852_2010hw3_Solutions - PHYS852 Quantum Mechanics II,...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: PHYS852 Quantum Mechanics II, Spring 2010 HOMEWORK ASSIGNMENT 3 Topics covered: Unitary transformations, translation, rotation, vector operators 1. [25]Symmetry: A quantum system is said to posses a ‘symmetry’ if the Hamiltonian operator, H , is invariant under the associated transformation. In other words, if H ′ = H , where H ′ := U † HU . (a) [5] Show that H ′ = H is equivalent to [H, U ] = 0 Start with U † HU = H Hit from the left with U and use U U † = I to get HU = U H . Put both terms on the l.h.s. to get HU − U H = 0 or equivalently [H, U ] = 0. (b) [5] Any hermitian operator can be used to generate a unitary operator via U = e−iGφ , where G† = G is the ‘generator’ of the symmetry transformation, and φ is a free parameter. Show that [H, G] = 0 is necessary and sufficient for H to be symmetric under U . If [H, G] = 0 then it follows that [H, f (G)] = 0 for any single-variable function f (x). As U is of this form, it follows that [H, U ] = 0 so that H is symmetric with respect to the U . (c) [5] Show that when [H, G] = 0, the probability distribution over the eigenvalues of G does not change in time. In QM this means that G is a ‘constant of motion’. Must a QM constant of motion have a well-defined value? Since [H, G] = 0 it follows that simultaneous eigenstates of H and G exist. We can label them |n, g so that H |n, g = En |n, g and G|n, g = g|n, g . The most general state is then |ψ (t) = n g cn,g (t)|n, g . From Schr¨dinger’s equation we find that o d n, g|ψ (t) dt i = − n, g|H |ψ (t) d cn,g (t) = dt = −iωn cn,g (1) so that cn,g (t) = e−iωn t cn,g (0), which gives |ψ (t) = n g cn,m (0)e−iωn t |n, g (2) The projector onto the subspace with eigenvalue g is I (g) = n |n, g n, g|, so that the probability for the system to be in an eigenstate of G with eigenvalue g is P (g, t)) = = ψ (t)|I (g)|ψ (t) n,n′ ,n′′ g ′ ,g ′′ c∗ ′ ,g′ (0)eiωn′ t cn′′ ,g′′ (0)e−iωn′′ t n′ , g′ |n, g n, g|n′′ , g′′ n c∗ ′ ,g′ (0)eiωn′ t cn′′ ,g′′ (0)e−iωn′′ t δn′ ,n δg′ ,g δn′′ ,n δg′′ ,g n = n,n′ ,n′′ g ′ ,g ′′ = n |cn,g (0)|2 (3) which we see is independent of time. (d) [5] If a system possesses ‘translational symmetry’ what operator is a constant of motion? i The translation operator is UT (d) = e− dP so that P is the generator of translation. Thus in a system with translational symmetry, momentum will be conserved. 