IPhO 2021 · Theory problem 3

Particles and Waves

Wave–particle duality, which states that each particle can be described as a wave and vice versa, is one of the central concepts of quantum mechanics. In this problem, we will rely on this notion and just a few other basic assumptions to explore a selection of quantum phenomena covering the two distinct types of particles of the microworld—fermions and bosons.

Part A. Quantum particle in a box (1.4 points)

Consider a particle of mass \(m\) moving in a one-dimensional potential well, where its potential energy \(V(x)\) is given by

\[ V(x)= \begin{cases} 0, & 0\leq x\leq L;\\ \infty, & x<0\text{ or }x>L. \end{cases} \tag{1} \]

While classical particle can move in such a potential having any kinetic energy, for quantum particle only some specific positive discrete energy levels are allowed. In any such allowed state, the particle can be described as a standing de Broglie wave with nodes at the walls.

A.1 — 0.4pt

Determine the minimal possible energy \(E_{\min}\) of the quantum particle in the well. Express your answer in terms of \(m\), \(L\), and the Planck’s constant \(h\).

The particle’s state with minimal possible energy is called the ground state, and all the rest allowed states are called excited states. Let us sort all the possible energy values in the increasing order and denote them as \(E_n\), starting from \(E_1\) for the ground state.

A.2 — 0.6pt

Find the general expression for the energy \(E_n\) (here \(n=1,\ 2,\ 3,\ldots\)).

A.3 — 0.4pt

Particle can undergo instantaneous transition from one state to another only by emitting or absorbing a photon of the corresponding energy difference. Find the wavelength \(\lambda_{21}\) of the photon emitted during the transition of the particle from the first excited state (\(E_2\)) to the ground state (\(E_1\)).

Part B. Optical properties of molecules (2.1 points)

In this part, we will study several optical properties of the cyanine Cy5 molecule—a widely used dye molecule, schematically shown in Fig. 1a. Its optical properties are determined mainly by the carbon backbone, composed of the alternating single and double bonds between carbon atoms, shown in Fig. 1b, while the influence of the rings at the molecule’s ends as well as radicals R is much smaller. Three of the four valence electrons of each C atom (and of N atoms) in the backbone form the chemical bonds, while the remaining valence electrons are “shared” and can move along the whole backbone. The net potential energy of each such electron is shown with oscillating thin line in Fig. 1c, with minima corresponding to the positions of the C and N atoms.

Figure 1

Figure 1. (a) Chemical structure of the cyanine Cy5 molecule (for simplicity, hydrogen atoms are not shown, and R denote some radicals). (b) The backbone of the Cy5 molecule, with mean inter-atomic distance \(l\). (c) Potential energy of the electron along the backbone (thin line) and its approximation by the step function given by Eq. 1 (thick line).

For simplicity, we will approximate this potential energy profile by a simple function given in Eq. 1 with the width \(L=10.5l\) (see thick line in Fig. 1c), here \(l=140\ \mathrm{pm}\) is the mean inter-atomic distance (see also Fig. 1b). As a result, we obtain the “electronic gas” composed of 10 electrons (7 from C atoms, 2 from the N atom, and 1 from the \(\mathrm{N}^{+}\) ion), moving in a one-dimensional potential well discussed in part A. In our evaluation, we can neglect the mutual interaction of these electrons; however, we should account for the fact that electrons are fermions and thus obey the Pauli exclusion principle. We also neglect the influence of other electrons as well as motion of the nuclei.

B.1 — 0.8pt

Evaluate the largest wavelength \(\lambda\) of the photon that can be absorbed by the Cy5 molecule assuming that the electron system is initially in its ground state. Express your answer in terms of \(l\), physical constants and some numerical prefactor, and calculate the numerical value.

B.2 — 0.4pt

Another dye molecule Cy3 has similar structure, but its backbone is shorter by 2 carbon atoms. Is its absorption spectrum shifted to the bluer or to the redder spectral region compared to the Cy5 molecule? Evaluate numerically the magnitude \(\Delta\lambda\) of this spectral shift. You can assume that removing two carbon atoms doesn’t change the molecule shape and only makes the backbone length shorter by two interatomic distances.

