1: \documentclass[12pt]{article}
2:
3: \usepackage{amssymb,amsmath,epsfig}
4:
5: \DeclareMathOperator{\sign}{sign}
6:
7: \newcommand{\pd}{\partial}
8: \newcommand{\vect}[1]{\mathbf{#1}}
9:
10: \title{Energy Levels of ``Hydrogen Atom'' in Discrete Time Dynamics}
11: \date{}
12: %\classification{03.65.-w,03.65.Bz,03.65.Ca,03.75.-b}
13: %\keywords{discrete time dynamics, quantum mechanics, Hydrogen atom,quantum interference}
14:
15: \author{Andrei Khrennikov\thanks{Andrei.Khrennikov@msi.vxu.se} and Yaroslav Volovich\thanks{Yaroslav.Volovich@msi.vxu.se}\\
16: ~\\
17: International Center for Mathematical Modeling\\
18: in Physics, Engineering and Cognitive science\\
19: MSI, V\"axj\"o University, S-35195, Sweden
20: }
21:
22: \begin{document}
23:
24: \maketitle
25: \begin{abstract}
26: We analyze dynamical consequences of a conjecture that there exists
27: a fundamental (indivisible) quant of time. In particular we study
28: the problem of discrete energy levels of hydrogen atom. We are able
29: to reconstruct potential which in discrete time formalism leads to
30: energy levels of unperturbed hydrogen atom. We also consider linear
31: energy levels of quantum harmonic oscillator and show how they are
32: produced in the discrete time formalism. More generally, we show
33: that in discrete time formalism finite motion in central potential
34: leads to discrete energy spectrum, the property which is common for
35: quantum mechanical theory. Thus deterministic (but discrete time!)
36: dynamics is compatible with discrete energy levels.
37: \end{abstract}
38:
39: \maketitle
40:
41: \section{Introduction}
42:
43: Discovery of discrete energy levels for atoms demonstrated that the
44: classical Newtonian model could not be used to describe this
45: phenomenon. Now days this phenomenon is described in the framework
46: of quantum mechanics. The advantages of the computational methods of
47: quantum mechanics are well known. They are widely used for
48: computation of energy levels not only of atoms, but also in
49: essentially more complicated situations. However, despite this
50: computational power, quantum mechanics induced many still
51: unsolvable phenomenological problems, see \cite{Heis}-\cite{Vax4}
52: for extended discussions. One of distinguishing features of quantum
53: mechanics (at least in Copenhagen interpretation) is the
54: impossibility to provide realist deterministic description of
55: quantum reality. In particular, here particles do not have well
56: defined trajectories.
57:
58: The impossibility to use the deterministic evolutionary model looks
59: rather counterintuitive. Moreover, it contradicts to many quantum
60: experiments where "trajectories" of particles are well observed,
61: e.g. in Wilson's camera. Of course, in modern quantum phenomenology
62: this problem is solved via the principle of complementarity.
63: Nevertheless, this strong deviation from our intuitive picture of
64: physical reality induces a rather general opinion that quantum
65: mechanics has a lot of mysteries \cite{FeyGib}.
66:
67: This unsatisfactory status of quantum mechanics induces new and new
68: attempts to create new models that would be closer to our physical
69: intuition. We just mention one model, Bohmian mechanics
70: \cite{BH,Hol}.
71:
72: Last years there were performed intensive investigations to
73: reconsider probabilistic foundations of quantum mechanics, see e.g.
74: \cite{Vax3,Vax4}. These investigations are of the great importance.
75: It is well known that "quantum probability" differs strongly from
76: "classical probability". Typically, see e.g. \cite{FeyGib}, it is
77: pointed out that quantum randomness is irreducible and fundamental
78: (in the opposite to classical randomness that could be reduced to
79: e.g. randomness of initial conditions). This irreducible quantum
80: randomness is deeply connected with the impossibility of realist
81: deterministic models that would reproduce quantum probabilities. It
82: seems to be impossible to imagine that quantum particles have
83: deterministic trajectories, since the existence of trajectories
84: should imply classical probabilistic rules. But these rules are
85: violated (for example, in the two-slit experiment \cite{FeyGib},
86: see also investigations on EPR-Bohm-Bell consideration and chameleon effect \cite{Bell,Acc,AIR}).