1 (e) [5] Consider a particle described by the Hamiltonian H= P2 + V (X ). 2M (4) What operator is the generator of translation? Show that H has translational symmetry only if V (x) = V0 . For H to possess translational symmetry required [H, P ] = 0. This then requires [V (X ), P ] = d 0. We know that [V (X ), P ] = i V ′ (X ), so translational symmetry requires dx V (x) = 0, or equivalently V (x) = V0 . 2 2. [25] Consider a system described by the Hamiltonian H= P2 1 + M ω 2 X 2 + M gX, 2M 2 (5) where g has units of acceleration. † † (a) [5] Show that UT (d)XUT (d) = X + d and UT (d)P UT (d) = P . We have † X ′ = UT (d)XUT (d) = † dxUT (d)|x x x|UT (d) = dx|x − d x x − d| = dx|x (x + d) x| = dx|x x x| + d dx|x x| (6) = X +d We know that P ′ = P because [UT (d), P ] = 0, given that UT (d) is a function of P . † (b) [5] Solve for d and E0 such that H ′ := UT (d)HUT (d) satisfies H ′ = E0 + P2 1 + M ω2X 2 2M 2 (7) We start from † H ′ = UT (d)HUT (d) † † UT (d)P UT (d)UT (d)P UT (d) 1 † † † + M ω 2 UT (d)XUT (d)UT (d)XUT (d) + M gUT (d)XUT (d) 2M 2 1 P ′2 2 + M ω 2 X ′ + M gX ′ = 2M 2 P2 1 = + M ω 2 (X + d)2 + M G(X + d) 2M 2 1 1 P2 (8) + M ω 2 X 2 + (M ω 2 d + M G)X + M ω 2 d2 + M Gd = 2M 2 2 Therefore the linear term will cancel for G d=− 2 (9) ω giving = H′ = = 1 P2 + M ω2X 2 + 2M 2 P2 1 + M ω2X 2 − 2M 2 so we see that E0 = − 3 1 M G2 M G2 − 2 ω2 ω2 2 MG 2ω 2 M G2 2ω 2 (10) (11) ′ (c) [5] Let |φ′ , n = 0, 1, 2, . . . be the nth eigenstate of H ′ , with corresponding eigenvalue En . What n ′ are En and φ′ (x) = x|φ′ ? n n We have ′ H ′ |φ′ = En |φ′ (12) n n Because H ′ is just an SHO plus a constant, we know that ′ En = ω n + 1 2 − M G2 2ω 2 (13) As a constant shift in the zero-point energy doesn’t change the shape of the wavefunction, we have 2 1 1 Hn (x/λ) e− 2 (x/λ) φ′ (x) = √ (14) n n n!λ π2 where λ = Mω . (d) [5] Show that |φn := UT (d)|φ′ is an eigenstate of H with eigenvalue En . What is the relationn ′ ship between En and En ? H |φn = HUT (d)|φ′ n † = UT (d)UT (d)HUT (d)|φ′ n = UT (d)H ′ |φ′ n ′ = UT (d)En |φ′ n ′ = En UT (d)|φ′ n ′ = En |φn (15) ′ with the definition H |φn = En |φn we see that En = En . (e) [5] What is the relationship between φn (x) := x|φn and φ′ (x)? What is φn (x)? n φn (x) = = = x|φn x|UT (d)|φ′ n x − d|φ′ n = φ′ (x − d) n this gives φn (x) = √ 1 π 2n n!