, and take the deri- vative: * (ese) = Sy), fay = frends = 4. y Integrate, to get f(y) = In (y) + C. Evaluating when y = 1, or a = b, we have JQ) =0 =n (1) =0 46, 0rd =0, 20 f(a, b) = In (b/a). 27. s=(Z)etarte... )= Lie -i(-)-0-e 28. (a) Dividing the generating equation F(z, t) = Dive He(zytt/k! by et and integrating over a closed contour in the complex t-plane enclosing the origin, we obtain nl fexp [at Stay) a umd A, (b) One forms the quantity OF , 92k 5,08 gee + tS — Ge and easily verifies that it is identically ceio, Using the expansion for F in terms of H,, this identity takes the form yy Hite) — 2eHta) + 2nH ey gor alt Hence Hy(z) — 22H(z) + 2nH,(z) = 0. (c) Differentiating the integral representation for H(z) we obtain dH, _ 2n{(n— DN ge w-t-2)1y de Qnt dae ae =2nH,.,(z). 29. Differentiating @ with respect to r, one obtains Taber eae — BP a86 MarHEmaticaL Puysics—SOLuTions differentiating @ with respect to gives Tater Frye EP o) Biiminating (i — Zar + r*y** between (i) and (2), one obiains E @ = rye Pile) = F Pie) which, upon equating coefficients of equal powers of r, gives Pix) = Pi_{x) + IP (x) 80. Dividing the equation eome-ien — FB I yyet by 2", and integrating along the unit circle in the z-plane, we obtain Tal) = gg 2 @ which, when anhstitnted in the ahove ex. efH™te-GAN dy, On this cirele however, 2 = pression, yields Tale) = x f cos [sin 9 — n6}d0. 31. The Green’s function for the problem is foynd from =VG(er, r') = 4x8(r — 1’) with G(r, r’) = 0 for z = 0. The solution to Vy = 0, with yp given on the plane z = 0, is We, we) = f de! dy bte', 29, Tt can be checked that. G=[e-2 + y—yh + e—2)T" Meet yy + e+ ey, which is found using the method of images, common in electrostaties. Thus 2 at sen gaya, 2 9/'HG + a ety yy 82. A Bessel function of nth order, Jq(z), satisfics TM) 4 The) + ot — aT) Therefore Ke(z) = Joliz) satisfies 2"Ky’ + 2K; — 2'K, = 0. By hypothesisMatTHEMATICAL PHysics—SOLUTIONS 87 K(x) = Jf exp (—x cosh $)d$, whence Ky= i cosh? exp (—z cosh $d and Ky = —[” cosh $ exp (—z cosh ) dp. When Ki is integrated by parts, the relation 2*Ki + 2K; —2*K, = 0 is easily verified. To find the asymptotic form K, = De-#/z'". make the sub- LSE for K and K-sef ap [-= cosh f] dy, and for large z make the expansion saree A + cosh fem = + $+ Oy'/2). ‘Therefore for 2 > 1 we have Kexti(evtdy and D=[ ewtdy=(. z Jo 38. By the divergence theorem, {r-dA = f dive-dV = 3 {dV = 3V where V is the volume of the torus. Let the inner and outer radii of the torus be of revolution. Consider a circle on the zy-plane, with center at (R, + R,)/2, and radius (R, — R,)/2. The volume enclosed, when this is rotated about the y-axis, is, by Pappus’ theorem, the area of the surface times the distance traveled hy the center-of-masa of the surface. or x(“@54) x on ($4) = TR + R,)(R, — R,)*. Hence the intogral is (3x*/4)(R, -+ R,)(Ry — B,)*. Note: The reader should verify Pappus’ theorem. Ry 34. The volume is given by V = { dz, dz, dz, dz, over the interior of the unit sphere. We know from three-dimensional spherical coordinates that88 MATHEMATICAL PHYSICS—SOLUTIONS dz, dz, dz, = 4mp* dp where p is the three-dimensional radius, p = r sin}. The volume integral then becomes V = 4zp' dp dz,. But by analogy with two-dimensional geometry we also know dpdz,=rdrdqy. Therefore V =f drdzr’ { dg, sin’ $y. The only problem now is to find the limits of the integration over $2. These may be determined from the constraint that p be of only one sign (by convention p > 0) so that the transformation between the 2’s and the is singlo-valued. Tho limits aro 0 < $2 < x. Thorofore = mee SS aoe ae at V4) dorsin® >.) rdr=F- ‘A more general approach calculates the volume of an n-dimensional sphere, with coordinate axes 2, .. . , using the integral, ecotoanb pee = feted dey = a8 = f, eB AR dR, be d a Let y — RY, and the integral is Apres youn indy = 4 ener dy = 4 1($): Hence 2n"?/T(n/2) and the volume of an n-dimensional sphere with unit radius is footy on 1 -f= JlARri aR 4 = ey ‘Speciat case: n = ZN (1.e. n even). Then 1(N) = (N — ji and V = This has the amusing property of going to zero as N — oo. 35. We take the density of air and helium at a point x in the pipe to be rls From the and lack of temperature gradients, we deduce that eal) na(2) + ma(z) = N = const; otherwise a pressnre gradient, wonld exist. in the pipe. The Aux of air in the f= pits + ng. ‘The term —D(dn,/dz) is due to diffusion of the air in He while the latter term represents the convection of air by He. In the atcady atate, the con- tinuity equation becomes V+ f 0 (ie. (2f-/@2) 0 in our case) and theMATHEMATICAL Puysics—SoLuTIONS 89 solution for na(2), which goes to zero as x —* —0o, is a(t) = Ner=*, We have taken n(0) — N since we are told m4(0) — 0. (z) na(z) + ma(z) 2/2. 36. (a) Vin + A*n=0. The iowest soiution is obtained by setting n = (sin kr)/r. But n(R) = 0, 80 kR = wp (p an integer). Therefore R = x/k. Solutions with p > 1 lead to negative densities, and must be ruled out. (b) The neutron density in the surface layer, n,(r), satisfying n,(R + t) = 0, is na(r) 4 sinh [u(r — R — #)]. In the interior of the pile we have a neutron density, n,(r), where m,(r) = (Bjr) sin (kr). The boundary conditions require: (1) Bain (eR) = —A sinh pit, (2) Bk cos (kR) — Ap cosh pt. Therefore ihe radius, 8, of the interior region is found from ptan (kR) = —k tanh (pt), which for 4 > k allows the approximation kR ~ x — (k/u) tanh (ut). Thus the new radius is shorter by the amount. tanh (yt)/y 87. Choosing a Cartesian coordinate system with z coinciding with the axis of symmetry, the solution to the steady-state diffusion equation —DV*n = Q8(r) is nle, v2) = gs BF cos =10%sec: a=75em: be = SQ + 1 + OE + TY, and the boundary condition n = 0, for 2 or y = -La, is used. On the z-axis the flux j, = —D(@n/dz) is given by a= Seyemn, It iu interesting to note that if z I? = 2Rt. 5. The range of a droplet sprayed from an angle @ is R = vf sin (2a)/g, and its variation with @ is given by dR = 2v§ cos (2c) dat/g. The amount of water falling on an annulus of radius R and thickness dR is proportional to p(a) sin @ da — eR dR, where c is a constant. If this is to be independent of a, sin (Act), sin (2) 0 0 if 3—(n +1)>0, jen <2, 9. From energy conservation, one has, in terms of the relative coordinate rand the reduced mass p = mym/(m, + m,), (1) (ar) _ Gaim _ _ Gru, \2u) \dt) T To The constant rp is obtained from the centrifugal equation F = Gmym,|r = po" |r; 8 Gon, tm (Kepler's Third Law) (2) that is,‘MEoHANICS—SOLUTIONS 95 From (1) the time ¢ at which the particles collide is ‘02 arta where 0 = a/ gq arc [a 040 = 205 Ht = Fu) - (2) Taz. Alternate solution: ‘The “tall to the center” is regarded as the limiting case ot elliptical motion. As the eccentricity of the orbit approaches unity, the orbit appears as shown. The collision point is marked by X, 0 the collision time ¢ is one-half the orbit time. One now uses the property of Keplerian urbiis that the periud is proportional ie (— where & = T + V is tie total energy. This may be verified from Eq. (2) of the preceding solution. In addition, when the particle is stopped in circular motion and allowed to fall to the center, the energy is doubled. This occurs because, when the from the 1/r dependence of the potential. Thus Z, = V,/2 for a circle, and when the particle is stopped the energy is all potential, and Z, = 2h. Finally, then, al~ 10. Assume that the man exerts constant force on the ground, until he leaves the ground. Assume also that when jumping from the surface of the moon he exerts the same constant force. Considering that the gravitational acceleration on the moon is ram al) = then, since the force is the same on earth as on the moon, the work done is the same, and Mgn(hn + 50) = Mg.(60 + 60). Hence hy = 6.1 m. 11. Let the center of mass of the rod have downward acceleration 2. Then W— F=me. a) The angular momentum equation implies WL _ zr (2)96 MEcHANICS—SOLUTIONS where I = moment of inertia about an end = }mZ*. One has 6 = 4/(L/2), a relation that is true for small 8, i.e., short times. From (1) and (2), F = W/4. 12. The vertical force F, exerted by the plane on one of the cylinders is fixed at the value 3W/2 where W is the weight of one of the cylinders. Since we wish to minimize the angle which the total force makes with the verticai, we shouid mini- mize the horizontal force, M',, which the plane exerts on the cylinder. From the force diagram we see that we must have: Verticai equilibrium: W + F, cos 30° + F, sin 30° Horizontal equilibrium: F, + F; cos 60° = F, + F, cos 30°. F,, has a minimum when F, = 0, for which we find F, = W/(4 + 2/3), and the tangent of the minimum angle is given by tan 6 — 4(2 + V3). 13. It rolls in the direction of the pull. When rolling without slipping, the bottom point P of the yo-yo has zero instantancous velocity. Rotation around a fixed axis passing through P is determined by torque about that axis, but only F has a nonvanishing torque. Therofore rotation is clockwisc, and the center of mass moves to the right. 14. Denote by B the point fixed on the disk; 8 is the angle through which the dog travels on the disk, and a the angle the disk travels. At t the ium of ihe disk is (337 F7/2)jed (uve tie parallel axis theorem), with respect to an axis through B, but fixed in space. The angular momentum of the dog is [& — (8/2)|[4mR? sin? (8/2)]. Setting the sum equal to zero we get da _ a8 4msin’ (6/2) ~ 2dt3M/2 + 4m sin? (6/2)° t ! t 1(* 4m sin’ (8/2) i 2 i} 46 3H/[2 + 4m sin® (8/2) We can write this in terms of y = 4/2. Then A the angle is given by the expression iB Dog I __4meosty _ 28 sin (8/2) f {73M 1 4m coy Bee] Note that the angle depends only on the maxs ratio of the dink and dog.MECHANICS—SOLUTIONS 97 15. Conservation of angular momentum implies Iva. = Ja with I, = 2M R/5 and I — (2M/5 + 2m/3)R*. Now 2M 8x 2m 8 am 5 iF? and 3g ae. Therefore aud henve (2 — 7o}/T'o = 5 dj DR. : cls? 16. Since the earth is rotating with angular velocity ©, the angular momentum expressed in a coordinate system on earth satisfies y au _y_ aN axl, / x where Nis the total applied torque. Looking down on the gyroscope at the equator, where Q is along the y-axis and the z-axis is vertical, one may write the foiiowing expression for i L, = Cosin $ = Cop) Ly = Co cos $ = Co L,= Ad. In this expression terms of order «/<1 have been dropped. Since there are no forces in the zy-plane, N, =0, and the equation for L, becomes —Ag = CoO. Thus ¢ oscillates with angular frequency v, where vt = CaM/A. } for /$1 <1, 17. When the total angular momentum is expressed in an inertial frame of reference it satisfies (dL/dt)nsra = 0. However, the principal moments of inertia are given in the hody-fixed frame of reference, rotating with an angular velocity Q with respect to an incrtial frame. The connection be- tween (dE /dijineninn aud (EL/dejoay is (AL dt) inertia = (4L/dt) cay + O x L = 0. In the body frame we then have the equations: a (2..81,) = 0, (1) d 3 Ga.) + 3 1.0,0,¢ 008 ot — @) £0) — 4 160,06 008 ot = 3) where J, 2mrt/5. Equation (1) has the solution 2, = Q4,/(1 + € cos at)98 Mecuanics—SoLvrions and Q, varies little with time since e< 1. From (2) and (3) we see that 0,4 + 0,9 rotates with an angular frequency w, = 3.0, 008 wt when dI,,/dt and dI,,/dt are neglected, i.e., when o < Q,. 18. For suifficientiy smaii dispiacemenis, motion aiong the direction of the rods is decoupled from motion normal to the rods. Denoting displacements along the rod by (21,22, 23) and those normal to the rod by (y:, 4293), We find that the Lagrangian is a straight line. ‘Motion in the z-direction is governed by the equations mi, + k(x, — a) = 0, — mes + k(ws — m4) = 0, 2mé, — K(2ry — x, — a) = 0. Conservation of momentum requires that Z, + 2%, + Z, = 0. This condition implies that there are only two normal modes for vibration in the x-direc- tion. The normal modes are a a — 2, with frequency (k/m)'* an x — 2xy +x, with frequency (2 k/m)'*. Motion in the y-direction is constrained by conservation of angular, mo- auenium, as Well as linear momentum. Consequen yy there is only one mode of vibration, with frequency (4é'/m)'*. The modes may be sketched as in the accompanying figure. 7 ¥ ua 19. Let the vertical displacements from the equilibrium positions be X,, X,. For the motion of the center of mass C, F = ma yields Fd + X= H+ Ey and the torque condition gives (for small X;, X;) $4, — y= Sas - x) @) and J, = moment of inertia about C = ML'/12.MECHANICS—SOLUTIONS 99 The gravity term, which merely determines the unextended spring lengths, can be transformed away. It does not affect the modes. From (1) and (2) there are obviously two modes: (a) a symmetric one, X,= 2, +2, with frequency o} = 2k/M, and (b) an antisymmetric one, X, = 2, — 2, of = 6k/M. 20. This problem was first studied by Daniel Bernoulli. Denote by » = m/L, the mass per unit length of the string. Application of Newton's, Sec- ond Law to & smail element dz of the string gives the equation a 24), ato pan" 0 where 7’ is the tension in the string. Because the rope is in equilibrium with gh, Me) = gf a/L) +}. The propriate boundary conditions for y(z, t) are HiL,0)=8, (2,0) =0, i Our po WOH — 6, Fal, i) — 0 ‘The second of these is satisfied by putting (x,t) = y(z) cos wt. Then y(z) satisfies an ordinary differential equation te |r) #| + oty =0. (2) With the ansatz y = f(x/T(@)), this reduces to the Bessel equation of order zero: with the general solution 2) ~ 45,(22,/7) + 2a, (22,/7), BING ) we} where N, is the zeroth Neumann function (i.e., Bessel function “of the second kind”, having a logarithmic singularity at the origin). The first and third boundary conditions yield the coefficients: y= [stem ete orem) (7B) J(22,/ 42), (22,/£(t-+-m))—Jo(22,/ Lt +m (22, /¥2) L Bs - 7 oN » 7 B/S The fourth boundary condition then gives transcendental equation for «.100 = MECHANICs—SoLuTIONs In the case that m < M, the arguments of the Bessel function are large, and one may apply the asymptotic expansions: Tule) —> os (vose + sinz), No(2) —> Fao —sin2z), as z increases without limit. In that limit, y = 5(z/L) cos at. In place of the fourth boundary condition, which is satisfied identically, we use the equivalent condition oy =~ = gf te to obt: gf 1E. 21, Waves on a stretched membrane obey the differential equation fy = TV. Letting y = u(x, y) cos ot gives —o*ou = T'V*u. Consequently, T [uViuda w= ft . a fudd ‘This may be computed tor various trial tunctions to obtain an approximate eigenfrequency. The trial functions u must be chosen so as to satisfy the boundary conditions, in this case u = 0 on the boundary. 22. On the the e: increased by @ slight amvund beyond ite value in vacuum because uf the small viscosity of air; the flywheel tends to drag air around with it. A watch made to run on the earth’s surface is adjusted to compensate for this effect. When removed to a high altitude, the moment of inertia is decreased ding to Professor effect was noticed by Professor I. Rabi, who was very proud of an extremely accurate watch, which ran slightly fast when he went to a mountain top for cosmic-ray experiments. Rabi proposed the problem to Professor Fermi while they were riding @ train during the war. as a diversion for Fermi. business on the train; they could not fly for security reasons.) Formi ox- plained the effect immediately, and in about an hour's time had worked out a complete quantitative theory of the phenomenon. 23. Young's Law requires AF/S = Y((AL)/L), where AL is the extension of the atring, and AF ia the foroo felt by the utring aa the masa decelerates.MECHANICS—SOLUTIONS 101 The maximum AL is determined by conservation of energy: mg(L + AL) = Sri (Azy. = 59+ VRS) + ae In order to prevent: breaking, we must have A F/S < 1; therefore ya (3 \2mg from which r\. ) 24, The train must be stopped before the spring is fully compressed. Other- wise, if the jump in the spring constant is large enough, the acceleration wohave F = ly — 1) = Ma, and a will be largest when 2 is smallest (i.e. when the train is stopped). At this point the energy equation tells us that }o(lo — 1) = }Mv*. Eliminat- ing ky we obtain since 1 < Is. Therefore the minimum value of Jy is given by 1o(0"/dmax)- 25. (a) We use Archimedes’ principle to put px Rh? — —paga Rtz. There- fore w" — paglph. (b) Consider a slab of fiuid of thickness dr and of area A parallel to the cylinder. The net force on this | five slab due to viscosity is which must oqual (p¢4 dr)(2u/2t). For a sinusoidal time dependence o*! one has with solution v = v,e~"*® where the damping length is given by = \/2y/pow. The total viscous force acting on the cylinder is thus F= AB = of afarr ery —u + LET sca The magnitude of F is | F| = ee lel. 26. Consider a section of the surface between the planes Z = Z, and Z = Z,, In order to have equilibrium in the vertical direction we must have (27 cos 6,)(2xr,) = (27 cos 6,)(2"r,) or simply r cos = const, where102 MECHANICS—SOLUTIONS cos @ = [1 + (dr/dZ)*]-"*. The equation for the surface is thus () =1+ (az) for which the general solution is r — r, cosh (Z/ro). (The origin of Z is taken midway betweon the two loops.) We calculate rp from the condition a=, cosh (d/re). Define z = (d/r.); then « satisfies the equation (a/d)z = cosh x, which has 4 solution for # only if (a/d) is greater than some minimum value. This can and cosh z have only a single intersection. Therefore the x corresponding to this value of (a/d) is found from x = coth z, for which we find x = 1.2, and (a/d) is calculated from (a/d)*(z* — 1) = 1. Therefore the maximum value of (dja) allowed for stahility is 0.66 27. Let the bending be defined by y as shown in the figure. One must first calculate the bending moment through a section of the strut at a height z. y “7 a Consider a plane scction of the strut at a distance Z from the mid-section. The length of this section, when the cylinder is bent through an angle 6, is (H + 2). ‘The strain is theretore ( + z)/, and the restoring moment = Ya“ . wale" (R+ ede = 7 The radius of curvature R is -#h+(@y”‘MxEcHANIcs—SOLvTIONS 103 We shall consider only infinitesimal deformations for which we may neglect (dy[dz) a8 compared to 1. Thus the restoring moment at height z becomes Yat dy. 12 dat In order to have equilibrium of the portion of the strut above this height, the moment must equal the torque of the weight about this point, W(e — y), where e = y(l). We therefore have the equation N= ag’ = We — y) Te ae = Me. for which the gencral solution is y = A sin (mz) + B cos (az) +e, where o* = (1217/ Ya"). The boundary conditions give conditions on A, B, and o: 0) 0+ B=, ¥'(0) =0— A y(Q) = € —> Asin ol + Boos ol = 0. The solution ia thna: y = e(1 — cos wz) with € cos wl = 0. For al < 7/2, the condition € cos o! = 0 can only be satisfied by € = 0, i.e. the vertical position is stable. However, when W increases to the point where ol = 7/2, an infinitesimal deformation is possible. Hence = (x* Ya"/48!*) ia the beginning of instability. ae, required io bend » segment of ace (du) aud longi de throught wn angie (d8) is Yau(d6) dujdz, and the total momént about # point P of those forces acting across the A B-plane is M— fs (dé) du ~ (45)2- For small bending, (26) = dz(2*h/22"), 80 x 3@) In order to have rotational equilibrium for the section of the beam to the right of a point z, one must have 2 Hi) = f poat (L — 2')dz’;104 MEcHANICS—SOLUTIONS hence oh ae The houndary conditions are h(0) = h'(0) = 0. Therefore hi: !2¥a) [6L%e? 412° | 2*] and A(L) = (3pgL*/2Ya). Note that the moment M and force @M/éz vanish at the free end. 6, SPL — 212 + 24), 29. Imagine the chimney separated at a distance | x, SY x above the ground. The rotation of the chimney | WA LZ as a whole about O satisfies the equation &y uL6 — Mgh sine, o (6= swan. ( a Likewise the lower purtin setisfies the equation Mx6 _ Mzigsing BL COL where F is the = force and the internal flexion torque at the point x. Similarly, ths the equation +2F ” (2) n of the upper portion ahont ita canter of masa satisfies M(L— 2) 5_(L—2)F pr 8s yt (3) Equations (i) through (3) may be soived to yield alt 2| _ Mgsin (L P= DL + 2) [ 4 and 41 Since the chimney is thin, i.c. w < L, whore w is its width, the forces respon- sible for y are very much greater than F. That is to say, y/w >> F. Thus the breaking is dominated by the maximum of y. ‘This occurs at x = L/3. 30. Consider a small volume element of liquid near the surface. The potential enorgy/volume is V = pau y The equation of the surface is obtained by setting the potential energy equal to a constant (in this case zero because ofMecuanics—Souvtions 105 31. From the expression for ¢ one finds the components of the velocity 0, = v4(1 — RY/r*) cos 6 vy = —v(1 + Rt/r*) sin 6. Because the door is open, the pressure F inside and ouiside musi be equai, with value P,. From Bernoulli’s equation for streamline flow, at any other point on the exterior surface, P+ Lov) = Py (P — Pa) = —L oof) gives the pressure at any angle 6. Then P= f(Pr— Pyedd = Lot (2 + Py fj sint on ao, Since the oscillations are incompressible, the pressure satisfies VP: everywhere. For boundary oscillation of the form 7 = @ exp [i(kz — at)], the oscillations cause slight changes in the pressure from its static value which damp ont. as one travela away from the interface P, = —pige + Bi exp [—ke + i(k — of] and P, = —pyz + Bz exp [kz + i(kx — ot), where &, and 8. are small constants which go to zero as 7 goes to zero. The boundary conditions are (Ql) Py = Pratze=q (2) From p(av/at) = —VP we have pj = —@P,/éz and prj = —@P,|dz. Remeubering that y is suall, we fad thet these yivid die eyuativus (1) —pge +B, = —pget + Br (2) —otpja =k, and apt = Kf ‘These equations are easily solved to yield (02 — pidkg, (02 + pi)106 MxcHaNIcs—SOLUTIONS Note that if , > ps the frequency is imaginary, and thus the oscillations are unstable. This is expected, since in this case the heavier air should fall. The phase velocity v — w/k is given by To relate p to the temperature one may use the ideal gas law, i.e. p = K/T, hence Lan(Pr + Tid 33. The velocity will be taken as irrotational, for which we have v = Vo and V-v = K8((z — a), (y — 6). Due to the presence a the walls the flow t 2 (0,4) — 0 will be satisfied if we consider the flow in the region a interest to be due to the original source and image sources of strength K located at (z, y) = (—a, 6), (—a, —b) and (a, —b). For a source of strength K the solution to Vid = KAllx — a). (y — BN) is Klog Sea + yO The velocity potential for the flow with boundaries (which has a velocity which goes to zero as the distance from (a, 6) increases) is $= Elo Ve= a FF + log Vea FUT OF + log We + a) + (y — BF + log le + a)? + (y + 5). ‘Use Bernouili’s equation to reiate the velocity to pressure; that is, F(VOF + P= Poy where P, is the pressure as vy — 0. Thus the pressure is given by =P. —t0o(%) [2-9 | : Pa P—w0(F) eta t eet] Om P=P,—49(2) [Sat wit)! on OA. 34. Cousider a body such as the muon, lucaied ax shown in ihe accompany- ing figure. The shape of the water surface is defined by the spherical coor- dinates (p, 6). In the center-of-mass coordinate system, which is rotating with angular velocity @, a unit mass of water on the surface is acted upon the pot V(p, 0) = Ve + Vn + VosMxcHanics—Souvtions 107 (gravitational potential due to earth), Vu= pe tp esa? — (eravitational potential due to moon), tnt V,=—27 (centrifugal force potential), with 7° = p* + £* + 2%p cos @. In order to obtain @ in terms of the given constants, note that, and tg — Man 7m Mote — OM tay It should be pointed out that in the above, the axis of rotation of the carth- moon system and the axis of the earth’s 24-hour rotation are taken per- pendicular to the plane of the drawing. The latter choice is made so that idal changes will be seen in a 24-hour period as defined by the Upon expanding Vm in powers of (p/r) and neglecting terms of higher order than (p/r)*, one has Vin = 2 ml — £0086 + £53 cost 9 — 1) + vel. In order to calculate p(6), one notes that the surface must be an equipotential. Thus, Vip, 6) = —Gm(ZMe + 3m) _ Gmmpl rg sagt g — 1) — Woe + m) 2 _ GMe ee r(M, + m) aS , ar) = const on the surface. Setting p = ro +h and neglecting terms quadratic and higher in h and terms of order ((M. + m)/M,](r3/r4), one obtains the equation for h 3 ri — ants cost 9 — 1) — Saha me mre ond = const.108 Mecuanics—SoLvrions Hence pea 4. Brnrb cost 6 b= hy + hs given by Ah = (0) — Alw| ID ratio of the tides duc to the sun and moon is thus (AR)son 7 apts. The presence of the cos* @ in h(8) shows that there are two tides in a 24- hour period. 