Being in the excited state, molecule can undergo a spontaneous transition to the ground state while emitting photon. The mean rate \(K\) of such events (i.e. the relative decrease of the molecules being in the excited state, \(\mathrm dN/N\), over time \(\mathrm dt\), \(K=\frac{1}{N}\frac{\mathrm dN}{\mathrm dt}\)) is determined by the wavelength \(\lambda\) of the emitted photon, the transition electrical dipole moment \(d\) (which is of the order of \(d\simeq el\), here \(e\) is elementary charge) as well as vacuum permittivity \(\varepsilon_0\) and Planck’s constant \(h\).

B.3 — 0.7pt

Using dimension analysis, determine the expression for the rate of spontaneous emission in terms of \(\varepsilon_0\), \(h\), \(\lambda\), and \(d\). The numerical prefactor for your expression is \(=\frac{16}{3}\pi^3\).

B.4 — 0.2pt

For Cy5 molecule, \(d\approx2.4\,el\). Evaluate the mean fluorescence lifetime of the lowest excited state of Cy5 molecule, \(\tau_{\mathrm{Cy5}}\), which is reciprocal to the rate of its emissive transition to the ground state.

Part C. Bose-Einstein condensation (1.5 points)

This part is not directly related to Parts A and B. Here, we will study the collective behaviour of bosonic particles. Bosons do not respect the Pauli exclusion principle, and—at low temperatures or high densities—experience a dramatic phenomenon known as the Bose–Einstein condensation (BEC). This is a phase transition to an intriguing collective quantum state: a large number of identical particles ‘condense’ into a single quantum state and start behaving as a single wave. The transition is typically reached by cooling a fixed number of particles below the critical temperature. In principle, it can also be induced by keeping the temperature fixed and driving the particle density past its critical value.

We begin by exploring the relation between the temperature and the particle density at the transition. As it turns out, estimates of their critical values can be deduced from a simple observation: Bose-Einstein condensation takes place when the de Broglie wavelength corresponding to the mean square speed of the particles is equal to the characteristic distance between the particles in a gas.

C.1 — 0.4pt

Given a non-interacting gas of \({}^{87}\mathrm{Rb}\) atoms in thermal equilibrium, write the expressions for their typical linear momentum \(p\) and the typical de Broglie wavelength \(\lambda_{\mathrm{dB}}\) as a function of atom’s mass \(m\), temperature \(T\) and physical constants.

C.2 — 0.5pt

Calculate the typical distance between the particles in a gas, \(\ell\), as a function of particle density \(n\). Hence deduce the critical temperature \(T_c\) as a function of atom’s mass, their density and physical constants.

To realize BEC in the lab, the experimentalists have to cool gases to temperatures as low as \(T_c=100\ \mathrm{nK}\).

C.3 — 0.6pt

What is the particle density of the Rb gas \(n_c\) if the transition takes place at such a temperature? For the sake of comparison, calculate also the ‘ordinary’ particle density \(n_0\) of an ideal gas at the standard temperature and pressure (STP), i.e. \(T_0=300\ \mathrm K\) and \(p_0=10^5\ \mathrm{Pa}\). How many times is the ‘ordinary’ gas denser? You may assume that the mass of the atoms is equal to 87 atomic mass units (\(m_{\mathrm{amu}}\)).

Part D. Three-beam optical lattices (5 points)

The first Bose-Einstein condensates were produced back in 1995, and since then the experimental work has branched out in diverse directions. In this part, you will investigate one particularly fruitful idea to load the condensate into spatially periodic potentials created by interfering a number of coherent laser beams. Due to the periodic nature of the resulting interference patterns, they are referred to as optical lattices. The potential energy \(V(\vec r)\) of an atom moving in an optical lattice is proportional to the local intensity of the light, and in your calculations you may assume that

\[ V(\vec r)=-\alpha\left\langle\left|\vec E(\vec r,t)\right|^2\right\rangle. \tag{2} \]

Here, \(\alpha\) is a positive constant, and the angle brackets indicate time-averaging which eliminates the rapidly oscillating terms. The electric field produced by the \(i\)-th laser is described by

\[ \vec E_i=E_{0,i}\,\vec\varepsilon_i\cos(\vec k_i\cdot\vec r-\omega t), \tag{3} \]

with the amplitude \(E_{0,i}\), the wave vector \(\vec k_i\), and the unit-length polarization vector \(\vec\varepsilon_i\).