λ (16) Hn (x + G/ω 2 )/λ e− 2 ((x+G/ω 1 4 2 )/λ 2 ) (17) 3. [10] Show explicitly that the momentum operator of a particle P is a vector operator with respect to rotation. Show that the operator P 2 = P · P is invariant under rotation about any axis (hint: chose a coordinate system where the axis of rotation is the z-axis). To show that P is a vector, we can consider an infinitesimal rotation about the z axis. ′ Px = e = i i ǫLz Px e− ǫLz i i 1 + ǫ(XPy − Y Px ) Px 1 − ǫ(ZPy − Y Px ) i = Px + ǫ [XPy − Y Px , Px ] (18) = Px − ǫPy ′ Py = e i ǫLz Py e− i ǫLz i i 1 + ǫ(XPy − Y Px ) Py 1 − ǫ(ZPy − Y Px ) = i = Py + ǫ [XPy − Y Px , Py ] = Py + ǫPx ′ Pz = e = i ǫLz (19) i Pz e− ǫLz i i 1 + ǫ(XPy − Y Px ) Pz 1 − ǫ(ZPy − Y Px ) i = Pz + ǫ [XPy − Y Px , Pz ] = Pz so that (20) ′ 1 −ǫ 0 Px Px ′ Py = ǫ 1 0 Py ′ 001 Pz Pz for a finite rotation this becomes ′ Px cos(θ ) − sin(θ ) 0 Px ′ Py = sin(θ ) cos(θ ) 0 Py ′ Pz 0 0 1 Pz (21) (22) which is the same as P ′ = M (θ )P P2 ′ 2 2 ′ ′ ′ = Px + Py + Pz (23) 2 2 = (cos θPx − sin θPy )2 + (cos θPy + sin θPx )2 + Pz 2 2 2 = cos2 θPx − cos θ sin θ (Px Py + Py Px ) + sin2 θPy + cos2 θPy + cos θ sin θ (Py Px + Px Py ) 2 2 + sin2 θPx + Pz 2 2 2 = (cos2 θ + sin2 θ )(Px + Py ) + Pz 2 2 2 = Px + Py + Pz = P2 (24) 5 4. [40/35] Consider an infinitesimal rotation about an arbitrary axis, described by the unitary operator UR (ǫ) = e− i L·ǫ i i i = 1 − L1 ǫ1 − L2 ǫ2 − L3 ǫ3 . (25) where L = j Lj ej and ǫ = j ǫj ej , with {e1 , e2 , e3 } being a right-handed set of orthogonal unit vectors. Using this notation, the angular momentum components are given by Lj = k,ℓ ǫj,k,ℓRk Pℓ , with ǫj,k,ℓ being the totally antisymmetric Levi-Cevita tensor, any index repeated 0; ǫjkℓ = 1; cyclic permutations of {j, k, ℓ} = {1, 2, 3} . (26) −1; cyclic permutations of {j, k, ℓ} = {3, 2, 1} The components of R and P satisfy the commutation relation [Rj , Pk ] = i δj,k . † ′ (a) [10] Evaluate Rj = UR (ǫ)Rj UR (ǫ) for each component of the position operator R = j Rj ej , and use this to deduce the 3 × 3 matrix, M (ǫ) that rotates an ordinary vector by the infinitesimal angle ǫ. In the infinitesimal limit, we have UR (ǫ) = 1 − i j ǫj Lj , so that ′ Rj = 1+ i ǫk Lk Rj 1+ i k = Rj + i ǫℓ Lℓ ℓ ǫk [Lk , Rj ] k = Rj + i ǫk ǫkℓm [Rℓ Pm , Rj ] k ℓm = Rj + i k ℓm = Rj + ǫk ǫkℓmRℓ (−i ) δm,j ǫk ǫkℓj Rℓ kℓ = δj,ℓ + ℓ ǫk ǫkℓj Rℓ (27) k This tells us that Mjk = δjk + ǫℓkj ǫℓ ℓ = δjk − so that ǫjkℓ ǫℓ ℓ 1 − ǫ3 ǫ2 1 − ǫ1 M (ǫ) = ǫ3 − ǫ2 ǫ1 1 6 (28) (29) (b) [5] Show that M (−ǫ) = M T (ǫ), then show that M T (ǫ) = M −1 (ǫ) by showing that M T (ǫ)M (ǫ) = I. Mjk (−ǫ) = δjk + ǫjkℓ ǫℓ ℓ = δkj − ǫkjℓ ǫℓ ℓ = Mkj (ǫ) T = Mjk (ǫ) M T (ǫ)M (ǫ) jk (30) T Mjℓ (ǫ)Mℓk (ǫ) = ℓ = δjℓ + m ℓ = ℓ δℓk − ǫjℓm ǫm δjℓ δℓk − = δjk − δjℓ ǫℓkn ǫn + ℓn ǫℓkn ǫn n ǫjℓm ǫm δℓk ℓm ǫjkn ǫn + n ǫjkm ǫm m (31) = δjk (c) [5] Now consider a finite rotation by δ = j δj ej , M N (δ/N ). described by the 3 × 3 matrix M (δ ). Clearly we must have M (δ ) = Take the limit as N → ∞, and use your result to part (a) to show that we can put M (δ ) into the form: M (δ ) = lim 1− N →∞ 1 Λ(δ ) N N = e−Λ(δ) where Λ(δ ) is a 3 × 3 antisymmetric matrix, whose components are given by Λj,k (δ ) = M (δ ) = M δ /N N (32) ℓ ǫj,k,ℓ δℓ . (33) where Mjk (δ/N ) = δjk − ǫjkℓ ℓ 1 = δjk − N = δjk − 7 δℓ N ǫjkℓ δℓ ℓ 1 Λjk (δ ) N (34) (d) [5] Show that the eigenvalues of Λ(δ ) are ω0 = 0, and ω± = ±iδ, where δ = |δ |. The eigenvalues of Λ(δ ) are solutions to −ω δ3 −δ2 det −δ3 −ω δ1 δ2 −δ1 −ω (35) =0 which gives the characteristic equation 2 2 2 −ω 3 − ωδ1 + δ1 δ2 δ3 − ωδ3 − δ1 δ2 δ3 − ωδ2 = 0 (36) ω ω 2 + δ2 = 0 (37) which simplifies to so that the solutions are ω0 = 0, and ω± = ±iδ. (e) Show that the eigenvectors of Λ(δ ) are u0 = u± = δ δ 2 (δ1 δ2 ± iδδ3 )e1 + (δ2 − δ2 )e2 + (δ2 δ3 ∓ iδδ1 )e3 2 2δ2 (δ2 − δ2 ) Λ(δ )u0 − ω0 u0 j = (38) (39) Λjk (δ )u0,k k = ǫjkℓδℓ u0,k kℓ 1 δ = = ǫjkℓ δℓ δk kℓ δ×δ j =0 Λ(δ )u± − ω± u± j = k = kℓ = (40) (Λjk − ω± δjk ) u±,k ǫjkℓ u±,k δℓ ∓ iδu±,j u± × δ ∓ iδu± j =0 (41) where we have used u± × δ = e1 (u±,2 δ3 − δ2 u±,3 ) + e2 (u±,3 δ1 − δ3 u±,1 ) + e3 (u±,1 δ2 − δ1 u±,2 ) 2 2 2 2 = e1 δ2 δ3 − δ2 δ3 − δ2 δ3 ± iδδ1 δ2 + e2 δ1 δ2 δ3 ∓ iδδ1 − δ1 δ2 δ3 ∓ iδδ3 2 2 + e3 δ1 δ2 ± iδδ2 δ3 − δ1 δ2 + δ2 δ1 2 2 = ±iδ e1 (δ1 δ2 ± iδδ3 ) + e2 −δ1 − δ3 + e3 (δ2 δ3 ∓ iδδ1 ) = ±iδu± 8 (42) (f) [5] Based on your result to part (e), show that = u0 (u0 · V ) + u− eiδ (u+ · V ) + u+ e−iδ (u− · V ) M (δ )V (43) where V is an arbitrary vector. We start from M (δ ) = e−Λ(δ ) (44) Λ(δ ) = |u0 ω0 u0 | + |u+ ω+ u+ | + |u− ω− u− | (45) M (δ ) = |u0 e−ω0 u0 | + |u+ e−ω+ u+ | + |u− e−ω− u− | (46) + u− eiδ u∗ · V − (47) + u− eiδ u+ · V (48) Using Dirac notation, we have so that switching back to standard notation, this gives M (δ )V = u0 u∗ · V 0 + u+ e−iδ u∗ · V + noting that u∗ = u0 , and u∗ = u∓ , this gives ± 0 M (δ )V = u0 u0 · V + u+ e−iδ u− · V (g) [5+5 bonus] Based on your results to parts (e) and (f), show that † V ′ = UR (δ )V UR (δ ) = M (δ )V = δ (δ · V ) δ(δ · V ) δ×V sin(δ) + V− cos(δ) + 2 2 δ δ δ (49) V ′ = M (δ )V = δ δ·V δ2 + cos(δ) u+ u− · V − i sin(δ) u+ u− · V u+ · V = u− · V = u+ u− · V + u− u+ · V 1 + u − u+ · V − u − u+ · V (50) 2 (δ1 δ2 + iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 − iδδ1 )V3 (51) 2 (δ1 δ2 − iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 + iδδ1 )V3 (52) 2 2δ2 (δ2 − δ2 ) 2 2δ2 (δ2 − δ2 ) = 2 (δ1 δ2 + iδδ3 ) (δ1 δ2 − iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 + iδδ1 )V3 2 2δ2 (δ2 − δ2 ) 2 (δ1 δ2 − iδδ3 ) (δ1 δ2 + iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 − iδδ1 )V3 2 2δ2 (δ2 − δ2 ) 2 22 2 δ1 δ2 δ3 − δ1 δ2 δ3 (δ1 δ2 + δ2 δ3 ) δ1 δ2 = V3 V1 − 2 V2 + 2 2 2 δ δ2 (δ1 + δ3 ) δ2 (δ2 − δ2 ) 2 22 2 2 δ1 δ2 δ1 δ3 δ2 (δ1 + δ3 ) − δ1 (δ1 + δ3 ) V1 − 2 V2 − 2 V3 = 2 2 δ δ δ2 (δ1 + δ3 ) δ1 (53) = V1 − 2 δ · V δ + 9 u+ u− · V − u− u+ · V = 1 − = = = u+ u− · V + u − u+ · V u+ u− · V − u − u+ · V u+ u− · V + u− u+ · V 2 2 3 2 (δ1 δ2 + iδδ3 ) (δ1 δ2 − iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 + iδδ1 )V3 2 2δ2 (δ2 − δ2 ) 2 (δ1 δ2 − iδδ3 ) (δ1 δ2 + iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 − iδδ1 )V3 2 2δ2 (δ2 − δ2 ) 2 2 δ3 (δ2 − δ2 ) δ2 δ2 + δ2 δ3 i V2 + i 1 2 2 2 V3 δ(δ2 − δ2 ) δ(δ − δ2 ) δ3 V2 δ2 V3 −i +i δ δ i (54) δ×V δ 1 2 2 (δ2 − δ2 ) (δ1 δ2 − iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 + iδδ1 )V3 2 2δ2 (δ2 − δ2 ) 2 2 (δ2 − δ2 ) (δ1 δ2 + iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 − iδδ1 )V3 − 2 2δ2 (δ2 − δ2 ) 2 δ2 − δ2 δ2 δ3 δ1 δ2 V2 − 2 V3 = − 2 V1 + δ δ2 δ δ2 (55) = V2 − 2 δ · V δ =− 2 2 (δ2 − δ2 ) (δ1 δ2 − iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 + iδδ1 )V3 2 2δ2 (δ2 − δ2 ) 2 2 (δ2 − δ2 ) (δ1 δ2 + iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 − iδδ1 )V3 + 2 2δ2 (δ2 − δ2 ) δ1 V3 δ3 V1 −i =i δ δ i (56) = δ×V δ 2 =− = 2 (δ1 δ2 − iδδ3 ) (δ1 δ2 − iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 + iδδ1 )V3 2 2δ2 (δ2 − δ2 ) 2 (δ1 δ2 + iδδ3 ) (δ1 δ2 + iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 − iδδ1 )V3 2 2δ2 (δ2 − δ2 ) δ3 (57) = V3 − 2 δ · V δ + u+ u− · V − u− u+ · V 3 = − = 2 (δ1 δ2 − iδδ3 ) (δ1 δ2 − iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 + iδδ1 )V3 2 2δ2 (δ2 − δ2 ) 2 (δ1 δ2 + iδδ3 ) (δ1 δ2 + iδδ3 )V1 + (δ2 − δ2 )V2 + (δ2 δ3 − iδδ1 )V3 2 2δ2 (δ2 − δ2 ) i (58) δ×V δ 3 10 Putting these pieces together gives V′ = δ δ·V δ2 + V − δ δ·V δ2 11 cos(δ) + δ × V sin(δ) δ (59) ...
View Full Document

This note was uploaded on 12/21/2011 for the course PHYS 852 taught by Professor Moore during the Spring '11 term at Michigan State University.

Ask a homework question - tutors are online