quant-ph0601056/ms.tex
1: \documentclass{appolb}
2: \usepackage{graphicx}
3: \usepackage{amsmath}
4: \usepackage{cite}
5: \headtitle{Quantum Brownian motion and the Third Law of thermodynamics}
6: \headauthor{Peter H{\"a}nggi and Gert-Ludwig Ingold}
7: \begin{document}
8: \title{Quantum Brownian motion and the Third Law of
9: thermodynamics
10: \thanks{This work is
11: dedicated to our colleague and  friend, Professor Peter Talkner, on
12: the occasion of his 60-th birthday.}}
13: \author{Peter H{\"a}nggi and Gert-Ludwig Ingold
14: \address{Institut f{\"u}r Physik, Universit{\"a}t Augsburg, 86135 Augsburg,
15: Germany}
16: }
17: \maketitle
18: \begin{abstract}
19: The quantum thermodynamic behavior of small systems is investigated
20: in presence of finite quantum dissipation. We consider the archetype
21: cases of a damped harmonic oscillator and a free quantum Brownian
22: particle. A main finding is that  quantum dissipation helps to ensure the validity of
23: the Third Law. For the quantum oscillator, finite damping replaces the
24: zero-coupling result of an exponential
25: suppression of the specific heat at low temperatures by a
26: power-law behavior. Rather intriguing is the behavior of the free quantum
27: Brownian particle. In this case, quantum dissipation is able to restore the Third Law:
28: Instead of being constant down to zero temperature, the specific heat now
29: vanishes proportional to temperature with an amplitude that is {\it inversely}
30: proportional to the ohmic dissipation strength. A distinct subtlety of
31: finite quantum dissipation is the result that the various thermodynamic functions of
32: the sub-system do not only depend on the dissipation strength but depend as well
33: on the prescription employed in their definition.
34: \end{abstract}
35: \PACS{05.70.-a, 05.30.-d, 05.40.-a, 05.40.Jc}
36: 
37: \section{Introduction}
38: The development of the theory of Brownian motion played  a pivotal
39: role --  and continues to do so -- in the development of statistical
40: mechanics and thermodynamics \cite{chaos05,physicaSI}.  Thermodynamics
41: together with relativity and quantum theory form three pillars on
42: which much of the entire structure of physics rests.
43: Tampering with the axioms in either of those theories is not a good
44: idea; doing so may well lead to contradictions with
45: the  other theories. In particular, the field of thermodynamics
46: bears consequences for many branches of physics. Its four laws are
47: well-known \cite{TD1,TD2}: the zeroth law guarantees that states of
48: thermal equilibrium exist which can be characterized by
49: a temperature $T$. The first law provides a balance among the
50: various contributions that make up the internal energy of a system
51: while the second law introduces the concept of  thermodynamic entropy $S$,
52: which notably is extensive and never decreases for a closed physical
53: system. In addition, the second law tells us that there exists an
54: absolute zero of temperature.
55: 
56: The Third Law is attributed to Walther Hermann Nernst (1864-1941)
57: and arose as the result of his seminal idea -- being guided by his
58: critical analysis of  chemical and electrochemical reactions at
59: lower temperatures -- that at low temperatures there occurs for
60: isothermal processes a perfect correspondence between the enthalpy
61: and the Gibbs free energy. Thereby, the approximate rule
62: hypothesized earlier by Marcelin Berthelot and Julius Thomson, becomes
63: a ``law'' at zero
64: temperature. Nernst announced this result already in his lectures in
65: 1905, terming it ``mein W\"armesatz'' (my law of heat)
66: \cite{ebeling1,ebeling2}. He took this result even further: He also
67: studied {\it how} fast the difference between the changes in the
68: enthalpy $\Delta H$ and the Gibbs free energy $\Delta G$, \ie
69: $\Delta H - \Delta G$ tends to zero \cite{Nernst}.  In fact, this
70: difference vanishes faster than linear in temperature implying that
71: the change of entropy itself must vanish at absolute zero. This in
72: turn implies identical, generally vanishing initial slopes for the corresponding
73: quantities $\Delta H$ and $\Delta G$ as a function of temperature.
74: An elucidating account of the history of the Third Law and the
75: controversies surrounding its acceptance is presented in the books
76: by Dugdale \cite{Dugdale} and by Wilks \cite{Wilks}. In its strict form,
77: as given by Max Planck
78: \cite{planck}, the Third Law reads: The entropy $s=S/N$ per particle
79: approaches at absolute zero a constant value $s_0$ that possibly depends
80: on the chemical composition of the system. This limiting entropy constant
81: $s_0$ can generally be set equal to zero.
82: 
83: The Third Law carries prominent consequences for quantum mechanics
84: and the field of low-temperature physics. First, the fact
85: that at absolute zero temperature the isotherm coincides with the isentrope (adiabat)
86: immediately implies that this absolute zero temperature is
87: unattainable by  use of a sequence of  isothermal and
88: adiabatic reversible operations \cite{nernst1912}.
89: Therefore, it has the consequence that
90: the efficiency of a Carnot engine, \ie a heat engine that cyclically
91: operates between two heat baths of different temperatures which are
92: never brought into contact with each other, can never reach $100\%$ for
93: any finite upper temperature. Moreover, the constant
94: value of the entropy at absolute zero is given by the degeneracy $g$ of the $N$-particle system
95: in the corresponding quantum ground state, \ie $S(T=0)= k_\text{B}\ln g$,
96: where $k_\text{B}$ is the Boltzmann constant. The limiting value of the
97: intensive quantity $s_0=S(T=0)/N$ in the thermodynamic limit of
98: particle number $N \rightarrow \infty$ will typically be zero,
99: so long as the degeneracy $g=g(N)$ does not grow with $N$ faster than
100: exponentially \cite{Leggett}. A well-known exception is the case of
101: noninteracting, independent particles carrying a non-vanishing spin $I$, yielding
102: $s_0=k_\text{B}\ln(2I+1)$ for the limiting entropy per particle.