35. A small but otherwise arbitrary deformation of the surface from a sphere will be represented by A(A) as shown in the figure. The function h(A) may be expanded in sphericai harmonics to 4(a) = K Dim AimY¥ in(ft). The gravitational energy is = 1 Gplryo(r’), u=- favav’ Seeiee ) = palt) + Splr), where p, is the density for a homogeneona sphore of radius R, and the onorgy becomes U=U+U,+0,, where =-% 71 Polt)polt”) u,=-$ favav adele) is independent of h, and v7, = —g [AV AV onBotr’ “J f serdar d¥" w=- =~) Spr'igir' ar’, Tr— FT where $(7’) represents the gravitational potential produced by a homoge- neous sphere dtr’) = @ fener _ _@ faV AV" 8plr) Bor’ = — 5 fee At thia point note that Sp(r) ia nonzero only in the shaded region, and that in these regions 8p = {p> depending, respectively, on whether h is positive or negative. Thus, for instance, the expression for U, is a (ee ar yaw wy ar pf By a Taylor expansion of the integral, keeping terms up to second order and dQdg’ W=—MEcHANIcs—SoLUTIoNs 109 in h, Us becomes GpiRt 1 h(a)h(a’) | SEE fanaa eet Similarly the expression for U; to second order in h is _ anager? et U,= —S26BF | a0(Ra +4). Finally, the total gravitational energy to sccond order in h is ve wom { aa( pe + 2) _ GR 0 39 aq Mahia’) U = Ue — pam J aa{Rs + 7} — SRE fans Te { Since the water is incompressible, one also has the equation en (°°? dQare = which, to second order in h, becomes Rf dann) = — Jaane) (2) is used, one finally obtains AxGptR> —1) gs v= + ee ys has mim Tt. should be noted that. because of incompressibilit Hs 0) m (Raq. (2)), the mode @ O}-mude, Otherwise, the (@ = 0)-mude dues occur; however, it is of second order in smallness. In addition, for the mode with 1 = 1, corresponding to uniform displacement of the sphere, there is no change in the gravitational energy (pu/2) fa 2 poRHD) { sagan) dO wl v = V¢, as in irrotational flow, and Y-v — 0 as for an incompressible fluid. To find } we note that be if, t) =v-n = aR), and expanding ¢ = ¥ bim(r/R)' ¥im(f) one finds Bim = RPA /t. The kinetic energy thus becomes110 MEcHANICS—SOLUTIONS The total Hamiltonian for small oscillations is the kinetic plus potential energies, : -y nda + [SOR — UN ara H= % (CE) Aten + [APCD] tnd) ‘The frequency of the mode A,» is thus ot — Sz — 1), " ‘B(21 + 1) or, in terms of the acceieration of gravity, = (42, p) 8 2 — Ig. = (Foor) fe = RF 1) The period of the lowest mode (I = 2) is r = 2x |, = x(5R/g)'* = 6640 sec. 36. From the relativistic transformation of coordinates, one may write tam tars = m( to); and te = 92(¢ — av»). But z, = 0 since the clock is at rest in S,. These equations are easily solved to yield ¢, — 9. (1 — o,e,¥. 87. At time ¢, as measured in the rest frame of the star, let the rocket have a velocity v. Consider the Lorentz transformation from this frame to an inertial frame moving with velocity v: 2 = y(z’ + uf’) and ¢ = y(t’ + v2’). - de _ vt (av, ana Fe [ae = 14+ a ee vf +o(& T we have the traneformat yt, We n law for accelerations along the time intervals dt’ = dt/y. The total time elapsed as seen by the moving clock is or » T= age oa), + a} gS ees esses celeration as seen by the rocket. To calculate v vs we notice that “the | total distance traveled is given by [Sp ¢ e=s[o44 uF Elimination of v, between these two equation gives 7” in terms of D and a’MxcHaANIcs—SOLUTIONS ill 38. ‘The period ris given by 7 = 4 {"de/v, where the velocity v is calculated from the energy equation met(1 VY, mort? _ pugs 4 mora? eT ey) te at In this manner the expression for the period becomes A(t __ (tot = xy} eye + -#)| T We expand the integrand in powers of o%a*/c?, and get, as the leading terms in the expression for the period, + = (2x/a) [1 + Bota"/16e | ---]. 38. Define the four-vevior p = py + Pn = Bp + Po, where, in the iad frame, p = (m, + mp, 0) = (M, 0). Solving for p, = p — p, and squaring, one finds . ‘mn = Ph = (P — Pr)” = (M — Hy) — pr = M* — 28M + my. Therefore Mi +m — OM 1.24 BeV. E,= The lab frame was chosen to evaluate the scalar product p-p,; since it is a Lorentz invariant quantity, any convenient inertial frame may be chosen. 40. (a) Any collection of particles has a center-of-mass velocity _ P (total) Et (total)’ as can be seen from the Lorentz transformation of momentum p= oP. — Bv); p', = p,. Choosing v parallel to p, we see p’ = 0 when v = p/E. In our case the result is v = (p, + p-)/(Z, + E-). (b) The totai energy and momentum constitute a Lorentz four-vector; hence the quantity ((, + B_)' — (p. + P- yA is an invariant. The bary- centric frame is the one in which p', +p. = 0; as m, = m_, one also has E', — E'_ in this frame. Thus 4(E.? = (E, + E_)* —(p. +d.) or finally i =)? = (ps ay p,- 5, -VEFEY= wey. (c) Consider the invariant J = (p, — p.)* — (£, — £_)*. In the rest frame of tho oloctron, p. = 0, E. =m, we find that EB. = m//I— of where112 Mrcuantcs—Souurions tea is the relative velocity. Thus 1 oo 2m* om? _lli->— =e ant and we hare n= [ wey 41. Choose a coordinate system with z-axis to the right and y-axis vertical. ‘The force equations are and 4 incin spesceee@i te ‘Thus p cos 9 = p, cos a is a constant of motion, while the second equation gives (d/dt) (p, cos at tan 6) = —eV /d. Weuse tan 6 = dy/dz and the approxi- mation d/dt = (d/ds) where c is the speed of light. This is a good approxima- tion for a relativistic electron since its speed is approximately ¢ throughout its motion. Thus we obtain i@)-aaiea * paced + (YT The general solution to this differential equation is y=4—A cosh Ale — a), B where A — eV [epsd cos . The constants A and aare calculated from y(0) = 0 and dy(0)/dv — tan a, 42, Let one photon have momentum p in the lab, and p’ in the pion rest frame. For a photon | p| = H. The Lorentz transformation connects p and p’: P' cos 6 = p(oos 6 — v), p' = pl — v0). By division, 1 sind tong = OE, and cos Probability is conserved, 80 P(6) dQ = P'(6) dQ’. Then Pip) = Pp) H(C089) _ pypy dE __ 1 : gee *? a (cos 8) wr FL — 0008 OF Ex — v eos Oy is the normalized probability density. 43. (a) One year contains 3.15 x 107 vec. Consequently laren = Yneatan = 10° y sec = 3.15 x 10’ sec => y = 3.15 x 10‘. As seen from the earth, the neutron energy is 3.15 x 10* times the neutron rest mass, or about 30,000 BeV.Mecuanics—SoLurions 113 (b) Let variables in the neutron rest frame be primed, and those in the earth frame be unprimed. The angles made by the decay product with respect to the neutron velocity (designated as along the z-axis) are @ and 6’ in the two systems. The relation w' sin 6” tno=15 cos Fe where w’ is the velocity of the decay product in question, is derived in the preceding problem. This relation can also be found from the Einstein addi- Both the neutrino and electron are ultrarelativistic in the primed frame. For the electron, d(tan 6)/d6’ = 0 implies cos 6” = —u'/v. Then tan Ong, = We = via Now ew, = 3.15 x 104, while Yur (electron) + 2.6. We see Onux ~ 10- rad. (c) The method of (b) breaks down when (u = 1) = v, since this would lead to cos 6° < —1. The largest angle @ 18 7; we obtain this for backward motion in the neutron rest frame. (d) In the neutron frame, the neutrino has maximum energy y{B’ + v-p'] = By(l—) =F x fw 1276 2y0m. 44, We begin with a crude order-of-magnitude estimate. From Kepler’s Lawa, v ~ E*?, where E is the energy excluding the rest mass. The first- 3 Elo! ) _ (,2,) mt Rt EB ~ 2 GM/2R~ \T6c!) (G28) _ 94am = 6, ~ 2nGM | Fe This result is not rigorous. The relativistic effects may be manifested through a speeding up of the particle in the nonrelativistic orbit, as well as through a rotation of the perihelion. In fact, this is just what happens for a circular orbit. For an orbit with eccentricity, the perihelion does precess, the rate of precession being roughly that given above. A more exact treatment fol- lows.14 MEcHANICS—SOLUTIONS The radial part of the force law is ad = 2 ov, F, Zi (my?) — meyr6* ‘The time is eliminated through ihe angular momentum equation Z = myr"6, with L a constant of motion, from which In terme of which is fu 1 y- ESM, when the energy equation Z = me'y + V is used. In 9 gravitational field (neglecting general relativity), V = —GMm/r = —ku, and the force equa- tion becomes ‘" _ ke w+ oft + | an with solution TE where & = (1 — Met!) = 1 — B/2ct Lt. In one revolution @ increases by 27’; thus the perihelion advances through an angle 8 = 2x(1 — a) = wk"/c*L* per revolution. For a nearly circular orbit of radius R this becomes [1 + econ (6 — 4)], nGM Re FS Alternate solution: In the nonrelativistic Kepler problem there are two well- known constants of the motion, namely the energy and the angular momen- tum. There is, however, an additional constant of the motion, namely Aaet EXP, km This may be verified from the identity d(#)/dt = L x r/mr? and the force law, from which p= —&?/r®. The necessity for an additional constant of the motion follows from the “accidental” degeneracy of the Coulomb potential; the Hamilton-Jacobi equation (in quantum mechanics, the Schrédinger equation) is separable in both spherical and parabolic coordinates. A has theMEcHANICS—SOLUTIONS 115 property of pointing always to the perihelion of the orbit; its length is the eccentricity e of the orbit; A is called the Runge-Lenz veetor.* In special relativity, A is no longer conserved. If K is a frame moving with the angular velocity 5 of the precessing perihelion, then —d@ x A. Since or LXt ge _ — pS@-) , (A) . FA = baa x t — pO =A) + (TP), ‘2mr' dt We have chosen S@ so as to make the average of (dA/d!)x equal to zero. When averaged over the nonrelativistic orbit, the term in p vanishes. ‘Hence HLGRY. = os In the limit that the nonrelativistic motion is circular, of radius R, and the precession is given by AO AO — xQM/RCt. GM/R; thus tion of the frame, by the equivalence principle, has the same physical con- sequences as a uniform gravitational field in the opposite direction. Hence we may imagine the container stationary on the earth’s surface. Then the bel . * W. Lenz, Z. Physik 24, 197 (1924); W. Pauli, Z. Physik 36, 336 (1926); D. F. Greenberg, Am. J. Phys. 84, 1101 (1966).ELECTROMAGNETISM 1, From the symmetry of the cube the currents in the resistors must be as shown in the figure. Conservation of current at the corners requires: I=2%+y, and y=2, ° where J is the input current. The requirement YL that the voltage between A and Bbe indepen. [— | 2 / dent of path yields the additional equation 2eR — 2Xy +2)R. These three equations have z the solution « = 31/8; y = I/4; and 2 = 1/8. The resistance between A and B is (2xR/1); therefore Rup = 32/4. ae 2. Tf 1 A is fed into A and taken out at infinity, then from symmetry, }A will flow in AC. Likewiso if 1 A is taken from C and fed in at infinity, } A will flow in AC. By the superposition of the two solutions, we obiain ihe solution to the given problem with the current in AU being 4 A. 3. Place the bars as shown at the right: If 1 is magnetized and 2 is not, there will be no net af bet the magnets, by metry. If 2 is magnetized and 1 is not, the bars : will attract one another, because of the poles induced in 1 by the field of 2. 4 and g, = 0,7 for the charges on the conductor and plate respectively when they are in contact, and V is their common voltage. Thus Gl = CIC: After the first contact g.=q and g,=@Q-—g, and ultimately g, —Q; hence g. — Qa/(Q — 9) const. 116ELECTROMAGNETISM—SOLUTIONS 117 5. The battery supplies the constant power, P = EI. The electrostatic energy of the capacitor, U = qE/2, is changing at the rate dU_E EI mh dt ~ % Thus the battery is doing twice as much work as is being stored in the capaci- tor. The difference appears as work done by the capacitor on the external agent that is causing the capacitance to change. & Let @ be the charge on ons plate; plate surface; let Z, J be the interplate current and current density respec- tively; let V be the voltage between the plates. to be the charge d EedA = £ f ad A (Gause’ Law) Hence RC = elg. 7. The linearity of Maxwell’s equations allows us to think of the magnetic field as arising from two current densities: (1) A current density j = J/x(b? — a*), carried by the cylinder of radius band (2) a current density —j carried by a cylinder of radius a. The sum of the current densities (1) and (2) is the current distribution of the bored-out cylinder. From Ampére’s circuital law § H-dl = (4z/c) f j-dA, one finds that (1) produces a magnetic field H = 2Id/c(b* — a*) at the center of the hole, while (2) produces no magnetic field at the center of the hole. The resultant magnetic field, H, is thus given by H = 2Id/e(8* — a’). 8 The solution is by the method of images. We choose image charges inside the conductor, with location and strength so as to make the conductor an equipotential. To make the surface of the boss equipotential, we must place an image q’ = —qa/p at distance a'/p from the origin, on the line joining118 ELECTROMAGNETISM—SOLUTIONS the origin with g. In order to keep the plane at constant potential, it is neces- sary to add additional images —q', —g at distances —a"/p and —p inside the conductor. The force on q is then 1 _gf_4atet 7 “(paar Say t weer apt] ~~ Let att * apt] 9, At the origin, the magnetic field is (foud's _ (dowt, Fe Tse a= wa where py = —V-M=0 and doy =M-dA act as magnetic volume and surface densities. On each surface, |r| is constant, and =doy| _ doy) _ _¥ con rob Fe ‘Upon substituting in (1), H Add dd. 10. From Gauss’ Law, E = 4xc, where o is the surface charge density. E is normal to the plane. When the disk rises an infinitesimal amount dz, the energy of the field is decreased by an amount equal to the volume of the excluded space times the electrostatic energy density: dU — —(E"/8z) (Adz), where A is the area of the disk. There is, therefore, a repulsive force Lal _ AR A(anoy yas dz ~ 8x a . The disk rises when this exceeds the weight of the disk. Alternate solution: The field has two sources—the charge on the disk and the charge on the plane. ‘I'he former cannot give rise to a net force on the disk. We calculate the field due to the charge on the disk, and subtract from the total field, as given by Gauss’ Law. The potential at: a distance 2 above the plane, on the disk axis, is {* 2x ordr ee a ght gt Jo ray = eo + — a, where X is the disk radius. Then Fi = = —2no[ a2 1] = 220. Wee Jao The E which acts on the disk is therefore 20, and the force on it is F =o [E-dA= 2n 0°. 11. The electrostatic potential $(r) is a function of the radius alone. ‘This is because $(R,, 6,4) = V, independent of @ and ¢, and the continuation of thia function into the dielectric must. also be independent of @ and ¢.ELECTROMAGNETISM—SoLUTIONS 119 Thus on a surface of radius R the electric field is radial and constant in magnitude over the entire surface. Applying Gauss’ Law to a sphere with radius R such that R, < R< Ry, one obtains fo-aa = 4xQ = —20 RE J" (eo + €; cost 6d (0086) from which E = 3Q/(3e + ¢:)R*. The potential difference between the plates is Thus — R.Ri(Beo + €:) o = Bi Re re. 12. We write H — —V¢. Then, since an infinite wire in empty space with current I has field By = 21/r (Amptre’s Law), we find that b = — 21 {Im part of In (x + iy)} to within a constant. Putting an image current J, inside the magnetic me- dium at distance a from the interface, we get o = — 2 In [(z — a) + iy) — 22, In [(z + a) + iy] — 9F, In [fe — a) + iy), where $ and ¢, are the potentials in vacuum and in the medium respectively. «@ ‘tangential component of Hf is continuous. (2) The normal component of B = yH is continuous. This gives two equations, from which =1#=) oa 1,=—22 wri) eri The force on J is just that due to the image current Iy. The field due to I,120 ELECTROMAGNETISM—SoLUTIONS at the wire which carries current J is path _ ww ret le (Z/a) on the wire. The force is attractive. 18, We use + to designate the medium where » = 1, and — to designate the medium where 1 = 2. The insertion ofthe loop on the interface evokes the strength, not the character of the B field. This is because the sources of B are the true current density j and V x M where M is the magnetization/cm*. However, M is proportional to H and V x H vanishes everywhere except where j # 0. Now §H-dl = Iiqe around any loop, where I 2 , is the enclosed current. In case A, J BY =n, =H, Be) LK B= yp. =2H., | x whereas in onee B, on BI=H, and BY=H. | Choose a line of constant | | as the contour tor pertorming the line integral in Ampére’s Law. We have =fara=f_ Beato BA dl Coe pra 2f 3 Hence B* = (4/3)B". If ¢ is the magnetic flux through the loop, 7, [Rida =! = effects of the inclusion are represented in V, where V =F Aumcies YG, 9). In this spherical harmonic expansion, the origin is taken in the interior of the inclusion. The problem now ix to find the firat nonzero term in thisELECTROMAGNETISM—SOLUTIONS: 121 expansion. It is easily shown that the term with 1 = 0 vanishes. This is because the integral of JE-da over a surface enclosing the inclusion is equal to 44y. If this did not vanish, then charge would be leaving the inclusion at the rate { j-dA = 0» [ E-dA = 470,Ao. This violates the steady- state nature of the problem. Thus A, = 0, and the leading term in the perturbation is V~ iE Gn YT). JE-EJot. r 15. Begin with the equation y [Ap cos (mO) + By sin (m6)]. Vor = a,8) = ¥ [Am 00s (mB) + By sin (mB). By, =6 by symmeiry since ¥(6) = ¥(—6). Aq may be caicuivied when V(q, 6) is multiplied by cos (n 6), and integrated over @ from zero to 27; then cn sen _ (°° 1, cosnadé + [" V, cosnode = 271 — V2) sin ("@) — A, in Jen m \z) and x(V, + Vi) = 204,. Thus yal + V) Vad Ps —V) $ z B35¢ zy" cos [(2n — 1)6]. Finally, it is worthwhile to note that the series expansion may be summed. This is aon by writing E(— Ut ge 08 2 — 198) = Ref” dy(-4) 5, (iye)?=—1, =x" for |z|<1, becomes: Re . 49( 53) =e 1 ml122 ELECTROMAGNETISM—SOLUTIONS Upon completing the integration, we obtain roo 22-1 e9s [(2n — 1)8] = Im log {FE Thus This closed-form expression can also be obtained through the use of Green’s- function techniques. 16. Combining Gauss’ Law, div E — 4p; Ohm's Law, J — cB; and the continuity equation 2p/at + div J =0, we find an equation for the time dependence of p: p= pret, The characteristic time for a charge to disappear is therefore (470)"'; for copper this is 1.5 x 10°" sec. This time is so short that a charge carrier would have to exceed the speed of light to travel even a very short distance. This indicates that the solution must be invalid. This may be attributed to a failure of Ohm’s Law. The charge cannot remain inside the conductor; if the conductor is completely insulated, the charge appears on the interface as a surface charge density. 17. The conducting sphere is the source of an electric field E = bV/r', which polarizes the dielectric. An induced dipole p has energy U = —} p-E in an applied field E. Howover, p = @E and the force on the dielectric is given by F = —¥U. Thus aU _ 2a? r,=-W__ : or e 18. P= N @ Ey = Na [Eme + 47 P/3], where E,,.. is the macroscopic 19, The field of the atoms outside the cavity is calculated from the average macroscopic polarization density P by assigning surface charge densities o = P-fi and a volume charge density p = —V-P to the medium, Thus B= Scans Ae + E+E,ELECTROMAGNETISM—SOLUTIONS — 123 The first term is the contribution of the surface charge density (—P-fi) on the cavity wall (A is taken outward from the cavity). The term E, represents the contribution from the other surface boundaries of the dielectric, and E, is the contribution from the volume density p = —V-P. The essential point now is to consider P as constant in the immediate neighborhood of the cavity. Then one may identify E + E, + E, with the average macrosopic field E,,,.; in addition, { (—P-i)(—fidd _ 40P Jeary r 3 Thus E + E’ = Ens, + (47/3) P. The electric field due to nearest neighbors located at x;, with electric dipole moment py, is = z 3(pi aa ip In case (a), two atoms have Inl=an while the other four have |x,| = 2. All dipoles have the same magnitude and direction along E; thus — pF) tee ee) Ke, a In addition, p = @E, and P — Na@E,; hence 0- iMa= The index of refraction n is found from (e—) . (— E,., with From the given lattices constant, ana finds N =: 1.9% ¥. 10*/em$, and from the value of » for the undistorted lattice, & = 0.83 x 10-" om?, Then in vase (a) vue has m= 2, which is lower than the undislurled value, In vase (b) E" = 6p(z, — x,)/2f, and is of half the magnitude of and opposite in sign to the contribution in (a); thus n = 2.1. P = Nak, 20. The critical angle is determined from Snell’s Law. This requires the calculation of the index of refraction n(X). This proceeds as follows: mk = +eE = cE, e™ = —mx wre, where w = 2nc{r, and the vlecirun is assumed iv uxt ie with the seme frequency as the x-ray, and with amplitude x, about a given ion. The maxi- mum dipole moment induced, assuming stationary ions, is qa)124 ELECTROMAGNETISM—SOLUTIONS of (1) would contain an additional term maf, representing the characteristic frequency of the electron bound to the ion. The polarization of the metal is P — —Ne*E,/ma*, and tho polarizability @ = P/E = —Ne'/mo*. Now D=cE=E + 42P = E(1 + 47a), from which =1 2 = sexe (less than 1!). cos @, (angles measured with respect to the or sing, — (A282) =e, mat where @, is the plasma frequency. For @ 6). Since > 0 for transmission, there is a minimum frequency given by @y — ex/a. The phase velocity is given by m= gaefi+ (BS + BS126 ELECTROMAGNETISM—SOLUTIONS while the group velocity is given by _ de a) (nt) m)\-"" wea +EE+R)} Similar resuiis are obiained for transverse magueiie modes; i, however, the cutoff frequency is higher. One notices also that »,v, 23. J is constant in time, since E = 0 every- fl where. Recanse the slab is infinite, H and J fee] bs equation curl B = one vbiains VB Ba =0, fe where we have used the identity curl curl = grad div — V*. A solution must be found obeying the boundary condition 5(-d) = Hy. The solution desired is i 4x3 af | T _— Ale te) _ g cosh (kz) Bee) = Hels ay = Bee ay where It — 4a/\e? The enrront density ia determined from _ __g0B _ __ guyz sinh (kz) = curl B= 85 = — Sho (ed) The field H has only the external currents as its sources and hence H = Hy everywhere. This is no contradiction since B = H + 4M, where M is the the magnetization per unit volume, and should satisfy cV x M = J. This is easily checked in this problem since V x H =0. Then Vx M = (1/4z) V x B— (1/4z) (4x Je) — Je, and everything is consistent. One sees that B is confined to a region of the surface of skin depth 1/k and that a super- conductor does not allow B to penetrate the interior. and 24. Consider Maxwell’s equations vV-D=0, qQ V-R=0, oy vxE-18, (3) vxH=12. (4) If P is the polarization in the rest frame of the cylinder, then to first orderELECTROMAGNETISM—SOLUTIONS 127 in v/c, one has P= af z with 3) (Gi este8) ant oP D=E+4xP and H=B-— 4xM. Thus to first order in v/c, one may write oe xE. (a) D=eE + (e— 122 (b) H=B+(e— 122 which may be rewritten in a form more convenient for our purposes: @) B= L{p— eX, (bh) HB 4 ES NYXD. Byuatious (a) aud (b) are the uomelativistic limit of Miukuwshi's cquativus for moving media (see Landau and Lifshitz, Electrodynamics of Continuous Media. Reading, Mass: Addison-Wesley, 1960). ‘We now look for circularly polarized solutions of the type D = Dil& + ieee? = with =D, = 0, B= Bik ify? with = B, = 0. (6) One sees that Eqs: (1) and (2) are satisfied. Substituting Eqs. (a’) and (b’) into Eqs. (3) and (4), we make use of the property that Vx(@x A Ox A when A = A(z, t), with 4, = 0 and v = Q x r, with © constant along the z-direction, and obtain: leyn €-lNoyn_ 12 ely eS e and vxB+&—Naxp= 12. (4) For circularly polarized solutions of the type in a. (65), these become @” $b, By $Me «) Tf this system of homogeneous equations for D, and By is to have a:solu- tion, the determinant of the coofficients of D, and B, must vanish. Thus 4 =f e— ny, 6)128 ELECTROMAGNETISM—SOLUTIONS A wave, which at z = 0 is polarized along the z-axis (ie. E(0) = Eo), becomes at 2 = L R(T) = Bepaieme 4 ott) 4 siete — ety Hence the plane of polarization is rotated through an angle @ = (k_ — k,) L/2 in the same sense of rotation as the cylinder. Using Eq. (6), we find yea where n = /"¢ is the index of refraction of the dielectric. This problem was first treated (incorrectly) by J. J. Thomson; then (cor- rectly) by E. Fermi in Rend. Lincei. 32 (I) (1923). Professor Fermi put the problem on the 1948 “Basic,” as a library problem. One student located ¥Fermi's originai paper, transiaied it, and was the only student to give a cor- rect solution. 25. Since Vs >> E.x, or Est», it is a good approximation to take the trajec- tory of a particle to he a straight, line except. in the immediate vicinity of the b: fo Sold whi bends the traj the equation V-E=0 at the center of the slit. There, (2£,/0x))~ (£, — E,)/t, where ¢ is the thickness of the lens. Expanding £, in the neigh- harhood of y=0, we write B,~ay. Thus V-R=0 implies a~ —(E,— £,)It. 8, =69 - 6) A particle of charge e entering the slit at height y is therefore acted upon by a force —e(Z, — £,)y/t. This force acts for a time t/v, where v is the velo- city of the particle. The net impulse given is then Ap = —e(E, — E,)y/v. by the angle Sh = —An/p. The particle will anf Lb. Substituting 6, = 5 — 6, and 6, = y/x, we obtain 2 yl dh=8) (2), z pe WV,ELECTROMAGNETISM—SOLUTIONS 129 26. Since the magnetic intensity changes gradually, it will be assumed that the motion in the plane perpendicular to the axis is approximately circular. The angular frequency of revolution is the cyclotron frequency, w=. «i me Vy In addition, the cylindrical symmetry of the solenoid guarantees that the component of angular momentum along the axis (taken to be the z-axis) nserved. TI Re(r x p) = a[r x (mv + ay] = mete +? and using Eq. (1) we obtain mre = mriw,, (2) where Jmrfo, is the angular momentum for B = By. In addition, for motion in a magnetic field, the kinetic energy is constant; thus ia: + }mrot = pmol, + $mrial. " 2 Eqs. (1) through 7 yield ee ee Since B < B,, the condition that reflection occur somewhere is ou Sue(BE— 1)". 27. Choosing a cylindrical enordinate system, the 4-component. of the force equation reads Ztmr6) = rH, and conservation of energy requires that }m(# + 7°62) — eV(r) = 0. The z-coordinate does not change. Equation (1) has the solution mr*6 = (eH /2c)(r? — a). When this expres- sion for 6 is substituted in the energy equation, we have dnfa t (4 -) (tat? } = Vir). Threshold is obtained when # = 0 at r = b: therefore Vz ~- eH — a2)?