87:
88: In papers \cite{Kh,Kh5,Kh6,Kh8} there was developed the contextual
89: approach to quantum probabilities that might be used to explain the
90: origin of quantum randomness in the classical (but contextual!)
91: probabilistic framework. Therefore the probabilistic constraint to
92: create realist deterministic models need not be taken into
93: account\footnote{Of course, there are also Bell`s inequality
94: arguments. But there arguments are strongly coupled to nonlocality.
95: And this problem is far from problems considered in the present
96: paper.}.
97:
98: What kind of realist deterministic models could be created for
99: quantum phenomena?
100:
101: As we have already mentioned, we could not directly use Newtonian
102: mechanics even for the simplest quantum system -- the hydrogen atom.
103: Thus Newton equations should be changed. The most straightforward
104: idea is to change classical forces (in the case of atom Coulomb's
105: law) and find new forces, quantum forces. Such forces should "drive"
106: particles along trajectories that reproduce quantum data. There is a
107: rather common view point that one of successful realizations of such
108: a program is given by Bohmian mechanics. However, this problem is
109: not simple, since Bohmian mechanics is, in fact, not mechanics, but
110: a field model. There exists an additional field equation for quantum
111: potential. We do not know any model that would give the realization
112: of the discussed program of "modernization" of Newtonian mechanics.
113:
114: In the present paper we consider discrete time Newtonian model. This
115: is a kind of classical physical model (in particular, realist
116: deterministic). The only difference is discreteness of time. Thus we
117: use the classical force -- interaction picture, but the discrete
118: time version of Newton's second law. One of the main advantages of
119: this model is its simplicity. We do not change phenomenology of
120: classical physics (position, velocity, force-interaction,
121: trajectories). We only change the mathematical representation of
122: time.
123:
124: On the other hand, discreteness is the main distinguishing feature
125: of quantum physics. In fact, M. Planck and A. Einstein obtained
126: Wien's law simply by assuming discreteness of energy. However, this
127: discreteness approach was not developed further to get a formalism
128: based only on the discreteness postulate. Discreteness of quantum
129: observables was reproduced by using an advanced mathematical
130: formalism based on the representation of the observables in the
131: complex Hilbert space. As we have already mentioned, the use of this
132: formalism (despite its computational advantages) induced many
133: phenomenological problems. We are trying to modify classical physics
134: by starting with one natural postulate:
135:
136: \medskip
137: \textbf{\underline{TD:}~~``Time is discrete''.}
138:
139: \medskip
140: Here, in particular, continuous Newton equations (differential
141: equations) are just approximations of real equations of motions,
142: namely, Newton's difference equations\footnote{We reverse the modern
143: viewpoint on the description of physical reality. Not (discrete)
144: difference equations are used to approximate (continuous)
145: differential equations, but the inverse!}. In the present paper we
146: use the discrete time model for the hydrogen atom. The discrete-time
147: postulate implies discrete orbits and energy levels. At the same
148: time we have deterministic motion along orbits. In the limit we get
149: continuous motion along circular orbit (a kind of Bohr's
150: correspondence principle).
151:
152: In the present paper we do not try to develop some statistical
153: theory (a kind of Born approach). Our investigation has some
154: similarities with the original paper of W.~Heisenberg (that was not
155: statistical one), \cite{Heis,Heis2}. We hope that starting with only
156: discrete time postulate we would be able to develop simpler
157: formalism to calculate spectra of quantum observables - without
158: using noncommutative calculus and without a cardinal change of the
159: phenomenology of classical physics.
160:
161: It is interesting to point that there is some similarity between our
162: approach and G. 't Hooft's approach \cite{Hooft1,Hooft2,Hooft3}
163: where a general scheme was proposed that maps states of quantum
164: field system to the states of a completely deterministic field
165: model. Although in this note we do not consider the field-particle
166: duality it would be interesting to study this problem.