103: Moreover, the Third Law also implies
104: that thermal quantities such as specific heats, the isobar thermal
105: coefficient of expansion, the isochor coefficient of tension, etc.,
106: all approach zero as $T \rightarrow 0$. Likewise, the (magnetic)
107: susceptibility becomes constant as $T \rightarrow 0$, so that
108: the classic Curie law must loose its validity at very small
109: temperatures.
110: 
111: Are there known exceptions of systems not obeying the Third Law? It
112: is known that many classical systems do not obey the Third Law. In
113: particular, noninteracting classical systems with their constant
114: values for the specific heat clearly violate the Third Law. A
115: well-known case is the classical ideal gas for which the entropy $S$
116: assumes the form $S=N[c_V\ln T+k_\text{B}\ln(V/N)+\sigma]$, where
117: $V$ is the volume and
118: $\sigma$ denotes the entropy constant. It clearly does not fulfill
119: the Third Law because it diverges logarithmically with temperature $T$
120: for a constant specific heat $c_V$. Even when we use for $c_V$ the
121: physically relevant low-temperature result, namely that quantum
122: mechanically the specific heat $c_V$ vanishes faster than $\ln T$,
123: we still find a dependence on the particle density which is not compatible 
124: with the formulation of Planck. This observation that the classical gas
125: does not obey the Third Law led Nernst to speculate that the
126: classical gases must undergo a ``degeneracy'', which has been
127: resolved with the quantum statistics for the ideal Fermi gas and the
128: ideal Bose-Einstein gas, which indeed do obey the Third Law
129: in the strict formulation by Planck.
130: 
131: Are there yet other remaining open problems with the Third Law?
132: Apparent difficulties with the Third Law occur for metastable states
133: that do not necessarily guarantee the sufficiently fast relaxation
134: within a finite time scale towards thermal equilibrium, the latter
135: being a prerequisite for the validity of the Third Law. In this
136: context, glasses provide a system class that can provide
137: detectable deviations from the Third Law at low temperatures which
138: likely are the result of frozen-in ordered excited states that have
139: not yet fully relaxed. According to common wisdom the known
140: deviations from the Third Law will all be cured by quantum mechanics,
141: quantum statistics, and interactions among particles.
142: 
143: In the following we shall investigate the thermodynamic low
144: temperature properties for open quantum systems that are coupled to
145: a heat bath of {\it finite} dissipation strength. Because both, the
146: thermodynamics of classical open systems and the quantum statistical
147: mechanics of open systems are strictly valid only for systems that
148: are only infinitely weakly coupled to a bath, it is
149: {\it a priori} not obvious how the Third Law lives up to the
150: sub-system in presence of finite quantum dissipation
151: \cite{leggett83,rise85,ingoldPR,HTB,schon90,dittrich98,grifoni98,weiss99,talkner04,chaosHI}.
152: The effect of the finite coupling of a sub-system to an environment
153: in fact induces several subtleties for quantum Brownian motion
154: \cite{chaosHI}. For example, the equilibrium density matrix
155: is no longer given by its standard canonical form $\rho_\text{S}=
156: \exp (-\beta H_\text{S})/Z$ where $H_\text{S}$ denotes the
157: system Hamiltonian and $Z=\Tr[\exp(-\beta H_\text{S})]$
158: is the partition function. Therefore, in clear contrast to the classical case
159: with dissipation, this reduced density matrix becomes also a
160: function of the interaction strength with the environment. This
161: being so, taking a gas of free independent quantum Brownian
162: particles that are coupled to a heat bath with finite dissipation
163: strength, one may speculate that the role of the interactions
164: of the free particle with the abundant bath degrees of freedom will
165: be sufficient to cure the shortcomings stemming from a classical gas
166: of free Brownian particles.
167: 
168: \section{Quantum harmonic oscillator revisited}
169: To start out, we first recall the standard results for a single
170: harmonic quantum oscillator that is infinitely weakly coupled
171: to a bath that establishes the temperature $T$. This
172: situation is reminiscent of the famous treatment of the specific
173: heat of a solid by Albert Einstein in 1907 \cite{einstein07},
174: where he found an exponential suppression of the specific heat
175: as $T \rightarrow 0$. This finding impressed the Berlin
176: school so immensely, in particular Nernst and his collaborators
177: (who in 1910 experimentally confirmed this salient first
178: prediction of quantum theory), that Nernst together with Planck were
179: able to bring the ``new Copernicus'' \cite{ebeling1,ebeling2}
180: into the exclusive circle of Berlin physicists in 1913.
181: 
182: \subsection{Partition function and entropy}
183: \label{pfe}
184: Let us consider an oscillator degree of freedom of mass $M$ and
185: force constant $f$, \ie its Hamiltonian $H_\text{S}$ reads
186: \begin{equation} \label{HO}
187:   H_\text{S} = \frac{p^2}{2M} + \frac{1}{2} fx^2\,.
188: \end{equation}
189: In terms of the angular frequency $\omega_0^2= f/M$ the quantum mechanical
190: energy eigenvalues read $E_n= (n+ \frac{1}{2}) \hbar \omega_0$, yielding
191: for the partition function $Z$ the well-known expression
192: \begin{equation} \label{PHO}
193:   Z= \sum_{n=0}^\infty \e^{-\beta E_n} =
194:   \frac{1}{2 \sinh[\hbar\beta\omega_0/2]}
195: \end{equation}
196: where $\beta=1/k_\text{B}T$ is the inverse temperature.