/8 metb?.130 ELECTROMAGNETISM—SOLUTIONS 28. Consider first the quadrupole lens The magnetic field H 1s derwvable from a potential, H = —grad U Since div H=0, () U must be a harmonic function, 16 a function of z — x + sy but nob of z* = x — or vice versa Now any polynomal m z18 a solution of Laplace's equation in two dimensions ‘The correct exponent 18 chosen on the basis of symmetry The quadrupole lens has symmetry under rotation by z rad, the potential chosen must. reflect thia symmetry Hence we choose 7) imaginary part embodies the boundary conditions given in the figure, name- ly, that the lnes (zy = const) represent equipotentials The forces are determined from the Lorentz equation F = q(v/c) x B Poaitively charged particles are foensed im the yz-plane, and defoensed in the z-direction, the opposite apples for negative particles On the other hand, a particle with no charge but magnetic moment #/ m a nonumform magnetic field experiences a force given by F = —V(—p-B) = (u-V)B To maintain focusing analogous to that for charged particles in a quadru- pole lens, one must have Fp pe neg azz rss OO aay, * On Ox" , 8 This suggests taking U/ = ket, and choosing the real part This aymmot This poten symmeti by writing in polar coordinates, z= re ‘The sextupole lens, shown in the figure, embodies this symmetry (Such lenses are actually used for focusing neutrons and nentral atoms A heam enmposed of some the x-axis, and of some others with anti parallel magnetic moments, may be polarized by such a lens, because one of the components 1s focused while the other 18 defocused ) 29. For a system of charges all moving with velocity v and generating an electric field E, the magnetic field 1s given by B =v x Ejc The force on a given charge 1s computed from the Lorentz force equation fosca ses) == =e Wa e{B + > x B}=e{1— (2) pe = ek The electric field E may be found from Gauss’ Law fu-aa = 4m (charge enclosed). yielding E = 2xper Thus there a net outward force, making the beam hla in tha ahannoa af focunng devicesELECTROMAGNETISM—SOLUTIONS 131 30. Magnetic monopoles are mtroduced mto electromagnetic theory by modifying one of the Maxwell equations to read div B = 4py An isolated monopole will produce a B field which can be written down m analogy with electrostatics An electrically charged particle of mass m and charge q then has the equation of motion The kinetic energy 18 conserved, as are the components of the Fierz vector [rxp+ 22] U r J 31. (a) The equation of motion of a charge 18 dpidt = (e/c) v x B Since v-F = 0, |v|18 constant, we shall write Pp pit vant and : a where de 1s an element of the path Under these conditions the force equa- tion becomes dr _edrig ds*~ peds ~ (b) For the equihbrium of a current-carrying wire we must have Zax p+ 7(#) —7(#) =0 Upon expanding (d*r/ds*), = (dr/ds), + (d*r/ds*), ds + — , we obtain ar lar a a Teas “® Se Ts ib . et de t tat Thus the current aust be chosen suurduny Wo =—7e, P + fhe system of clevtruns has an additional motion over that in no magnetic fieid This motion 18 a rotation about the nucleus of angular velocity 2 = —eH/2me (known as Larmor precession) Q 1a determined by transforming to a rotating frame where the magnetic 82. In a weak umgueie field (Banna (Ff)a, 1 9132 ELECTROMAGNETISM—SOLUTIONS and (3 ean = CZ) + 29 * (Finan +2 % (OX 2 ‘The foree (evje) x H may be canceled by the Coriolis foree if ane chnases Q = —cH/2me. Thus the electron cloud rotates. (Terms of magnitude Q? are neglected, i.e, weak field limit.) These rotating electrons produce a magnetic field at the origin given by: Lrix( 5) = an=+/fiXt Say where j=py=pixr; hence an=—1/00 ext gy = 1 [OF He), ay, cy e * But p(t) is spherically symmetric, which means that the integral of pr (t-Q)/r? equals that of (Q pr')/r’. Finally am = — (sii) J = — (gees) AH/H = ro is the classical electron radius, e*/mc’. For AH 3 Spy lo. 33. Outside the charge distribution, the electric field is longitudinal (by symmetry) and constant (Gauss’ Law). Hence no experiment performed outside the charge radius will detect the pulsations. The apparatus must probe inside the charge distribution. The apparatus required will depend ‘on the extent. of the distribntian, and the frequency @ For a very large radius and small «, one might use an electroscope. For a microscopic system, iL might be nevessary Wo use o charged urbiling particle, whuse inveractivus with the electric structure of the nucleus could be made to act as a pulse detector. 34. There is no radiation. This is obvious, because the charge and current distributions are constant in time. 35. Blvirie-dipule radiation is proportivual iv [dBjdi|', where Bis the electric dipole moment. But D = e(r, + F,), and this has zero time deriva- tive because it is proportional to the center-of-mass vector. The magnetic moment. is As : Sie xT = 2S ex vd = SOL, ; Zc Ps Ime ‘Thua M is proportional to the angular momentum of the system, whichELECTROMAGNETISM—SOLUTIONS 133 is constant. Magnetic-dipole radiation is proportional to |dM/dt|?, and so is equal to zero. 36. Let the incident radiation be monochromatic: E = Ey e~‘'. Then the induced dipoie moment is F E, e “*, if f is a unit vector in the direc- tion of the observer at large distance R, then the electric field is xf) x8 a@eo(E, x fi) x 8 E(R, a) = CX _ _ coe. , The total time-averaged power radiated by the scatterer is |B |?/3ci = ao! B3/3c3. Let the ratio (energy scattered per unit time)/(incident energy flux) be o: 8x, o = En, where ry = e*/me* is the classical clectron radius. 37. The voliage developed in ihe wire witile rotating with angular frequency ois =-+5 & (magnetic flux). f= Haat Hora*O gin (cot) — IR, where Ris the electrical resistance and Z the current in the ring. The average power loss to Joule heating per cycle is = cpp) = Hittatot P= PR) = —SaR The only source for this heating is the kinetic energy of the ring, Ia*/2, where J is the moment of inertia about the axis of rotation, I = ma*/2. Conservation of energy requires that d mao? Hix'ato* a4 2cR The solution to this equation is @ = we", with + = mRc*|(xHH ya)’. But m= 2nad p and 2najoA; hence 7 becomes = 4pc!/oHt = 1.6 sec.134 ELECTROMAGNETISM—SoLUTIoNs 38. The potential energy of the charge and its image 18 é ele aa — LF oe V=-a q—D + for smal tions The kinctic o: ce rey quency 18 given by w* = e*/4mL(d — Ly When calculating the energy loss due to radiation, one must include the radiation field of the image so that the electric field will be normal to the conducting plane. Choosing the origin of a spherical coordinate system to be on the conducting plane midway between the source and its mage, we find that the radiated electric field in the upper half-plane, m the direction A, 18 proportional to Be peta xin x ae ay Bw De’ a x (a x aia) ai) the sum bemg over both the source and its image (note p(source) = —plimage) = ea). After squarmg and integrating over }, we find that the average power 18 proportional to Jenin 6446 { cost 6+ 58 nt (Eh cos 6) ‘Upon change of variables x — cos 8 and usc of the identity sin® 2 = (1 — cos 22)/2, the power 1s given by P= leap f' de(t + 2t)[1 — 0s (2kh2)). . aiity K, since if the interference term proportional to cos this expression must reduce to the power radiated by a single dipole, 1¢ @*e*a® P+ 2 Hence K = at/4c’. Of course one could have used Eq. (1) with the mul- tuplicative factors given by radiation theory, and the result would have been the aame. Although the integral may now be completed, we shall be contentELECTROMAGNETISM—SOLUTIONS 135, with the approximation 1 — cos (2khz) ~ 2kh*2", valid for kh 1 When use 1s made of the expression for w, the condition kh<1 becomes e/L <4me? The power 1s then p _ 4éota"h? eat ~ 15 — 240m LH” 39. For a single antenna at rj, R-_hay @y noe” aya @x May Here fi 1s a unit vector from rf to the field pomt, p, 1s the electrie dipole moment, k = 2x/\ and R 1s the position veotor to the field pomt The fields from the antenna add coherently W2-¢ yy 47 6/7 y cos 6 In the radiation zone, — B= HBX GX Pun 5 tka ny, a ‘The sum of the phases 18 and aint & ain @ cos: 4) sin? (F sin 6 cos 4) a) The average power radiated into solid angle dQ 1s Ly eva np136 ELECTROMAGNETISM—SOLUTIONS ©) The ratio is 49: 1. For the array of 7 antennas, the greatest power is radiated along the y-axis; for a single antenna, any direction in the zy- plane. 40. in order to caicuiate the power radiated by a relativistic particie, one may make use of the fact that the power radiated is an invariant; P = dB/dt, and after transforming from rest frame to a moving frame, dE! — ydB and dt! — ydt. Hence dE'/dt' — dB dt. Nonrelativistically, the where p represents the momentum. The Lorentz-invariant generalization reducing to this in the nonrelativistic limit is P= GN) dt Ndr where dr = dth aE = Pap, ak — Rap, is the proper time, and p, is the four-momentum. Since awe (at) — (22)'], For circular motion dp/dr = 0, while p = 2yfotpt _ ‘3m* ‘The anergy radiated por rev (28) 8 6("E) noite of met P-(7) cA in unite of mye. 41. The mirror sees a smaller frequency (longer wavelength) given by »! —v,(1 — v). There is no change of frequency upon reflection. The ob- server sces a still smaller frequency (—») v= wl — 0) = vo ‘The intensity of the beam is the energy crossing unit area in unit time. pe 02 ==O> > Oo So This may be computed in two ways.ELECTROMAGNETISM—SOLUTIONS 137 Method A views the light beam as a free electromagnetic wave, with |E| —|H|. If E is the electric field in the source frame, then the mirror sees B — y(B +v x B)= Vl — ol +0) E. But E’ is unchanged on reflec tion, whereas B’ is reversed in direction. The observer sees The energy density is proportional to the intensity ; it depends quadratically on E. Hence J = I,{(1 — »)/(1 + »)}. Method B regards the light beam as a beam of photons, each having energy proportional to v, and separated by a distance z, from one another along the length of the beam. If at time ¢ — 0, a photon is reflected from the mirror, the next photon reaches the mirror at a time ¢ determined by ¢ = 2 + vt, or t = 29/(1 — v). ‘The distance separating the two photons is now ¢ -+ vt = zp (1 + »)/(1 — »). The rate at which photons reach the observer is inversely proportional to their spacing, and therefore depends on the velocity accord- ing to (1 — »)/(1 +»). The frequency (and hence energy) of the photons transforms in ihe same way, so thai the intensity is reduced according io T= Tf — "(1 +o") 42, In the rest frame the force per unit length F is given by F = XE, where E is the electric field at. one wire produced by the other. This ia easily found from Gauss’ Law, J B-aA = 4 (charge enclosed). ‘Thus EZ = 2d/a, and the force F = 2M'/a (repulsive). In a frame in which the rode are seen to move with velocity v, there is a magnetic ficld B = v x E'Je, in addition to the clectrie ficld B’. Tho total force per unit length F’ is then F=n(B'+43)-v(1-4)e. However, E’ = 24'a where X’ is the charge ae scen in the new frame (\’ = yA because of the Lorentz contraction of lengths). Thus The fact that F’ = F may be seen easily by an alternate argument. If in ite rest frame, one of the rods is allowed to move under the action of the force FL onit, it would gain momentum dp = FL dt, while in the frame in which the rods move, the gain is dp’ = F’L'dt’. But dp = dp’ because momenta138 ELECTROMAGNETISM—SOLUTIONS normal to the direction of a Lorentz transformation are invariant under such a transformation, and dt’ = ydt. Hence LF = yF'L'. In addition L’ = L}y, due to Lorentz contraction; hence F = F’. 43, If we choose a cyiindricai coordinate system, the Lagrangian equations of motion for r and @ are 4 fme(1— BY") = mrbt(1 2) "22049, a) 4 bpeglg — Ny ered _ aa) tet Since B, = 0, only A, has been taken nonzero with B,=12 (ra) and B= — 0. If F = 0, then Eq. (1) reduces to () rey ea msi(1 - tt) + £24) = = c or a) mri(1 — ney" =-t4, (3) Elimination of the term proportional to m between Eqp. (1’) and (3) yields 2irdy = Ae () From § A-dl = { B-dA, it follows that =Ae(aep ae and Eq. (4) becomes Bir) —B2, 6) where (B,) is the average of B, over the area of the orbit, and ry is the orbit radius. In addition to this condition, one must also have B, = 0 in the plane of orbit to guarantee 2 = 0. The angular frequency 6 is found from Ea. (1') to be 6 (6) where @ is the cyclotron frequency, @ — eBy|me. The kinetic energy of the particle is given by Bano BY") mB)ELECTROMAGNETISM—SOLUTIONS 139 In order to calculate the radial oscillations, assume a solution for r(t) de- fined by rofl te(t)] with |el where K = const. thea hence The constant K is found from the requirement that a = [jeden 4s], m0. Thus K = Bur} (n — 1)/(n — 2). Upon expanding in e, 2rd — Ae = — Baron — Ue. With these small oscillation approximations, Eq. (1) becomes me,(1 + met)" é=— Barby De. Using Eq. (6) for 6 and the definition of the cyclotron frequency, this equation becomes a4 wl me _ + Rat]? which has bounded sinusoidal solutions only if» < 1. For this case the radial frequency is given by , where of = =») 1) Terre For amall oscillations in the z-dircetion, we have the equation mi(1 \ rely". ~£ Berd, (8)140 ELECTROMAGNETISM—SOLUTIONS obtained from the force equation dp; _ ¢ OP: _ & yy doe ‘To find B,(z) we nvie thai V x B = 6 in the vicinity of the orbit and thus Hence OB, _ OB, _ nb a or To Thus B, = —mB,z/ro and when this is substituted in Eq. (8), one obtains : 4 eine Ptr ae 0. Hence 2 no* 24 ota gt = qe gee) md ott ola bh. 44, (a) Let , be the four-momentum, with the property PuPs =P — EF = — mi. ‘The force dp/dt gencralizes naturally to the four-vector dp,/dz, where 7 is the proper time. This is to be set equal to # four-vector function of the electromagnetic field tensor F,, = 2,4, — 0,4, and p,. Une also wants an expression linear in the electromagnetic field. This is guaranteed by setting dy _ de CP Pe + BoM? Peep. Here e***” is the compietely antisymmetric tensor with part of this equation becomes ap FE = am(B + v x B) + 2ifm(B — v x B) . The spatiai Thus one must choose a = e/m and 8 = 0 to obtain the Lorentz force equa- tion. The term proportional to 8 would be the force on a magnetic monopole, if one existed. Since, under parity, or space inversion p—> —p, E—> —E, B— B, @ must be a scalar, even under parity, and @ a pseudoscalar, in Lorentz force equation is dye rr aw Pv (b) We invent a four-vector S,, with nonrelativistic limit (0, 8). Note that S, p, = 0 in the rest frame, and therefore is zero in any frame. The covariantELECTROMAGNETISM—SOLUTIONS: 141 generalization of dS/dt is dS,/d7. Thus, in analogy to the procedure of part (a), one sets aS, _ ge Te = amg Fee + bry. The term linear in p, is necessary to guarantee p,S, — 0. Thus from fa = p, Se 4 g,%s — dpy _ ¢ ae (SoPe) = Pegs t Suge =O and R= 7 FP one finds & = e(g — 2)(PyF wS,)[2m*. ‘The covariant equation for the spin is thus as, e 2) } ay _ 8 . Be = oF 8, + 2p, FaaSaPa It is of interest to write explicitly the equation for the component iS, — iS-p/p. — iS-v, in the case E — 0. Then a a ‘The helicity S-v is couserved when g ‘This is indicaied in ihe figure, where the double arrow indicates spin direction. For both the electron and the muon, the g value differs from 2 by a small amount. As a result, the helicity is not quite conserved. This fact. was used in the CERN experiment to measure accurately the muon g value. = a — 2)B-(v x 8). q “@ ’ Aa. jit It is worth noting that, given the components of the spin in the rost frame of a particle, (0, 8), one may determine the components in a frame moving with velocity v = p/# by using the two Lorentz-invariant relations p,S, = 0 and 8,8, = — $*. Note that in any frame, the number of independent com- ponents of the spin is three, due to the subsidiary condition p,S, = 0, which imposes a constraint on the components of the four-vector. This is necessary if the spin iv iv be desuribed by w three-vevive in the resi frame, since die number of independent components should not depend on the frame of the observer.ELECTRONICS 1. The circuit will fail to operate as a generator when @ is so great that an electron cannot respond to the changes in grid voltage, i.e. or > 1 where is the electron’s transit time in the tube. ‘This corresponds to @ > 3000 Mc. 2. We solve the more general problem, in which L and € are replaced by Z, and Z,, arbitrary impedances. Let Z be the impedance of the network. Th th c with solution ga AANEEEEE z The plus sign is used for the square root, since Z, = 0 implies Z = Z,. In the problem given, Z, = joL, Z, = —jo0. Then igs tall [eee = B= St gy er + After passing through n stages, an input voltage is reduced to V, = V(l — Z,/Z)" = arV, where a=(1_ 4) = (Sat vMt a) _ job + Va oT BING ENR FAD as! ~ Goh + V4L0— oD It is easily seen that |a| — 1 if @ < 2w» but |a| <1 if © > 2a». Thus the system acts as a low pass filter without attenuation for @ < 20) = a. 3. The presence of Cy introduces frequency dependence in the amplification (= Vee! Vin); this frequency dependence is to be eliminated. This can be done by placing capacitors (’,, C,, and C, across resistors R,, R,, and Ry, pro- vided the values of the capacitances are chosen properly. The impedance Z of a parallel combination of R and C is Tessa 1 Sear Z = Ht sol = Fl + jokey. 149ELECTRONICS—SOLUTIONS 143 It is evident that if the capacitors are such that R,C, = R.0, = R,C, = R,Cy, each resistor-capacitor combination will have the same frequency dependence for its impedance. The amplification, a ratio of impedances, will then be independent of frequencies, and equal to the amplification when the circuit is purely resistive. In other words, the effect of C. on the amplification is nullified by the introduction of the other capacitors. 4. The cquivalent circuit, using a potential-source representation for the triode, is shown in the figure. Because of the hint, the grid resisior and grid capacitor have been omitted. fo) beh c Now 1,2, =IpZcand I, +I; =I, from which I = 1,(1 + Z;|Zc). Because of the mutual inductance, Z, = 1,Zy = joMI;, and from Ohm’s Law, Md Ba) = wBy — Mp + 1.2, — Le Ro(1 + 32) + 24]: Hence the impedances are related by = ar play = Ro(1 + 38) + Zr. (ay But Z, jhoO and Z, But 2, — —jla0 and 2, — yields two equations: 1 R _ (RRC + D =z(1+#) and a = BRC . a ‘The latter condition on M is actually a condition for the onset of instability. For M > (RRpO + L)/p, oscillation occurs. This can be seen by solving Eq. (1) with = wg + Joy. Since the time dependence is taken to be exp (j@#), one obtains exploding solutions (i.e., oscillations are possible) if @, <0. Substituting this expression for w in Ea. (1). one solves for os. _o, — HM = RRC HL 7 2RpLO . Thus w, <0 if M > (RR-C + L)/p. The circuit will fail to oscillate when M < (RRC + L) #144 ELEcTRoNIcs—SoLUTIoNs Open ® Open eee | Vp const y, ‘Open: Mn 2p, re 7 Nana Senne YpeoX | 7 Vg const State e [re Vg No geometrically; thot is, Zp —Vp ~ W/o, Y4Vo, IBV, 3/10Np, ete. 5. The operation falls into three phases, which we label a, 8, y. During the first quarter-cycle, V, increases. V, > Vp > 0. The diode D, is hence open, and Vn» > 0 allows the diode D, to remain closed. In this phase V V,>0¢ D decreasing, s0 D, opens. Since there is nowhere for the charge on the capaci- tor to go, Vp remains constant, and Vyp remains constant at V9/2, for the same reason. This phase continues until Vp reaches 0. But Vp <0 is not possible, so Vn = 0. The first half-cycle has been completed. Now V, again increases from —V,, and again V,» is constant at —V, (we are again in the A-phase), and V,> Vo» stays at Vo/2 until Vp rises to moot Vz. When Vp = Vm the B-phase has concluded, and a new a-phase begins. Each cycle goes through the phases a — 8 — y — A, in that order; the circuit is not periodic in time, however, as may be seen from inapecting the graph. Vz approaches 2V, geometrically; the circuit ia known as a voltage doubler.OPTICS 1. Call the object distance for the first lens surface S,. Then the object distance for the second lens surface is —S{, where Sj is given by Snell’s law at small angles: Also, Adding these two equations gives mm ats The focal longths are obtained by ectting S, = co and S; = oo, respectively. Thus fi m ; (n = m)/R, + (ta — 0)/ Re and h-@& Soe Ma aaa eae (n — m,)/Ry + (ry — 2) It follows that f/S, + f2/S: = 1. Special case: When m, = nz, then f, foand 1/8, + 1/8, = Uf. 2. Ifthe pattern of incident, reflected, and transmitted waves is as is shown in Fig. 1, then time reversal invariance requires that the pattern in Fig. 2 is also possible. In addition, the superposition principle allows Fig. 2 to Whos kak Fig. 4146 Optics—SoLvTIoNs be considered as the sum of Figs. 3 and 4. Equating the incoming and out- going waves of this sum to those of Fig. 2 yields the Stokes’ relations 4m =1 and rt =0. 3. If the amplitude of the incident light is denoted by E, the amplitude of the light upon the first reflection from A is defined to be Er,, and that of the transmitted light, Ht,. The transmitted light undergoes multiple reflections, and some fraction emerges again into the vacuum, the amount being governed by the reflection and transmission coefficients r., r,t», ta ta defined in the figure. The total amplitude reflected into the vacuum is B, = Brat Bedirge™(1 + (ere) + (rat +] a = B+ = pltat ree td — rita) 1 = rirget* In this expression for E,, k= 2zn,/A and X is the vacuum wavelength. ath 7 ata tag Et ararth EB, _ [rat rete] ET trre™) We wish now to relate the coefficients r, and r to the indices of refraction, nm and 1. For this it is necessary to use the boundary conditions required by Maxwell’s equations. Thus at the interface A the tangential components of the clectric and magnetic ficlds must be continuous across the interface. E+En=EBy of l+t=te a) H+#H, @) Now for a plane wave H = EB, H, = —E, (the change of sign is due to a reversal of the dircetion of motion for the reflected wave), and H, = mE. Thus By. (2) becomes i— ry = mig and one finds r4 = (1—m,)/(1 +m).Oprics—Sotvrions 147 Similarly Tp = (m — M4)/(m + M2). The condition that the reflected wave vanish is t=, re = tp In terms of the indices of refraction, these conditions become Ww { n. 1) or @) { ne = ni ) Inap = peel Iman = ep + yf where p is an integer. It should be mentioned that £,/E may also be cal- culated by postulating « standing wave clectromagnetio field in the dieleo- tric m,, and soiving the system of equations required by matching boundary conditions at the interfaces A and B. Thus for the electric field these equa- tions are at A,E,+B=E,+E,; at B, Bye + Bye “= By. at A, E—E,=n,E, — nF: at B, n,E,e%* — n,E.e™ nE,. When #,, #; and £, are eliminated between these four equations, one finds the above expression for H,/E. 4. Standing waves of light are formed in the emulsion. Each wavelength 2 will therefore develop the film at depths (2n + 1)/4 (n — 1,2...) in the emulsion (where the intensity has maxima). When white light is incident on a particular iayer of silver, a smaii amount is refiected; most is trans- mitted. Light of a given wavelength will be reflected coherently only from the layers with spacing 1/2. incoherently from all other layers. thus reproducing the initial colors. A serious limitation to the method is the color distortion caused by developing the emulsion and by temperature effects. A quantum fieid theoretic treatment of the Lippmann process may be found in E. Fermi, Revs. Mod. Phys. 4, 87 (1932). I Film j aH 4 5. Consider a source of light P on the sun. The image of this source is spread over a region of lincar dimension &¢ = D + 1(A/D), where D is the pin-hole diameter. The first term in 3z is due to the apread of raya over tho pin-holo148 Oprics—SoLvrions opening. In addition, due to diffraction, there is a spread in angle of approx- imately A/D upon passing through the opening, and hence a further increase in 8z of about (A/D). The total 8 has a minimum for D* — M which for = 5000 A and I = 10 cm gives D = .22 mm. 6. The Rayleigh criterion asserts that two lines are barely resolvable if the maximum intensity of one coincides with the minimum of the other. Principle maxima in the mth order occur for d sin@ = md. Here, md re- mum from two adjacent siits, If the grating nas W siits, the difference in path length is Nm), between rays from opposite ends of the grating. Suppose the angle @ is increased slightly, until the path lengths from the opposite ends of the grating differ by Wah +2. The pattern then has zero intensity, th path length by 2/2, and hence they cancel; similarly amplitudes from other pairs of slits cancel, giving zero intensity. To satisfy Rayleigh’s criterion, these minima mnt, aeenr for the same angle as the maxima in = a mNA | X= MNO FAN or N= Ty A grating requires at least 982 slits to separate the sodium doublet in first order (m= 1). Note — that the result is independent of the grating | spacing. The corresponding angular separation is Groting 10-5 rad. 7. Ray i changes its phase when it reflects from the surface of the lake. Therefore the difference in phases between the two rays arriving at the detector is 1 by — Gy = 4 EAL = 208 28), ‘When the star first appears over the horizon, the difference is 2; the waves are out of phase, and the intensity at the detector is minimum. As 6 increases, the intensity increases; a maximum is reached when sin@ = X/4d; this heppe: happes a, : ee 4 loveOprics—Sotutions 149 8. Each atom of the ribbon radiates independently of the other atoms of the ribbon. Hence each point produces its own diffraction pattern, and these overlap on the screen. The interference pattern is expected to disap- pear when the first maximum from the pattern due to A coincides with the first minimum from the pattern due to B (the Rayleigh criterion). Light from A has interference pattern centered on the observation screen at 6, = —W/2L; the first minimum is at 6, -+ /2d. Light from B has pattern centered at 0. — W/2L. The pattern vanishes when 6, + 2/2d,< 6, . Hf, as d is increased, the pattern vanishes at d = de, then In Wa 9. Let a be the amplitude of light coming from a single slit, and let 8 = (2x/A)d sin 8 be the phase difference between rays of light from adjacent slits, The total amplitude is Andi 4 herg Long shee =at=h -"U= pe) The intensity is jaye — atl + 14* — cos N62") seete ee (F— cos 8) 16. A =al[(1 +a) + (1 — ae® + (1 + ae + (1 — ae “+ (= ayer"), is that of the preceding problem, and we have taken N even. The series is finite; we may rearrange terms: Not x A=ald e™ + ay (—1te™] Ho Be nan ‘sin (N8/2) ___ sin (N8/2) acto 10 | The intensity is proportional to AA* = at [8102 (N6/2) 42 Sin? (W6/2)), Usin® @72) ~ cos* (8/2) J which may be written as __, sin? (N8/2) te 8 Pah gra (1 + a? tan 4) The effect of a ia most detectable when 8 ~ (2k + 1), when the inten- sity in J — ad,150 Oprics—SoLvTiIons 11. The amplitude of the wave on the positive z-axis which is transmitted through the opening is given by yA f[ee, where do is an element of surface area of the opening and r is the distance from do to the observation point. This expression follows from the simple version of Huygen-Fresnel theory where the obliquity factor is taken constant (valid for faJ2) <}). Using de — {a/z} <1). Using de dp — eae find that the amplitude integrates to ap — —iAMexp (ka? + #) — exp (ikz)}, which has zeros for Va? +2 =z +d; n is a positive integer, thus 2 = (a? — ntX3)/2nd. 12, After passing through a polarizer whove relative angle of deviation trom the preceding polarizer is 9, the intensity is decreased by the factor cos* @. The attenuation of this polarizer is thus (1—cos* @). The probability that the relative angle is between 8 and @ + d0 is Be~** d0, where Bis chosen so that the total probability is unity. The average attenuation per polarizer is thus A= = (1 — cos? 6) Be-** 48. Actually the limits of integration are from —z to +2, but a> 1 and one may extend the limits to infinity to a good approximation. In addition, the integrand is large only in the neighborhood of § = 0. allowing the ap- ion cos @ = 1 — 6. The average attenuation then becomes Af" ooze ——B Sf” cman, As a function of a, j*. e-** 4 = Ua", where U is some constant, (U = frud9 e*), Thus B = a"?