Quantum physics the bottom up approach
0 likes469 views
The document summarizes key concepts from elementary quantum physics that will be built upon in the text, including: 1) The time-dependent and time-independent Schrodinger equations, which describe the wave function and energy levels of quantum systems. 2) Observables in quantum physics are represented by operators, and the measurement of an observable leaves the system in an eigenstate of that operator. 3) The Heisenberg uncertainty principle limits the precision with which conjugate variables like position and momentum can be known simultaneously. 4) Angular momentum is quantized and can be decomposed into orbital and spin components, with associated quantum numbers and eigenstates. Operators for total and z-component angular
Quantum physics the bottom up approach
- 1. Chapter 1 Recollections from Elementary Quantum Physics Abstract We recall the prerequisites that we assume the reader to be familiar with, namely the Schrödinger equation in its time dependent and time independent form, the uncertainty relations, and the basic properties of angular momentum. Introductory courses on quantum physics discuss the one-dimensional Schrödinger equation for the wave function Ψ (x,t) of a particle of mass M moving in a poten- tial V i ∂Ψ ∂t = − 2 2M ∂2Ψ ∂x2 + V Ψ. (1.1) Therein = h/2π is the reduced Planck constant. The function Ψ is understood as a probability amplitude whose absolute square |Ψ (x,t)|2 = Ψ ∗(x,t)Ψ (x,t) gives the probability density for finding the particle at time t at position x. This probability density is insensitive to a phase factor eiϕ. With the Hamilton operator H = − 2 2M ∂2 ∂x2 + V, (1.2) the Schrödinger equation reads ˙Ψ = − i HΨ. (1.3) The dot denotes the time derivative. In this text, we print operators and matrices in nonitalic type, like H, p, or σ, just to remind the reader that a simple letter may represent a mathematical object more complicated than a number or a function. Ordinary vectors in three-dimensional space are written in bold italic type, like x or B. The time dependent Schrödinger equation reminds us of the law of energy con- servation E = p2/2M + V if we associate the operator i ∂/∂t with energy E and the operator ( /i)∂/∂x with momentum p. Take as an example the plane wave Ψ (x,t) = Ψ0ei(kx−ωt) of a photon propagating in the x-direction. The energy opera- tion i ∂Ψ/∂t = ωΨ then relates energy to frequency, E = ω, and the momentum operation ( /i)∂Ψ/∂x = kΨ relates momentum to wave number, p = k. D. Dubbers, H.-J. Stöckmann, Quantum Physics: The Bottom-Up Approach, Graduate Texts in Physics, DOI 10.1007/978-3-642-31060-7_1, © Springer-Verlag Berlin Heidelberg 2013 3
- 2. 4 1 Recollections from Elementary Quantum Physics The probability amplitude Ψ (x,t) for finding the particle at time t at position x and the amplitude Φ(p,E) for finding it with energy E and momentum p turn out to be Fourier transforms of each other. Pairs of Fourier transforms have widths that are reciprocal to each other. If the width x in position is large, the width p in momentum is small, and vice versa, and the same for the widths t and E. The conjugate observables p and x or E and t obey the uncertainty relations p x ≥ 1 2 , E t ≥ 1 2 . (1.4) The exact meaning of x, etc. will be defined in Sect. 3.4. For stationary, that is, time independent potentials V (x), the Hamilton operator (1.2) acts only on the position variable x. The solution of the Schrödinger equation then is separable in x and t, Ψ (x,t) = ψ(x)e−iEt/ . (1.5) The probability density (in units of m−1) for finding a particle at position x then is independent of time |Ψ (x,t)|2 = |ψ(x)|2. The amplitude ψ(x) is a solution of the time independent Schrödinger equation − 2 2M ∂2ψ(x) ∂x2 + V (x)ψ(x) = Eψ(x). (1.6) In operator notation, this reads Hψ = Eψ. (1.7) For particles trapped in a potential well V (x), the requirement that total proba- bility is normalizable to +∞ −∞ ψ(x) 2 dx = 1 (1.8) leads to the quantization of energy E such that only a discrete set of values En, with the corresponding wave functions ψn(x), n = 1,2,3,..., is allowed. n is called the main quantum number. The mean value derived from repeated measurements of a physical quantity or observable A is given by its expectation value A(t) = +∞ −∞ Ψ ∗ (x,t)AΨ (x,t)dx. (1.9) For a stationary state Eq. (1.5), the expectation value A does not depend on time and is insensitive to any phase factor eiϕ. For example, the mean position x of a matter wave is given by the weighted average x = +∞ −∞ x ψ(x) 2 dx. (1.10)