197: 
198: Using familiar relations we find that the internal energy  $E$ reads
199: \begin{align} \label{HOE}
200:   E &= -\frac{\partial}{\partial\beta}\ln(Z)\\
201:     &= \frac{\hbar\omega_0}{2} +\frac{\hbar\omega_0}{\exp(\hbar\beta\omega_0)-1}
202: \end{align}
203: and, correspondingly, the entropy is given by
204: \begin{align} \label{HOS}
205:   S &= k_\text{B}\left[\ln(Z)-\beta\frac{\partial}{\partial\beta}\ln(Z)\right]\\
206:   &= k_\text{B}\left[\frac{\hbar\beta\omega_0}{\exp(\hbar\beta\omega_0)-1}-
207:        \ln\big(1-\exp(-\hbar\beta\omega_0)\big)\right]\,.
208:   \label{HOSe}
209: \end{align}
210: For low temperatures, $\hbar\beta\omega_0\gg1$, the entropy
211: approaches zero like
212: \begin{equation}
213:   S = \frac{\hbar\omega_0}{T}\exp\left(-\frac{\hbar\omega_0}{k_\text{B}T}\right)\,.
214: \end{equation}
215: 
216: The specific heat can now be derived either from (\ref{HOE}) as
217: \begin{equation}\label{CE}
218:   C=\frac{\partial E}{\partial T}=-k_\text{B}\beta^2\frac{\partial E}{\partial\beta}
219: \end{equation}
220: or from (\ref{HOS}) as
221: \begin{equation}\label{CS}
222:   C=T\frac{\partial S}{\partial T}=-\beta\frac{\partial S}{\partial\beta}\,.
223: \end{equation}
224: In both cases, one obtains for the specific heat
225: \begin{equation}
226:   C=k_\text{B}\left(\frac{\hbar\beta\omega_0}{2\sinh(\hbar\beta\omega_0/2)}\right)^2\,.
227: \end{equation}
228: Its low-temperature behavior
229: \begin{equation}\label{CLT}
230:   C=  k_\text{B} \left(\frac{\hbar\omega_0}{k_\text{B}T} \right)^2
231:   \exp\left(-\frac{\hbar\omega_0}{k_\text{B}T}\right)
232:   \,,
233: \end{equation}
234: is not analytic in temperature and corresponds to Einstein's result
235: for the low-temperature behavior of the specific heat of a solid
236: \cite{einstein07}. For high temperatures one finds
237: \begin{equation}
238:   C = k_\text{B}\left[1-\frac{1}{12}\left(\frac{\hbar\omega_0}
239:   {k_\text{B}T}\right)^2+O(T^{-4})\right]\,.
240: \end{equation}
241: As this result shows, the specific heat for a free particle cannot simply
242: be obtained by taking the limit $\omega_0\to0$ of the harmonic oscillator.
243: Such a procedure will not properly account for the reduced number of degrees of
244: freedom which within the equipartition theorem will lead to a high-temperature 
245: specific heat of only $C=k_\text{B}/2$ for the free particle.
246: 
247: 
248: \section{Quantum harmonic oscillator: The role of quantum dissipation}
249: \subsection{Harmonic oscillator coupled to an environment}
250: 
251: We now couple the harmonic oscillator of the previous section to an
252: environment consisting
253: of an infinite number of harmonic oscillators forming a heat bath. In
254: contrast to the previous section, the coupling strength will not be
255: kept negligible here. In addition, system and bath are infinitely
256: weakly coupled to a superbath which has the purpose to provide
257: the temperature $T$.
258: 
259: The total Hamiltonian $H$ does not need to account for the superbath
260: and therefore consists of three parts
261: \cite{ingoldPR,HTB,dittrich98,chaosHI,mag,connell}
262: \begin{equation} \label{total}
263:   H = H_\text{S}+H_\text{B}+H_\text{SB}
264: \end{equation}
265: where $H_\text{S}$ is given by (\ref{HO}), the bath Hamiltonian reads
266: \begin{equation}
267:   H_\text{B} =
268:   \sum_{i=1}^\infty\left(\frac{p_i^2}{2m_i}+\frac{m_i\omega_i^2}{2}x_i^2\right)
269: \end{equation}
270: and the coupling is bilinear in the coordinates
271: \begin{equation}
272:   H_\text{SB}= -q\sum_{i=1}^{\infty}c_ix_i + q^2\sum\frac{c_i^2}{2m_i\omega_i^2}\,.
273: \end{equation}
274: 
275: We note that quantum systems that are coupled to an environment of finite strength
276: are rarely exactly solvable. The dissipative quantum oscillator  becomes exactly solvable
277: with its bilinear coupling to a bath because of the inherent quadratic structure of the total
278: Hamiltonian in (\ref{total}). This fact holds true even for the case of time-dependent,
279: parametrically driven dissipative quantum harmonic systems \cite{zerbe,kohler}.
280: 
281: In order to describe the influence of the environment on the system
282: oscillator, it is sufficient to know the spectral density of bath
283: oscillators defined by \cite{leggett83,ingoldPR,HTB,grifoni98}
284: \begin{equation}
285:   \label{jom}
286:   J(\omega) = \pi\sum_{i=1}^{\infty}\frac{c_i^2}{2m_i\omega_i}
287:      \delta(\omega-\omega_i)\,.
288: \end{equation}
289: For later purposes, we introduce the Laplace transform of the
290: damping kernel, which generally depends on frequency, thereby
291: causing memory-friction \cite{talkner80,ingold85}, \ie,
292: \begin{equation}
293:   \hat\gamma(z) = \frac{1}{M}\int_0^{\infty}\frac{\text{d}\omega}{\pi}
294:   \frac{J(\omega)}{\omega}\frac{2z}{\omega^2+z^2}\,.
295: \end{equation}
296: The important special case of strictly ohmic dissipation is characterized by
297: $J(\omega)=M\gamma\omega$ and $\hat\gamma(z)=\gamma$ which leads to a
298: memoryless damping of strength $\gamma$.