/C, and 13. For a perfect grating with grating spacing a, the amplitude of the dif- fracted wave is Ca per where @ is the amplitude of the wave diffracted by a single slit opening, andOprics—SOLUTIONS 151 8 = (2xd/A) sin 6 = kd sin 6. The intensity is JApt = at int (8/2) _ | 4,2 sin* B sin (6/2), sin? (6/2) Bain? (5/2) direction 0 = 0. When a given slit is displaced by an amount z from equilibrium, the amplitude from that slit is fer Avett’ du = An (2 22) guzsme, Soames } Then A= 5S Ayettanne ime sinf, The average intensity observed over a time long compared to the period of oscillations is T=QAP) =| Aol? ee 2 ef(m—m)8 (lt ln Oana), e* , where ¢ is the root-mean-square displacement, one has Getter s0 Ay = Bag + A Baal F degdge atte eH tanta) = 8nn + (1 — SandeBr aint, 14, The velocity of light in the lower conduit is ¢/n. As seen in the laboratory, the velocity in the upper conduit is (using the relativistic addition formula for velocities) rr temo ~ [+] -e]~ 540 - Then Xf = c/n and Xf = c/n | u(l — 1/n*). Upon emerging from the conduits, the two rays differ in phase by ¢$ =2nl(b - +) =2n(2)(+) or -1). The first minimum occurs when $ — =, or152 Oprics—SoLuTIons 15. In a birefringent crystal, the electromagnetic waves polarized parallel and perpendicular to the optical axis travel with different phase velocities. Thus a linear polarized wave Eo(% + 9)//2, after leaving one polarizer and traveling a distance 2, becomes ln y parallel to fi = (& +5 "After traveling the complete length, 2d, of a crystal, the amplitude which is transmitted through a polarizer is E-A, where Ba =2 2 exp (in,k2"d) + exp (in_k2"d)] = n exp [ikd2"""(n, + n_)] cos (2%), and ¢ = (n, — n_)kd/2. Therefore, after passing through § elements, the intensity is reduced by the factor T = [00s (f)-cos (24)... cos (25-4) Using the identity cos @ = (sin 26/2 sin 6) repeatedly, we find the transmis- sion factor to be in (284) Examination shows that T() primarily for g — px, where p ~ T($ +x), and that transmission occurs integer. The width 8p ~ 2/2. In terms A= (n, — n_jdjp and of width baja ~ Zip:QUANTUM MECHANICS i. If the operaiors B and C aniivommuie, ihe corresponding eigenvalues 6 and c must have zero as their product: bc = 0. Since C* = 1, we have ¢ = 1; hence 6 = 0. Thus a state may be an eigenstate of charge con- jugation only if the baryon number is zero. 2, The matrix M = (M,, M,, M,) represents an angular momentum matrix, because of the commutation rules. It is evident that the matrices do not represent: irreducible representations; rather they represent several irredu- cible representations. If a state with spin J is represented, 37, has (27 + i) eigenvalues, ranging in integral steps from +J to —J, each appearing once. Hence there are no states of spin greater than 2, only one of spin 2, and eight of spin 3. One of the 28 entries of +1 is accounted for by the J = 2 state; there are, therefore Each eigenvalue of M* corresponding to spin J has value J(J + 1); to each representation there are (2J + 1) such values, We then construct the following table: Number of | Number of J J+) (27+ 1) | representations | entries in M? | 2 6 5 1 5 F t v 3 ae 4 8 32 1 2 3 27 BL ; : 2 48 96° ° 0 1 42 42 153154 QUANTUM MECHANICS—SoLUTIONS, 3. Repeated use of the commutator [L,, z, te,% and the identities [AB, C] = A[B,C] + [4, CIB, and eineinn = Binben — Bydem Yields dae = 2(°8u — aim}, where = ¥ aa. If € is an eigenvector of this matrix with eigenvalue 2, then 2rte — 2r(r-e) = de. Thus one may choose two degenerate eigenvectors perpendicular to r and with eigenvalues 2r’. The eigenvector parallel to r has eigenvalue X = 0. 4. Taking matnx elements of Hq. (1) between states of ditferent m, one obtains is to be nonzero. Likewise from Eq. (2), CjmolU lm’) is i from Eq. (2), — xQuanruM MECHANICS—SOLUTIONS — 155 We now introduce a complete set of intermediate states: O|2Hz|0> = ¥ Olz|n>Eq = Elo PB a, GO| Ha | = C0280) = ¥ COl7|m> nl 2] 0B, = S| and the identity a % Lena's — By) = Fe 2m 7. In the absence of sources, the Maxwell equations take the form: — 10E —_10H -E= vxna=lS vxE=-24%, v-B=0=v-B. In torms of F and F* (which are linearly independent), these become iF S====221- EY vxF=i%, and Vx im, div F*. Noting that (VX Fla = €apyOsF, = —OptsarF ,, and div F = where we have used the notation dg = 0/dxg, the equation for F becomes , ah (80 9)( —iepay) Hy = =~ The momentum operator is pp = —id,; this therefore becomes iar, Peter? — oF Now for fixed 8, —iegey is a 3 x 3 matrix, Siia,.)- The equation for F then takes the form aS. = (p+ 8)F = 4 This is called the Dirac form of the Maxwell equations because of similarity with the Dirac equation for leptons: (Poa + Amy =i When the Dirac field is massless (i.e. m = 0), the two equations have the same form. Using the standard definition of ¢,,, in which (even\ \odd / 0 otherwise,156 QUANTUM MEcHANICS—SoLUTIONs we obtain a representation for the matrices, 00 0 OOor 0 —1 0) S,-i/00-1}), 8, 000}, and S,-if1 00}. Yl Of \-1 0 0/ \o ovo We can verify easily that the S’s obey the angular-momentum commutation relations § x § = iS, and furthermore, that This shows that the equation describes a particle of spin 1, for which Si, = SS + UI. 8. From the rule 6 dz = nh and the enerey equation (p*/2m) + max = B. we obtain 2 rn dz[2m(B — mgz)]"" = nb. Thus # = (¥mg*h*n*/3z)'’*, where n is an integer. 9. A state of the three-dimensional oscillator is uniquely specified by the set of three numbers (n;, ns, 3) with m,, ns, my > 0 and m ++; =n. The numbers 7, m;,, are the harmonic-oscillator quantum numbers for excitations along the a-, y-, and cases respectively. Fur fixed ny and n, the number of pairs (n,, 2) for which m + nz =n — 7, is lam +0). Finally summing over n, we obtain the total degeneracy, d,: (n+ 2)(n +1), d= Ent ny) = ln $1 — met) 10. The Hamiltonian of this system is given by H = Lt/2I. where I = 3mr* and L, is the angular momentum perpendicular to the plane of the three masses. The cigenfunctions are ¢''* with J an integer; thus the energy levels are E = hl*/2I. However, the three particles are bosons, and the wave function must be invariant under rotations of 120°. Hence | = 3n with nan integer, and E, = 9n*h?/21. 11. The proton and antiproton magnetic moments are given by 1, — 24.8, and [#; — —2j0Ss. Choosing the axis joining the two particles for the z-axis, one has 2 v= Bl-48, +8: + 128,,8p:].Quanrum Mzcuanros—SoLvrions 157 Now 281.82. = (Si — 4) 28, - 8, = SS + 1) — The eigenstates are S = 0,8, = OandS = 1,8, = 1,0, —1. The correspond- ing energies are } for spin $ particles. v Vee Ma 32, The total apin operators S,— (ol! + olM/2 and St — to + aei/t commute with the Hamiltonian, as is easily checked. In terms of these operators, the Hamiltonian may be rewritten H = 2AS, + B(28* — 3), where oi =o} =o! = 1 has been used. Thus for a singlet state St=8,=0 and K,=—3B, while for the triplet states St = 2 and E, = 24S, + B, with S, = +1, 0. 13. Choosing the magnetic field B along the 2-axis, we may write the Hamil- tonian H= 21% 6, + 0) B- 2S Mo, — 0) B+ Jor- 0%. Upon going to the representation where S, and S? are diagonal with 8 — 4(¢, + 9%), the Hamiltonian becomes H=—(a + ABS, + Zs — 6) — D Blow. — 012). The first two terms are diagonal in this basis, and in addition the triplet, S=1, and singlet, § = 0, hoth have definite parity under exchange of particles (1) and (2) (oven and odd respectively). Thus the only nonzero matrix element of the last term (which is odd under exchange) is between the singlet and triplet states. Une calculates directly (on, — on)( ABE) Pua) — 9 (an6l8) + BUYAIE)), <10| | 00> = — (a — A)B. S=1,8,=41 EB. . Fa+ AR +d,158, QuaNntUM MECHANICS—SOLUTIONS while for those states with 8, = 0, the Hamiltonian may be represented by a matrix _ J (ap) H(S, = 0) ( wear ”), where the off-diagonal elements are due to triplet-singlet mixing. The eigen- values of this matrix are E.(S;, = 0) = —J + [4J* + (@ — BY BY. 14. By solving Schrddinger’s equation, one finds the energy #(K). Now suppose the sphere expands uniformly by a small amount. The work done is dW = Pav = 4x RP dR = —a5(R) = — 2B ap. Hence P = —(dE/dR)/4xR*. For the lowest s-state, sin (kr) oa from which P = xh*/4mR'. The lowest p-state is (le) de Here, the conditiun y(t) = 0 implies # cut (kA) = i. Tt iv easy iv svive this numerically, obtaming kK ~ 4.5, and P = (4.0)'h'/4emH*. The nu- merical solution is managed by determining the smallest nonzero, positive value of y for which y = tan y. This is the intersection of the two graphs, as seen in the figure. We first guess y — 3/2 — 4.71 and then “zero in” by solving ian ating. One oF two iterations are sufficient if the initial guess is good, and the functions vary smoothly near the intersection.QUANTUM MECHANICS—SoLUTIONS — 159 15. Since there is no angular momentum, the Schrédinger equation reduces to -E\S4 2ahy— Ey (outside the well) and wie 22 = Re 22y =(Vo+H)p (inside the well). With the introduction of U = rp, the wave equation is u"—au=0 (r>a), sod 4 0" 4 QUT Oa), p= BB SFr FP 10 000 8 etm = SEE Oy + Busine) Since <&j7 jk) = 0 and has no matrix elements connecting degenerate states, the first-order perturbation correction to the energy levels vanishes. The second-order shift in energy is given by az, — 3 lVIDUY In, However, from the rules of degenerate perturbation theory, one must dia- gonalize ae Ci F HDF fn) En — By with respect to degenerate states n and n’. For the problem discussed here, the only nonzero Hy with E, = By and nn’ occurs for n = +1, and n= —1, on = —1 and n’ = +1. Therefore to (correctly) obtain the shift in the energy of the states with n = -+1, one must diagonalize the matrix Va, ' _ ae Ft ( 3) v,, 6 3 1 The ot Be? 19, For those states |) for which || #1, the energy shift is given by + D4 > =—Da-— Ag, = eV ln + $+ 11P Ia + Gl lne 1G = 10 |n> Ban® 1) ay 21, Taking the electric field to be along the z-axis, the perturbing interac- tion V is given by V = e#'2. The first-order correction to the energy vanishes, i.e.,(18 |z| 18> = 0, and one must go to second order; —— lap _ ylAS|V [alm a eS) where | nlm) is an eigenstate of the hydrogen atom with principal quantum n, angular momentum J, and J, =m. The energy levels of the hydrogen atom, E,, are given by E, = —met/2h¢n?. Since |(19|V |nlm)[? is alwaysQuanrum MzcHANics—SoLurions 163 positive, and E, is monotonically increasing, it follows that -5 BIAS] s|almyp < F< ees E[K1S|=|nlm> In setting the iower limit, the contribution from the continuum states 18 assumed negligible. Since the states |nlm) form a complete set (neglecting the continuum), ZX K1S|z| nlm> = (18 | 2718). Expressing the state | 1S) in spherical coordinates it foliows that: <18|2*|18) = a re trlady = at, = E, = 2'/82, 0 (16/3)a’. The experimental value is a = 4.5’. 22, The total spin of the system commutes with the Hamiltonian of the system, thus allowing the total wave function, ~p, to be written oe = Gir re) xUi, 2), where ¢ is function of the space variables r,, r,, and x is the spin wave function of the two particles. However, the total wave function, y, must be antisymmetric; thus by choosing the total spin to be S = 1, (I, 2) is totally symmetric, and hence ¢$,(r,,1:) must be totally antisymmetric. Similarly for S = 0, x(1, 2) is antisymmetric, and ¢$s(r,, t) must be sym- metric. Therefore, for the purposes of energy calculation, there are two types of spatial wave functions: ds and oy. In the absence of interaction between the particies, the wave function $(r1, 7%) is given by symmetrized products, $n(ti)da(ts) + dalti)dn(Fs)s where 4s) = ("sn (22) am (2) an (294) with energy equal to x*h*(n? + m*)/2Md? where d = 10-® cm and m,, n, are integers, n = (m1, ms, m3), cte. The interaction between two particles is 3|[>@ a=10-" em 0 for |r, — aeeceee (G0 222 ==eeaeen ee i-% for jr. —mi+|LZ,=—1>. The two eigen- functions yp, are separated by an energy difference proportional to (A — B). Hence, in the general case, i.e. A # B, the functions yr, and yy_ are non- degenerate and one may conclude that = 0. for theseQuantum MecHantos—SoLutions 165 24, Defining new variables 2 = the Hamiltonian separates: H= san (2 ' Ba) — Beosy. Unperturbed eigenfunctions (corresponding to B 1 “hon = sex () exp (#3 with energy eigenvalues An Em =F E+ 1 +62 and y = 6, — 6s, one finds that 0) are () where [and m are integers which are either both even or both odd to guaran- tes th as of yp; that is WO, + j2m, 0, + kx) = WO, 62), with j and & arbitrary integers. (a) Here we treat V = —Beosy as a perturbation, the matrix elements h lm! | ¥ |m> = ai 46, J" 49, cos (8, — 8) ay exp (0, + 6) + S20, —,)}, where Am=m—m' and Al=1_—l, These matrix elements are nonzero only if Al = 0. in which case im! |V |im> = {8(m' — m — 2) + 8(m' — m+ 2)}. ) To procecd further with first-order perturbation theory, one must diagonalize is nonzero only if m = +1 as seen from Eq. (2). Thus for m# 41, the shift AF,, = = 0, which follows from the fact that the ground state has definite parity, i.e. ro = (1/a) cos (2/2) con (sry/2a).Quantum MEcHANIcs—SoLvTions 167 (2) Matrix elements tro = are real. This follows from choos- ing real eigenfunctions. Now if 249% 0, then yxo=0, and vice versa. This must be true, for if x70, then the y-dependence of yr, must be cos (7ry/2a), and hence yxo = 0. Finally then Cap’ |2|qh’> = CF, and ap’ |y |b) = CF; where = in ChE, asin (: tema = Ff", and 26. (a) The total Hamiltonian is we H= #)+teta =-E(S+S)+ t+ a+ om Defining new variables £, 7 by we express the Humilvonian ws pegss te (Oe ape Oe) apeite 241 E— oy? H= mae =e mm) + Fk + OE + Fle on The exact energy levels are thus given by E = nheo, + nahors + (4)lo + 0s) where n, and n, are positive integers, and ==, gg _b+0). nm m (b) Treating 11’ as a perturbation, we write the unperturbed eigenfunctions as Inne> = Uo(2)U-,(22). where the U's are simpie harmonic-osciiator wave functions, the energy levels of which are Egy, = holm +m +1), where ot = E. The first excited state of energy 2ho is two-fold degenerate, the states cor- responding to (n,m) = (1,0) or (0,1). The Hamiltonian matrix of the168 QUANTUM MEcHANICs—SoLUTIONS perturbation, taken with respect to the unperturbed states, is H= (’ °, where a — {10|H"|01> — cX0|2;|1)<1 [22 0). Rather than calculate a by doing an integral ({ zU,Usdz), we may write a = cQ0|z|1><1| 2/0) = eC0|2*|0>. This is so since = ZX <1| x10). The matrix element of z* may be deduced from the quantum-mechanical version of the virial theorem, which in this case yields that the expectation value of the potential energy is one-half the total energy. Thus clei) =$ (44). 10> +] 01 be De elo are Bho (I ck e/Ak) to first order in c. Notice that these energies are equal 27. The three lowest eigenstates, labeled by quantum numbers n and m, are antes = seep =a eo ki Nu LZ J ho = {= uexp [-t@ + v)> a =4/Zzep[-Le ty) with corresponding energies Hoo = 1, E,,0 = Eo, = 2. The perturbation Hamiltonian is H' — (8zy/2)(z* + y*). The (0, 0) level is shifted by an amount <0, 0|2'|0, 0> = 0 by symmetry under reflection. The (10) and (01) levels are degenerate; therefore the Z x Z matrix ae sae) <01|H'|10> 01] H"|01> must be diagonalized. ‘The diagonal elements vanish by symmetry under reflection; the off-diagonal elements are each equal to 38/4. After diagonali- zation, H’ has as diagonal elements, 35/4 and —38/4. The perturbation,Quantum MEcHANIcs—SoLvTions 169 therefore, splits the degenerate levels, whose energies are now -2488, with corresponding wave funciions yp. = (ij 2 jor = Wi,0)- 28. We take as the Hamiltonian of the electron, H = pBo,, where jo represents the magnetic moment of the electron, and B the magnetic field strength. Tn addition, we choose the representation sca A Rl) ace The spin wave function satisfies the equation, io =H — where p= A(t) vei ) + co s(_ i) A and C are to be determined subject to the constraint A(0) = e, C(0) This guarantees that for t = 0 the spin is in the eigenstate 8, = }. The equations of motion give iA(t) = u»BC and iC(t) = uaBA. and the appro- priate solution is A(t) = e* c08 (1 Bt) CW) = ie sin (u BE). = +4) =lAOF = cos* (uoBY (b) (Probability that 8, — —4) —|C()[? = sin? (u,Bt) (c) (Probability that S, = +4) = }. vperains obey their classival cyuatious of mutiva (Ehrenfest’s theorem). We therefore have the equation d¢S)/dt = x B, where = 24,8 is the electron magnetic moment. Subject to the initial conditions that <8.) = 4. <8, =0, and <8.) =0 at time t = 0, this equation has the “GS. — 4 208 2468); <8 — — Bt); (8) — 0. To find the probabilities, we need only use the definition (S,) = (P, — P_)/2 and P, + P.=1 with similar expressions for (S,> and . The solution obtained in this manner agrees with the previous result. 29. An electron with energy in the kilovolt region moves away from the daughter ion so quickly that its influence on the atomic wave function may be noglocted. If we assume the decay takes place in negligible time, then the “sudden approximation” is valid, and the wave function of the bound elee- tron remains unchanged during the transformation. If the initial state is n 1, Z2=1, the amplitude for finding the clectron in the state n = 2, % = % immediately after decay ia A =n -- 2, Z=2|n - 1, Z= 1).170 QUANTUM MECHANICS—SoLUTIONS The normalized states are Z = |n =1, Z=V = (xa) exp (=), and [n= 1, Z=2) = (aa) 4(1— Len, where a is the Bohr radius. Thus the amplitude is given by ("ane ar(1 — L)em = 1. B acces z The probability is given by P =|A* = 4. 30. Let the initial B field define the z-axis; the final B field, the z-axis. In the approximation that the change in field occurs in negligible time, the wave functions just before and just after the change must be equal. Initially y = (}), with respect to the z-axis. However in the basis with J,, diagonal cos (6/2) \_ sin (6/2), ain (/2)/ This result is obtained by requiring that 1 be an eigenstate of o-f, where Ai points in the 2-direction; i.e., one solves [corp —sing\, \-sin —cos $/* ~ * The probabilities for finding m} = 4 or mj = —} are, respectively, cos* (#/2) and sin? ($/2); ie, 3 and 4. Fi ihe Gvid is applied, the expeciution value of by P SWPIE-# LY) = 0094 pide ly> = 4 con g = Po =P with P, + P_=1. Here P,, P_ are the probabilities of finding m; = +4 along the z’-axis, Thus: 31. Conservation of angular momentum would require that the emitted photon have total angular momentum zero. Now a photon has spin | (des- cribed by the polarization vector e) and therefore must be in an (L = 2)- state (described by ita momentum vector k) in order to form aJ = 0 atate.Quantum Mecuanics—Sonvtions 171 Thus the transition amplitude is proportional to the scalar combination e-k, which vanishes because of the transversality condition V-A — 0. Hence it is impossible to put a single photon in aJ —0 state, and the transition is forbidden. Radiative decays may therefore proceed only with the emission of at least two photons. A well-known example of a forbidden zero-zero transition occurs in the hydrogen-like atom. The transition from the 28-state to the ground state is forbidden via single photon emission; however, the decay may proceed via two photons. The two photons have actually been seen in coincidence in the decay of the 25-state of singly ionized helium IM. Lipoles, et al., Phys. Kev. Letters 15, 690 (1965)]. Another example of a zero-zero transition proceeding, with the emission of two photons, is the decay ® — 1 + yy. ‘Transitions in which J is large proceed slowly because the transition matrix clement goes iv zero ai least as fasts as (H/Aj"—! as J geis large, where is the wavelength of the emitted radiation and R is a typical radius of the emitting system. This can be seen from a multipole expansion of the transition matrix element, , where j.(r) is the electromagnetic current: of the 82, The transition rate in perturbation theory is given by T = 2x|H',[? x (phase space), where H'’ is the perturbation Hamiltonian. For electromagnetic processes, the perturbing Hamiltonian is obtained by the principle of minimal elec- tromagnetic coupling (Ampére’s Hypothesis) p—>p — A/c. Then the Hamiltonian is —H,-,2@- 2me* P) to first order in e. In the coordinate representation, p is the gradient operator. One may choose a gauge for A in which div A = 0; hence the perturbing potential may be taken as H’ = —(e/mc)(A-p). ‘This is taken between the initial state Z\*" \-zrnae wa (2) er and the final state, assumed to be a plane wave e'?**, Then exhibiting only Z.depondont factors. Z.dependont factors, Hie~ a" fate exp {a where q — p — k, and k is the photon momentum. Completing the integral, we obtain 7 gen W~ Gt Day172 QUANTUM MzcHANIcs—SoLUTIONS At high photon energies one may neglect Z*/4a3 < q* and the Z-dependence of H, is H', ~ Z*. In addition, when recoil and binding energy are ne- glected, the final density of states is Z-independent. Finally then, the cross section is proportional to Z*. 33. The spherical square well may be treated as a one-dimensional square well, provided the potential is replaced by the effective potential V(r) = —Vo + (Hm?) L(L + 1) inside the well, and V(r) — (K/2mr*) L(L + 1) the wei, is given, in WKB-approximation, as T 2 (pe) ete, where SE = 7 — By ° = (by) ep {-2 f ar[ EY — 2". Here, 6 is the point at which the radicand vanishes. Each time the particle, which may he nictnred as honneing hack and fort! ‘il the [2/my z+ ¥.)}""/2a) ‘nw 30 (E+ Vo)}'",, ~ ref! 2ma* | thus, Upon change of variables the integral in the expression for 7’ becomes Geese mat 2 T+1 ae(5—1)", here - (PRES! VUEFN | ae( where ¥ = | TFT), The exucd integration is somewhat involved; however, in the cave <1, one has [fae ne (2) and the expression for + becomes Gye 34. The second-order perturbation amplitude for the transition is The interaction Hamiltonian is taken as H = — (cA - p)/me, and the dipole approximation is used. Both initial and final states have L = 0, while the only intermediate state contributing has L = 1. As the matrix clements . However theQuaNToM MECHANICS—SoLUTIONs — 173 initial and final states (L = 0) are 30) 5 spherically symmetric, so M x $6, + € Ip, and for these partial waves, 7, = 1; ie, there is no scattering because there is no interaction. However, for 1a), Uilr) = Arloos 8,4.