- 3. 1 Recollections from Elementary Quantum Physics 5 Sometimes the Hamiltonian H can be divided into two parts H = H1(x)+H2(y), one depending on one set of (not necessarily spatial) variables x, the other on a disjoint set of variables y. Then, as shown in standard textbooks, the solution of the Schrödinger equation is separable in x and y as ψ(x,y) = ψ1(x)ψ2(y). As in Eq. (1.5), we must attach a phase factor to obtain the time dependent solution Ψ (x,y,t) = ψ1(x)ψ2(y)e−iEt/ . (1.11) The probability densities for the joint occurrence of variable x and variable y then multiply to Ψ (x,y,t) 2 = ψ1(x) 2 ψ2(y) 2 . (1.12) This is consistent with classical probability theory, where probabilities multiply for independent events. The probability that two dice both show 5 is 1 6 × 1 6 = 1 36 . If, on the other hand, the quantum system can evolve along two mutually exclu- sive paths ψ1 or ψ2, as is the case in the famous double-slit experiments, then the two probability amplitudes add coherently, and the total probability is |Ψ |2 = |ψ1 + ψ2|2 = |ψ1|2 + |ψ2|2 + 2|ψ1||ψ2|sinδ, (1.13) with phase angle δ. Only if phase coherence between the partial waves ψ1 and ψ2 is lost, does the total probability equal the classical result |Ψ |2 = |ψ1|2 + |ψ2|2 , (1.14) with probabilities adding for mutually exclusive events. The probability that a die shows either 5 or 3 is 1 6 + 1 6 = 1 3 . Figure 1.1 shows the setup and result of a double-slit experiment with a beam of slow neutrons. What we see is the self-interference of the probability amplitudes ψ1 and ψ2 of single neutrons. The conditions for the appearance of quantum interfer- ence will be discussed in Sect. 6.6. For most students the Schrödinger equation is no more difficult than the many other differential equations encountered in mechanics, electrodynamics, or trans- port theory. Therefore, students usually have fewer problems with the ordinary Schrödinger equation than with a presentation of quantum mechanics in the form of matrix mechanics. In everyday scientific life, however, spectroscopic problems that require matrix diagonalization are more frequent than problems like particle waves encountering step potentials and other scattering problems. Furthermore, as we shall see in Sect. 19.1.2, the Schrödinger equation can always equally well be ex- pressed as a matrix equation. Therefore, matrix mechanics is the main topic of the present tutorial. Instead of starting with the Schrödinger equation, we could have started with the Heisenberg equation of motion (discovered a few weeks before the Schrödinger equation) because both are equivalent. We postpone the introduction of Heisenberg’s equation to Chap. 10 because students are less familiar with this approach.
- 4. 6 1 Recollections from Elementary Quantum Physics Fig. 1.1 (a) Double-slit experiment with neutrons: The monochromatic beam (along the arrow), of de Broglie wave length λ = 2 nm, λ/λ = 8 %, meets two slits, formed by a beryllium wire (shaded circle) and two neutron absorbing glass edges (vertical hatching), installed on an optical bench of 10 m length with 20 µm wide entrance and exit slits. (b) The measured neutron self-in- terference pattern follows theoretical expectation. From Zeilinger et al. (1988) To begin we further assume acquaintance with some basic statements of quantum physics that are both revolutionary and demanding, and our hope is that the reader of this text will learn how to live and work with them. The first statement is: To any physical observable A corresponds an operator A. The measurement of the observ- able A leaves the system under study in one of the eigenstates or eigenfunctions ψn of this operator such that Aψn = anψn. (1.15) The possible outcomes of the measurement are limited to the eigenvalues an. Which of these eigenfunctions and eigenvalues is singled out by the measurement is uncer- tain