299: 
300: \subsection{Specific heat of a damped harmonic oscillator}
301: \label{cdho}
302: We next discuss the specific heat of the damped harmonic oscillator
303: by following two routes. First, we  start from the energy $E$ and
304: employ the common relation in (\ref{CE}). As an alternative route
305: we shall in Section~\ref{evspfdho} determine the
306: entropy from the partition function by means of (\ref{HOS}) from which (\ref{CS})
307: allows one to evaluate the specific heat in the case of strictly ohmic damping.
308: 
309: The energy of the damped harmonic oscillator is given by
310: \begin{equation}
311:   \label{EHO}
312:   \langle E\rangle = \frac{\langle p^2\rangle}{2M}+\frac{M}{2}\omega_0^2
313:   \langle q^2\rangle
314: \end{equation}
315: where the expectation value of an operator $O_\text{S}$ acting in the Hilbert
316: space of the system is defined with respect to the canonical density matrix of
317: system plus environment as
318: \begin{equation}
319:   \langle O_\text{S}\rangle = \frac{\Tr\left[O_\text{S}\exp(-\beta H)\right]}
320:   {\Tr\left[\exp(-\beta H)\right]}\,.
321: \end{equation}
322: For ohmic damping, the second moments of position and momentum can
323: be expressed as \cite{ingoldPR,talkner84,riseborough85}
324: \begin{equation}
325:   \langle q^2\rangle=\frac{\hbar}{M}f_0(T)
326: \end{equation}
327: and
328: \begin{equation}
329:   \langle p^2\rangle=M\hbar f_2(T)
330: \end{equation}
331: where we have introduced a temperature-dependent function
332: \begin{equation}
333:   \label{fnT}
334:   f_n(T)=\int_{-\infty}^{+\infty}\frac{\text{d}\omega}{2\pi}
335:   \frac{\gamma\omega^{n+1}}{(\omega^2-\omega_0^2)^2+\gamma^2\omega^2}
336:   \coth\left(\frac{\hbar\beta\omega}{2}\right)\,.
337: \end{equation}
338: For $n=2$, \ie when evaluating $\langle p^2\rangle$, the integrand
339: decreases only with $1/\omega$ and a finite value can only be obtained
340: by introducing a high-frequency cutoff in the damping kernel
341: $\hat\gamma(z)$. However, this divergent term gives rise only to a
342: temperature-independent contribution to $\langle p^2\rangle$ and thus to
343: the energy (\ref{EHO}). When evaluating the specific heat according to
344: (\ref{CE}), this constant term will disappear and a finite result is
345: obtained even for ohmic damping. After some algebra, one finds for the
346: specific heat
347: \begin{equation}
348:   \label{CHO}
349:   \frac{C}{k_\text{B}}=1-\frac{\hbar\beta\gamma}{2\pi}+
350:   \lambda_+^2\psi'\left(1+\lambda_+\right)
351:   +\lambda_-^2\psi'\left(1+\lambda_-\right)
352: \end{equation}
353: where
354: \begin{equation}
355:   \label{lambda}
356:   \lambda_\pm = \frac{\hbar\beta\omega_0}{2\pi}\left[
357:   \frac{\gamma}{2\omega_0}\pm\sqrt{
358:   \left(\frac{\gamma}{2\omega_0}\right)^2-1}\right]
359: \end{equation}
360: and $\psi'(z)$ is the trigamma function. At low temperatures, the
361: specific heat thus assumes the form
362: \begin{equation}
363:   \frac{C}{k_\text{B}}=
364:   \frac{\pi}{3}\frac{\gamma}{\omega_0}\frac{k_\text{B}T}{\hbar\omega_0}+
365:   \frac{4\pi^3}{15}\frac{\gamma}{\omega_0}\left[3-\left(\frac{\gamma}{\omega_0}
366:   \right)^2\right]\left(\frac{k_\text{B}T}{\hbar\omega_0}\right)^3+O(T^5)\,.
367: \end{equation}
368: This result differs significantly from the expression (\ref{CLT}) in
369: the absence of dissipation. While in the latter case, the presence
370: of an energy gap led to an exponential suppression of the specific
371: heat, we now find a linear increase with temperature. This behavior indicates
372: the existence of a finite density of states even at small excitation energies \cite{hanke}. Even
373: at high temperatures the effect of dissipation can be detected, albeit in
374: a less spectacular manner. The leading correction in the
375: high-temperature expansion reads
376: \begin{equation}
377:   \frac{C}{k_\text{B}}=1-\frac{\hbar\gamma}{2\pi k_\text{B}T}+
378:   \frac{\hbar^2(\gamma^2-2\omega_0^2)}{24(k_\text{B}T)^2}+O(T^{-3})
379:   \,.
380: \end{equation}
381: Thus, the leading correction depends on the damping
382: strength $\gamma$. This finding is in clear contrast to the behavior
383: of the quantum escape rate \cite{HTB,ingold85}: There, the leading
384: quantum correction to the escape rate always enhances the classical
385: result and is {\it independent} of the dissipation strength.
386: 
387: \subsection{Energy versus partition function for a damped harmonic
388: oscillator}
389: \label{evspfdho}
390: 
391: Another prescription to obtain the specific heat starts out from
392: the canonical partition function
393: \begin{equation}
394:   \label{ZHO}
395:   Z = \frac{\Tr\left[\exp(-\beta H)\right]}
396:   {\Tr_B\left[\exp(-\beta H_\text{B})\right]}\,,
397: \end{equation}
398: where $\Tr_B$ denotes the partial trace in the Hilbert space of the bath.