(kr) — sin 8:m(kr)], where the spherical Bessel functions are all obtained from the recursion relation and from the solutions for 1 = 0: + cbr) — Sin (kr) _ 008 (kr) Jolker) = SE and goler) = — Since U,(a) = 0, tan 5, = j,(ka)/7,(ka) gives the phase shifts. As we are told to neglect D-waves, only two phase shifts need be computed: 5, ~ tan 5) = —tan (ka) tan (ba) __(ka)*, b= tan}, = He ~ eh The differ ee ernaa section ia = prl Bet sin (6) 21 + 1) Py(cos 0) = LI ,-tea gin (ha) + e-t90"2 gin [ET ana al? I Bees us aft — Ua" + 2(ba)? cos8 + terms of higher order in (tay). The total cross section is a= [a0 86 = 4nat(1 — UD +...)176 Quawrum MxcHanrcs—SonuTIoNs 39. In case (a) we are dealing with low-energy scattering for which the S- wave dominates. The radial wave function U = sin (kr + 8,)/r must vanish ‘on the surface of the sphere; thus 8) = — ka. The scattering amplitude f(0) then reduces to /(6) = —a, and the total cross section o = {dQ f[? = 4xa*. In case (b) we are dealing with high-energy scattering and all partial waves contribute. Since classically all particles having impact parameter greater than a are unaffected, one expects 5, = 0 for Ih > hka. However, those particles for which /k < hka would be climinated from the forward direction (ciassicaiiy speaking). This intuitive ciassicai argument suggests that one take (ay eh 1 for 1 > ka 0 for < ka] for purposes of caicuiation in ihe forward direciion (oniy). Actually these arguments can be made more rigorous, and in so doing we show that for seattering on a hard (totally reflecting) sphere, we obtain a diffraction peak in both the forward and backward directions, while for a black (totally ab- 4 the forward dircct the condition: U,(a) = cos 8,5, (ka) — sin 8,9,(ka) = 0. (2) For ka > 1 we make use of asymptotic expansions; thus (a) for >> ka > 1, one has j,(ka) ~ 0 and n,(ka) > 1, and Eq. (2) is solved by taking 8; = 0 in this case. (b) for ka > 1, one has i~esin(ka—S) and m~— ia” \ zy Eq. (2) io then aolved by taking 8, - ka | Ix/2 ‘Thus a somewhat more rigorous approach than our original one yields pea thi {i (—Ie*#2 for ka > 1 This expression for eis to be substituted in the formula for the scattering amplitude, (3) $0) = sip & 2 + 1(e — 1)P,(c08 4) 1 = (2 + 1)(e** — 1)P;,(cos 6). Near the forward direction P,(cos@) ~ 1 for all 1, and it makes little dif- ference whether one uses e”®' =: 0 or (—1)'e-***, because the latter choiceQuaNTuM MECHANICS—SoLUTIoNs 177 alternates in sign, thus making its contribution neglible compared to the sum 5/2, (21 + 1). Thus, for either choice, JO = 5p $ (1+ 1)P,(cos 0) for small 8. For ka > 1, one may further approximate this sum by an integral: ioe S0)~ =, J, tdxPz(cos 6), ‘Thus do/dQ = | f|* has the approximate value a‘k* in the forward direction and decreases rapidly in the angular region 0 < @ < 1/ka. Consider now the backward direction. For a black sphere, e*® ~ 0 for 1 < ka, and the scattering amplitude is sum tend to cancel, giving no peak in the backward direction. However, for a hard sphere we have shown that e%® ~ (—1)'e~**; thus, in the back. din ono has ha ka TE (2 + WPi(cosp), where $ = x — 4. Sn(6) = Hence for a hard sphere one obtains a diffraction pattern of the same size and shape in both the forward and backward directions. 40. The wave equation inside the well is REGS 1 (Veay8lary — By, ) while outside the well, a We use the parameter (Va) with dimensions energy-times-length to de- scribe the strength of the potential. Because the introduction of a delta- funetion in the wave equation makes a direct: solution of (1) difficult, we first climinate the delta from (1) by an integration. We integrate (1) over z, from —e to +, and then let e — 0. We have [ako HH=9)] + (rear) Bf wayaz >o (1)178 QUANTUM MEcHANIcs—SoLUTIONS Outside the well, the wave function is We) =a'e* — forz>0 ater for s <0, normalized to {*2 ||? dz = 1. Substituting this solution in (1’) relates a to the potential strength, & — —m(V,a)/A?. Then (2) gives the bound-state energy (for V, attractive) E= {Fea}? ‘Thus the delta-function potential has only one bound state. It is striking that if we choose Vy — —e'/a, we obtain the energy of the ground-state hydrogen atom, and the wave function for z > 0 becomes that of the hy- drogen atom (radial part). To solve the scattering problem, we introduce, for x <0, an incident wave and a reflected wave, while for x > 0 we have a transmitted wave. ‘Thus ple <0) = e%* + Hes, He > 0) = Tees, Substituting into (1'), with the condition plz — 07) = We —0*) which requires 1} R=, we obtain —(MMLR)/m 4 (Voa)(1 +R) =0, from which soersccicee 2— (—mVaay’ and = 1 : 2 1 + imVsajkht) The reflected intensity is therefore fs 1 . IRU = map while the transmitted intensity is 1 = : ITP= T+ (mV.a/n7ky? As the strength of the potential well increases relative to the energy E = h'k*/2m of the incident: beam, the intensity of the reflected heam ap- proaches unity. On the other hand, as (Vya) —> 0, the transmission coeffi cient approaches unity, ie. the potential has no effect, as expected. It is interesting to note that | | and |‘7'| do not depend on the sign of Vs and that the pole in R or 7 at k= —imVa/h? yields the bound-state energy E = MH2m = —m(Vea)#/2h*. Also see Problem (11-21).Quantum MecHantcs—Sotutions 179 ‘An alternate solution to the bound-state problem is obtained by con- sidering the wave equation in momentum space. The wave equation in coordinate space, for all x, is HO) + Fyudierpie) = Ba). Upon multiplying by e'”* and integrating over all z, we obtain, assuming that p(x) and dap(z)/dz —+ 0 as|z|— 0 in the integration by parts for the first: term, the wave equation in momentum space E op) + Voarp(x = 0) = Ep(p), where = fe ada) ed. Thus taba = 0) _ _ 2mVoap(z = 0), $9) = —2mi ~~ p+ am where W = —E =|E|> a We now obtain -19: dp = — = We) = hy ft pde 2 dp = — 2 Vomfe = 0) [ pm Complex p plane The integration may be performed in the complex plane, closing the contour beluw when « > 0, and above when « <0. We obtain Wa) = — ect, where e(z) = 1(—1) when x > 0 (< 0). The requirement that yo(z— 0*) = ap(z — 0-) = yx = 0) gives a self-consistency condition — mete Wa VO ig AJ 2mW 2a Notice that @ bound state is possible only when Vo is negative.THERMODYNAMICS 1. The thermodynamic potentials are defined by: U (internal energy), H=U + PV (enthalpy), A =U — TS (Helmholtz. function),* @ = H — TS (Gibbs function). ‘The quaniities aU = TdS — Pav, dH = TdS + VaP, dA — —PdV —SaT, dG = VdP— Sat, are all exact differentials. Equations (1) follow immediately. One also obtains such identities as and To establish (2), expand ras = 1(25) av + 7(28) ar = 1(2%) av +c,a7, using (1). To establish (3), TdS =dU + Pav, (5), - P+ (ap), - 24). 2. In a reversible, infinitesimal change of state, TdS =a + Pav, * We will sometimes denoto the Helmholtz function (freo energy) by A; F is also used for this function. Both notations are common in tho literature. 1x0THERMODYNAMICS—SOLUTIONS 181 from which 3. Consider infinitesimal changes of state along the phase boundary. (a) dV, =dV;, or Trapt ray\ey rayyt ayy on), — (ap) l¢T = dP| (55) — (sp) | LNOP ie NEBr) LNOP/r OP /r 4 Calling @ = (1/V)(aV/27)» the coefficient of volume expansion and & = (1/V\(@V/aP); the bulk compressibility, we have (b) dH, = dH, implies dS, = dS, in a phase change at constant pressure. Then (28) — (28y Jap = (28) — (28) Jar. L\OP/r \OF/ rh L\OT/P \OT/P\ o easy as\' ar (G51-G5)_ ee ap _ Worle Noe} _ _ aP = [/asy a8 ar ary (sp), —(ap),] 7 Lar) — (or). | ‘These two expressions for dP/d’ are known as Ehrenfest’s equations. 4. Ou the assumption that ihe body of water iv taken down with nu heat exchange (1.e.,48 = U), the change in temperature is AY’ = (A7'/OP)sP where AP — pg Ah — 9.8 x 10" dyne/em?. Using the identity © = 300 x .0013 em’ deg/cal eog/dyue. Henco T = 1°C.182 ‘THERMODYNAMICS—SOLUTIONS 5. If the two adiabats intersect, they may be jomed by an isotherm Now consider a cycle along the closed curve The work done 1s equal to the enclosed area, and heat 1s absorbed along the isotherm Thus the cycle absorbs heat and does work, without giving off any heat The efficiency of the process 18 one hundred percent in violation of the Second Law of Thermodynamics 6. (a) We have 8 = Cp tn (228) — + cp n (328) 4 from which n = 01 mole (b) The heat removed from the first mole 13 =C,AT, + = (18 male deg a) 25 deg) + 1438 — 1888 cal, while the heat absorbed by the second mole 1s Q: = Cp AT, + nd’ = 2320 cal ‘The work done by the refrigerator 18 Y — Y; = 430 cal 7. Since the final entropy does not depend on how the final state 1s reached it will be calculated as if 1t were reached ssobarreally This 1s possible hecause TdS—C, dT AS,=CplogZ and AS, = Cp log Zt. Fr, P But 1; = (1, + Ty)/2 and Cp = (6/2) Nk Therefore AS = AS, + AS, = ENtlog (qo T T, rh) sNblog (Tit To ° 2ST T, J? which vanishes if T, = T as 1t should 8. (a) Mechanical equhbrium requires V/V, = 1/3.att =Oand V,|V,=1 at tf = co (b) The final temperature 1s 71, = Ty = 27, The process 1s 1sobanic, there- fore T dS = Cp dT for each compartment Integrating, one obtaims = Oplog (2 and AS, = Cp log (?®) The total entropy change 18 thus Clog (8) #R low (4),‘THERMODYNAMICS—SOLUTIONS 183 (c) If the transfer had been reversible, we would have had AS=0, or fora? 1 RY 0 Grin (ie) + Rin (G4) =o. Plugging in V, = (V4 + Vs)/2, Va = Vs/2, Va = 3V 2, one obtains QTRi4)-Rice _ gtyar-2a Ty = OTS RY — BTR $y or The amount of useful work possible would have been W =U, — Us = 0,(4T, — 27) = 3R(2T, — Ty). 9. Ina free expansion, there is no heat exchange with the outside universe, and no work is done. Hence dU = 0. But au = ($9) ar + (S9) av =crar + SP } =a = dilja. ar——(q4n) @¥<0, oF M=MH(T Pe A free expansion is not a reversible process; to compute the change in en- tropy, one must find a reversible process connecting the initial and final states. Then one computes AS for this process; the result is independent of the process, as di is an exact: differential. A convenient process consists of (a) expansion of the gas at constant temperature to the final volume, then (b) cooling at constant volume, to the final temperature. ‘Then, asm (89 — (QUEM, PHY | ("age the first term contributing only during the expansion, the second only dur- ing the cooling. Inserting P = RT/(V — b) —a/V*, and (a1/aV)p=a/V? in the first term, we have As— Orin Be + wn {P= 4} Rinft, H=U+PV and a. However, AU — 0; hence = 40 + A(P7). 1 l)a 2 roe: 10. Tho process described is throttling, in which tho enthalpy, H =U |- PY, romaina constant. ‘The initial enthalpy of | gm of water AH = P,V,— P,V, = R(T, — T,) = (184 ‘THERMODYNAMICS—SOLUTIONS is H, = U, + P,V. If, in the final state, a fraction, x, is converted to steam at the boiling point of water 7’, = 100°C, then the final enthalpy is given by Hy = (U, + O(T;— 7) + PV] + 2b. brackets is the enthalpy of 1 gm of water at P; — } atm and 7; = 100°C. The last term is the enthalpy change upon a change of phase of x gm. Here L is the latent heat of vaporization. ‘Thus OT, — 1) + 2L = (P,— PV, where C = 1 cal/gm - deg: L = 540 cal/gm: V = 1 em*. Then (P; — P;) ~ 10 atm ~ 10" dyn/em?; 1 cal = 4.18 x 10" erg: From this equation one finds x = 0.3. change on cooling is given by T dS =C,dT + Pav, which, combined with the ideal gas equation, yields 2, P, ASe = (Cr + B) log 7 — Blog 5. The entropy change upon liquefying is AS,=—L/T,. However, (AS), + (AS), = 0 by hypothesis; thus P= Pa exp (zn) = 100 atm. 12. Suppose that the center of the compression at any time has a higher by a distance 4/2 where 2 is the wavelength of the o 7/2 (where 7 is the period) one has approximately [Heat flow from crest to valley] = x($3)4($)> [Heat flow necessary to raise temperature by AT] = pa(> Joar. Oscillations tend to be isothermal if the heat flow is sufficient to equalize the temperature in time 7/2. Thus Karst S> Apu @ (isothermal condition) or in terms of the frequency f = 1/r, 1<3k. a However, the frequency and wavelength are related by fX = /Y/p. Thus‘THERMODYNAMICS—SOLUTIONS 185 the above relation may be written, after eliminating 2, > = 5 Xx 10! sec"! 13. Let n, and n, be the concentration of CO in the first and second vessels, respectively. The current of CO from the first to the second vessel is then 7 Dim mA, L which must equal — V (dn,/dt). However, n, + n, = 7 is constant, since the total amount of CO is conserved. Thus, dn, 2DA\, | DAn & =~ (Cap) +a The solution to this equation with ny(0) =m (all CO initially in the first vessel) is m= BE etoreny, and since the partial pressure is proportional to the concentration, P. ~sapeer} p= Pg ecanery, 14, The heat fiow equation V- iH + Cp (67 /ét) = 6, with H = —KVT, has the spherically symmetric solution for t > 0, Ter, t) =F Agere marlA), — with — TR, where a = (K/Cp R*) and R = radius of sphere. The coefficients A. are found from the initial condition Tr,0)= T.= 100°C forr< RK. Thus J A, sin (nzr/R) = Tor, and from the orthogonality relations foarsin 2" sindat — pO one finds Therefore mmr (-tt mr mt om R T(r, t) = 2LER § (=U exp (—n?x%at) sin 22 and (0, t) = 27> Da-1 (—1)"*! exp (—n*x* at). When t = 15 min, x*at = 2.85 x (900)/(20)' 4.16. It in therefore a good approximation to keep186 ‘THERMODYNAMICS—SOLUTIONS only the first term in the sum T(0, 15 m) ~ 200°C x exp (—4.16) = 3.1°C. 15. (a) The temperature satisfies the heat equation aT 42 = x VER a =o where a =a. In spherical coordinates with T only a function of the radius r, one has Ut (eyatlaete, and the for the tempe: emp Ty.) = T+ Zee», with @(r, t) satisfying the equation Q) The boundary conditions on @ are QRe>o— 7, 7, In terms of the Green’s function G(r, t\r’t’) defined in the region r and r’ > R and satisfying the equations : 20 + 42 = a6 — ypu —09, @) Gr, t\r,)=0 when >t @) and Gir, HB, ¢) = 0, ® the solution is 7,1) = far oR, 1) 2S tine. This may be proved by multiplying Eq. (1) by @, Eq. (2) by @ and inte. nd f, m: Baa. (2) and (4). To calculate @, con of Eqs. (2) and (4). To caloulate O, con- are sider the function Gp defined to satisfy Eqs. (2) and (3): Gop — ryt —t) = J akflk,t — yer”. Then kf + (1Ja?\aflat) = A(t — t')2r with solution yo e-Hant-t9 S(k, t — #1) = 5 Ot — the > where for t>e, “8 for t R. Finally then (= Rot —¢') (r— Ry 2an/ a(t ty" woe {- at —?) igh) and the temperature is given by 11,8) = T+ RE = tye = By dle T-BALL) ¢—t)* ? a !) = aR tle.) which can be simplified by a shane of variables to read 2RUT, — 1) om T(r,t) = 1) + = de, ta dere mnann (b) In the limit as t —> 00, the integral approaches fleeea 23 thus 7 — 7, + (7, — )(R/r), as would be expected from a static problem with the boundary conditions T(R) = 7, and T(0o) = 16. (a) We start from the energy equation with U = EV, av) — Taz), — P- The pressure is given hy P = F/3. Hence (5), aot) = oT}, 3 aV)r Substituting these relations in the above equation yields 42/T — aE /aT. This has solution B = kT". Alternate solution: Consider an infinitesimal Carnot cycle operating on the photon gas. In an infinitesimal displacement, dP — d#/3. The work done is the volume enclosed in the PY-plane, or W — (dP)(dV) — dB a¥/3. The heat adsorbed is @ = E dV + PdV = (48/3) dV. The efficiency of the engine is W_dE_ aT 7~Q 4h T°188 THERMODYNAMICS—SOLUTIONS This has solution # = kT‘. The equating of 7 to dT'/T follows, since all reversible heat engines have efficiency d7'/T. The constant cannot be ob- tained from thermodynamics. It can be calculated from quantum-statistical mechanics; alternatively, it can be related to the Stefan-Boltzmann constant o. Imagine * a small hole in the cavity, from which a energy is emitted. The energy flux is PT P-aP,T-dt (cB) _ (kT) _ yy _ gu 4 4 , from which k = 4o/c. a v 17. The power absorbed by the satellite is (47 R*aT%)(xr'/4xD*). The power it. radintes is @T"-4zr?. This gives an equilibrium temperature determined by T* — T4Rt/4D* — ot*T4/4. The data in the problem give T — 288°K. 18. The probability of observing a magnetic-moment density between M and (M + dM) is proportional to dM e~*"", ‘Thus the most probable magnetization M, is the one for which A(M, 7) has a minimum as a function of M. In addition, when one neglects the effects of fluctuations, the average magnetization, m=(au Mes | ( aM“, qa) is equal to Mz. This is because the distribution function e~“"" is large only in the vicinity of Ms. From the condition that A(M, 7) be a minimum, one finds To obtain an estimate of the effect fluctuations have on the calenlation of #, t fe) <7, and (b) 7, < 7. (a) Define a new variable x by er i", J M=M+al, where Mo=[Si7,— 17]. Lép dt Then, expanding A(M, 1) about the value. My, one finds at . ag A= Ao — Ga(Pe— 7 +er(e4+3 +i) with B= (aT. — T)/B kT)".THERMODYNAMICS—SOLUTIONS 189 With this new variable, one finds from Eq. (1), Exact evaluation of these integrals as a function of B is not possible. How- ever, in the limit B > 1, one has, to lowest order in 1/B, that [" are ¢ Lt oe og Vetm oe ag, | ,74ze ‘ a) rede ~ Zl] ve “az. Thus, in this limit B >> 1, one finds then ‘ A=4,+ er(y +4), and Sluayewenm [ vd ie yy TTS) Payee , where B is as defined in (a). In the limit B > i, one obisins 8= [Fra] In the presence of a magnetic field, one adds a term —M - H to A. Thus above the Curie temperature. the average magnetization is given by ke inf ydve 00 = [apt as)" fae ‘ where tarms of order 1/R < 1 hava heen neglected. Thna to first order in H, vB [era(T—7.)1) (H) = H se Wek aa{T 7) Curie-Weiss law). ‘The susceptibility is given by 1 — om oo X= OM|y-. FATT)190 ‘THERMODYNAMICS—SOLUTIONS Note that this result could also have been obtained from the thermody- namie relation H = 24/0M = 2a(T — T.)M + 48M*. Thus 1=par— 7.) | emt which implies om 2H = 1 jun~0 2a(T — T.)” 1%. The sound veiocity is caicuiated from the adiabatic puik moduius at constant magnetic field H: We begin by showing that the familiar relation OP) _ Cp.n(@P’ (Pan O46), is true even in the presence of a magnetic field, the specific heats being cal- culated at constant H. Of course these specific heats will be function of H, and it is because of this fact that the sound velocity is changed. From the identities: © Gp). = Gr). + Gla u(FP)..» = (eal! + GP)eal Fn elBP ea ® Pe alet w= — Grew ® (a al)eu =~ (Four one finds by substituting (b and (c) into (a) that (ran (Peal! + cr a(or)s alr) where we have in addition used the result that Cy, = T(OS/OT ya. Le addition, the definitions and‘THERMODYNAMICS—SOLUTIONS 191. yield the relation OFF aa Sng = Cre — Ore (=P)au = Gr), lt + oy] = B64(5P ew We must now calenlate Cp, z/Cy, 2. This is accomplished by calculating Thus Cr.w=T(28/2T)r.2 and Cy = T(aS/OT vn From the T dS relation, T dS =dU — aM + Pay, one finds that the thermodynamic potential ¢ = U — TS — HM satisfies dp = —M dH — SdT — Pav. () In addition, we are given M = yHI/T’; thus, integrating the above equation at constant temperature and volume, one finds = yt" O(T, V,H) = oT, V) — Fr» where $o(7, V) is independent of H. The entropy is then found from 27 Finally, then, where C2, C2 are the specific heats in the absence of a magnetic field. As- suming that the magnetic field does not affect the equation of state relating P, V,and T (e.g. PV = RT for an ideal gas), then (@P/2V)r is independent of H, and one has ot) = MOL + ATOR) OU IO)” OT, yields oH) — v(0) _ —vH? (C% — OF) “3 O* ‘0(0) 2Cr, Op 20. Because M ~ 0 for the normal state, the Gibbs function in the normal siaie may bv assumed to be independent of Gy(P, HM) ~- Gy(T, H,). a)192 THERMODYNAMICS—SOLUTIONS Furthermore, there is no discontinuity in @ across the phase boundary: G(T, H.) = Gx(T, H.). (2) Inside the superconductor, B = 0 = If + 4M, from which M — —H/4x. Therefore, at constant temperature, dG, = H dif/47. This may be integrated from H < H, to H., holding the temperature constant at T < T'., to give Hm G(T, H) — GA, H) = However, frum (i) and (2), G(T, H) = Gy(T, 2). This gives G(T, H) — Gy(T, H) = H?/8x — H?/8z. This is negative below the critical field, and positive above it. This shows that the superconductor is energeti- cally favored, when H < H,,and the normal state is favored when H > H. Only the state (ei free energy at a ven temperature and magnetic field is stable. In order to compute the latent heat of transition, the analog to Clapey- ron’s equation is derived. For a change of H and T' along the phase boundary, the changes in the Gibbs function for the normal and superconducting states the th dQ, —4Q, with dQ — Sa? — Mau, Th dH,JaT = (Sy — 8)|(M, — My) along the phase boundary. But the latent heat is given by L = (Sy — S,)T, and hence b= TM, — My (He) = HP) (ZY. Var) = 2az) [1 (x) | ‘The discontinuity in the ae heat is given by pa (L Oy — 0, — PAi8y — 8) = PEF) in pe = aT! —3(z) | Note that for H, = 0, that is 7’ = 7',, the latont heat vanishes while the discontinuity in the specific heat is nonzero. This indicates that the phase transition is second order at H = 0. 21. Any difference between P, and P.. must be due to surface tension y. Fora process in which P and 7 are given, the Gibbs function (is @ minimum u. G= Mg. + Mag. + 4a". Here gj, 92 are the Gibbs functions per unit mass. The last term represents the surface effects; it arises because the energy of a droplet is Ue tar’, the first term giving the energy when surface effects can be ignored.‘THERMODYNAMICS—SOLUTIONS 193, Setting 5G = 0 and insisting that mass be conserved, one has =0= aot |= =p el 84. =0 = [Bho — 99) + Bor Ae = Bat[o, —o + 2], where p is the mass density of the droplet. Hence 9. — 91 = 2y/or in equili- brium. Now for a given phase: ag) _ seceaes (38), = see Solution (7-1)1 so (2a/@P)» = 1/o. Hence. differentiating with respect to P at constant tamper 11 __ 24a Prayer Parp = POP On the assumptions that (1) pierre K Prarop)» and (2) the vapor is a perfect (i), = ~ Gr) When integrated, (Q) where M is the mass of a single molecule of the substance. At a given tem- perature and pressure, only drops of radius r given by (1), are in equilibrium. Droplets that are too small will evaporate and disappear, in an attempt to increase P aud reach cyuilibriuu. Bui this only further decreases 7. Similarly, large drops tend to become larger. 22, The universe is regarded as a gas of stars; in the long run, the average for each species of “molecule” will be approximately the same. Hence ay of Thna y= (y—1) =5 x 10", and vjctet © € 23, Statistical arguments show that ortho-hydrogen may have J = 1, 3,5,... ; and para-hydrogen may have J = 0, 2,4,.... The relative popu- lation of (J = 2)- to (J = 0)-levels is IT + 1h) {ort new [er Nha} or Salag Sexp| ‘site| l=: m where194 THERMODYNAMICS—SoLUTIONS If we consider 1 mole of gas, the total number of molecules is No. Then yao + Myar = No(l — 2), and _ No(1 — 2) I+ Fexp [6h"/mR'kT) (a) Take Ey.) = 0. Then £, = 2 h*/mR* and E, = 6 h*/mR?, and the total rotational energy is given by my B, ~ BM pee + 30(1 — ayer), gf crest cs aR, _ AY owmnner | t Cy, = Fit = 180R0 —»( r, the expression for f(r, t) reduces to (st — r)/4z, which is what one would expect. Finally, the solid angle dQ. is related to an element of surface area, do, on the collector, through dO. = da cos 6/r*. the maas collected per uni in time t ie gi aM _ xmPa* cos‘ £4( h ) do WET cos 6STATISTICAL PHYSICS 1. At each corner the man may go either up or to the right. Therefore a particular path is specified by a sequence (u, u, r, u,...,1) where the total 1dr’ lw Fos maple: the path sh. in the figure has the sequence (r, u, u, ¥, r, rw). Tho numaber of distinguish- able ways of writing such a sequence and hence the total number of paths is (m + n)!/ntm!. n and m respa 2, Assume that the radiation consists of standing waves enclosed in an N-dimensional eube of side a. Then the field vectors have the spatial varia- tion ql sin (kx) = fi a (2), with n-2 Here, the , are integers. Imagine an N-dimensional occupation space, in which each mode is assigned # point with coordinates m,..., ny. The den- sity of points in this space is unity. The number of states with frequencies between @ and do, regardless of direction, is then the differential element of volume in acenpation space, integrated over angle. This mnst. he doubled, because of the two polarization modes. Also the mode obtained when ny —1 is nol independent of that form; hence only 1/2” of @ spherical shell is included. Then dv 2-4 ARYAR, where A is a constant, determined in Problem 1. 34, and (nt = 2h) =S5 2A(alzxc)* @*~‘ dw. ‘he energy is cf - _2da"h (* _o%do z =f Boy = Broy® J aor 1 = ABET) (kT\* (> 2% dex (exe)? \h) Joe aT ‘Then dV Hence K ON EL 197198 STaTisTIcAL PHYsics—SoLvTIoNs 3. The number of states per unit volume, in bandwidth df, at frequency fis 4xp' ,, _ 8xf? -2_M 2x S2P dp —8f'a, whore p—- 2M, and the extra factor of 2 comes from the two possible polarizations of the photon. For those modes such that Af < kT (well satisfied for 3-cm radia. tion at these temperatures), the average energy is given by # — kT. There- af with fore, the energy density of the radiation field in the band Mf ) — srr — > + op folk) = in addition, the energy U is given by 5-CF)- (2) 2 Thus £8) =P eath( BOY and I) = 2 oath (FO) we \2kT) we 2. \2kT) In the classical limit, kT’ >> hw, these expressions reduce to y=2 and )*> = CE’ — 2EXE) + CE)*), we have (E — (E>)*) = — CH)*. In addition LHe 1 ae {BY = AS a = Z 28” where Z, are the energy states of the system, Z=Dem and (3) =ET. Similarly emer 1 og «BY = Eso - aR thus cB — (nyt = 2 (1.92) — 2m, aohz a) = — 28 but eno ED _ w- Mg ond RP =O. Finaily one obtains ((# — T, is as = [242 _ a —in9y Jo FF ) In order to calculate Cnax, an independent calculation of AS is needed. This is furnished by the Boltzmann formula for entropy, S = k In W, where W is the number of distinguishable states of the system. Because of the ferro- magnetic property. all the spins are lined up at 1 = 0°K. Hence W(0°K) = 1. However, for a temperature greater than 7’, we see that the system has maximum entropy, since it can no longer absorb heat. Thus all spins arc uncorrelated and W(T > 7.) = 2”, where N is the number of spins (Avo- gadro's number for one mole). Finally, then, one has AS = Nkla2 = R In 2, which, when combined with the previous expression for AS, yields Rin2 Cae = Za 15. Consider a state described by @ complete set of quantum numbers. with energy eigenvalue e. If the state is occupied by p noninteracting par- les, the energy is pe. Since Fermi statistics allows at most only one par- having a given set of quantum numbers, the partition function for this state is Z, = 1 +e", On the other hand, Bose statistics allows an unlimited number of particles in a given state. Hence, Zy= Di-0e-"!7 = 11 — 7). That form of tic204 SratisticaL PHysics—SoLvTions para-statistics which allows two particles to a given state has Zp =1 te pe mar, The energy is found by computing = (ery22llog Z) 0 = I Ey One finds The bracketed factor in each case is the statistical factor. (One should em- phasize that no examples of para-statistics have as yet been discovered; nature so far seems to prefer the two extreme cases.) 