until the measurement is actually done. If we regard the ordered list ψn as elements of a vector space, called the Hilbert space, the operation A just stretches an eigenvector ψn of this space by a factor an. The best-known example of an eigenvalue equation of this type is the time indepen- dent Schrödinger equation Hψn = Enψn. (1.16) H is the operator for the observable energy E, ψn are the eigenfunctions of energy, and En are the corresponding eigenvalues. In this text, we shall only treat the sim- plest case where the spectrum of the En is discrete and enumerable, as are atomic energy spectra, and in most cases limit the number of energy levels to two. Another basic statement is derived in Sect. 3.4: Two physical quantities, de- scribed by operators A and B, can assume well defined values an and bn, simultane- ously measurable with arbitrarily high precision and unhampered by any uncertainty relation, if and only if they share a common set of eigenfunctions ψn, with Aψn = anψn, Bψn = bnψn. (1.17)
- 5. 1 Recollections from Elementary Quantum Physics 7 An important operator is that of the angular momentum J, which we shall de- rive from first principles in Chap. 16. Let us beforehand recapitulate the following properties of J, as known from introductory quantum physics. The components Jx, Jy, and Jz of the angular momentum operator J cannot be measured simultaneously to arbitrary precision. Only the square magnitude J2 of the angular momentum and its component Jz along an arbitrary axis z are well defined and can be measured simultaneously without uncertainty. This means that the phase of J about this axis z remains uncertain. The operators of the two observables J2 and Jz then must have simultaneous eigenfunctions that we call ψjm, which obey J2 ψjm = j(j + 1) 2 ψjm, (1.18a) Jzψjm = m ψjm, (1.18b) with eigenvalues j(j + 1) 2 and m , respectively. Angular momentum is quantized, with possible angular momentum quantum numbers j = 0, 1 2 ,1, 3 2 ,.... For a given value of j, the magnetic quantum number m can take on only the 2j + 1 different values m = −j,−j + 1,...,j. For j = 1 2 we have m = ±1 2 and J2 ψ1 2 ,± 1 2 = 3 4 2 ψ1 2 ,± 1 2 , (1.19a) Jzψ1 2 ,± 1 2 = ± 1 2 ψ1 2 ,± 1 2 . (1.19b) If we arrange the 2j + 1 eigenfunctions of the angular momentum ψjm into one column ψj = ⎛ ⎜ ⎜ ⎜ ⎝ ψj,j ψj,j−1 ... ψj,−j ⎞ ⎟ ⎟ ⎟ ⎠ , (1.20) we shall print this column vector ψj in nonitalic type, as we did for the operators. The corresponding row vector is ψ† j = (ψ∗ j,j ,ψ∗ j,j−1,...,ψ∗ j,−j ), where the dagger signifies the conjugate transpose ψ† = ψ∗T of a complex vector (or matrix). The components Jx and Jy of the angular momentum operator J do not have simultaneous eigenfunctions with J2 and Jz. If we form the linear combinations J+ = Jx + iJy, J− = Jx − iJy, (1.21) these operators are found to act on the angular momentum state ψjm as J+ψjm = j(j + 1) − m(m + 1)ψj,m+1, (1.22a) J−ψjm = j(j + 1) − m(m − 1)ψj,m−1, (1.22b)
- 6. 8 1 Recollections from Elementary Quantum Physics with J+ψj,+j = 0 and J−ψj,−j = 0. The J+ and J− are the raising and lowering operators, respectively. For j = 1 2 we have J+ψ1 2 ,− 1 2 = ψ1 2 ,+ 1 2 and J−ψ1 2 ,+ 1 2 = ψ1 2 ,− 1 2 . Angular momentum may be composed of the orbital angular momentum L and of spin angular momentum S, which can be added to the total angular momentum J = L + S. The orbital angular momentum quantum number l can only have integer values, spin quantum number s can have either integer or half-integer values. The triangle rule of vector addition tells us that the (integer or half-integer) angular momentum quantum number j has the allowed range |l − s| ≤ j ≤ l + s. (1.23) The total angular momentum quantum number is m = ml + ms, (1.24) going in unit steps from m = −j to m = j. For example, two angular momenta with l = 1 and s = 1 2 can be added to j = 1 2 or j = 3 2 . For l = 0, j = s we have m = ms, and for s = 0, j = l we have m = ml. For the sake of simplicity, in these two special cases we shall always write m for the respective magnetic quantum numbers ms or ml. References Zeilinger, A., Gähler, R., Shull, C.G., Treimer, W., Mampe, W.: Single- and double-slit diffraction of neutrons. Rev. Mod. Phys. 60, 1067–1073 (1988)