399: In the absence of a system-bath coupling, this expression would correspond
400: to the partition function of the system alone. For a damped harmonic
401: oscillator, the partition function becomes \cite{talknerFN}
402: \begin{equation}
403:   \label{ZHOe}
404:   Z = \frac{1}{\hbar\beta\omega_0}\prod_{n=1}^\infty \frac{\nu_n^2}
405:   {\nu_n^2+\nu_n\hat\gamma(\nu_n)+\omega_0^2}
406: \end{equation}
407: with the Matsubara frequencies $\nu_n=2\pi n/\hbar\beta$. In view of the
408: divergence for strictly ohmic damping mentioned above, we allow here for
409: a possible frequency dependence of $\hat\gamma$.
410: 
411: Following the standard procedure of statistical mechanics, we can obtain the
412: energy from the partition function by means of
413: \begin{equation}
414:   \label{EZZ}
415:   \langle E\rangle_Z = -\frac{\partial}{\partial\beta}\ln(Z)\,.
416: \end{equation}
417: Inserting (\ref{ZHOe}), one obtains
418: \begin{equation}
419:   \langle E\rangle_Z = \frac{1}{\beta}\left[1+\sum_{n=1}^\infty
420:   \frac{2\omega_0^2+\nu_n\hat\gamma(\nu_n)-\nu_n^2\hat\gamma'(\nu_n)}{\nu_n^2+
421:   \nu_n\hat\gamma(\nu_n)+\omega_0^2}\right]
422: \end{equation}
423: which in general differs from the expression
424: \begin{equation}
425:   \langle E\rangle = \frac{1}{\beta}\left[1+\sum_{n=1}^\infty
426:   \frac{2\omega_0^2+\nu_n\hat\gamma(\nu_n)}{\nu_n^2+
427:   \nu_n\hat\gamma(\nu_n)+\omega_0^2}\right]
428: \end{equation}
429: obtained by evaluating the integral (\ref{fnT}) by residues. The only
430: exception is the special case of strictly ohmic damping where
431: $\hat\gamma(\nu_n)=\gamma$ is constant.
432: 
433: This generally non-vanishing difference does not come as a
434: surprise if one only takes a closer look at
435: the partition function (\ref{ZHO}): The  evaluation of (\ref{EZZ}) yields
436: \begin{equation}
437: \begin{aligned}
438:   \label{EZvsE}
439:   \langle E\rangle_Z &= \langle H\rangle-
440:                         \langle H_\text{B}\rangle_\text{B}\\
441:                      &= \langle E\rangle + [\langle H_\text{SB}\rangle
442:                          +\langle H_\text{B}\rangle
443:                          -\langle H_\text{B}\rangle_\text{B}]
444: \end{aligned}
445: \end{equation}
446: where the index ``B'' denotes an average with respect to the bath
447: Hamiltonian $H_\text{B}$ only. This result differs from
448: the energy $\langle E\rangle=\langle H_\text{S}\rangle$ by the
449: term in the brackets which, generally, vanishes only in the absence of a
450: system-bath coupling. The coincidence between $\langle E\rangle$ and
451: $\langle E\rangle_Z$ for the harmonic oscillator subject to strictly
452: ohmic damping should therefore be considered as exceptional.
453: 
454: Nevertheless, we briefly sketch how one would obtain the specific heat from
455: the partition function because this will give us as a by-product an expression
456: for the entropy of the damped harmonic oscillator. For strictly ohmic damping
457: the product (\ref{ZHOe}) does not converge and in principle a high-frequency
458: cutoff for $\hat\gamma$ should be introduced. However, this divergence can
459: again be traced back to an infinite energy shift due to the environmental
460: coupling. We may shift the energy by an arbitrary amount $\Delta$ by
461: multiplying the partition function by $\exp(-\beta\Delta)$ without changing the
462: entropy or the specific heat. After performing an appropriate energy shift, we
463: arrive at an expression of the partition function valid even for strictly
464: ohmic damping \cite{lnp}, reading
465: \begin{equation}
466:   \bar Z = \frac{A}{\hbar\beta\omega_0}\left(\frac{2\pi}{\hbar\beta\omega_0}
467:   \right)^{\hbar\beta\gamma/2\pi}\Gamma\left(1+\lambda_+\right)
468:   \Gamma\left(1+\lambda_-\right) \,,
469: \end{equation}
470: where $\Gamma(z)$ is the gamma function, $\lambda_\pm$ are defined in
471: (\ref{lambda}), and $A$ is a constant whose precise
472: value is irrelevant for the following. By virtue of (\ref{HOS}) we
473: obtain for the entropy the result
474: \begin{equation}
475:   \label{SDHO}
476:   S=k_\text{B}\left[1-\ln(\hbar\beta\omega_0) +
477:   \frac{\hbar\beta\gamma}{2\pi}+
478:   g\left(\lambda_+\right) +
479:   g\left(\lambda_-\right)\right] \,,
480: \end{equation}
481: where we introduced the abbreviation
482: \begin{equation}
483:   g(z) = \ln[\Gamma(1+z)]-z\psi(1+z)
484: \end{equation}
485: with the digamma function $\psi(z)=\Gamma'(z)/\Gamma(z)$. In the absence of
486: damping, \ie $\gamma=0$, this reproduces the  result (\ref{HOSe}) for the
487: entropy of an uncoupled harmonic oscillator.
488: 
489: For very low temperatures, the entropy (\ref{SDHO}) vanishes like
490: \begin{equation}
491:   \label{Slt}
492:   S=\frac{\pi}{3}\frac{\gamma}{\omega_0}\frac{k_\text{B}^2T}{\hbar\omega_0}
493:   +O(T^3)
494: \end{equation}
495: as required by the Third Law of thermodynamics. By means of (\ref{CS}), the 
496: expression (\ref{CHO}) for the specific heat is recovered identically.
497: 
498: The low-temperature behavior (\ref{Slt}) is in agreement with the
499: expression derived by Ford and O'Connell on the basis of the free energy
500: \cite{ford05}. For the damped harmonic oscillator these authors 
501: found that the entropy vanishes in the limit of zero temperature also for 
502: more general forms of the bath density of states.