16. Assume, io simpiify the argument, that iongitudinai and transverse modes of vibration propagate with the same speed c, and neglect dispersion (ie., ¢ does not depend on frequency). In the Debye theory with boson- phonons; q) ‘The upper limit is found from ho,, = 1, where @ is the Debye tempera ture. If the phonons were fermions, we would have AV pm atdo ay wel, [FLT “ At high temperatures (1) gives @ constant specific heat, whereas (2) gives a specific heat going to zero. At low temperatures, the simple transformation 2 = ha|kT shows ¢ ~ T? in both cases. 17. The number of particles per unit volume in the velocity interval d'v is = 2m gs, 1 7 © exp {imv"/2) — EVRY + 1” where Zp is the Fermi energy. The current density leaving the cathode and entering the plate is _ —oe(™)* [7 - vedvs J = Kats) = 2 (#) ie dv, do, [ exp (eno) — By kT + 1)’ where mu!/2 = E, + + eV is the total kinetic energy an electron must have in order to reach the plate, Hero ¢ is the work-function and V is theSratisticaL Paysics—SoLvrions 205 retarding voltage. Assuming e( + V)>> kT, the current density becomes: . 7 - J= 2e() efrttt [age mitetdy, [dv do, eomebrohiar 18. The Fermi energy, computed from the equation V [7° 2-4xpt dp = N, s\n #,=(aem) ” where £ 1s the radius of the star. If the star 1s sufficiently dense, the Fermi energy will be much larger than the average thermal energy, and few elec- trons will be excited above the Fermi level. Thus the star may be treated as a degenerate gas. The ground-state energy is Sa U= [Pp peydp = SNE, and the pressure is P = esa) = NE,/4V. But P can also be obtained by requiring hydrostatic equilibrium: if the star expanda, the decrease in gravitational potential energy mut he the work done by the pressure. Hence P-42R* ~ GM*/R* (order of magnitude). Here is the gravitational mass; it is different from zero because the neu- trinos have a total kinetic energy Mc? = U = 3N#;/4. For equilibrium, the two expressions for pressure must be equal, and this leads to the condition 3¢NE,/4Rc' = 1, or, in terms of the mass, GM Ret The considerations developed here may be extended to the case in which the leptons have a rest mass (Chandrasekhar's theory of white dwarf stars), and it may be shown that there is an upper limit to the mass of the star.* * 8. Chandraokhar, Introduction to the Study of Stellar Structure. U. of Chicago Preaa, 1939, Chap. XT.206 SratisticaL Prysics—SoLvutions 19. For case (a), the net magnetic moment is for MH<&ET. Hence For case (b), M(et#"7 — ¢-Ma0T) MH OD = “Camera ~ gr for small H. The magnetic susceptibility is y = M*/kT. 20. The energy of a particle whose magnetic moment is parallel (snti- parallel) to 17, is given by Since the energy levels of the system are populated according to the dis- tribution function i IO) = 0 6a 1’ and the density of levels is given by (4xV/h*) p'dp, the total number of particles N is given by and the magnetization/volume is SE [apotif(U.) — 110.0. 2) Equation (1) may be solved for £ in terms of N, T, and H, and £ may then be substituted in Eq. (2) to determine M/V as a function of N, 7, and H. Upon defining a new variable of integration # = p*/2m and using the low-temperature expansion formula given, we find that Eq. (2) becomes pee e+ wal + F(ee a) | eal s (ep) which, after expanding in powers of H and keeping only the leading term,SratisTicaL PHysics—SoLvTIoNs 207 becomes (4 M) _ Sep om gin afi- (2) + a 4 terms of order H? 0) =9 =H poner SCE) + Solving for £, one obtains where & is the Fermi energy at 7’ = 0°K, and £3? = 3h*n/16x(2m')'?. ‘The susceptibility then becomes =- (eh Ey +] x= aH (He) BCE) +}: 21. Consider the force equation for the particle in one dimension, ydix ott. Bay a Map PG t Fe ay where 6 = 6xRn (Stokes’ law) and F(t) is the force due to fluctuations of the molecular collisions. Equation (1) may be rewritten as M@ oy ata — spe) — Me = A (2) + oF) When an average is taken over a large number of drops (ensemble average), this becomes $d) FENG») @ where (xF(t)> = 0 because x and F(t) are uncorrelated and the average Fey Fey Integrating Eq. (2), one finds a ART ata, (% v)= at Cee (3) Neglecting the transient term. we find that Eq. (3) has the solution a) = 2kTH/6x Ry. Since = kT tx Ry. Substituting 7’ = 300°K; ¢ = 10 sec; 7 = 189 x 107° P; and K = 10™* cm, {ADVt = 2.7 x 10-8 em.208 = SraTisicaL PHysics—SoLUTIONS 22. The plasma is electrically neutral, as a whole. Nevertheless local devia- tions in density appear. Consider the electrical potential $(r) in the vicinity of a particular ion. The energy of another ion, of charge e, in that potential, is e(r). Therefore the density near the ion has the dependence n(r) = neon, ( The constant n must be the average density of the plasma, because the influence of the potential energy is expected to disappear as the thermal dae Q) iy Fe . Another relation between g(r) and nq is provided by Poisson’s equation relating the potential to the charge density: Vig(r) = —4x & cata. (2) On the assumption that the plasma is very hot, we may write, for Eq. (1), ne(t) = na{} - 4) We substitute in ‘Hq. (2) and obtain the Helmholtz equation Vibe) = 4nd eld =e, which has the solution: Plt) = Cro The effects of the electromagnetic interaction are therefore limited to a sphere of radius k~! (Debye-Hiickel radius).ATOMIC PHYSICS 1, The energy levels of the hydrogen atom are given by In the absence of interactions among the electrons, the energy, levels of helium for single-electron excitation would be those of the hydrogen atom, with nuclear charge Z = 2. However, interactions modify this. The strongest of these is the Coulomb repuision e*/r,, between the electrons. The quaiita- tive effect of this is seen by taking the expectation value of e%/r,; between unperturbed wave functions. The effects of spin are taken nto account only so far as statistics are concerned. rls 13.6] —n=1 -24.47|——-n=1 Hydrogen atom Parchelium — Orthohelivm ‘The (n = 2, | = 0)-level [that 18, one electron in the (n = 1, J = 0)-level, the other in the (n = 2, ! = 0)-level] has, in the absence of Coulomb repul- sion, two degenerate spin states: § = 1 and S = 0 The Coulomb repulsion lifts the degeneracy. giving the energy shifts (J + K). (J — K), where Ta [Prd rane dual )aelea uel) and K = [ d*red*raus(rsdetealre)(£ \ wsalreduentrads J Na the former correxponds to the xpatially symmetric, (8. 0)-atate and the 20210 Atomic Puysics—SoLutions latter to the spatially antisymmetric triplet, (S = 1)-state. Since K > 0, the triplet lies lower. The physical origin of this is obvious: the Coulomb repulsion implies higher energy for states in which the electrons have a high probability for being found close to one another. The P levels split similarly, as do the (n = 3)-levels, ete. In addition, the radiation interaction commutes with the total spin of the electrons; hence, radiative transitions between singlet and triplet levels are absolutely for- bidden, in the absonce of spin-orbit coupling. Thus there are two independent specira for helium, 2. Carbon has six electrons. The Pauli Principle requires a ground-state configuration 1s 2s 2p, The two electrons in the p shell may combine to form 8 = 0,1 and L = 0, 1,2. The spin-zero state is antisymmetric under : . : : : antisymmetric; the allowed states are "D, with L L=1,8 =1, and '8, with L=8 =0. The lowest terms are given by Hund’s Rules: 0, *Po,1,2 with (3) minimize J. The first two rules serve to minimize the Coulomb repulsion of the electrons, while the,last minimizes the energy due to spin-orbit coupling. We ignore the latter: this amounts to neglecting the energy differences among the *P states. In order of increasing energy, the lowest terms are *P, 'D,, 'S.. It remains to find the wave functions. First the orbital part of the wave function will be constructed. Consider the D state. We have to combine two (L = 1)-states to get L = 2. There will be five wave functions, one for each of the possible values of m, = 0, +1, +2. In general 12, mip =D OZ, mz | mr, ma) H(1, ma) PCL, ms), where the coefficients C' must be determined, and where $ and wp are single- particle wave functions. Some of the coefficients can be written by inspec- tion. The (L = 2) wave function must be symmetric, so 12, 2> = @(1, Ly, 1), 12, 2) = $(1, “Ny, —D), =aase Vigees care ae ee aes ae 12,07 = pid, HG, H+ GE, HG, OH}, 12, = frien, Opa, —1) + (1, 1)ae(1, 0)} L\JI,M+D=Atomic PHysics—SoLuTions 21 Then 12,0) = /¥ $0, Yl, —l) + VF $C, OU, 0) + VF HU, D1). Jet spin wave function, which in the absence of LS-coupling simply (2 )fee(1) B(2) — (2) 8(1)}- ‘We repeat for the P states. Here we must, add 1 and 1 to get 1. This is simple heeanse the sta I) = Dota oy) (1, Lp(1, 0}, 11,0 =e £60, DK, I) — 64, —D 40, UD), 1 I, =D = 5 (60, OWL, 1) — 60, —1)W11, 0). Note that the $(1, 0) states are not used in constructing |1, 0), because there 18 no way of using $(1, U)4(1, 0) in the construction and, at the same time, maintaining the antisymmetric character of the state. The |0, 0> state must be symmetric: we may write 10, OY = aG(1, 11, —1) + BG (1, 0) (1, 0) + a6(1, —1ap(, 1). Also [0, 0> must be orthogonai to |Z, 0); this shows 6 = —a. Therefore 10, 0> = Agf6(L, DY, —1) — $41, OH, 0) + 64, — DY, D). The |0, 0) = state combines with the (9 — 0)-spin state to state. The three P states must be combined properiy with the three tripiet- amplitudes 11,1) = aa), ji, 6) = yim a) + wizisia, and 11, -1) = AMA) to make the correct J states. This is again the addition of 1 plus 1 to get 0, 1, 2. For example, the 'P, state with M, = 2 is just |1,1>|1,1) and 80 on. 8. Nitrogen has the ground-state configuration 1s*2s*Zp*. ‘The various terms arise from the three p electrons, each having / = 1, (m, = 0, +1) ands=4 (m4 4).212 Aromic Puysics—SoLvrions Each electron may be in one of six states, labeled by (m:, m,). These are a= (1,4), b= (1-4), c= (0,4), d= (0, —4), e=(-L4), f=(-1,-»- A particular state is obtained by combining three of these to form a state with quantum numbers (M;, Ms) =( My, E M,). The exclusion prin- ciple is satisfied by choosing no two states having identical quantum num- bers. One obtains wtb een Be, wtetd= OH) wer = OH) a+b+e=(I,4), (at+d+e)= (0,3) O+e+) =H) @+e+f)— 04. States with negative values of M, or Ms have been omitted as uninformative —they give nothing new. ‘The presence of a (2, }) state, and the absence of states with higher valucs of Hi, or Bg indicaies a °D term. Assovisied with it iy une siaie having quantum numbers (1, $) and one with (U, 4). ‘he other state (1, $) implies the existence of a *P term, which accounts for another (0, $) state. Remain- ing are the levels (0, 3), (0, 4), explained by a 4S state. Hund’s Rules imply E@P) > ECD) > E(‘8). 4. Each level labeled by principal quantum number n has sublevels labeled by J, which takes on values from 0 to n — 1. Each sublevel is allowed 2(21 + 1) electrons by the exclusion principle. This information is sufficient to build up the atoms specified. The configurations are ar: 48*4p° 4d* and Hf: 5s? 5p* 5d! Each has four electrons in an unfilled d shell: chemically the two elements are very similar because these levels are so similar. 5. (a) The energy of a 4123-A photon (expressed in om) is 24,300 om=!. Since a photon must be absorbed in its entirety, there will be no excitation of the atom. (b) The energy of a 3.3-eV electron (expressed in em”) is Elke = 28,000 m=. However, the electron may give up only part of its energy, and the transi- tions 3e + 3p, and 3¢ — 4¢ will take place. There is not enough energy to excite the atom to higher levels.Atomic Puysics—SoLvuTions 213, 6. The electrostatic potential energy of a point electron and » uniform charge distribution of radius R, is [+ when r > R, T= —e¥ini=—e 9 : e - [240 — im) when and — interaction of electron with point nuclear charge), when r < R, when r > R. Hence Also, since R= 10" em <4 x 10"%em =a, we may take e”/* = 1. Therefore ~ _2eRy AR~ 2e(R +t Rt r)a22(2) = 4 4 x (18.5 eV) x (2 x 10-5) 244 X10 CV. 7. In spherical coordinates, Vi) = —£ + Flat (8 — @) cost 6), where 6 is oe polar angle. The inequalities 0 . This expression for a follows from the perturba- of parity or because of conservation of J. As a consequence, the levels |2P, m, = £1 are unshifted. Upon diagonalizing this matrix, one finds that the 2P level is shifted by the amonnt — (A? + 4a 2 while the shift of the 2S level is [(A* + 4a*H*)'* — AJ/2. Note that for strong fields, a£ > A, one obtains a lincar Stark effect, ie. shifts proportional to Z. However, for weak fields, af < A, the shifts are quadratic in E.Atomic PHysics—SoLUTIONS 215 9. The energy due to the presence of an orbital angular momentum in a magnetic field is found from the principle of minimal electromagnetic cou- pling, which provides the rule p — (p — A/c) in the Hamiltonian. To first order in the field, the additional energy is (e/me)A-P. But for a uniform field, one may take _Bxr a=Bx, 80 : AR = © mR x wep = So Re x ny = Rm, Tel XP = geRele XB) = i where B is the axis of quantization for L,. Electric dipole transitions obey the selection rule Am, = 0, +1, according to which the components of the normal Zeeman effect are separated by Ao = eB/2me = 8.8 x 10° sec”'. This is, in fact, the separation observed here. The argument given above must be augmented for transitions in which spin plays a role, In that case the Zeeman effect is called anomalous. . The inveraciion with tie mmgueiie feild respousivie for the iinear Zeeman ettect is H=-—p-B, where w= 2y,(8* — 8°); 8* and 8~ are spin operators for the positron and electron respectively. The spin wave functions of the unperturbed state are: a(+)a(—), Triplet: | CHAI) + a(—)8(+) J, . 2 LB(+H)A(—); Singlet [2CHB=) — a8), The shift in energy duc to a lincar Zcoman effcet is ] H |1p). 11. In terms of the raising and lowering operators J. L, = Lz + iLy, the Hamiltonian may be written H= (es.0. + pal Le + 28,8) 4+ £(6,L. + 9.L,). — 8, 4 i8,, and216 Atomic Prysics—SoLvtions If Bis taken along the z-axis, J, commutes with H and may be used to label states. We will denote by m, the eigenvalues of J,. For the cases m, = +4, matrix elements of the second term vanish, and the first term gives only diagonal matrix elements. The energy levels for the two cases are Esa = ¢/3 + 2poB. For the cases |m,| = 4, there are two states, (2) ~ (re) for each my. Here L, and S, are chosen diagonal and y? is the orbital wave function for a P state. In the basis | (i). matrix elements of the first term are purely diagonal while the last term has no diagonal matrix clements. The off-diagonal elements are given by Sioa Hig = Saini He) = G SeriS 1D, — IZ. imy + ¥), which, upon using the matrix elements of the raising and lowering operators, only to the state iJ, Bf + 1) (jd, 2¢ — 1). All other matrix elements of J, vanish. The value of the nonvanishing matrix element is determined by taking the matrix element > By: faep(ey (wep — (BV) V9}.218 Atomic Puysics—SoLutions Note at this point that, since | y(r') is spherically symmetric, one finds that $(t) is spherically symmetric. Then taking p(r) to be spherically symmetric, we integrate as follows: f dr pir) JOT aa ae, Thus H = 41th f dr ply V*S, and when one uses V6 = —4zj(r)j*, this reduces to H= Saw j d*r p(r) | wr). In addition, jie)! is slowly varying over ihe dimensions of the proton and 8x 3 8% Hobe 3 aimee ce H He|yp(0)[F — (Fermi’s formula) where n is the principal quantum number, and a is the Bohr radius 1#/me*), with pp = 2.79 wy, and ft} triplet) [—3. singlet] The energy difference between the singlet and triplet is (aa) = 8X (2.79)/.) sig 3 \mJso 13. Due to the interaction of the spin angular momentum 8 = 4 (L = 0) with the nuclear spin I = }, the total spin state of the hydrogen atom ground state is shifted by 4Ay,u.8,-8,, while AB(D) = 2App8p-8.. ‘The thas ABUD = 4A sin & thus AE(H) = 4A jinjte 33 3A ppp. Taking the ratio gives, for the hyperfine splitting in deuterium, AE = 0.33 x 10° eps. 34. The eiecironic anguiar momenium is J = §. Uf 7 is the nuciear spin, the multiplicity is (2/ +1) or (2+ 1), whichever is smaller. But 27 +1 = 6, and since only four terms are found, 2 + 1 = 4, and I = 3. The energy shift. in hyperfine structure arises because the nuclens inter- acts, through its magnetic moment, with the magnetic field created by the clevicun. This suggests AE ~ U.S = FF +1)— MI +1) J +1),Atomic Puysics—SoLutions 219 where F = I + J takes on integral values from 1 to 4. One finds AR(F =1)=2+0, ARF 640, AnF iz +6, AR(F = 4) = 20 +0. ‘The differences are 4, 6, 8; this suggests the interval ratios 2:3:4 in the hyperfine quadruplet. 15. The Hamiltonian for the electron in the presence of both the external magnetic field and the Coulomb field of the nucleus is y—-@aAly & 2m r Choosing a gauge such that A = $B x r, where B is the constant magnetic field, and defining H, = p'/2m —e*/r, the Hamiltonian becomes H = H, + Hy, where H, — —¢(h+B) , &(B Xr), 2me | Bmet * and L =r x p. Treating H, as a perturbation, the shift in the ground-state energy is given by : . AE = (18|H, [18> = SOSLB x 18>, when use is made of the relation L| 19) = (a = Bohr radius), |. Upon using | 18) — (a!)""%e-rle OO f anrte ti dr = 2B'a Taxmoa J mc! The diamagnetic moment is given by 16. (a) ‘he Doppler shift in the limit of small velocities 18 »’ = v(1 + 8); hence Ad/A = 28. The average value of f is determined from kinetic theory : At room temperatures & = 1.4 x 10-*: thus AX = 0.7 x 10°°A. (b) From the uncertainty principle, (Aw)(Av) ~ 1. Due to collisions in the gas, Ar = Lv, where Z is the mean free path L = 1/4xpR*; so Ad/A? = 2PR*B/k1. This equals the Doppler broadening when P = kT'/R*A. 17. Each molecule is homonuclear; those which have nuclei with integral (half-integral) spin obey Bose (Fermi) statistics, and must have total nuclear220 Avomic Paysics—SoLvTions wave functions which are symmetric (antisymmetric) under exchange of nuclei. Hence the product of rotational and spin wave functions must be symmetric (antisymmetric). The vibrational part of the wave function is symmetric in all cases, since it is a function of the magnitude of the nuclear separation only. ‘We treat the integral spin case first. Nuclear states with total nuclear spin N = 0,2,4...are symmetric and thus require a rotational angular momentum J = 0, 2,4... in order to have a totally symmetric wave func- tion. Likewise states with N = i,3,5...are antisymmetric and require J =1,3,5....Of the (2£ + 1)* nuclear spin states, (2/ + 1)(f + 1) are symmetric while I(2I + 1) are antisymmetric. Remembering that a rota- tional state of angular momentum J has degeneracy (2J + 1), and that the intensity is proportional to the degeneracy (ic. e-¥*7 = 1), we have Intensity (2J—>2J—2) _ (I+ 1)(4 +1) Intensity @J—1—>27— 3) IJ —1) For the half.integer case exactly similar arguments yield Inter (QJ —1—+ 97 — 3) _ For large J values, which are abundant at high temperatures, these ratios reduce to (I -+1)/I and I/(I + 1) respectively. (a) (I), Protons have I — }. The ratio is 1:3. (b) (2): Deuterons have I = 1. The ratio is Z: 1. (c) (He’), I= 4. The ratio is 1:3. (a) (He'), Bach nucleus is an at particle; I — 0. Here there are no antisymmetric nuclear wave functions and only J = 0,2,4, are allowed. ‘Thus every other line in the spectrum is absent. 18. With the introduction of the auxiliary function u(r) = rap(r), the wave equation is Bde, [-ay(4 — 1) 4 MEM), = ay, apart \p ip Zpa*p* J where p is the reduced mass (4 = M,M,/(M, + M,)). The effective potential is the term in brackets; it has a minimum when _LL+i_,. p t22= Spat Se see seeee P= p= Expanding about this point, V(p) = —Voll + BY" + Voll + BY*(p — po)’. The wave equation then takes the form vy + ye + Vol + By"! — Vo(l + B)-"(p — po)"}u = 0.Atomic Puysics—Souutions 221 This has the form of the one-dimensional harmonic oscillator equation; hence, 4 BV, 7 1), B+ Vall + By = by + B) (n+4) and for small B, T+ 1)h* 1 WL(L+1 B= — 1, + HELM 4 na (ny 2) — 2 MEE Me td) Hake /5¥Taal. Fe wg — BV of pat 19. The far-infrared spectrum is rotational. The rotational energies are —MI +1) ~ Sart The lines are then separated by an amount h/yr* = 2xhcA(1/\) which gives r=14 x 10% em. Ey where p is the proton mass. 20. The energy of a diatomic molecule may be approximated by E=—A +ho(n + 4) + BU +1) + higher order terms. The dissociation energy is the difference between the energy of the ground- state molecule (n = 0, J = 0) and the energy of the two noninteracting atomic ayatems. Thus the dissociation energy is given by E’ = +A — tho. The coefficient A depends only on the internuclear separation and the respec- tive charges of the two nuclei. Indeed, in the adiabatic approximation, one may neglect the motion of the nuclei. In that approximation, the masses of the nuclei are irrelevant. As that is the only difference in the properties of the two nuclei, it follows that A(H) = A(D). Therefore E((D) — EH) $F hlo(D) — o(H)}. Now the frequencies are given by @ = /R][yi where k is the force constant, depending on the charge distribution, and pis the reduced mass of the two heo(H) (1 — «(D)/o(H)] _ heo(H) (1 — 1/2) _ pO et) soc =U?) = 0.08 eV, and the zero-point energy of H, is given by ha(H)/2 = 0.27 eV. 21. An accurate solution requires the use of computers*; we shall therefore be content with dimensional arguments. The nuclei in a diatomic molecule are bound by electrostatic forces. This guarantees that the internuclear distance be of the same magnitude as the atomic radius of the valence elec- * W.Kolos, C.C.J-Roothaan and R.A.Sack, Rev. Mod. Phys. 8%, 178 (1960).222 Atomic Puysics—SoLutions trons. Using the Bohr model to estimate this, we have r(u) _ me) 1, re) m(z) 200 dimensions of an ordinary molecule. Similarly, the ratio of the electronic energies (for identical quantum numbers) is Fale) _ mH) ~ 200, 2} - Bale} The vibrational energies of the molecule may be described by a “force con- stant,” whose magnitude is given by simple arguments as approximately k=cr, Therefore k(u)/k(c) = (relr4) = 8 X 10%, and Ey(y)/Ed(e) = E(w) /e(e) = 2.84 x 10° is the ratio of zero-point energies. ‘Thus r(z) = 5 X 10" cm and #,(u) = 400 eV. The data given tor Hy allow the energy spectrum for Hj to be written U = —A(e) + Ex(e)(2n + 1) + smaller terms, with Ale) = 2.84e6V and R,(e) = 0.14 eV. Since A arises from electronic 4 Jed field of the two nuclei, as well as @ contribution from the energy due to the Coulomb repulsion of the two protons, one has A()/A(e) ~ 200. The energy spectrum for the mulecule is then — Aly) + Bolwh(2n + 2), with A() = 568 eV. The binding energy is therefore D(u) = A() — Bal) = 168 eV.10 SOLID STATE 1. Let 4, and Ay, represent the lattice distances for the two cases From simple geometry, Yee Are 2 Now, the face-centered crystal has four molecules per unit cell, while the body-centered erystal has only two Since there 1s no volume change, we for the h pe 11a P 14h = 1A De|Dye = VFB (2)! = 1 029 2. Suppose a sample of the metal 1s subject to an acceleration —a From the viewpoint of the metal, the electrons experience a backward acceleration a, and hence an equivalent electric field E = ma/e This produces a current density J = maa/e which can be measured Since o 1s known from electrical measurements, one obtains a value for e/m gas, giving rise io hihium chionde The reaction may be broken down into several steps (Born-Haber cycle) Lina | D > Liyapors Liane + NC —> Li, + N electrons, zon + (F)E— a. Cl+ N electrons —> Cl. + NF, Li, + CL — LiUhuig + A Here N 1s Avogadro's number We add Lena + (Fe. +D+NO4 (5)2 > LiCheng + NP +A Also. Linn t (F)et => Choa 1B.224 Sorry SrarE—SoLvutions Subtraction yields NF = B— A + D + NC + B/2. The addition can be performed only when all the quantities are expressed in the same units. Converting calories to electron-volts through 1 cal = 4.2 x 10" ergsand 1 eV = 1.6 x 10- ergs. with V = 6 x 10", one finds F = 4.07 V. 4. Consider current flowing in the y-direction with a magnetic field along the z-axis. The fact that germanium shows no Hall effect means that the The velocity of electrons and holes in the y- gi and », = pin2. Because of the magnetic field these charges experience a force, on electrons and holes respectively, of F, SH Py = ce) = along the z-direction, thus inducing transverse velocities H = pn = +HE ond 0, = whe, and prodneing a current, I’ along the z-axis, equal to: T T' = enyvh, — engy, = eHE (nmi, — epi). This vanishes when * Myf = Nope (ly ‘The total current in the y-direction is I = e(yym, + prom)H, and the fraction of this current which is due to electrons is Using Eq. (1) this becomes f=(1 +e) =e mw) 7 5. Imagine the two phases in equilibrium at temperature 7’. For equilibrium. the chemical potentials of the two phases must be equal, ie. Hesotution — Mee We assume that carbon is perfectly soluble in the fee phase. but not soluble at all in the bee phase. We assume also that the carbon and iron atoms are chemically inert with respect to one another. Then j.. is obtained from the Gibbs function for pure bee iron, but pion has an additional term due to the presence of carbon. In order to compute this extra term, it is first necessary to find the entropy increase due to the carbon. If the total number of iron atoms is N, and n isSotip StaTE—SoLUTIONS 225 the number of carbon atoms, then the number of ways in which the carbon atoms in the fee phase may be arranged is —(N+n)! nt Nt The entropy is then given by the Boltzmann formula, S=kInP = MN +n) log(N +n)—NlogN] for N>n. Now, in the absence of impurities, jije = pice, at equilibrium, However, introducing the impurity shifts the equilibrium temperature from 7’ to (7 + AT), changing the chemical potentials from y to [u + (au/2T) AT). Therefore ar(f),.