PHY6095/PHZ6166: homework assignment # 4: Solutions due Monday, April 22 2 1. Problem 1 Fill in the blanks (complete proof is required!) pseudoscalar]: [v ≡ polar vector, p ≡ pseudovector, s ≡ scalar, a ≡ a) v × p = v b) p × p0 = p c) v · v0 = s d) p · p0 = s e) s(v · p) = a f) ap = v g) v · (p × p0 ) = a Under reflection in a plane with normal n̂, polar- and pseudovectors v = v · n̂ + v − (v · n̂)n̂ p = p · n̂ + p − (p · n̂)n̂ transform as v → v1 = v − 2(v · n̂)n̂ p → p1 = 2(p · n̂)n̂ − p The cross-product transforms into v1 × p1 = (v − 2(v · n̂)n̂) × (2(p · n̂)n̂ − p) = −v × p + 2(v · n̂)(n̂ × p) + 2(p · n̂)(v × n̂) Check the component of v1 × p1 parallel to n̂: (v1 × p1 ) · n̂ = −(v × p) · n̂. OK, looks like a polar vector. To be sure, check the component ⊥ to n̂: n̂ × (v1 × p1 ) = −v(n̂ · p) + p(n̂ · v) + 2(v · n̂)(n̂(n̂ · p) − p) + 2(p · n̂)(v − n̂(v · n̂) = n̂ × (v × p) Therefore, v × p is a polar vector. The rest of the examples is done in the same way. 2. Problem 2 Consider the Hamiltonian for decay of K0 and K̄0 into pions HKπ = g1 π † π † K0 + g2 π † π † K̄0 + g3 π † π † π † K0 + g4 π † π † π † K̄0 + H.c. • Show that CPT requires that g1 = −g2∗ , g3 = g4∗ . Recall that C(π) = π; P (π) = −π; T (π) = π C(K0 ) = K̄0 ; P (K0 ) = −K0 ; P (K̄0 ) = −K̄0 ; T (K0 ) = K0 ; T (K̄0 ) = K̄0 . (1) We work in the rest frame of K0 s, when the momenta of all π’s lie in the same plane. In this case, P and T do not reverse the signs of the momenta. Also, recall that T (aA + bB) = a∗ T (A) + b∗ T (B). Applying CPT to H, we find CP T (HKπ ) = −g1∗ π † π † K̄0 − g2∗ π † π † K0 + g3∗ π † π † π † K̄0 + g4∗ π † π † π † K0 + H.c. Therefore, CPT requires that g1 = −g2∗ ; g3 = g4∗ . (2) 3 • Derive the transformation properties of the short-living (S) and long-living (L) superpositions of K0 and K̄0 1 KS† = p g1 K0† + g2 K̄0† |g1 |2 + |g2 |2 1 KL† = p g2 K0† − g1 K̄0† . |g1 |2 + |g2 |2 Using (2), KS† = p KL† = p 1 |g1 |2 + |g2 |2 1 |g1 |2 + |g2 |2 g1 K0† − g1∗ K̄0† −g1∗ K0† − g1 K̄0† . 1 CP T (KS† ) = p CP T g1 K0† − g1∗ K̄0† |g1 |2 + |g2 |2 1 ∗ † =√ −g1 K̄0 + g1∗ K0† = KS† ... 1 CP T (KL† ) = √ g1 K̄0† + g1∗ K0† = −KL† . ... 3. Problem 3 In certain semiconductor heterostructures, a two-dimensional gas contains holes rather electrons. Strong spinorbit interaction locks the angular momenta and spins of holes into the total momentum J = 3/2 with projections Jz = ±1/2, ±3/2. The lowest energy state happens to be that with Jz = ±3/2. These states can be thought of as having effective spin 3/2 rather than 1/2. The wavefunction of spin 3/2 state is symmetric only under three-fold rotations, i.e., the Pauli matrices transform under rotations as σ± → σ± e∓3iθ . Suppose that a layer of holes is sandwiched between two different materials such that the top-down symmetry is broken, but the system is still invariant under a) rotations about the normal by an arbitrary angle; b) reflections in a plane that contains the normal; and c) time-reversal. Construct the Rashba Hamiltonian for this system and find the eigenstates and eigenergies. Rotational invariants are p3+ σ− and p3− σ− . Under reflection in a vertical plane that contains the x axis, px → px , py → −py ⇒ p± → p∓ while σx → −σx ; σy → −σy ⇒ σ± → −σ∓ . Therefore, the two rotational invariants must enter with a relative minus sign. Also, HSO must be Hermitian, which requires a factor of i up front: α 3 HSO = i p+ σ− − p3− σ+ 2 . The full Hamiltonian p2 ˆ H= I + HSO = 2m p2 /2m −α(py + ipx )3 3 −α(py − px ) p2 /2m (3) Eigenenergies E= p2 ± αp3 2m Eigenvectors |p, si = 1 −ise3iφ ; s = ±1 4. Problem 4 The symmetry group of a square, D4 , consists of eight elements: rotations by 0◦ , 90◦ , 180◦ , and 270◦ about an axis perpendicular to the plane of the square and passing through its center (chosen as the z−axis) plus rotations by 180◦ about the lines that bisect opposite sides of the square (chosen as the x and y axes) plus rotations by 180◦ about the square diagonals. 4 • List the effect of all operations of D4 on the coordinates x, y, z. – – – – – – – Identity: x, y, z → x, y, z C4 about z (π/2 rotation): x, y, z → y, −x, z C43 about z (3π/4 rotation): x, y, z → −y, x, z C2 = C42 about z: x, y, z → −x, −y, z C2 about x (U2x ): x, y, z → x, −y, −z C2 about y (U2y ): x, y, z → −x, y, z C2 about y = x (U2y=x ): x, y, z → y, x, −z – C2 about y = −x (U2y=−x ): x, y, z → −y, −x, −z • Construct a table, similar to that we constructed for group D2 in the class, for the effects of all these operations on the following functions, which can represent 3d-orbitals: u1 = z 2 (r), u2 = (x2 − y 2 )f (r), u3 = xyf (r), u4 = xzf (r), u5 = yzf (r), where r is the distance from the origin. Functions u1−4 and z form 1D representations, because they transform into themselves up to a possible sign change. Subjecting these functions to all operations listed above, we obtain characters (= ±1) for the representations. Functions u4 and u5 transform as components of a 2D vector. For example, on C4 , u4 = xz → yz = u5 and u5 = yz → −xz = −u4 , which means that the transformation is described by the matrix 0 1 (4) −1 0 with Tr= 0. Likewise, C43 corresponds to a matrix 0 −1 1 0 (5) also with Tr= 0 ,while C2 corresponds to −1 0 0 −1 (6) with Tr= −2. The table looks like this: D4 u1 = z 2 f (r) z u2 = (x2 − y 2 )f (r) u3 = xyf (r) {u4 , u5 } = {xzf (r), yzf (r)} A1 A2 B1 B2 E I 1 1 1 1 2 C2 1 1 1 1 −2 2C4 1 1 −1 −1 0 U2 1 −1 1 −1 0 U20 1 −1 −1 1 0 • Show that the functions u1 , u2 , and u3 satisfy Pα un = cαn un, where Pα is each of the eight symmetry transformations, and thus that u1 , u2 , and u3 generate onedimensional representations of D4 . Show also that u4 and u5 generate two-dimensional representations, i.e., that they transform as vectors. 5. Problem 5 Find which transitions between the states u1 , u2 , and u3 can be induced, to first order in the field, by an electric field directed along the z axis. Electric field along z transforms as A2 of D4 . Choose u1 (A1 ) as the initial state. Then A2 ⊗ A1 = A2 , which is not one the u1−3 states. Choose u2 (B1 ) as the initial state. The characters of A2 ⊗ B1 are shown in the table below 5 D4 2 u1 = z f (r) z 2 u2 = (x − y 2 )f (r) u3 = xyf (r) {u4 , u5 } = {xzf (r), yzf (r)} I A1 1 A2 1 B1 1 B2 1 E 2 A2 ⊗ B 1 1 A2 ⊗ B 2 1 C2 1 1 1 1 −2 1 1 2C4 1 1 −1 −1 0 −1 1 U2 1 −1 1 −1 0 −1 −1 U20 1 −1 −1 1 0 1 −1 Without doing any further calculations, we see that A2 ⊗ B1 = B2 . Likewise, A2 ⊗ B2 = B1 . Thus, transitions u2 ↔ u3 are allowed.