503: 
504: \section{Free quantum Brownian motion coupled to a heat bath: Is the
505: Third Law obeyed at zero temperature?}
506: 
507: According to the equipartition theorem the specific heat of a free
508: particle is $k_\text{B}/2$. In the limit of an infinite ``box'' this
509: represents even the correct quantum value, because the classical
510: and the quantum partition function become equal. In view of the fact that
511: the specific heat of an ideal gas thus remains non-zero down to the lowest
512: temperatures, one becomes curious to investigate the specific heat
513: of a free particle coupled to an environment when the coupling
514: strength is not assumed to vanish: Does quantum dissipation help to
515: restore the Third Law?
516: 
517: Free quantum Brownian motion has been addressed in
518: earlier work \cite{aslangul85,hakim85,gsi87,schramm87} wherein the main
519: focus centered on the role of free quantum diffusion
520: \cite{ingoldPR}. Interestingly enough, for ohmic dissipation the
521: quantum diffusion remains ``classical'', being proportional to time
522: $t$, except at zero temperature itself, where one finds a
523: logarithmic behavior in time $t$ \cite{aslangul85,hakim85,gsi87,schramm87}. At
524: finite temperatures this quantum behavior is observable at
525: intermediate times only \cite{aslangul85,jung}. One is thus tempted to 
526: conclude that
527: finite quantum dissipation will not be sufficient to cure the classical
528: behavior for the specific heat. Therefore, we shall next investigate
529: the behavior of the specific heat for free quantum Brownian motion in closer detail.
530: 
531: As already mentioned at the end of Section~\ref{pfe}, simply taking
532: the limit $\omega_0\to0$ in the results for the damped harmonic
533: oscillator is not without problems.  We therefore start our calculation
534: from the energy
535: \begin{equation}
536:   \langle E\rangle = \frac{\langle p^2\rangle}{2M}=
537:   \frac{1}{2\beta}\left[1+2\sum_{n=1}^\infty
538:   \frac{\nu_n\hat\gamma(\nu_n)}{\nu_n^2+\nu_n\hat\gamma(\nu_n)}\right]
539: \end{equation}
540: which can be obtained from the second moment of momentum $\langle
541: p^2\rangle$ by evaluating the integral (\ref{fnT}) by residues. Proceeding
542: as in Section~\ref{cdho}, one derives the specific heat of the free
543: particle in the presence of strictly ohmic damping
544: \begin{equation}
545:   \label{CFPo}
546:   \frac{C}{k_\text{B}}=\frac{1}{2}-\frac{\hbar\beta\gamma}{2\pi}
547:   +\left(\frac{\hbar\beta\gamma}{2\pi}\right)^2\psi'\left(1+
548:   \frac{\hbar\beta\gamma}{2\pi}\right)
549: \end{equation}
550: which for either $T\to\infty$ or $\gamma\to0$ yields $C=k_\text{B}/2$, as
551: expected from the equipartition theorem. On the other hand, for low
552: temperatures the specific heat tends to zero as
553: \begin{equation}
554:   \label{CFPlin}
555:   \frac{C}{k_\text{B}}=\frac{\pi}{3}\frac{k_\text{B}T}{\hbar\gamma}-
556:   \frac{4\pi^3}{15}\left(\frac{k_\text{B}T}{\hbar\gamma}\right)^3+O(T^5)
557: \end{equation}
558: in agreement with the Third Law of thermodynamics. The specific heat
559: (\ref{CFPo}) together with its linear low-temperature behavior are depicted
560: in the main part of Fig.~\ref{cfree} as full and dashed line, respectively.
561: Temperatures $k_\text{B}T\gg\hbar\gamma$ much larger than the damping
562: strength are required in order to restore the classical result
563: $C/k_\text{B}=1/2$.
564: 
565: \begin{figure}
566:   \centerline{\includegraphics[width=0.8\textwidth]{cfree.eps}}
567:   \caption{The specific heat $C$ of a free quantum Brownian particle is shown as a
568:   function of the temperature $T$ for strictly ohmic friction of strength
569:   $\gamma$. The dashed line indicates the linear low-temperature behavior.
570:   In the inset, the modification due to a finite cutoff frequency
571:   $\omega_\text{D}$ is depicted for different cutoff scales, $\omega_\text{D}/\gamma=0.01,0.1,1,
572:   \text{and }\infty$ from the upper to the lower full line. The dashed line
573:   indicates again the linear low-temperature behavior.}
574:   \label{cfree}
575: \end{figure}
576: 
577: As the damping strength already serves to set the temperature scale, it is
578: instructive to introduce a cutoff in the density of states (\ref{jom}) of the bath
579: oscillators in order to study how a reduction of the environmental
580: influence changes the dependence of the specific heat on temperature.
581: To this end, we introduce a high-frequency cutoff $\omega_\text{D}$ by
582: choosing the Drude model where
583: \begin{equation}
584:   \hat\gamma(z)=\frac{\gamma\omega_\text{D}}{z+\omega_\text{D}}\,.
585: \end{equation}
586: This environment leads to a specific heat
587: \begin{equation}
588:   \frac{C}{k_\text{B}}=\frac{1}{2}-\frac{\hbar\beta\gamma}{2\pi}
589:   \frac{1}{\sqrt{1-4\gamma/\omega_\text{D}}}\left[z_+\psi'(1+z_+)
590:   -z_-\psi'(1+z_-)\right] \,,
591: \end{equation}
592: where we have introduced the abbreviations
593: \begin{equation}
594:   z_\pm = \frac{\hbar\beta\omega_D}{4\pi}
595:   \left(1\pm\sqrt{1-\frac{4\gamma}{\omega_\text{D}}}\right)\,.