~47(é4),. + #: where p’ is the chemical potential due to the mixing: aN, “oN ~ Thus o(().~ Gp).]--9 0 In addition, ~ de ~ av en) & and this is the entropy per molecule. If we multiply Eq. (1) by Avogadro's number, we obtain AT[Siee — Spec] = —cRT, the S’s being molar entropies. But the latent heat is defined by L = T(Sre — Spee]. Thus opt 7 ar=— wherec = 3 = 10". The transition temperature is therefore lowered by 11°C. Ah i ~ = nha Note that we neglect the zoro-point energy as a nonessential complication. The probability for a given level n to be occupied is _ TrofkT) St “MwA Le 0 1226 © SoLID STraTE—SoxvTions Each molecule has three modes of vibration; hence 3Nho BY = . (E> = ao At high temperatures, 2 — SNAP — Sa RT, where w is the number of moles and R the gas constant; while at low temperatures E = 3Nhwe"™*", Hence, at high temperatures, =3R ~ 6 calfmole-deg. At low temperatures, c= 3K (BBY e-tenr, er) At high temperatures, the specific heat is in agreement with the empirical law of Dulong and Petit; at low temperatures C — 0, in agreement with observation. However, the temperature dependence of ' for small 7’ pre- dicted here, is not in agreement with experiment. A more sophisticated calculation (Debye model) gives better agreement. 7. The energy due to lattice vibrations is given by vy af —herlorde _, J exp GafkT) —1 where g(«) is the density of states for phonons. At low temperatures the low frequency dependence of (~) is all that is needed to calculate the energy PS _ ak _ 4nktdk _ snot deo (0) do = BS = “ome = omit”? where c is the speed of sound. Thus g() ~ o*, yielding an energy proportional to Thand as eat pray to 73. However of two-dimensional crystals (as graphite is), then g(@) ~ , yielding a T? dependence for specific heat. Thus the 7* dependence indicates that this phase of carbon is composed of two-dimensional crystals. 8. The conductivity is given by o = e(n.u. + myn). where na. Ma (Mu. Un) are the density and mobility of the electrons (holes). Moreover, in a pure semiconductor, my = 7, with 1 j Po . ® J exp(Q ¥ p2m, — kT} +1 The probability p,() for a hole to occupy a state of energy E is equal toSony StaTE—SoLvtions 227 the probability that an electron does not occupy that state; i.e. px(B) = 1 — p(#), where one must take # — —p*/2m,. Therefore, anes p* dp : “RJ exp ((p772m, + H/kTY +1 In the above expressions A is the energy gap (0.1 eV) and £ is the Fermi energy. If (A — £)/kT > 1 and £ > KT, as is the case in this problem, the ahove expressions simplify to 1 (A + p?/2m, — £) = al it exp [ : kT } and wo [orton [lg The equation n, = ny implies peas A (m,\**_ 2 + kT iog (™) é The density of conduction electrons is then ny = ATI#e 37, re a nally as a function of temperature is ¢ = oy"? exp (—A/2K7), and ot) _ (22) exp [A=]. of) \T,, RTT, Converting A to a temperature, one finds A/k = 1160°K. Hence o(T2)/o(T,) = 6.5. 9. In vacuum, the electron velocity is determined by $mvj = 25 eV, while in the crystal, the velocity is increased, according to 4mvi = 25 + $. Thus the quantity ./(25 + ¢)/25 acts as an index of refraction for the electrons. Upon emerging from the crystal, the electrons are diffracted according to Snell's Law, which gives an 6, =, | gin 1607 sin 6, = 4/ 5-4 Sin (60°), and 6, = 26 from the geometry. Diffraction from a set. of crystalline planes with Miller indices (h, , 1) is governed by the von Laue equation, : : sin’@ = ae +B +R = MA.228 Soip StarE—Soxurions However, the appropriate wavelength is not that of the electrons in vacuum. Rather, rical values of the problem, von Lauo’s equation becomes +). Now ae eee 6006) = 1 — 2sin? = 1 — peo Then the restriction cos* , + sin* 6, = 1 gives the condition 2+ 6=sa— ae Since ¢ is positive, we must have 8A > 25, or A > 3. Further, the condition 124? — 25(84 — 25) > 0, or (equivalently), (6A — 25)(24 — 25) > 0, gives rise to the conditions A < 28 or A > 12.5. Thus the smallest. ring occurs for an A which represents reflection from the (200) plane. The cor- responding depth of the potential well is — 17/7 eV. The fact that (100) reflection did not occur (if it did, it would be at a smaller angle) is explained if the crystal is face-centered cubic; then the maximum separation between adjacent planes is half the lattice constant. 10. Conservation of energy implies de = eE-dx. Thus e(t) — e(0) = eE+(x — x). and eft) — e(R(y. 50 and et) {R{t}}. Also Ak— 1 Vjem. BL. (a) Let one of the photons feall it photon (2)} = make an angle @ with respect to the velocity of the electron being annihilated, assuming the positron to one be at rest: (see figure). If the velocity v of the electron = ye a the two photons to be Mel, cach 7! ar the two photons to be antiparalicl, cach photon having a momentum p= mc. For v0, there is a momentum Ap = mv sin #, perpendicular to photon (1), which must be carried by the other photon. Thus to first order in v/e, the photons deviate from collinearity by the angle 3 =AP =” sing. peas One must now estimate the magnitude of v/c to be expected. An electron bound in an atom has a mean kinetic energy of <}mv*) equal to the binding energy, Z*at*mc?/2n* where a = e*/hc ~ 1/137 and n is the principal quan- tum number. Taking Z—1, since we are assuming that the outer shell electrons dominate the annihilation, we obtain Smx = v/e ~ a/n ~ 1/137. An alternative method is to calculate the Fermi velocity, vp, in terms of the density of conduction electrons, with 3mx ~ vz/¢. Since this caleulation of vy is needed later, it will be done here. For a free-electron gas at 7’ = 0°K,230 Sonip StarE—SoLvutions the density of electrons is given by ~ eal = ou : a) As an estimate for vp we choose n = 1/a’ where a the Bohr radius. Then v,/c = (677)'*a = a@ = 1/13 (b) The transition rate in a metal, I’, is given by the product T' =o x flux x phase-space density. The product ( x flux) does not depend on velocity, so I" depends only on the phase space. Since the proba- bility of a given angular separation depends only on the probability of an electron having the appropriate velocity in the direction bisecting the angle between the photon momenta, the direction of Ap makes a convenient choice for the z-axis. Then W(8) depends only on the probability distribu- tion p(v,). Hence we write vt = 04, | o, where v, is in the plane normal io v,. Roie thai we huve wssumed implicitly thei wil messurcmenis ace made in a plane. ‘This 1s mm fact necessary for accurate measurements, as 6 1s small. The phase-space factor is */me?; that is, a is dv, dv.p(v1, 0.) = vido, dgdv. p(v,, $, 0), > fe, e 2 eae angular correlation depends only on v,, we compute wd DA. dv, p(v.) = de, fv, do, ddp(v1, dv). Ch. Gy are counters At T =0°K, p is constant inside the Fermi surface. Therefore p(v,) is pro- portional to the area of a section of the Fermi surface normal to the diree- tion of v,. Thus the angular correlation, which measures p(v,), gives informa tion about the Fermi surface. For an isotropic metal, lv.) = 4 f odo, = 4 J” vdv = Aes — 09. Thus W(8) — const x (Xs — 84), with 8 Fermi distribution is = 1 . PW) = Same? — oh) RET] ET At ordinary temperatures this differs trom the previous distribution only at velocities v= vy. One thus expects the following qualitative be-Soup StarE—SoLvtions 231 havior for W(8) as a function of temperature: T=0%K T#0% we) NX we) BN 5 v,/e we In reality, however, thermal motion of the positron is the dominant factor. 6 42z. Consider the long-wavelength plasma oscillations of the lattice. ‘ihe motion of an individual ion satisfies Mv = ZeE, which, in terms of the current density of the ions j = ZeNv, reada In addition, the motion must satisfy the charge conservation equation V-i+p=0. @) Combining Eqs. (i) and (2), we obiain BAN y-B) + p=0. @) ‘To proceed further, we must represent the screening effect of the electrons in the dielectric constant. Thus, for a plane wave E = E(k, o)e*-* and lattice charge density p = p(k, w)et"**-", we have V-E = (4z/e(k, «))p, which when substituted in Eq. (3) yields @ _ oO an ot = 422 ee eb) eee ee For a heavy ion, we expect the electrons to follow the ion motion adiabati- cally, and thus o* = Q#/e(k, 0). The sound velocity is defined as the limit Thus One must now calculate the dielectric constant e(k, 0). Consider a static external charge density ps(z) = pse** embedded in the electron gas. Then the total charge density is p)/c, and the induced electron charge density is expressed a8 pp = po(l — e)/e. The electrostatic potential satisfying V'V = —4p,/e is then given by _ 4m Y= p85 ®232 SoLip SraTE—SoLvtions In order to solve for e, we need another relation between V and p,. This can be obtained by the Fermi-Thomas method where the density of electrons is taken as ala) = =, [ dp, Gray s° ? expip' jam — Ve) — Bryer ti In the degenerate gas limit Zy > kT, we obtain (a) n(x) [Hy + eV(2))", while in the classical limit, appropriate for a hot plasma, {b} ala} oo emp (ePQaik2}. Expression (b) is not appropriate for electrons in a metal, which behave like a degenerate gas; it was discussed in Problem (8-22) in connection with the Dehye-Hiickel radius. pats) ~ —eBn(a) — — (ele When this equation is combined with Eq. (1), we obtain Bretn ee, = 1+ Finally, then, the sound velocity is given by 2)" = (2m m (Note n= ZN.) Since m/M <1, the sound ay will in general be much less than the Fermi velocity.11 NUCLEAR PHYSICS 1. Cerenkov radiation appears when the velocity of the particle exceeds the phase velocity of light in the material, c/n, when n is the index of refrac- tion. Therefore the 4’s alone will give Cerenkov radiation for those y such that civ, > n > c/v,. The velocity pe(p* + m*c*)-"* of the w and yw are cf 2 and c/[1 + (106/140)?]"” respectively. Therefore the desired range of nis 1414 = /2 >n>[1+(4 1? = 1.26. 2. According to the uncertainty principle, the momentum and position of a particle are not completely well defined: Az-Ap ~ 4. When a meson is emitted by one nucleon and absorbed by another, its position during transit is uncertain by an amount equal to the range of the force. ‘Ihus for an order-of-magnitude estimate, Ac=r and Apxme, where r is the range. Hence r = ijme. 3. Let N = number of protons/cm’; then in terms of A = 1/No and d, = 1/No,, the probability that the antiproton is elastically scattered in the intervals dz, and dz, is ha ygdar,)e-B 2-2 (gd je“ A-™, where 0 <2, x + 7 is forbidden merely by con. servation of parity and angular momentum. 7. Describing the initial spin- state by a spinor, yp = (g), the most generai expression for the probability (summed over the final spin projections of Band C) that a given particle is emitted in the direction A is yp May. M is a (2 x 2) Hermitian matrix because the probability must. be real. Since any (2 x 2) matrix may be expanded in terms of the unit matrix and the o matrices, M must be of the form M = A + Bova, where A and B are real constants, independent of f so as to guarantee rotational invariance. If the initial state were polarized along the z-axis, then P(i) = A + B cos 6. However, since the decay proceeds via electromagnetic or strong inter- actions, parity is conserved and P(A) must be even under parity. Thus B=0 since p*a-fp is odd under parity (A + —A and ¢ — @; hence @-ii —> —-i), and the decay is isotropic. In weak decays, where parity is not conserved, the coefficient B will not vanish, and asymmetry in the decay will be observed. A typical example is the weak decay A —+ px-. 8. ‘Tho relative parity ofthe (x”d) systom is P(x” d) — (—1)"«(—1)"*(—1}' has positive parity; hence Pix d) = P(x ). The totai anguiar momentum of the initial state is J(a- d) = 8, +8, +1, from which J(x~ d) = Since the reaction conserves angular momentum, J(nn) = 1. The resultant spin of the two neutrons may be S — 0, or 1, and when senbined with the the only possibilities are 98, °P,, "Pi, and "Dy. In addition the two neutrons, being fermions, must be in a state antisym- metric under interchange of the neutrons. The spin states (S = 0) and (S =1) are antisymmetric and symmetric respectively under exchange. Similarly S and D states are symmetric while P states are antisymmetri Thus, of the above possibilities only *P, is allowed on the basis of Fermi statistics. ‘The parity of this final state is (—1), since the product of the intrinsic parities of two identical particles is even, The parity of the allowed236 © NuciEar Puysics—Souvtions final state *P, is odd, and since parity is conserved in this reaction (as in all strong interaction processes), one concludes that the parity of the pion is negative, ic. P(e) = —1. 9. The isospin of A is 0, while the px_ and nz” states are jinear combinations of states with definite isospin; that is, they are not themselves eigenstates of the operator I?, The nucleons have isospin 4, while the pions have isospin 1. The pion-nucleon system may have those values of isospin resulting from isospin (or angular momentum): i @ $= $ tule, the A decays only to the (I = $)-state. We may expand |. —1) = alpa-> + blazx’y, —® = blpx-) —a|nz, where a and } are the amplitudes for the px~ and nz® states in the (I = 3)- combination. ‘The corfciente in the expansion of the (I = ¥).state have mined, then give the ratio 4 —Rate(A—pr-)_ [bP Rate(A nz) — |al? ‘The coefficients a and 6 depend not on the detailed properties of the ele- mentary particles, but only on the isospin properties. We emphasize this by introducing the notation Ipx-> = 14, 11, 1» and |nz> =|4, — 11,0. In this notation, 1% —D=4lb 11, 1D + blk, —D K1L O. Now the coefficients, depending only on the transformation properties of isospin, cannot depend on whether the basis states are “elementary” composite. For example, the coefficients cannot depend on whether we describe a stale by writing |1,—1> or |4,—4>]4,—H. We therefore write each state with J = 1 as the appropriate linear combination of (I states: IL-1 =], DHL —P = A280), 1,0) = Fell DH PD +s PIE DI = Fz laeee) | ee)aa)).‘Nucigark Puysics—SoLvuTions 237 In this decomposition, 13, 4) = aaa) +P eayta@ee) | e@)a)), where the index (1), (2), or (3) labels the states. We recall, at this point, the result of Problem (6-5), that when n particles of spin 4 combine to make a total spin n/2, then all (2n + 1) states are symmetric under interchange of any of the spin-} particles. The same result applies to addition of states with isospin $. Now the |$, —})-state written above is already symmetric under the interchange of particles (2, 3). The requirement that it be symmetric under the interchanges (1, 2) or (1, 3), imposes the condition a = b/,/2 . Therefore A t/a? — 2. ‘The wansitions x-p— K°A, x‘n— K°A go via strong interactions; therefore they conserve isospin. Since the x~p and xn systems may have I=}, or 1 =$, the K° must have I = } or I = $. In the former case, the ratio is expressed as p_(ep— KA) _ 9 (xn — KA’) while in the latter case, B = $. 10. For low excitations. one would expect. from the shape of V(r), that the levels of the three.dimensional harmonic oscillator approximate those of V(r). The onergy levels of a threo-dimensional harmonic oscillator are given by E = ha(m, +m, + m, +4). The ground state corresponding to ‘m, = m, = m, = is spherically symmetric and, therefore, has orbital angular momentum L=0. The first excited state corresponding to m, + m, + m, = 1 is threefold-degenerate corresponding to the (2L + 1) dogeneracy of an ZL —1 atate. One also aces that the pin orbit coupling gives a lower energy to states with S and L parallel; therefore the Py: is lower than the P,q. ‘The level scheme of the first few states is thus Energy Degeneracy Pin 2 Pas 4 Sy. 2 According to the model given, the neutrons and protons independently fill these levels. (a) The two neutrons completely fill the Siz level, combining to form a J, = 0. The proton in the $1 has a J, = 3, and the total spin and parity of HP ie J” = 4¢238 NucieaR Puysics—SoLvtions (b) The 8,2 is completely filled with protons and neutrons. In the Ps there is one proton and two neutrons. Since the neutron wave function must be antisymmetric, the only allowed J, values are J, — 2 or 0. The angular momentum of the protons is J, = $ and the total spin and parity is J* = 4°, $, ¥, ¥-. If one also uses the fact that the neutrons in the same level pair off to form J,, = 0 in the ground state, then one has J” = $~ for the ground state. (c) The six neutrons completely fill the S12 and Pyjz levels while the five protons fill the 8,,; evel and have a “hole” in the Py, ievei. Thus the neu- tron angular momentum is zero, while that of the protons is }. The angular momentum and parity is J? = 3°. (a) The eight neutrons completely fill the S12, Pi and P,,; levels, while there is only one proton in the P,/,. Thus J? = 11. The maximum energy of the total 8-spectrum may be measured with a Geiger counter and an energy spectrometer consisting of varying thickness of aluminum shiclding (the range-energy relation being known). This deter- nines the presence of 67 with maaimum euergy 1.54 MeV. The preseuve of two y's may be detected with a scintillation counter and the energies mea- sured. The maximum energy of 8, may be determined as was that of Bx, but by counting only those f’s which are in coincidence with the 1.3-MeV ‘y-ray, thus eliminating the 8, background. To complete the scheme it should be shown by the abuve procedure that the y of 2.1 MeV is in coincidence with B:. 12. From the fact that there are two 8 groups, the level scheme follows. Type of Photo. electron ; electron Transition _ejected ee A IL K =| “ev RB i L } o I K 11d 0.261 Mev D I L 2s====- = ig di K 13. Consider the transition in the rest frame of the c*e™ pair. By definition, there is no net momentum in this frame; hence the photon has no momentum in this frame. But, for a photon, energy and momentum are proportional. Thus the photon has neither energy nor momentum, and hence docs not exist. Thus all transitions to states having massive particles are forbidden.NucLEar Puysics—SoLvtions 239 14. The x* has the same charge as the charge (Ze) of the nucleus. Thus upon disintegration the x* is given additional kinetic energy (of magni- tude ~ Ze*/R) due to Coulomb repulsion. One would therefore expect few low-energy * mesons. 15. From charge independence of nuclear forees, the only difference in mass between ,,Si*” and ,,A" is due to Coulomb energies. Taking a nucleus to be a homogeneously charged sphere of radius R and charge Ze the Coulomb Therefore the expected mass difference between these two mirror nuclei is eh aay — coy — 1. Bplay’ — (13)") = Assuming that R = 1A", this energy becomes 27e*/5ro. Interpreting the positron energy to include the rest mass, we find rp ~ (24)r., where 7, is the classical electron radius, r, = e/me = 2.8 x 10- cm. 16. = 0'* +n+T, + 3.72 MeV, and F = O" 4 1.72 MeV. (Recoil corrections to the energies are neglected.) U'°* may be an excited state of O'*, and the problem is to find the kinetic energy 7’, of the neutron. ‘These equations, together with Eqs. (1) through (3), yield T, = 1.99 MeV — (O'** — O'*). ie gig > transition to the O'* ground state, and neutrons are emitted with only one energy, 1.99 MeV. ees owe NP. ---- ove oe 5.08 MeV | ont” Tummy (b) and (c) The data given allow the construction of the above level diagram. (The dashed levels are the levels expected on the basis of charge indepen. denee.) Tho fact that the energies of 0!" and F"” are go close ia due to charge240 Nuctear Puysics—So.vtions independence, O" and F" forming an isospin doublet. The energy difference is due to Coulomb repulsion, which associates a higher energy with higher Z. As @ partner to O'*, one expects an excited state of F'’ with an energy F'* — O'"* = 1.72 MeV. The N'" has a T, = —3; thus at least three more states are expected to form a 7’ = 3 isospin multiplet. These three states are another excited state of 0”, F"", and the ground state of Ne". 17. To calculate the energy of the alpha from element Y, it will be assumed that the lifetime of au alpha cuibier and the cuergy uf the devay are related by the Gamow theory in the limit of large Coulomb barrier; i.e. 1/7 = where @ and f are taken to be constants. Actually, 6 is proportional to Z, but we are given no information in this problem about Z, so it will be as- sumed that the fractional change in Z from one emitter to another is small form for the p Jem is to write log (r) = A -- BE-"®. Then the data given for X and Z allow A and B to be determined from 1o=4 +B) "* and 3=A4+ B(0)', () A+ BE"#=0. (2) From Eqs. (1) and (2), one finds Z = 14.8 MeV. ‘The hinding energy of a nentron is eqnal ta the center-of-mass kinetic energy of the n, N'“ system at threshold, i.e. B = (14) 10.6 MeV = 10 MeV. The N“ nucleus is assumed to is at rest in the laboratory. The center-of- T = (14) 14.8 MeV = 11.5 MeV. Thus the reaction N'(a, an)N" is energetically possible. It should be re- marked that there is a Coulomb barrier of the order (4 x 7)e!/R ~ 14 MeV, and the rate is inhibited by this barrier, but nevertheless the reaction is possible, since in quantum mechanics particles may tunnel through such barriers. 18. The reaction cross section for the neutron is simply o, = za*, which is the cross-sectional area of the target. If a neutron in the beam is within this area, it will strike the target nuclous and thus cause a reaction. However, @ proton which is initially far removed from the target and within this cross- sectional area may not strike the nucleus because of the Coulomb repulsion. Thus the reaction cross section will be smaller than za’. To calculate this reaction cross section, consult the figure. Here 6 is the largest impact para-Nuctgsr Puysics—So.vtions 241 meter possible which will allow the proton to strike the stationary target. For this 6 the proton will strike the nucleus tangential to the surface, and thus the flight path is perpendicular to the radius drawn to this point. If Z ia the total kinetic energy of the protons at infinity, then conservation of energy requires that Ze @ Ka = fme, = B at the point a. In addition, conservation of angular momentum about the point O re- quires that the initial angular momentum L = pb = b »/2mE be the same as the angular momentum when the particle is at a, that is, L = a ./2mK,. ‘Thus a J/2mK, = b /2mE. This equation, together with the energy conservation equation, allows one to solve for 6, Thus = 2 = wat(y — 26 baa /i— and xb? = xa’ (1 an): Finally then (1.22 me = py 2, a, _ | ak @ = | 0 for <2. 19. The grain density increases as the electron slows down; this is not so for the nucieus of charge Z = 11, because the nucieus begins to pick up electrons as it slows; these screen the nuclear charge, diminishing the ioniza- tion rate. EA Zerlet. Therefore the foree equation in the nuclear rest frame is dp _gegvxr dt er242 Nucigar Puysics—SoLurions For small-angle scattering, the right side of x Eq. (1) is computed by taking (v,0,0) and (vt, a, 0). do, Zeg__av dt ‘me (a? + vty?” doe. dry. 9 a dt Integrating, subject to v.( 0, we obtain __ Beg vt 1 = — 281 + oa: _ 2Zeg mca —24eg ‘moav This provides @ unique relationship between the impact parameter a and the scattering angie 0. Now if f is the incident flux of particles, then conservation of particles requires that f-2nada =~; 10 on dcos 6). The left side represents the number of particles incident per unit: time on an annulus of radius a and width da, while the second is the scattering rate into solid angle dQ. = —2nd (cos 6). We solve for the scattering cross sec- tion, to obtain doa a |da| (2)'2, dQ ~ sin wall 6 |d6| ~ = “ES (Rutherford). les lose energy primarily to the electrons, due tu the small aged pe electronic mass, and the same is true for monopoles. The momentum transfer to an electron accompanying » small-angle scattering is 2eg/ca, and the corresponding energy transfer is (Ap)*/2m = eg|meay. The number of ol For an adequate discussion of the determination of a... ANG Gy, the reader should consult the textbooks in which energy loss by charged par- ticles is treated.Nuctzar Puysics—SoLutions 243 21. Consider the s-wave Schrédinger equation «” + (k* — 2mV) u = 0, where the wave function yp = u/r. If V is a short-range potential, then in the region outside the potential, the solution is sin (kr + 8) where 8 = 0 if V =0. The scattering amplitude is given by f = (e*® — 1)/2ik; and the differential cross section, do/dQ2 = sin*2/k*. ‘To study the effect of a nearby bound state on the scattering, write the solution w in the form lke” 21 with u taken real, that is o(k)* = 6(—K*). When k = in/2mB = iy. cor- responding to a hound state of energy, — R, we know that $(—iy) must be zero, in order that a solution u —> e-”, as r —+ co, exist. Taylor-expanding (—&) about this point, we have (—k) ~ ia(—k + iy), where a is real. [This follows from $(k) = *(—k*).] Thus the phase shift is given by —Imgh)_ _ ik tend = Beg) ~ —y . which, for small k, reduces to 3 = —k/y. Defining a scattering length, a, by 8 = —ka, we find a = I/y. The total cross section is 22, From the preceding problem, Eee o as 0. Here _2mErouna =5 x 10" cm” hence o = 2.5 x 10°* cm* = 2500 barns. In this solution the correction for the range has been omitted, since rey < 1. 23. One expects only s-wave scattering to be important at this energy; therefore int Taher with og =4esin'8 From the data given, Nt = 5 x 10* atoms/em*; thus a = In 2/Nt = 14 barns. Also kt = 2m = 5 x 10* cm-?, Finally, then, sin 8 = £0.23. 24. Lead is a doubly-magic nucleus; therefore one expects the eross section for inelastic scattering to be small. Assuming only elastic scattering and s-wave dominance, the cross section is ¢ = 4xR?, estimating R= (14 x 10°? em)"; with A = 208, we obtain o ~ 8.6 x 10° em*, Tho244 Nuctgar Puysics—SoLvrions number of lead atoms/cm’ is given by nome where N,=6.02% 10% and p= 11 gmjem’. I= I, exp (—not) = I, exp (—0.55) = 0.58 Ip. Hence 42% of the incident neutrons are scattered out of the beam. In this treatment the effects of multiple scatterings have been neglected. 25. Define the following quantities: area being irradiated; $ = flux of particles = i/4e, where e is the charge of the deuteron; total number of Mn‘* present at time ¢; (log 2)/T1,2. The initial number of Mn* nuclei with which the deuteron beam may interact within a range R is N.s = RA/M. where M is the mass of Mn**. Note that duction of Mn are aN — go Ny — Noo, GNss_ dN s6 a dt The solution to these equations satisfying N..(0) = RA/M, and N,o(0) is Nast) — HAD py-got gy, Nel) — aH — Gay” Numerically, 3 x 10'*/sec, x 10*/om?, “y= (log 2)/(2.6 hr) = 7.4 x 10-* see. Note that for reasonable areas, ic. A > 1 cm, one has X>> do and that for t= Bhr, gat = 0. This means physically that for this relatively short- time de time the doplotion of ¥ the doplotion of 3 Nea(5.2 hr) = " AAge( = t) =3.7 x 10", 26. Consider first the scattering of neutrons on atomic hydrogen. The scat- iering ampiiiude way be written (considered in the limit & — 0) Aa US + o,-0,)Nucigar Paysics—SoLvtions 245 Here a, and a, are the scattering lengths for the triplet and singlet states of the n-p system. Note that in the triplet (singlet) state ¢,-0, = 1 (—3). The differential cross section for unpolarized neutrons on uncorrelated hydrogen nuclei is thus 4: o= FEAL SA liye = where the sum is over the initial and final spin states of the neutron only. Tsing the properties Tr {a,(n) = 0 and Tr {a,(n) a,(n)} = 2 Bij, one finda © _ (By +4) , (a — a)? oy-0, 16 16 3a, — ai)* _ (Sai + a) gfe eles ct 2 ceecee: x Tr{AA‘}, _ (Ba + 16 ‘Thus a measurement of the cross section on a single hydrogen nucleus does not determine the relative sign of a, to a,. However, the scattering from ‘molecular hydrogen is coherent, since the neutron wavelength as k —> 0 is very much greater than the nuclear separation in H;. Then the total coherent- scattering amplitude from the two protons is (neglecting recoil effects) (aban (or aenr(ort 02) a) A, The total cross section is given by o 1 + _ (3a, + 4,)* (a, — a,)*8* £- pms, = Satar 4 (ays, where § is the total nuclear spin of the two protons, Thus, by measuring the cross section for ortho- ($ = 1) and para- (S$ = 0) hydrogen, the relative sign of a, to a, may be determined, since these cross sections depend on the terms (3a; + a,)* and (a, — a,)* respectively; these are sensitive to the reiative sign. Actualiy Eq. (i) for A, is not quite correct, since it neglects an overall factor due to a reduced-mass effect arising from the fact that the mass of the H, molecule is twice that of a single proton. However, this overall factor cancels when we consider the ratio Gorany yy 24 — a)? ara " Ga Fa,)?” which, together with a knowledge of aj and aj obtainable by fitting the low-energy total cross section on single protons, 0 = lee amy + Pe ap)” in nufficient to determine the relative sign of a, to ay.246 ‘Nuctgar Puysics—SoLvtions 27. Let the original reaction be given as A+B—C+4+D, @ and the inverse reaction as C+D—ATB. (uy) Omitting constant factors, we have do _ y, |Matrix clement |*5°(p4 + Pp — po — Po) d*ped* Py dO. shite Tncident flux ‘We shaij assume thai the mairix eiemenis for reactions I and Ii are invariant under the time reversal operation, i.e. the operation which flips spins and reverses the directions of momenta. Hence the amplitude for the process (Par Pa» Sa 88) —> (Pe> Pos Sc, 8p) is equal to the amplitude for the process (—Pe, —Po» 8c, —8p) —> (—Par — Pa —Se, —8p)- In the cross-section measurement, the spins are not detected. On the assump- tion that. initial heams are unpolarized, the spin snm includes a snm over the Sna! epi d 1 hho ini ino, But & taken also; hence every s taken in the sum, —s BME = 2 IME. fh. Because uf this, the ratio of the cruss sevtiuus dues aut depend un thie details of the interaction. For the original reaction the incident flux is given by (palE4 + Pal»), While for the inverse reaction it is (po/Ec + Polo). The phase space, for the original reaction, is proportional to prahf = ——_\bi___, GB (peiiie) + (Poin) where 1 =|pel= Py =Ipel =|Pol- Similarly, Pp} = |pul = [pal For the inverse reaction the phase space is (pt eal + PalEo Because of the average-over-initial spins, the spin degeneracy of each final state must be included. Then doa _ (v))" se +1)(2sn +1) _ 3p doy|d (PAP 284+ YRse+ 1) € Pe 34 — mi, — mi) — anim). ~4 E'(E® — 4m)NucigeaR Prysics—SoLurions 247 28. We shall use units 4 = c = 1, and treat the scatterer as stationary. The Hamiltonian of the system (to lowest order in v) is H = —u-B where B = —v x E. The magnetic moment of the neutron is B= oo, where @., dy, a, are the Pauli matrices. The Hamiltonian may be conven- H = —po9-(E x p)/m. The scattering amplitude is A= (lH liy = Pa. | axbey(+) x (iV Wi where on= and |%,> are the initial and final spin states. Finally A reduces to idwZe pu. walk ele Ui ie ‘The definition of differential cross section is do/dQ. = a/@ where $ = flux = i,m and w is the transition rate into dQ: @ = 2x| A p(#). The density of states is 1 ak, mk, p(E) = Gaya = ex : Thus $m eee} lose der x BOE. ki Summing over final spin states and using |k, — k,|? = 2K? (1 — eos 4), do _ (1 geyt208" (6/2) Fey = (We Ze) ey 29. The interaction Hamiltonian consists of a nuclear and a magnetic term. ‘The nuclear scattering can be described by a constant scattering length, a. For very slow neutrons, where s-wave scattering dominates, this is a good approximation. However, the magnetic scattering will be calculated in the Born approximation. Thus the total scattering amplitude, with & = 1, is given by fq) a— (#) f arettaiere™, where q-= ko — ky,248 Nucigar Puysics—SoLvtions and H(t) is the magnetic interaction (magnetic moment of neutron = 4): Hale) = —{ @yoe + ye x [HI]. = ali + 2D yt x (x a) where y— faregteyter and tty — Cf. The initial and final spin states of the neutron are |i) and | f) respectively. Since 9(0) = 1 and the range of g being considered is such that |q| x (Bohr radius) < 1, the approximation g(q) ~ 1 will be used. The spin-flip amplitude then becomes Buel x lin Wh 2 = 32 2 a, = [anise = 2am’. The no-spin-flip amplitude is ix)? fa =a aret xa), and the total no-spin-flip cross section is Om = 4a? + Saee + 30. The matrix element for the transition is = = fig Seetiaz M = OMHIS> = OS arpel OV zap & * O-utous| The modulus-squared of this matrix element, summed over the spin popula- tion of the A in the final state, is aeth® > utio-k Rar ae \ BanV CMs PME) otea, WOR % ehuauitere x chu. When using the completeness relation) u,uz = 1, which may be verified directly:Nuctgar Prysics—SoLvrions 249 and the property (o-A)* = At, valid for any vector A for which A x A = 0, the sum reduces to 2 _gteth? (ke x €)? ods, AE 2aV (Me + My’ independent of the initial =° polarization. Because of the transversality condition k-¢ = 0, one has (k x €)* = k? = w*/c?. The phase-space factor is p(B) = (V/(2nh)*\(d"p/dE). Because the decay is isotropic, one may per- form the trivial angular integration, and write OR) = sas aR Conservation of enorgy and momentum require te — OB = Me | IMs thus MY do__do (M& + MR) dE ~ d(Mx*) 2M The transition rate is given hy the Golden Rule, r= oe) S [ae 2, where the sum is over the A polarization and the two polarization states of the phoion. Combining the factors above, one finds dere e(, a (es M3)(My, — M3)*c* 7 4 \aerhe, A MR(Mz + M,)* This gives r ~ 3 x 10-'* sec, upon using & = 6.6 x 10-*? MeV/sec. Note that we have treated the A nonrelativistically. by using Pauli spinors: this, however, is a very good approximation because the kinetic energy of the A io only 2.5 MoV. 31. In nonrelativistic Fermi theory of 8-decay, the weak interaction matrix element forn — p +e" +3 is Hy = 9 f AV pena h2enula). If the nucleons are in nuclei, then the appropriate generalization is Hy = Ef dix oo dx abtle, + aallry + indy «++ ade ele) =9 fae Mayr, where (7, -|- i7,)/2 converts a neutron into a proton (i.e. isotopic spin raising operator) and yx, and 4, are the initial and final nuclear wave functions.250 Nucugar Paystcs—Sovrions An allowed transition is one for which M= f dM (2x) 0. rovess Z—+(Z— 1) betty, _ 2) dN aD, — lanl where dN/dW is the deusity uf fiusl states ‘p.d°p.V? : an = eae endee?= 1a FS + pe| : (NR) Therefore, neglecting nuclear recoil, we have when the wave functions of the electron and neutrinos are plane waves, that is y = e'**/,/V. The total rate is then given by Py ___ 89° |My Zm) ~ (ex) (L0a)en m* 4g ME otin (Ww edo TN where p, is the maximum momentum of the emitted electrons. For the process Z + e~ — (Z—1) + and the electron is captured from a 18 hydrogen-like state; the matrix element is given by where (0) = (Z*/za*)" is the value of the clectron wave function at the nucleus. (Here a = f#/me? is the Bohr radius.) The density of final states is 4x pidp,V dx piV (xh) w= (@xhyc’ where p, is the momentum of the emitted neutrino; p, is found from the of (a) positron eumissivn, Pal2me + pow + mee? ; aud (b) K-capture, E,+me— B= EH; +pe, where £, and 2, are the initial and final nuclear energies and Bis the binding energy of the electron in a 18 level. Thus B— 5, = 2 Be 2 pe = Imet — B+ pe + PE x Ime.Nucigar Puysics—SoLuTions 261 The transition rate is given by Te = 22 Hata = PE)L ME gO) Pax] oxny'et Hence Py goog ye(*t) (me! Pio 42002 3) (#2) , which, in terms of W = pi/2m, may be written Es = a20n01()" r (ze) 32, The transition rate is given by the Golden Rule as r=) mp. ‘The dynamics is contained in M; for the purposes of this problem it is a constant. The final factor, dn/dE, is the density of states in phase space. Momentum need not be conserved among the emitted pat les; the infinite f th beorb energy. Henve E = (pi + mi)!? + (pi + mi”. The differential volume in phase space is d'n = 16x" pi dp, pi dp./h*. aN However, to compute dn/d£, it is more convenient to choose, as variables, not p, and p;, but H and p,. Holding p, fixed, we have from which PM = l6x'ptp/ BE mA = Nort pip LE — Vee me o din _ 16x*pt 2 2] 59 — Gea — vot + mi) [(e — vot + ml) — ma)". This exhibits the energy dependence of the decay rate. In the case of 8. decay (og. n—+p tem +i), the decaying nn finitely heavy compared to the leptons. Also, ay he regarded as i ‘my = m(9) Then252 Nucigar Paysics—SoLvurions This suggests the introduction (due to Kurie) of the * variable 1 din \" m=0 =(447 )" «eR, ( pidE dp, ) met / Ca : . ‘ — E. This gives a straight-line plot as shown in the accom- . panying diagram. One can see the effect of a small neutrino mass m, from (1): when — pt mi is small, dn/dZ becomes smaller for m, # 0. This would eause a dip near the endpoint. of the spectrnm, indicated with dotted lines in the figure. Before any inferences could be made from such a plot con- cerning possible nonzero rest mass of the neutrino, corrections for Coulomb distortion must be made. ‘These corrections apply mainly to the low-energy end of the spectrum, however. 83. (a) The number of states for fixed p, is given by dn = V/(2xh)? d*p,,. In order to calculate dnjdw, express d*p,, in terms of |p,,| — py, 008 8 and w, where w = (Ipe +1 P| + 1Pu De 0080 = perPrilPe Pr and py, is determined by the momentum conservation equation Pa + Pn + Pe = 0. In this way we obtain dn im — amv sin ao pi, he ag _ (Mey 2{2( Mc — p.)( pe cos — pe) + Me’ = 2xV sind do MLD (Mc — 2p,) (POE poe 2086 — pat Me MP et et ‘} where the total energy ae W = Met (M is the muon mass). 8, we obtal spdp 1 — Fete (Oxy 6c The number of electrons emitted in the momentum interval between p and p+dp is thus dn, = Ap(3W? — 6Wp + 2p*)dp, with a maximum attainable p = Me/2. % (3M*ct — 6Mep + 2p). An alternative solution, due to Fermi, and involving less algebra, is to note that n(W) is proportional to f d°p,, with p,, +p, + P.e=0 and [pul +1 Pu! + [Pel = Wie. Hence n(W) is proportional to the volume of the ellipsoid in the figure above. This ellipsoid has principal axesNuctgar Puystcs—SoLvTions 253, = (W — p)/2,b=¢ = 4/W? — 2Wp,. The volume is Rabe = E(w — p(w — 2Wp) ww — p.)\W* — 2Wp,)] = (3W* — 6Wp. + 2p%). (b) Integrating the above expression for I' over the allowed range of Py < pe < W/2} we vbiain gt L _ 1Mteg 7 7680x*HT This leads to g ~ 3 x 10-* erg = cm’. 34. Denvie ihe heavy pariicie of mass 37 vy p and ihe iighier one of mass ‘m by e. To obtain the threshold frequency, wo, for the reaction e + y — pt, the center-of-mass energy of the photon and electron must be M."Therefore, from the Tarentz, invariance af (p, + p,)*=—M", one obtains (Pere ‘The hey tv finding jue-reversal invariance. Denuie by S,,.s,.«(# — © + 7) the matrix element for the decay of a y at rest (pola- ization s,) into an electron and photon of polarization s, and e respectively (Fig. 1). Also denote by S,,.,,,.(¢ + y — 2) the matrix element for the inverse process (Fig. 2). ‘Time-reversal invariance requires fen | = ; ZS, nel e+ yr = & WSruse(e +y—> mw P. Ov, Ova, The lifetime of a u is given by = any [GD p pg yet aif be fr 7 3, + Tied Qx) dE sad Fig Fig. 2 where p is the momentum of one of the final particles, and E — M is the total energy, 2 = p + /p* + m*. Thus Hfesce== = (“a me (ae +m wa) 7 x(Qs,+1)\ 20 2M For the inverse process in the rest frame of the 1, the unpolarized rate is given by J [Sssnele— © + 7) Pe aps 222k! * 2(28. + 1) dee where dN/d# is the density of occupied initial states in the center-of-mass frame and E is the center-of-mass energy. Thus from and Bos’ + Sm? +07, and do'/do = /(1 — v)/(1 -F 0), where ow! is254 Nucigar Puysics—So.vuTions the frequency in the center-of-mass frame, one finds — (Mt + m) Uo) | aE 2M a for dV occurs because the density of photons transforms like the fourth com: ponent of a four-vector on going to the center-of-mass frame (just like the Doppler shift in frequency). This is because the total number of photons is invariant, nnder Lorentz. transfarmations. Now the differential element of in three-space, and the same is true in four-space of the appropriate unit of volume d‘z = d°x dt. Thus dN 1 dz zy ap are showing that the photon density transforms like the time-like component of a four-veetor. Note that the same argument may be applied to obtain the transformation law: Combining Eqs. = gt (2% +0) Deve) Ty = an! enor = 7 @ 7 However, this is the rate in the center-of-mass frame, and must be corrected for time dilation to obtain the rate in the lab. Hence r,=f 87"(25, + 1)M*m Uo) 1, = 7 ~ Ba FOF = my Fm) 7 For s, — se = 4, we have (28, | 1)/(2se | 1) = 1; when 9, = $, % then (25, + Uii2s_. +1) = 2. When this problem was proposed, the problem was of interest because of the conceivable decay 4 — e + +. This decay is now known not to occur. The problem is still of interest today, however, in connection with the photo- production of high-spin (>>3/2) baryonic resonances. 35. Let n be the neutron density and © the total volume of the pile. Then in time dt there will be n(o di/A) collisions and hence a net increase in the number of neutrons by n Qv(k — 1)/AN per unit time due to collisions. However. in this time nvA atla neutrons strike the surface and escape of change of the neutron density is dn dn — no Ko} 4 > AAN/Ak — 1).NuctgaR Puysics—SoLvutions 255, For a cube 0 = L* and A = 6L’. Therefore we need 1 > &X 10 x 100 = q Tx O0E = 275 x 10tem 36. (a) Consider a neutron scattered at an angle 6 in the center-of-mass frame (nonrelativistic). The originally stationary proton has a recoil energy s(t cos? 8 + a — cos | of a neutron of energy £ with a stationary proton as Ep. Therefore after n collisions (lay = Fad, H..y = Gee? Hence (B,) = ()" Be To caloulate p,(B) we first note that in a single collision a neutron of energy E is scattered isotropically in the center-of-mass frame, i.o. the probability thai a neuiron is scaiiered inio a solid angie d& is dG/4x ; hence the proba- bility p(#)d# that a scattered neutron has an energy between # and (E + dB) is dB/E', where 0 < B < E’. This may be proved by differentia ting the expression Hf = Ri(1 — cos §)/2. After n collisions the probability of obtaining a neutron of energy # is found from _dE (dB, dE, dB, eae — FSS FE’ where the integration is over the region E,< £,..< +: <£,< Ey. However the integral is a symmetric function of Hy-1, En-2,--., E the limits of each integration may be taken as E, < E < E, after di by (n — 1)!, which is the number of ways the sequence E,1, 2-2, may be arranged. Thus [r-mail dB (dB on) — aa ee (b) In the steady state the number of neutrons per unit energy is inde- pendent of time, thus the number entering the interval d# per unit time must equal the number leaving: NalBa\voy 2 +N {™ pu \oe ae Bt), F Nf(E)(o.(£) + o.(£) WE. (09) In this expression W is the total density of scatterers; n(Z,) is the density of neutrons at energy #; f(#)di ia the flux of neutrons that are in the in- terval (E, E + dE). The problem is to calculate this flux per unit energy, given that neutrons of energy E. are produced at tho rato g/cm! sec.256 NUCLEAR PHysics—SoLvTIons Differentiating Eq. (1) with respect to EB, we obtain the differential equation f (eer (E) _ Ets myo ny) +O) where ar=a,+0, (2) 4 = od" AE, = o, dE" 108) = pox [— J SF] = pls ome | + nee) The constant of integration ia found by substituting in Eq. ( thet 4 — nlf) 2, of EYE, thet 4 — n{B,) vy of By)/ Eu. for A to q. This is done by the requirement that the number of neutrons produced per unit time, of energy Eo, equal the number that leave, i.e. g — Nn(B,) v[o,(Eo) + oa(Es)]- Therefore A = q o,(Ey)/EyNox(E,). |, where it. ia 37. Seance ee 7) Fe te Assume that each proton acts independentiy, and that the nucieus is point- like; then in calculating the ground-state muon wave function, we find Ir. = Zr. A factor of Z* comes from the normalization of the muon wave function, and one factor of Z for the total number of protons with which the muon may intcract. Then Ur_ = 1fto + Z/t, while 7, = 7 since the j:* is not absorbed. In zine, 7, = 2.1 x 10-* sec, r. = 2.45 x 10°'* sec. A more accurate ealonlation wonld inelnde the finita size of the nnclens hy using a Conlamh 38. In the limit uf zeru range, the deuteron wave function ix e-7"/7, where of = mB (B is the binding energy of the deuteron and m is the neutron mass). In the rest frame of the deuteron this wave function gives a distribu- tion function for the neutron momentum proportional to d°p/(p* + ¥)?. For y < po, where po — [4m(200 MeV)]"” is the deuteron momentum, the angle which the neutron makes with the original beam direction is small. If p; is the transverse momentum which the neutron possesses when the proton is removed, then this angle is given by 6 ~ p,/p0. From the distribution function for the neutron momentum, d*p/(p* + 9*)?, the distribution in transverse momentum is found by integration over the longitudinal momentum. In this manner the distribution in transverse momentum is found to be p, dp,/(pi + *)*”, or, in terms of the angle 6, the normalized distribution is (y/po) 0d 6/(@? + y7/p3)*. In the small angle approximation used here, 6 = sin @ and thus 6d 6 = dQ/2z, where dQ isNucLgar Puysics—SoLvutions 257 the increment of solid angle. Finally the probability that a neutron goes into the solid angle dQ is ¥ dQ aP= : (oxo FoF] pay? From the formula it is seen that neutrons are confined to a cone in the for- ward direction of opening angle 8 ~ y/po. In this problem y/p, ~ z's. 39. In Problem (2-35). we studied the free oscillations of a liquid. dro th In this prob the Coulomb energy is obtained from the gravitational potential energy of Solution (2-35) by the substitution —Gp}— (3Z¢/4R°)*. In addition, the surface tension 7’ gives rise to a potential energy, which must be esti- mated. When the drop undergoes a deformation (A), in the notation of Solution (2-85), the arca changes by a small amount dA, and the potential energy of the deformation is u = [ TA. We proceed to calculate the change in area resulting from a small de- formation. We begin by noting the relation between solid angle and area, dA = r*dQ|(n-f). The unit normal n is obtained from the equation of the surface, f(r, 8, 6) = const. Here it is most convenient to use sphericai coordinates. Then, using the fact that @ and @ may be regarded as the independent variables, with r=1(6, $), we write -fYoat 1 fg or 7 6 rain and Hence and U= f ran R + May hl +4 (2) + ta) Because (27/20)? = (2h/08)? in already second-order in h, we may take (2r/20)"/r*—(ah/26)"/R*, and perform a Taylor expansion of the xquare258 Nuctgar Puysics—So.vtions root. Then, to second order in h, A=4,+ 42+ 4s, where 4=fanwel +2 { (2 / 2h ia) {! + om[(20) + sinrolag) and Ay + As = [ dQ QRh + #4). Nuclei may be regarded as incompressible. Hence a0 Ra = —f mao, as in Solution 2-85. Therefore A,+4A,= —faow Collecting terms, we have 4 — [40 (ahy 1 (ah)? ons J-2 (N26) | sinta\a8) We integrate by parts to obtain 200 auth — 2n4), ” | J 6A where @reeeeln=03/=,-=5 (0! 1 oa aU = Shaolin 959) + aro oF This operator has the familiar property L? ¥;,m = ii +1) Yim. Then, since h= RY Aum Yim; we have oe 0 APE M+) — Mint — TE UH Yt Ayan g the Hamiltonian, we use the (2-36) to obtain H= 3 {2 dtadin + [BEC ne + 2) — 2D attain} energy a5 gi The frequency of the mode A, » is thus + _ dnl! — 4 ___3(Ze)* ot = Sam, (2+ 2) — sere ne: Sinee, for | = 2, m ranges from 2 to | 2, taking integral values, the lowest miude hus five degrees uf freoduun, The energy fur the t= 2 mude ix therefore E = hax (m +--+ +5 +$) = has (n + §).NucigaR Puysics—SoLvrions 259 That is, the (J = 2)-mode behaves like a five-dimensional harmonic oscillator, with level spacing _ 3 (Ze) \" a My z[4 AT eR The oscillations will remain bounded so long as w is real. However, we notice that when, TU + A) < 2a a +z NR’ 5 oy pound ag g to unbounded If we use the incompressbility condition to write R=R,A"? with Rp = 1.2 x 107 om, and also write T = w/4x Ri, the condition for fission becomes 3(Ze)"/10R» > ued. Upon using the data given, one finds stability provided by Z*/A < 39. For an unstable nucleus such as »,U**, Z*/A ~ 36. This is good agreement, considering the simplicity of the modelList of Notations Gravitational constant Speed of light Hlectronic charge Planck’s constant Fine-structure constant Muon mass Pion mass Proton mass Bobr radius Bohr magneton Nuclear magneton Boltzmann constant Step! oltzmann constant Avogadro's number Gas constant Proton magnetic moment. Neutron magnetic moment Units of measurement: 1 Angstrom = 1A l year = x X 10° see Appendix USEFUL INFORMATION 6.67 x 10-® dyn em? gm-* 3 x 10" cm sec"! 4.8 X 107 esu 16 xX 10°C 6.625 x 107” erg sec hj2x = 1.05 x 10-* erg sec a — etfhe — 1/137 m(x) = 270 m, = 140 MeV M =16 x 10-“ gm = 938 MeV a, — h*/met = 0.53 x 10-8 em fe — emo” — 2.82 x 10" om 0.927 x 10°" erg gauss * 0.505 x 10-* erg gauss“ 6.02 x 10" mole Nok = 2 cal-mole™ = 8.3 x 10" ergs-mole~! °K"! 279 nm 1.91 nm ae 10° cm T electron volt = 1.8 ¥ 1071? erg 1 calorie = 4.25 1 atmosphere = 14.7 Ib in-* = Stirling's Formula 10° dyn em"* nl = 2m ni (n/e)* for large n 201262 APPENDIX—USEFUL INFORMATION Spherical Harmonics (for L = 0, 1, 2) w= /2, cos 8, [fz = fe r= Fa/ gz sn dere = Fa/ 32008 O sin 0 e='* Yi= vie cos? @ — sin®@) yj? = Bint erm Hydrogen Atom Wave Fam The normalized wave functions are defined by fn,..m = Rar) ¥7. The lowest levels are described by: Ry(t) = (2)"erer, Rat) = B1)_-er, ect, 3/ Zz (az) "(e- Rain) =(ZY"(_2 \2itg) \Ggw/ Vector Identities VF-G) = (F-V)G | (G-V)F | Fx (Vx @) +6 x (Vx F), OV-F +F-V¢, GV XE) — FV x &), 6V xX F+V6xF, Vv x (F x @) = K(V-6) — G(V-F) + (G-V)F — F-V6, — W(V-F) — V'R, fea for Baad) av = [uraa. Various familiar identities may be obtained from the above formula by an appropriate choice of F. For example, if F = V; is a vector, it follows that fv-vav =f vena. Another identity is obtained by choosing F = ee, V,; then the above for- mula yields f¥x Vay = [nx vad. Units for Electromagnetism There is considerable arbitrariness permissable in the choice of units for electromagnetism; as a result, a variety of ayatems ix in use. We have usedAPPENDIX—UsEFUL InForMATION 263 Gaussian units or rationalized MKS units, whichever seemed most suitable to the problem at hand. In these two systems of units, the fundamental equations of the theory are System | | He DH Lorente force DB} aP i | | . eee H=B-4M ) D=«E+P Ration- alized | 2°*| ae x 10-7 B ah +¥xB) MKS Byes One may convert physical quantities from one set of units to the other through the use of the following table, which expresses unit amount of the quantity measured in MKS units, in the equivalent amount of Gaussian Quantity ‘MKS units Geussian units Charge 1 coulomb (C) 3X 10° eeu. Potential | 1 volt (V) | 1/300 ‘Magnetic flux 1 weber (Wb) 10° maxwells (Mx) ‘Magnetic intensity (B) 1 weber/m* 10! gauss (G) ‘Magnetic field (H) 1 amp-turns/m 4x 10 oersteds(Oe) Capacity 1 farad (F) 9 x 10“ em ABCDF4087