Figure 2

Figure 2. (a) Three-beam optical lattice: three plane waves with wave vectors \(\vec k_{1,2,3}\) intersect and interfere in the area indicated by the grey circle. (b) Symmetries of a regular hexagon: solid and dashed lines show two sets of symmetry axes. (c) Saddle point: a point on a surface where the slopes in orthogonal directions are all zero, but which is not a local extremum of the plotted function. Travelling along the trajectory marked by the full line one encounters an apparent minimum. Additional analysis of the perpendicular direction (dashed line) is needed to distinguish a true minimum from a saddle point (shown).

Your task is to study triangular optical lattices that are produced by interfering three coherent laser beams of equal intensity. A typical setup is shown in Fig. 2a. Here, all three beams are polarized in the \(z\) direction, propagate in the \(xy\) plane and intersect at equal angles of \(120^\circ\). Choose the direction of the \(x\) axis parallel to the wave vector \(\vec k_1\).

D.1 — 1.4pt

Using Eqs. 2 and 3 obtain the expression for the potential energy \(V(\vec r)\) as a function of \(\vec r=(x,y)\) in the plane of the beams.

Hint: the result can be neatly expressed as a constant term plus a sum of three cosine functions of arguments \(\vec b_i\cdot\vec r\). Please write your result in this form and identify the vectors \(\vec b_i\).

D.2 — 0.5pt

The resulting potential energy has a sixfold rotational symmetry axis, i.e., the potential distribution is invariant with respect to a rotation by a multiple of \(60^\circ\) around the origin. Provide a simple argument to prove that this is indeed the case.

The above observation of symmetry simplifies the analysis of the two-dimensional potential distribution \(V(\vec r)\). As shown in Fig. 2b, a regular hexagon has symmetry lines that, respectively, connect opposite vertices (solid lines) and midpoints of opposite edges (dashed lines). Therefore, in our situation one does not need to produce and study two-dimensional potential plots as many insights can be deduced by focusing on the coordinate axes \(x\) and \(y\) that run along the symmetry lines.

D.3 — 1.2pt

Derive the behavior of the potential \(V(\vec r)\) along the coordinate axes, i.e., determine the functions \(V_X(x)\equiv V(x,0)\) and \(V_Y(y)\equiv V(0,y)\). Identify the locations of the extrema of \(V_X(x)\) and \(V_Y(y)\) as functions of a single argument. As these functions are periodic, please include in your lists only one representative from each family of periodically repeated minima and maxima.

We are interested in determining the locations of so-called lattice sites, i.e., the minima of the full two-dimensional potential \(V(\vec r)\). The obtained minima of single-argument functions \(V_X\) and \(V_Y\) identify their suspected positions but still have to be checked to eliminate the saddle points. As shown in Fig. 2c, when studied along a single line, saddle points may disguise as minima but they are not.

D.4 — 0.8pt

Review your results in the previous question to determine actual minima of the optical lattice: Identify all equivalent minima nearest to (but not coinciding with) the origin. What is the distance \(a\) between the nearest minima, in other words—the lattice constant of our optical lattice? Express the answer in terms of the laser wavelength \(\lambda_{\mathrm{las}}\).

Charge neutrality of ultracold atoms suggests that their interactions become relevant only when two or more atoms occupy the same site of an optical lattice. However, experimentalists are also able to explore setups that sustain long-range atomic interactions. A possible approach relies on creation of the so-called Rydberg atoms that are physically large and feature other exaggerated properties. Rydberg atoms are excited atoms with one electron promoted to a state with a very high principal quantum number \(n\). The size of a Rydberg atom can be estimated by calculating the radius of the classical circular orbit of that electron with the orbital angular momentum \(n\hbar\), here \(\hbar\) is the reduced Planck constant.

D.5 — 1.1pt

Calculate the value of \(n\) that corresponds to the radius of the Rb Rydberg atom comparable to the wavelength of laser light \(\lambda_{\mathrm{las}}=380\ \mathrm{nm}\). Give your answer in terms of \(\lambda_{\mathrm{las}}\) and physical constants and find its numerical value.

Official English text of the problem.