596: \end{equation}
597: Independently of the value of the cutoff frequency $\omega_\text{D}$, the
598: specific heat will go to zero linearly as stated in (\ref{CFPlin}).
599: However, for temperatures larger than the cutoff frequency, the suppression
600: of the specific heat due to the environmental coupling will be ineffective.
601: A reduction of the environment thus tends to restore the classical value
602: of $k_\text{B}/2$ for the specific heat. This effect is depicted in the inset
603: in Fig.~\ref{cfree}, where the cutoff frequency takes the values
604: $\omega_\text{D}/\gamma=0.01, 0.1, 1$ and $\infty$ from the upper to the
605: lower curve. The dashed curve represents the linear low-temperature
606: behavior which still dominates at temperatures below
607: $\hbar\omega_\text{D}/k_\text{B}$. In contrast to many phenomena in quantum
608: dissipation, here an increase of the coupling to the environment does not render the system
609: more classical. On the contrary, a stronger environmental coupling  makes the
610: dissipative quantum system behave more quantum mechanically and thus helps to
611: ensure the validity of the Third Law of thermodynamics,
612: \ie the vanishing of the specific heat with decreasing temperatures.
613: 
614: \section{Conclusions}
615: With this work we have explored the behavior of the specific heat for a
616: quantum system that is coupled to a heat bath with finite coupling
617: strength. Our findings are contrasted with the Third Law of
618: thermodynamics which generically predicts a vanishing of the
619: specific heat at low temperatures. For a harmonic oscillator the
620: presence of quantum dissipation  changes the well-known
621: Einstein-like behavior of an exponentially fast approach towards
622: zero specific heat into a power-law behavior with a
623: slope that increases with increasing coupling strength.
624: Even more intriguing is the behavior for a freely moving quantum
625: particle: While the quantum treatment in absence of dissipation
626: simply coincides with the classical behavior, \ie the specific
627: heat takes a constant value $C= k_B/2$, the role of finite quantum
628: dissipation is able to restore the Third Law, yielding a leading
629: linear temperature dependence. Quite counterintuitively,
630: its approach to the classical value occurs the faster the weaker
631: is the dissipation strength.
632: 
633: In contrast to common quantum statistical mechanics which
634: intrinsically is based on a vanishingly small coupling to the
635: environment, the finite coupling strength between the sub-system and
636: the bath causes some subtleties that must be recognized. As
637: made explicit in Section~\ref{evspfdho}, the thermodynamic
638: quantities depend on the procedure invoked in their definition:
639: the commonly used expression based on the partition function provides
640: results that generally do \textit{not} agree with the result obtained
641: from the corresponding quantum expectation value.  Interestingly
642: enough, in the strict ohmic limit (\ie in the absence of a high-frequency
643: cutoff) the specific heat for the damped harmonic oscillator
644: does not depend on the prescription employed. For the case of memory
645: friction, however, where a finite cutoff frequency is present, the
646: thermodynamic quantities depend both on the value of this cutoff
647: and the prescription used in their evaluation.
648: 
649: The results obtained for the low-temperature behavior of the specific heat of
650: simple quantum systems are not only of academic interest, but may turn out to
651: be relevant for experiments in nanoscience where one tests the quantum 
652: thermodynamics of small systems \cite{fqmt} that are coupled to an environment 
653: with a finite coupling strength.
654: 
655: \section*{Acknowledgment}
656: 
657: This work has been supported by  the Deutsche Forschungsgemeinschaft
658: (DFG)  (PH, SFB 486). Both authors like to congratulate Peter
659: Talkner for his first 60 years and his fine scientific career. We
660: both have heavily profited repeatedly and continue to strongly
661: profit from his insight and breadth of knowledge. May our present
662: ongoing  fruitful collaborations with him blossom further and even
663: intensify.
664: 
665: 
666: \begin{thebibliography}{99}
667: 
668: \bibitem{chaos05}
669: P. H\"anggi and F. Marchesoni, Chaos {\bf 15}, 026101 (2005).
670: 
671: \bibitem{physicaSI}
672: P. H\"anggi, A. Alvarez-Chillida, and M. Morillo, Physica A {\bf 351}, XI (2005); special issue.
673: 
674: \bibitem{TD1} C. J. Thomson, {\it Classical Equilibrium Statistical Mechanics}
675:  (Oxford University Press, 1988); chapter I.
676: 
677: \bibitem{TD2} H. B. Callen,
678: {\it Thermodynamics and an Introduction to Thermostatics}, second
679: edition (John Wiley, 1985).
680: 
681: \bibitem{ebeling1}
682: W. Ebeling, in: {\it Zur Gro{\ss}en Berliner Physik} (B. G. Teubner,
683: Leipzig, 1987).  See: ``Zur Geschichte der Thermodynamik in Berlin'',
684: pp: 71-88;  and private communication.
685: 
686: \bibitem{ebeling2}
687: W. Ebeling and D. Hoffman, Eur. J. Phys. {\bf 12}, 1 (1991).
688: 
689: \bibitem{Nernst}
690: W. Nernst, Nachr. Kgl. Ges. d. Wiss. G\"ottingen {\bf 6}, 1 (1906);
691: see also: W. Nernst,  Sitzungsber. Preuss. Akad. Wiss. {\bf
692: 13},  311 (1911).
693: 
694: \bibitem{Dugdale}
695: J. S. Dugdale, {\it Entropy and its Physical Meaning} (Taylor and
696: Francis, 1996).
697: 
698: \bibitem{Wilks}
699: J. Wilks, {\it The Third Law of Thermodynamics} (Oxford University Press, 1961).
700: 
701: \bibitem{planck}
702: M. Planck, {\it Vorlesungen {\"u}ber Thermodynamik}
703:   (Veit \& Comp.,  1917).
704: 
705: \bibitem{nernst1912}
706: W. Nernst, Sitzungsber. Preuss. Akad. Wiss. 134 (1912).