167:
168: Some statistical consequences of the postulate (TD) were
169: investigated in our previous papers \cite{KV1,KV2,KV3}.
170:
171: \section{Discrete Time Dynamics}
172: \label{sec:ddyn}
173:
174: In classical mechanics a dynamical function
175: $A=A(p,q)$ (here $p$ and $q$ are momenta and coordinates
176: of the system) evolves according to the following well known equation
177: \cite{Dir}
178: \begin{equation}
179: \label{dyneq} D_t A=\{A,H\}
180: \end{equation}
181: where $H=H(p,q)$ is a Hamiltonian of the system and in the right
182: hand side is a Poisson bracket, which could be presented as
183: \begin{equation}
184: \label{cposs} \{A,B\}=\frac{\pd A}{\pd q}\frac{\pd B}{\pd p} -
185: \frac{\pd A}{\pd p}\frac{\pd B}{\pd q}
186: \end{equation}
187: The left hand side of (\ref{dyneq}) contains a continuous time derivative
188: $$
189: D_t A = \frac{dA}{dt}
190: $$
191: As it was mentioned earlier we are interested in construction
192: dynamics with discrete time. This is done with the help of
193: \textit{discrete derivative} which is postulated to be
194: $$
195: D^{(\tau)}_t A = \frac{1}{\tau}[A(t+\tau)-A(t)],
196: $$
197: where $\tau$ is the discreteness parameter. This parameter is finite
198: and is treated in the same way as Plank constant in quantum
199: mechanical formalism. In particular if $\tau$ is small relative to
200: dimensions of the system then classical approximation with
201: continuous derivative might work well (although this could not be
202: the case all the time in the same sense as there are examples when
203: quantum formalism is reasonable even for macroscopic systems, for
204: example in superfluidity).
205:
206: Summarizing, the discrete time dynamical equation is postulated to
207: be
208: \begin{equation}
209: \label{dyneq-post} D^{(\tau)}_t A=\{A,H\},
210: \end{equation}
211: where $A(p,q)$ is a real-valued function of real-valued dynamical
212: variables and in the right hand side there is classical Poisson
213: bracket (\ref{cposs}). The equation (\ref{dyneq-post}) could be
214: solved in the sense that we can write
215: \begin{equation}
216: \label{dyn-sol} A(t+\tau)=A(t)+\tau \{A,H\}
217: \end{equation}
218: thus providing the evolution of any dynamical function $A=A(p,q)$.
219:
220: Note that in our model the coordinate space is continuous.
221:
222: \section{Motion in Central Potential}
223:
224: \begin{figure}
225: \centering \epsfig{file=datom.eps,width=12cm} \caption{First three
226: trajectories.} \label{fig:traj}
227: \end{figure}
228:
229: Here we will study the properties of motion in central potential
230: $U=U(r)$ in discrete time formalism. As we will see discreteness of
231: time enriches mechanics with some new properties which are usually
232: thought as having quantum nature. In particular as it will be shown
233: below in discrete time mechanics stationary orbits (i.e. finite
234: motion) have discrete energy spectrum. We point that the phase space
235: is assumed here to be a continuous real manifold.
236:
237: Following the general approach described in the previous section we
238: start from the classical Hamiltonian and then write the dynamical
239: equations. In polar coordinates $(r,\varphi)$ the Hamiltonian of the
240: system with mass $m$ in central potential $U(r)$ is given by
241: \begin{equation}
242: \label{HamiltU}
243: H(r,p_r,\varphi,p_\varphi)=\frac{p_r^2}{2m}+\frac{p_\varphi^2}{2mr^2}+U(r),
244: \end{equation}
245: where $p_r$ and $p_\varphi$ denote momenta corresponding to $r$ and
246: $\varphi$ -- radial and angular coordinates respectively. Using
247: (\ref{dyn-sol}) let us write the dynamical equations. We obtain
248: \begin{align}
249: \label{r-dyn}
250: r(t+\tau)&=r(t)+\tau \frac{p_r}{m}\\
251: \label{pr-dyn}
252: p_r(t+\tau)&=p_r(t)+\tau \left(\frac{p_\varphi^2}{mr^3}-\frac{\pd U}{\pd r}\right)\\
253: \label{phi-dyn}
254: \varphi(t+\tau)&=\varphi(t)+\tau \frac{p_\varphi}{mr^2}\\
255: \label{pphi-dyn} p_\varphi(t+\tau)&=p_\varphi(t)
256: \end{align}
257: The equation (\ref{pphi-dyn}) corresponds to angular momentum
258: conservation in central field -- this is a direct analog of the
259: angular momentum conservation law in classical (continuous time)
260: dynamics and is a consequence of the fact that our Hamiltonian does
261: not depend on $\varphi$, (it is a so called \textit{cyclic
262: variable}).