Answers

Here \(m_e\) denotes the electron mass. For Part D, define the common field amplitude \(E_0\), \(U=\alpha E_0^2>0\), and \[ k=\frac{2\pi}{\lambda_{\mathrm{las}}},\qquad P_X=\frac{4\pi}{3k},\qquad P_Y=\frac{4\pi}{\sqrt3\,k}. \]

  • A.1: \(\displaystyle E_{\min}=\frac{h^2}{8mL^2}\).
  • A.2: \(\displaystyle E_n=\frac{n^2h^2}{8mL^2}\), \(n=1,2,3,\ldots\).
  • A.3: \(\displaystyle \lambda_{21}=\frac{8mcL^2}{3h}\).
  • B.1: \(\displaystyle \lambda_{\mathrm{Cy5}}=\frac{882}{11}\frac{m_ec\,l^2}{h}\simeq648\ \mathrm{nm}\).
  • B.2: A blue shift: \(\lambda_{\mathrm{Cy3}}\simeq519\ \mathrm{nm}\), and \(\displaystyle \Delta\lambda=\frac{1580}{99}\frac{m_ec\,l^2}{h}\simeq129\ \mathrm{nm}\).
  • B.3: \(\displaystyle K=\frac{16\pi^3d^2}{3\varepsilon_0h\lambda^3}\).
  • B.4: \(\displaystyle \tau_{\mathrm{Cy5}}=\frac{3\varepsilon_0h\lambda_{\mathrm{Cy5}}^3}{16\pi^3(2.4\,el)^2}\simeq3.3\ \mathrm{ns}\).
  • C.1: \(\displaystyle p=\sqrt{3mk_{\mathrm B}T}\), \(\displaystyle \lambda_{\mathrm{dB}}=\frac{h}{\sqrt{3mk_{\mathrm B}T}}\).
  • C.2: \(\displaystyle \ell=n^{-1/3}\), \(\displaystyle T_c=\frac{h^2n^{2/3}}{3mk_{\mathrm B}}\), within the stated transition estimate.
  • C.3: \(\displaystyle n_c\simeq1.59\times10^{18}\ \mathrm{m^{-3}}\), \(\displaystyle n_0\simeq2.41\times10^{25}\ \mathrm{m^{-3}}\); the ordinary gas is \(\displaystyle 1.52\times10^7\) times denser.
  • D.1: \(\displaystyle V(\vec r)=-U\left[\frac32+\sum_{i=1}^3\cos(\vec b_i\cdot\vec r)\right]\), with \(\displaystyle \vec b_1=k\left(\frac32,-\frac{\sqrt3}{2}\right)\), \(\displaystyle \vec b_2=k(0,\sqrt3)\), and \(\displaystyle \vec b_3=k\left(-\frac32,-\frac{\sqrt3}{2}\right)\).
  • D.2: The six vectors \(\{\pm\vec b_1,\pm\vec b_2,\pm\vec b_3\}\) form a regular hexagon. A \(60^\circ\) rotation permutes them, and cosine is even, so \(V\) is unchanged.
  • D.3: \(\displaystyle V_X(x)=-U\left[\frac52+2\cos\left(\frac{3kx}{2}\right)\right]\), and \(\displaystyle V_Y(y)=-\frac U2\left[1+2\cos\left(\frac{\sqrt3\,ky}{2}\right)\right]^2\). Modulo \(P_X\), the \(x\)-section has a minimum at \(0\) and a maximum at \(P_X/2\). Modulo \(P_Y\), the \(y\)-section has minima at \(0,P_Y/2\) and maxima at \(P_Y/3,2P_Y/3\).
  • D.4: \(\displaystyle a=\frac{2}{3}\lambda_{\mathrm{las}}\). The six nearest minima are \((\pm a,0)\) and \(\left(\pm a/2,\pm\sqrt3\,a/2\right)\), with independent signs in the latter pair.
  • D.5: \(\displaystyle n=\sqrt{\frac{m_e e^2\lambda_{\mathrm{las}}}{4\pi\varepsilon_0\hbar^2}}\simeq84.7\), corresponding to \(n\approx85\).