707: 
708: \bibitem{Leggett}
709: A. J. Leggett, Ann. Phys. (N.Y.) {\bf  72}, 80 (1972).
710: 
711: \bibitem{leggett83}
712: A. O. Caldeira and A. J. Leggett,  Ann. Phys. (N.Y.) {\bf 149},
713: 374 (1983).
714: 
715: \bibitem{rise85}
716: P. S. Riseborough, P. H{\"a}nggi, and E. Freidkin, Phys. Rev. A
717: \textbf{32}, 489 (1985).
718: 
719: \bibitem{ingoldPR}
720: H. Grabert, P. Schramm, and G.-L. Ingold, Phys. Rep. {\bf 168}, 115
721: (1988).
722: 
723: \bibitem{HTB}
724: P. H\"anggi, P. Talkner, and M. Borkovec, Rev. Mod. Phys.  {\bf 62},
725: 251 (1990).
726: 
727: \bibitem{schon90}
728: G. Sch\"on and A. D. Zaikin, Phys. Rep. {\bf 198}, 237 (1990).
729: 
730: \bibitem{dittrich98}
731: T. Dittrich, P. H\"anggi. G.-L. Ingold, B. Kramer, G. Sch\"on, and
732: W. Zwerger, \textit{Quantum Transport and Dissipation} (Wiley-VCH,
733:  1998).
734: 
735: \bibitem{grifoni98}
736: M. Grifoni and P. H\"anggi,  Phys. Rep. \textbf{304}, 229 (1998).
737: 
738: \bibitem{weiss99}
739: U. Weiss, \textit{Quantum Dissipative Systems}, second edition
740: (World Scientific,  1999).
741: 
742: \bibitem{talkner04}
743: K. M. F. Romero, P. Talkner, and P. H\"anggi, Phys. Rev. A {\bf 69}, 052109 (2004).
744: 
745: \bibitem{chaosHI}
746: P. H\"anggi and G.-L. Ingold, Chaos {\bf 15}, 026105 (2005).
747: 
748: \bibitem{einstein07}
749: A. Einstein, Ann. Phys. (Leipzig) {\bf 22}, 180  (1907); A.
750: Einstein, Ann. Phys. (Leipzig) {\bf 22}, 800 (1907), erratum; A.
751: Einstein, Ann. Phys. (Leipzig) {\bf 35}, 679 (1911).
752: 
753: \bibitem{mag}
754: V. B. Magalinski\u\i, Sov. Phys. JETP {\bf 9}, 1381 (1959) [J. Exp.
755: Theor. Phys. {\bf 36}, 1942 (1959)].
756: 
757: \bibitem{connell}
758: G. W. Ford, J. T. Lewis, and R. F. O'Connell, Phys. Rev.  A {\bf 37},
759: 4419 (1988).
760: 
761: \bibitem{zerbe}
762: C. Zerbe and P. H\"anggi, Phys. Rev.  E {\bf 52}, 1533 (1995).
763: 
764: \bibitem{kohler}
765: S. Kohler, T. Dittrich and P. H\"anggi, Phys. Rev. E {\bf 55}, 300 (1997).
766: 
767: \bibitem{talkner80}
768: H. Grabert, P. H\"anggi, and P. Talkner, J. Stat. Phys. {\bf 22}, 537 (1980).
769: 
770: \bibitem{ingold85}
771: P. Hanggi, H. Grabert, G.-L. Ingold, and U. Weiss,  Phys. Rev. Lett.
772: \textbf{55}, 761 (1985).
773: 
774: \bibitem{talkner84}
775: H. Grabert, U. Weiss, and P. Talkner, Z. Phys. B \textbf{55}, 87
776: (1984).
777: 
778: \bibitem{riseborough85}
779: P. S. Riseborough, P. H\"anggi, and U. Weiss, Phys. Rev. A
780: \textbf{31}, 471 (1985).
781: 
782: \bibitem{hanke}
783: A. Hanke and W. Zwerger, Phys. Rev. E \textbf{52}, 6875 (1995).
784: 
785: \bibitem{talknerFN}
786: The authors of reference \cite{talkner84} introduced two different
787: partition functions for a damped harmonic oscillator, see their
788: Eqs.~(4.19) and (5.11). The one which corresponds to our Eq.~(\ref{ZHOe})
789: is denoted as $Z'$ and is defined in their Eq.~(4.19).
790: 
791: \bibitem{lnp}
792: G.-L. Ingold, Lect. Notes Phys. \textbf{611}, 1 (2002).
793: 
794: \bibitem{ford05}
795: G. W. Ford and R. F. O'Connell, Physica E \textbf{29}, 82 (2005).
796: 
797: \bibitem{aslangul85}
798: C. Aslangul, N. Pottier, and D. Saint-James, J. Stat. Phys.
799: \textbf{40}, 167 (1985).
800: 
801: \bibitem{hakim85}
802: V. Hakim and V. Ambegaokar, Phys. Rev. A \textbf{32}, 423 (1985).
803: 
804: \bibitem{gsi87}
805: H. Grabert, P. Schramm, and G.-L. Ingold, Phys. Rev. Lett. \textbf{58},
806: 1285 (1987).
807: 
808: \bibitem{schramm87}
809: P. Schramm and H. Grabert, J. Stat. Phys. \textbf{49}, 767 (1987).
810: 
811: \bibitem{jung}
812: R. Jung, G.-L. Ingold, and H. Grabert, Phys. Rev. A \textbf{32},
813: 2510 (1985).
814: 
815: \bibitem{fqmt}
816: V. {\v S}pi{\v c}ka, Th. M. Nieuwenhuizen, P. D. Keefe, Physica E
817: \textbf{29}, 1 (2005).
818: 
819: \end{thebibliography}
820: \end{document}
821: