EE201/MSE207 Lecture 15 Perturbation theory (Ch. 6) (time-independent, nondegenerate, Sec. 6.1) Usually solving TISE πΌπ = πΉπ is too complicated; need approximations. Perturbation theory is one of approximate methods to solve TISE. Idea: separate Hamiltonian into simple and small parts (if possible) πΌ = πΌ 0 + πΌ 0 π = πΉπ is simple (can be solved), πΌ 1 πΌ 1 is small where πΌ = πΌ 0 + π πΌ 1 then power series in π βͺ 1 , and then π = 1 Trick: (0) + ππ π (1) + π 2 π π (2) + . . . π π = π π notation π = |πβͺ (0) + ππΉ π (1) + π 2 πΉ π (2) + . . . πΉ π = πΉ π 0 + ππ π 1 + . . . ) = (πΉ π 0 + ππΉ π 1 + . . . )(π π 0 + ππ π 1 + . . . ) πΌ 0 + π πΌ 1 (π π 0 = πΉ π 0 π π 0 . . . order π 0 : πΌ 0 π π order π 2 : (solvable exactly) . . . 1 + 0 = πΉ π 0 π π 1 + πΉ π 1 π π order π 3 : 0 order π 1 : πΌ 0 π π πΌ 1 π π . . .
First-order perturbation theory ( π 1 ) 1 + 0 = πΉ π 0 π π 1 + πΉ π 1 π π 0 πΌ 0 π π πΌ 1 π π β β ππ¦ , or (the same) β©π π 0 π¦ 0 |. . . Multiply by π π and integrate, ββ 1 βͺ + β©π π 0 | 0 βͺ = πΉ π 0 β©π π 0 |π π 1 βͺ + πΉ π (0) | 1 β©π π πΌ 0 |π π πΌ 1 |π π equal notation π 1 πΌ π 2 = β©π 1 | πΌ π 2 βͺ 1 = β©π π 0 | 0 βͺ πΉ π πΌ 1 |π π First-order correction to energy is just the average (expectation) value of πΌ 1 in the unperturbed state (very natural)
First-order perturbation theory (cont.) 1 + 0 = πΉ π 0 π π 1 + πΉ π 1 π π 0 πΌ 0 π π πΌ 1 π π (1) to wavefunction Now find correction π π 0 ) π π 1 = β( (1) ) π π 0 ( πΌ 0 β πΉ π πΌ 1 βπΉ π Rewrite (1) = πβ π π π (π) π π (π) = 0 (0) π π π π Expand in zero-order eigenstates from normalization (0) β πΉ π (π) π π 0 ) π π 0 = β( (1) ) π π 0 πβ π (πΉ π πΌ 1 βπΉ π 0 β πΉ π 0 ) π π π = β π π 1 π ππ 0 0 (0) | : (πΉ π πΌ 1 π π + πΉ π Multiply by β©π π (1) (another way of derivation) For π = π we obtain the previous formula for πΉ π 0 0 π = π π πΌ 1 π π For π β π : π π 0 β πΉ π 0 πΉ π 0 0 π π πΌ 1 π π (1) = (0) π π π π 0 β πΉ π 0 Rename π β π πΉ π πβ π
First-order perturbation theory: summary πΌ = πΌ 0 + πΌ 1 1 = β©π π 0 | 0 βͺ πΉ π πΌ 1 |π π 0 = πΉ π 0 π π 0 πΌ 0 π π (0) + π π (1) + β¦ 0 0 π π πΌ 1 π π π π = π π (1) = (0) π π π π 0 β πΉ π 0 (0) + πΉ π (1) + β¦ πΉ π πΉ π = πΉ π πβ π (0) = πΉ π (0) (i.e. when degeneracy). Then Remark. Correction to π π is not good if πΉ π the formalism is a little different. Usually degeneracy is lifted (disappears) due to perturbation. For example, in hydrogen atom there is fine structure (due to relativistic correction and spin-orbit) and hyperfine structure (due to magnetic interaction of electron and proton). Second-order perturbation: similar but lengthier 2 Result for second-order correction 0 0 π π πΌ 1 π π (2) = to energy of n th level: πΉ π 0 β πΉ π 0 πΉ π πβ π
WKB approximation (Ch. 8) (Wentzel, Kramers, Brillouin, 1926) β β 2 π 2 π π¦ πΉ can be discrete + π π¦ π π¦ = πΉ π(π¦) ππ¦ 2 2π or continuous If π π¦ = const , then easy to solve If π π¦ varies slowly, then modify solution for π π¦ = const . Idea: Two cases: πΉ > π(π¦) (classical region) πΉ < π(π¦) (tunneling)
WKB approximation, classical region, πΉ > π(π¦) 2π(πΉ β π) If π π¦ = const = π , then π π¦ = π΅ π Β±πππ¦ , π = β π¦ π π¦ β² ππ¦ β² Then for π(π¦) we expect π π¦ β π΅ π¦ exp Β±π 2π π ππ β πΎ = πβ ππ¦ β π β ππ From conservation of the probability current ππ¦ 1 π΅ π¦ β we obtain Therefore π(π¦) π¦ π π¦ β const π π¦ β² ππ¦ β² 2π[πΉ β π π¦ ] exp Β±π π(π¦) = π π¦ β π slightly different in the textbook, Β± β π π¦ ππ¦ π€ π¦ , so π 2 β Remark 1. 1 π(π¦) β 1 1 π€(π¦) , as it should be. Remark 2. If π(π¦) (as in SiGe technology), then π΅(π¦) β π(π¦) π(π¦) . Remark 3. WKB approximation works well only for slowly changing π(π¦) .
WKB approximation, tunneling, πΉ < π(π¦) 2π(π β πΉ) If π π¦ = const = π , then π π¦ = π΅ π Β±ππ¦ , π = β Similarly π¦ π π¦ β const π π¦ β² ππ¦ β² 2π[π π¦ β πΉ] exp Β± π(π¦) = π π¦ β WKB approximation is often used to estimate probability of tunneling (coefficient of transmission) through a strong (almost βopaqueβ) tunnel barrier π(π¦) Tunneling probability π ππ π ππ£π’ π out 2 π in 2 π β πΉ π β 1 π π π¦ ππ¦) 0 π β exp(β2 π 0 (very crudely; we neglect all pre-exponential factors, which usually are within one order of magnitude, while exponential factor can typically be 10 β3 β 10 β10 )
WKB, connection between the two regions assume smooth π(π¦) πΉ (different result for an abrupt potential) π(π¦) π¦ 0 (classical turning point) WKB approximation does not work well in the vicinity of π¦ 0 , we need a better approximation near π¦ 0 (linear potential, Airy functions). Result: π¦ π· π π¦ β² ππ¦β² , exp β π¦ > π¦ 0 π(π¦) π¦ 0 π π¦ β π¦ 0 2 π· sin π π π¦ β² ππ¦β² , 4 + π¦ < π¦ 0 π(π¦) π¦
Variational principle (Ch. 7) Only Sec. 7.1 Theorem: For an arbitrary |πβͺ , the ground state energy πΉ π satisfies inequality πΉ π β€ π πΌ π = β© πΌβͺ Proof is simple. Let us expand π = π π π |π π βͺ . Then since πΉ π β₯ πΉ π , we get πΌ = π π π 2 πΉ π β₯ πΉ π π π π 2 = πΉ π This theorem can be useful to estimate πΉ π (or at least to find an upper bound) Idea: Use trial wavefunctions |πβͺ with many adjustable parameters and minimize β© πΌβͺ . Hopefully min β© πΌβͺ is close to πΉ π . Extensions of this method can also be used to find |π π βͺ , first-excited state energy and wavefunction (using subspace orthogonal to |π π βͺ ), second-excited state, etc.
Recommend
More recommend