Solution

The numerical evaluations use \[ \begin{aligned} h&=6.62607015\times10^{-34}\ \mathrm{J\,s}, &c&=2.99792458\times10^8\ \mathrm{m\,s^{-1}},\\ m_e&\simeq9.10938\times10^{-31}\ \mathrm{kg}, &e&=1.602176634\times10^{-19}\ \mathrm C,\\ \varepsilon_0&\simeq8.85419\times10^{-12}\ \mathrm{F\,m^{-1}}, &k_{\mathrm B}&=1.380649\times10^{-23}\ \mathrm{J\,K^{-1}},\\ m_{\mathrm{amu}}&\simeq1.66054\times10^{-27}\ \mathrm{kg}. \end{aligned} \]

A.1

The wave must have a node at each wall. The longest nonzero standing wave fitting into the well has one half-wavelength across its width: \[ L=\frac{\lambda_{\mathrm{dB},1}}2. \] The de Broglie relation therefore gives \[ p_1=\frac{h}{\lambda_{\mathrm{dB},1}}=\frac{h}{2L}. \] Using nonrelativistic kinetic energy, with zero potential energy inside the well, \[ \boxed{E_{\min}=\frac{p_1^2}{2m}=\frac{h^2}{8mL^2}}. \] A zero-wave-number state cannot satisfy both boundary conditions unless the wave is identically zero, so it is not an allowed particle state.

A.2

In the \(n\)-th standing-wave mode, \(n\) half-wavelengths fit between the walls: \[ L=n\frac{\lambda_{\mathrm{dB},n}}2,\qquad n=1,2,3,\ldots. \] Consequently, \[ \lambda_{\mathrm{dB},n}=\frac{2L}{n}, \qquad p_n=\frac{nh}{2L}, \] and \[ \boxed{E_n=\frac{n^2h^2}{8mL^2}}. \]

A.3

Energy conservation equates the emitted photon energy to the decrease in the particle’s energy: \[ \frac{hc}{\lambda_{21}} =E_2-E_1 =\frac{(4-1)h^2}{8mL^2}. \] Hence \[ \boxed{\lambda_{21}=\frac{8mcL^2}{3h}}. \]

B.1

Each spatial energy level can hold two electrons with opposite spins. Thus the ground state of the ten-electron system fills levels \(n=1,\ldots,5\), two electrons per level.

The smallest possible excitation promotes an electron from \(n=5\) to the first unoccupied level, \(n=6\). Its energy is \[ \Delta E_{\mathrm{Cy5}} =E_6-E_5 =\frac{(6^2-5^2)h^2}{8m_eL^2} =\frac{11h^2}{8m_eL^2}. \] The smallest absorbed photon energy corresponds to the largest wavelength: \[ \lambda_{\mathrm{Cy5}} =\frac{hc}{\Delta E_{\mathrm{Cy5}}} =\frac{8m_ecL^2}{11h}. \] With \(L=10.5l\), \[ \boxed{ \lambda_{\mathrm{Cy5}} =\frac{8(10.5)^2}{11}\frac{m_ec\,l^2}{h} =\frac{882}{11}\frac{m_ec\,l^2}{h} }. \] Substituting \(l=140\times10^{-12}\ \mathrm m\), \[ \boxed{\lambda_{\mathrm{Cy5}}\simeq6.48\times10^{-7}\ \mathrm m =648\ \mathrm{nm}}. \]

B.2

Removing two carbon atoms removes two shared electrons and shortens the well by \(2l\). Therefore Cy3 has eight electrons and \[ L_{\mathrm{Cy3}}=8.5l. \] Its ground state fills levels \(n=1,\ldots,4\). The lowest excitation is \(4\to5\), so \[ \Delta E_{\mathrm{Cy3}} =\frac{(5^2-4^2)h^2}{8m_e(8.5l)^2} =\frac{9h^2}{8m_e(8.5l)^2}. \] Thus \[ \lambda_{\mathrm{Cy3}} =\frac{8(8.5)^2}{9}\frac{m_ec\,l^2}{h} =\frac{578}{9}\frac{m_ec\,l^2}{h} \simeq519\ \mathrm{nm}. \] This is shorter than the Cy5 wavelength, so the shift is toward the blue. Its magnitude is \[ \begin{aligned} \Delta\lambda &=\lambda_{\mathrm{Cy5}}-\lambda_{\mathrm{Cy3}}\\ &=\left(\frac{882}{11}-\frac{578}{9}\right) \frac{m_ec\,l^2}{h}\\ &=\frac{1580}{99}\frac{m_ec\,l^2}{h}. \end{aligned} \] Therefore \[ \boxed{\Delta\lambda\simeq1.29\times10^{-7}\ \mathrm m=129\ \mathrm{nm}}. \]

B.3

The dipole moment has dimensions of charge times length: \[ [d]=\mathrm{C\,m}. \] Because \[ [\varepsilon_0]=\frac{\mathrm C^2}{\mathrm{J\,m}}, \] we have \[ \left[\frac{d^2}{\varepsilon_0}\right]=\mathrm{J\,m^3}. \] Dividing by \(h\), whose dimensions are \(\mathrm{J\,s}\), and by \(\lambda^3\) produces the required inverse-time dimension: \[ \left[\frac{d^2}{\varepsilon_0h\lambda^3}\right]=\mathrm{s^{-1}}. \] Using the supplied dimensionless prefactor, \[ \boxed{K=\frac{16\pi^3d^2}{3\varepsilon_0h\lambda^3}}. \]

Here \(K\) is the positive decay rate. For the number \(N(t)\) of molecules remaining excited, the decay equation is \(\mathrm dN/\mathrm dt=-KN\); the statement’s “relative decrease” is interpreted as the positive loss.

B.4

Under the given model, the lowest excited state emits a photon of wavelength \(\lambda_{\mathrm{Cy5}}\) when it returns to the ground state. Its dipole moment is \[ d=2.4\,el \simeq5.38\times10^{-29}\ \mathrm{C\,m}. \] The spontaneous-emission rate is \[ K=\frac{16\pi^3(2.4\,el)^2} {3\varepsilon_0h\lambda_{\mathrm{Cy5}}^3} \simeq3.01\times10^8\ \mathrm{s^{-1}}. \] For exponential decay, the mean lifetime is \(1/K\): \[ \boxed{ \tau_{\mathrm{Cy5}} =\frac{3\varepsilon_0h\lambda_{\mathrm{Cy5}}^3} {16\pi^3(2.4\,el)^2} \simeq3.3\ \mathrm{ns} }. \]

C.1

For the temperature estimate specified in the problem, equipartition gives the mean translational kinetic energy in three dimensions: \[ \frac12m\langle v^2\rangle=\frac32k_{\mathrm B}T. \] The typical momentum associated with the root-mean-square speed is therefore \[ \boxed{p=m\sqrt{\langle v^2\rangle}=\sqrt{3mk_{\mathrm B}T}}. \] Its de Broglie wavelength is \[ \boxed{\lambda_{\mathrm{dB}}=\frac hp =\frac{h}{\sqrt{3mk_{\mathrm B}T}}}. \]

C.2

At number density \(n\), the mean volume per particle is \(1/n\). The characteristic linear spacing is consequently \[ \boxed{\ell=n^{-1/3}}. \] Applying the problem’s transition criterion, \[ \lambda_{\mathrm{dB}}(T_c)=\ell, \] gives \[ \frac{h}{\sqrt{3mk_{\mathrm B}T_c}}=n^{-1/3}. \] Squaring and rearranging, \[ \boxed{T_c=\frac{h^2n^{2/3}}{3mk_{\mathrm B}}}. \] This is the critical-temperature estimate obtained from the stated wavelength–spacing criterion.

C.3

For rubidium, \[ m=87m_{\mathrm{amu}}\simeq1.44467\times10^{-25}\ \mathrm{kg}, \qquad T_c=100\ \mathrm{nK}=1.00\times10^{-7}\ \mathrm K. \] Inverting the result of C.2, \[ n_c=\left(\frac{3mk_{\mathrm B}T_c}{h^2}\right)^{3/2}. \] Numerically, \[ \boxed{n_c\simeq1.59\times10^{18}\ \mathrm{m^{-3}}}. \] The corresponding spacing is approximately \(8.57\times10^{-7}\ \mathrm m\).

For the ordinary ideal gas, the ideal-gas law gives \[ p_0=n_0k_{\mathrm B}T_0. \] Using the temperature and pressure specified in the question, \[ n_0=\frac{10^5}{k_{\mathrm B}(300)} \simeq2.41\times10^{25}\ \mathrm{m^{-3}}. \] Thus \[ \boxed{\frac{n_0}{n_c}\simeq1.52\times10^7}. \] The ordinary gas is about fifteen million times denser.

D.1

Let all three field amplitudes be \(E_0\). Choosing the wave vectors consistently with Fig. 2a, \[ \vec k_1=k(1,0),\qquad \vec k_2=k\left(-\frac12,\frac{\sqrt3}{2}\right),\qquad \vec k_3=k\left(-\frac12,-\frac{\sqrt3}{2}\right), \] where \(k=2\pi/\lambda_{\mathrm{las}}\).

All polarizations are along \(z\), so the fields add as scalars in that direction: \[ \vec E(\vec r,t) =E_0\hat z\sum_{i=1}^{3}\cos(\vec k_i\cdot\vec r-\omega t). \] The required time averages are \[ \left\langle\cos^2(\phi-\omega t)\right\rangle=\frac12, \] and \[ \left\langle \cos(\phi_i-\omega t)\cos(\phi_j-\omega t) \right\rangle =\frac12\cos(\phi_i-\phi_j). \] Therefore \[ \left\langle|\vec E|^2\right\rangle =E_0^2\left[ \frac32+\sum_{i<j}\cos\big((\vec k_i-\vec k_j)\cdot\vec r\big) \right]. \] Define \[ \begin{aligned} \vec b_1&=\vec k_1-\vec k_2 =k\left(\frac32,-\frac{\sqrt3}{2}\right),\\ \vec b_2&=\vec k_2-\vec k_3=k(0,\sqrt3),\\ \vec b_3&=\vec k_3-\vec k_1 =k\left(-\frac32,-\frac{\sqrt3}{2}\right). \end{aligned} \] Since cosine is even, the sign chosen for each difference is immaterial. With \(U=\alpha E_0^2\), \[ \boxed{ V(\vec r)=-U\left[\frac32+ \cos(\vec b_1\cdot\vec r)+ \cos(\vec b_2\cdot\vec r)+ \cos(\vec b_3\cdot\vec r)\right] }. \] Combining the first and third cosines gives the useful coordinate form \[ \boxed{ V(x,y)=-U\left[ \frac32+ 2\cos\left(\frac{3kx}{2}\right) \cos\left(\frac{\sqrt3\,ky}{2}\right) +\cos(\sqrt3\,ky) \right] }. \]

D.2

The three \(\vec b_i\) have equal magnitude \(\sqrt3\,k\) and are separated by \(120^\circ\). Including their negatives gives six vectors at successive angular intervals of \(60^\circ\).

Rotating the position vector by \(60^\circ\) is equivalent, in each dot product, to rotating the corresponding \(\vec b_i\) by \(-60^\circ\). This sends every \(\vec b_i\) to another member of the set \(\{\pm\vec b_j\}\). Since \[ \cos(-\vec b_j\cdot\vec r)=\cos(\vec b_j\cdot\vec r), \] the sum is unchanged. Thus the potential has the stated sixfold rotational symmetry.

D.3

Setting \(y=0\), \[ \boxed{ V_X(x)=-U\left[\frac52+2\cos\left(\frac{3kx}{2}\right)\right] }. \] Its fundamental period is \[ P_X=\frac{4\pi}{3k}=\frac{2\lambda_{\mathrm{las}}}{3}. \] The derivative is \[ V_X'(x)=3Uk\sin\left(\frac{3kx}{2}\right). \] Within \(0\leq x<P_X\), the minimum is at \(x=0\), and the maximum is at \(x=P_X/2\).

For the \(y\)-axis, introduce \[ q=\frac{\sqrt3\,ky}{2}. \] Then \[ V_Y(y)=-U\left(\frac32+2\cos q+\cos2q\right). \] Using \(\cos2q=2\cos^2q-1\), \[ \boxed{V_Y(y)=-\frac U2(1+2\cos q)^2}. \] Its fundamental period is \[ P_Y=\frac{4\pi}{\sqrt3\,k} =\frac{2\lambda_{\mathrm{las}}}{\sqrt3}. \] Differentiation gives \[ V_Y'(y)=\sqrt3\,Uk(1+2\cos q)\sin q. \] Hence the stationary points satisfy either \(\sin q=0\) or \(\cos q=-1/2\).

One representative of every family of extrema is:

Section Coordinate in one period Character along that axis Potential
\(V_X\) \(x=0\) Minimum \(-9U/2\)
\(V_X\) \(x=P_X/2\) Maximum \(-U/2\)
\(V_Y\) \(y=0\) Minimum \(-9U/2\)
\(V_Y\) \(y=P_Y/3\) Maximum \(0\)
\(V_Y\) \(y=P_Y/2\) Minimum \(-U/2\)
\(V_Y\) \(y=2P_Y/3\) Maximum \(0\)

Thus \(V_X\) is a sinusoid about \(-5U/2\). Over one period, \(V_Y\) rises from a deep minimum to zero, falls to a shallower minimum, rises to zero again, and returns to the deep minimum at the next period boundary.

D.4

First check the apparently shallow minimum at \[ (x,y)=\left(0,\frac{P_Y}{2}\right). \] Here the second derivatives of the full potential are \[ V_{xx}=-\frac92Uk^2,\qquad V_{yy}=\frac32Uk^2,\qquad V_{xy}=0. \] The curvatures have opposite signs, so this point is a saddle, not a lattice site.

At the origin, \[ V_{xx}=V_{yy}=\frac92Uk^2,\qquad V_{xy}=0, \] so the origin is a true minimum.

To locate all true minima, write \[ t=\frac{3kx}{2},\qquad q=\frac{\sqrt3\,ky}{2}. \] The stationarity conditions are \[ \sin t\cos q=0,\qquad \sin q(\cos t+2\cos q)=0. \] If \(\cos q\neq0\), stationarity in \(x\) requires \(\sin t=0\). Positive curvature in \(x\) then requires \[ \cos t\cos q>0. \] Since \(\cos t\) and \(\cos q\) have the same sign, \(\cos t+2\cos q\) cannot vanish. Thus \(\sin q=0\). The minima therefore have \[ t=\pi r,\qquad q=\pi s, \] where integers \(r,s\) have the same parity. The remaining stationary possibility, \(\cos q=0\), also requires \(\cos t=0\); its Hessian has negative determinant, so it gives only saddles.

Writing \(r=2j+s\), the lattice sites are \[ \vec r_{j,s} =a\left(j+\frac{s}{2},\frac{\sqrt3\,s}{2}\right), \qquad j,s\in\mathbb Z, \] with \[ a=\frac{4\pi}{3k}. \] At these points all three interference cosines equal \(1\), and \[ V_{\min}=-\frac92U. \]

The squared distance from the origin is \[ |\vec r_{j,s}|^2=a^2(j^2+js+s^2). \] For nonzero integer pairs, the smallest value of the positive integer \(j^2+js+s^2\) is \(1\). It occurs at six sites: \[ \boxed{ (a,0),\ (-a,0),\ \left(\frac a2,\frac{\sqrt3a}{2}\right),\ \left(-\frac a2,\frac{\sqrt3a}{2}\right),\ \left(\frac a2,-\frac{\sqrt3a}{2}\right),\ \left(-\frac a2,-\frac{\sqrt3a}{2}\right) }. \] They form a regular hexagon around the origin. Therefore \[ \boxed{a=\frac{4\pi}{3k}=\frac23\lambda_{\mathrm{las}}}. \]

D.5

For a highly excited outer electron, the nucleus and inner electrons act approximately as a core of net charge \(+e\). We use the stated classical circular-orbit model, neglecting core-penetration effects and the motion of the much heavier core.

Coulomb attraction supplies the centripetal force: \[ \frac{m_ev^2}{r} =\frac{e^2}{4\pi\varepsilon_0r^2}. \] The prescribed angular momentum is \[ m_evr=n\hbar. \] Eliminating \(v\), \[ \frac{n^2\hbar^2}{m_er^3} =\frac{e^2}{4\pi\varepsilon_0r^2}, \] so \[ r=\frac{4\pi\varepsilon_0\hbar^2}{m_ee^2}n^2 =a_0n^2, \] where \[ a_0\simeq5.292\times10^{-11}\ \mathrm m. \] Taking the estimated radius to equal \(\lambda_{\mathrm{las}}\), \[ \boxed{ n=\sqrt{\frac{\lambda_{\mathrm{las}}}{a_0}} =\sqrt{\frac{m_ee^2\lambda_{\mathrm{las}}} {4\pi\varepsilon_0\hbar^2}} }. \] For \(\lambda_{\mathrm{las}}=380\ \mathrm{nm}\), \[ n=\sqrt{\frac{380\times10^{-9}}{5.292\times10^{-11}}} \simeq84.7. \] Thus the corresponding integer principal quantum number is \[ \boxed{n\approx85}. \]

The solution was checked against the official solution.