263:
264: Let us limit ourself to circular stationary periodic orbits. In this
265: case we should have $r(t+\tau)=r(t)$ and thus using (\ref{r-dyn}) we
266: see that the radial momentum should be zero,
267: \begin{equation}
268: \label{pr-zero} p_r(t)=0
269: \end{equation}
270: From (\ref{pr-dyn}) and (\ref{pr-zero}) we obtain the following
271: condition
272: \begin{equation}
273: \label{pu} \frac{p_\varphi^2}{mr^3}=\frac{\pd U}{\pd r}
274: \end{equation}
275: Let us finally come to the angular coordinate $\varphi$. For stable
276: motion the following \textit{periodicity} condition should be
277: satisfied (Fig. \ref{fig:traj})
278: \begin{equation}
279: \label{percd} \varphi(n\tau)=\varphi(0)+2\pi,
280: \end{equation}
281: where $n=1,2,\ldots$ (note that discreteness of $n$ is a consequence
282: of discreteness of time). Using (\ref{phi-dyn}) and (\ref{percd}) we
283: get
284: $$
285: n\tau\frac{p_\varphi}{mr^2}=2\pi,
286: $$
287: or
288: \begin{equation}
289: \label{p-phi} p_\varphi=\frac{2\pi m r^2}{n\tau}
290: \end{equation}
291: From (\ref{pu}) and (\ref{p-phi}) we get the following equation for
292: the radius of the $n$-th orbit
293: \begin{equation}
294: \label{rn} \frac{4\pi^2 m r}{n^2\tau^2}=\frac{\pd U}{\pd r}%,~~n=1,2,\ldots
295: \end{equation}
296: If potential $U(r)$ is known then from equation (\ref{rn}) we can
297: find $r=r_n$. If we consider physical potentials, i.e. potentials for which
298: the force, $-U'(r)$, is smooth, negative, and is strictly monotonically decreasing in absolute value as
299: $r$ grows, vanishing on infinity, then solution of (\ref{rn}) always exists and unique.
300: Now upon substituting $r_n$ to the Hamiltonian (\ref{HamiltU}) we obtain energy levels $E_n$.
301:
302: For circular periodic orbits the original Hamiltonian
303: (\ref{HamiltU}) due to (\ref{pr-zero}) and (\ref{pu}) simplifies to
304: the following form
305: $$
306: H=\frac{1}{2}r\frac{\pd U}{\pd r} + U(r)
307: $$
308: Thus the expression for energy $E_n$ of the $n$-th stable periodic
309: orbit in terms of its radius $r_n$ is given by
310: \begin{equation}
311: \label{En} E_n=\frac{1}{2}r_n\left.\frac{\pd U}{\pd
312: r}\right|_{r=r_n}+U(r_n),~~n=1,2,\ldots
313: \end{equation}
314: As we see if potential $U(r)$ allows stationary periodic motion
315: and equation (\ref{rn}) has unique positive solutions $r_n$ then the
316: energy spectrum is discrete. This situation is directly analogous to
317: quantum mechanics where for finite motion we might expect discrete
318: energy levels.
319:
320: \section{Energy Levels of Hydrogen Atom}
321: \label{sec:hydr}
322:
323: Our task in this section is to study whether the discrete time
324: dynamics in central field described can lead to energy levels of
325: hydrogen atom. We treat energy levels as given (measured) quantities
326: and our task is to find the corresponding potential. We restrict
327: ourself to the simplest case when the atom is unperturbed by
328: external electric or magnetic fields and thus currently we do not
329: study splitting of the energy levels (note that in order to observe
330: ``degenerate'' levels in quantum mechanical treatment one needs to
331: somehow perturb the system in such a way that the levels split).
332:
333: We start from the following energy spectrum for hydrogen atom
334: \cite{Bohm}
335: \begin{equation}
336: \label{Enatom} E_n=-\frac{\gamma}{n^2},~~n=1,2,\ldots,
337: \end{equation}
338: where $\gamma\approx 13.6~eV$ is ionization energy of the hydrogen
339: atom. The task is to find such $U=U(r)$ that leads to the spectrum
340: (\ref{Enatom}). Using (\ref{rn}), (\ref{En}), and (\ref{Enatom}) we
341: get
342: \begin{equation}
343: \label{Enrna} \frac{1}{2}\xi
344: \frac{r_n^2}{n^2}+U(r_n)=-\frac{\gamma}{n^2},
345: \end{equation}
346: where the constant $\xi$ is given by (note that $\xi$ depends on
347: discreteness parameter $\tau$)
348: \begin{equation}
349: \label{xidef} \xi=\frac{4\pi^2m}{\tau^2}
350: \end{equation}
351: and $n=1,2,\ldots$ Let us rewrite equation (\ref{Enrna}) in the
352: following form
353: \begin{equation}
354: \label{Enrn2} U(r_n)=-\frac{1}{n^2}\left(\frac{1}{2}\xi
355: r_n^2+\gamma\right)
356: \end{equation}
357: We want to find $U(r)$. The idea is to obtain dependence of
358: $r_n=f(n)$ on $n$ and then inverting it substitute $n=f^{-1}(r_n)$
359: in (\ref{Enrn2}), here $f^{-1}$ denotes function inverse to $f$. As
360: a result of this procedure we will get rid of explicit dependence of
361: the right hand side of (\ref{Enrn2}) on $n$, it will depend only on
362: $r_n$. Then we interpolate the result for any $r\geqslant 0$. The
363: resulting $U(r)$ we check by substitution to (\ref{rn})-(\ref{En}).
364:
365: \begin{figure}
366: \centering \epsfig{file=U-atom.eps,width=7cm} \caption{The shape of
367: potential (\ref{U-atom}) which leads to energy levels of the
368: hydrogen atom (we put $\beta=\xi=\gamma=1$).} \label{fig:U-atom}
369: \end{figure}
370:
371: Let us proceeding as described above. First, we have to find
372: dependence of $r_n$ on $n$. To do this let us assume that $n$ is
373: continuous and take the derivative in $n$ of both sides of
374: (\ref{Enrn2}). This gives
375: \begin{equation}
376: \label{Ueq11} \left.\frac{\pd U}{\pd r}\right|_{r=r_n}\frac{d r_n}{d
377: n}= \frac{\xi r_n^2}{n^3}-\frac{\xi r_n
378: r_n^{\,\prime}}{n^2}+\frac{2\gamma}{n^3}
379: \end{equation}
380: Now, from (\ref{rn}) we find
381: \begin{equation}
382: \label{Ueq12} \left.\frac{\pd U}{\pd
383: r}\right|_{r=r_n}=\xi\frac{r_n}{n^2}
384: \end{equation}
385: Substituting (\ref{Ueq12}) to (\ref{Ueq11}) we obtain the following
386: differential equation for $r_n$
387: \begin{equation}
388: \label{rnatom} 2\xi n r_n r_n^{\,\prime} - \xi r_n^2 - 2\gamma=0
389: \end{equation}
390: Solving (\ref{rnatom}) and taking into consideration only positive
391: solution we obtain
392: \begin{equation}
393: \label{rnsolat}
394: r_n=\sqrt{\frac{-2\gamma+e^{2\beta\xi}n}{\xi}},
395: \end{equation}
396: where the constant $\beta$ is due to integration. Inverting
397: (\ref{rnsolat}) we get
398: \begin{equation}
399: \label{n-sol} n=e^{-2\beta\xi}(2\gamma+r_n^2\xi)
400: \end{equation}
401: Substituting (\ref{n-sol}) to (\ref{Enrn2}) we obtain
402: $$
403: U(r_n)=-\frac{e^{4\beta\xi}}{4\gamma+2r_n^2\xi }
404: $$
405: or performing interpolation, i.e. putting $r$ in place of $r_n$ we
406: finally get
407: \begin{equation}
408: \label{U-atom} U(r)=-\frac{e^{4\beta\xi}}{4\gamma+2r^2\xi }
409: \end{equation}
410: Now, substituting (\ref{U-atom}) to (\ref{rn})-(\ref{En}) we come to
411: the expected energy levels (\ref{Enatom}) of hydrogen atom.
412:
413: The form of the potential (\ref{U-atom}) is presented on
414: Fig.\ref{fig:U-atom}. It is interesting to note that unlike
415: Coulomb's potential it is nonsingular at $r=0$.
416:
417: \section{Spectrum of Harmonic Oscillator}
418: \label{sec:osc}
419:
420: Procedure described in the previous section could be used to obtain
421: potentials corresponding to arbitrary energy spectrum. Here we
422: deduce the potential which results in linear energy levels of the
423: homogeneous two dimensional quantum harmonic oscillator.
424:
425: In quantum mechanics homogeneous two dimensional harmonic oscillator
426: is a system described by the potential
427: \begin{equation}
428: \label{Vharm} U(r)=\frac{1}{2}m\omega^2 r^2,
429: \end{equation}
430: Solving the Schr\"odinger equation in potential (\ref{Vharm})
431: results in the following energy spectrum
432: \begin{equation}
433: \label{ELosc}
434: E_\Lambda=\hbar\omega(\Lambda+1),~~\Lambda=0,1,\ldots
435: \end{equation}
436: The simplest way to get this relation is to note that we deal with
437: two uncoupled oscillators (indeed, if $x$ and $y$ are Cartesian coordinates
438: then equation (\ref{Vharm}) takes the form $U=\frac{1}{2}m\omega^2(x^2+y^2)$),
439: each having energy
440: $E_k=\hbar\omega(k+\frac{1}{2})$, then the total energy is just the
441: sum of two such terms and we get (\ref{ELosc}).
442:
443: Let us first rewrite (\ref{ELosc}) in terms of $n=1,2,\ldots$, we
444: have $n=\Lambda+1$ and thus
445: \begin{equation}
446: \label{Enosc}
447: E_n=\alpha n,~~n=1,2,\ldots,
448: \end{equation}
449: where $\alpha=\hbar\omega$. Our task is to find potential $U=U(r)$
450: which result in energy levels (\ref{Enosc}). Proceeding as in the
451: previous section we assume that $n$ is continuous and write the
452: differential equation
453: \begin{equation}
454: \label{Enalpha}
455: \frac{dE_n}{dn}=\alpha
456: \end{equation}
457: Now, using (\ref{rn}) and (\ref{En}) equation (\ref{Enalpha})
458: reduces to
459: \begin{equation}
460: \label{rnosc}
461: 2\xi \frac{r_n r_n^{\,\prime}}{n^2}-\xi\frac{r_n^2}{n^3} = \alpha,
462: \end{equation}
463: where $\xi$ is given by (\ref{xidef}). The positive solution of
464: (\ref{rnosc}) is given by
465: \begin{equation}
466: \label{rnOSCsol}
467: r_n=\sqrt{\beta n + \frac{\alpha n^3}{2\xi}},
468: \end{equation}
469: where $\beta$ is a constant due to integration. This leads us to a
470: cubic equation in terms of $n$, which has one real solution
471: $$
472: n=\frac{\mathcal{V}(r_n)}{3\alpha}-\frac{2\beta\xi}{\mathcal{V}(r_n)},
473: $$
474: where $\mathcal{V}(r)$ is given by
475: $$
476: \mathcal{V}(r)=3^{1/3}\left(9r^2\alpha^2\xi+\sqrt{3}\sqrt{27r^4\alpha^4\xi^2+8\alpha^3\beta^3\xi^3}\right)^{1/3}
477: $$
478: We finally get the expression for potential
479: $$
480: U(r)=\frac{\mathcal{V}(r)}{3} -
481: \frac{2\alpha\beta\xi}
482: {\mathcal{V}(r)} -
483: \frac{9r^2\,{\mathcal{V}(r)}^2\,
484: {\alpha }^2\,\xi }{2\,
485: {\left( {\mathcal{V}(r)}^2 -
486: 6\alpha\beta\xi\right)
487: }^2}
488: $$
489: If we put $\beta=0$ in (\ref{rnOSCsol}) the potential takes simple
490: form
491: \begin{equation}
492: \label{U-osc} U(r)=\frac{3}{2}
493: \left(\frac{r^2\alpha^2\xi}{4}\right)^{1/3}
494: \end{equation}
495: Substituting (\ref{U-osc}) to (\ref{rn})-(\ref{En}) we get correct
496: energy levels (\ref{Enosc}). It is interesting to provide the
497: expression (\ref{U-osc}) when all constants are substituted, we have
498: $$
499: U(r)=\frac{3}{2}
500: \left(r\,\frac{\hbar\omega\pi\sqrt{m}}{\tau}\right)^{2/3}
501: $$
502:
503: \section{General Case of Arbitrary Spectrum}
504: \label{sec:arb}
505:
506: One can find explicit relations for the radii of the orbits for
507: arbitrary spectrum $E_n$. Indeed, From (\ref{rn}) and (\ref{En}) we
508: want to find $U$ as a function of $r$ in terms of a given energy
509: spectrum $E_n$. As we did above we assume the continuity of the parameter
510: $n$ and take derivative of both parts of (\ref{En}) in $n$, we have
511: $$
512: E^{\,\prime}_n=
513: r^{\,\prime}_n\frac{\xi r_n}{n^2}
514: -\frac{\xi r_n^2}{n^3}
515: +r^{\,\prime}_n\left.\frac{\pd U}{\pd r}\right|_{r=r_n},
516: $$
517: where prime denotes the derivative in $n$. The use of (\ref{rn})
518: allows us to get a differential equation for $r_n$ in terms of only
519: known quantities. We have
520: $$
521: 2r_n r^{\,\prime}_n-\frac{1}{n}r_n^2=\frac{n^2}{\xi}E^{\,\prime}_n
522: $$
523: Introducing a new variable $\rho=r^2$ we obtain a linear
524: differential equation which could be rewritten as
525: $$
526: \frac{d}{dn}\left(\frac{\rho_n}{n}\right)=\frac{1}{\xi}n\,E^{\,\prime}_n
527: $$
528: which can be integrated to obtain
529: \begin{equation}
530: \label{nr}
531: r_n=\sqrt{\frac{1}{\xi}n \left(nE_n - E_1 - \int_{1}^n E_k dk + \epsilon \right)}
532: \end{equation}
533: Equation (\ref{nr}) expresses $n$-th radius in terms of $n$, i.e. it
534: has the form $r=f(n)$ now if we invert it we relate $n$ in terms of
535: $r$, $n=f^{-1}(r)$, which if substituted to (\ref{En}) gives an
536: equation for $U$ in terms of $r$ only (actually in terms of $r_n$,
537: but we perform interpolation effectively ignoring the fact that the
538: relation strictly holds only for orbit radii).
539:
540: Note that in (\ref{nr}) we have a constant $\epsilon$ (having units of energy)
541: arising due to the integration, this means that we have a set of potentials
542: resulting in the same energy spectrum. As we see from (\ref{nr}) the constant
543: $\epsilon$ could be determined if for example the smallest radius $r_1$ is known.
544: In sections \ref{sec:hydr} and \ref{sec:osc} the situation was the same resulting in the
545: constant $\beta$ (see (\ref{rnsolat}) and (\ref{rnOSCsol})).
546: Since $\epsilon$ is more interesting from the point of view of physical interpretation we
547: provide the expressions for $\beta$ in terms of $\epsilon$. For the case of
548: energy levels of hydrogen atom we have (see (\ref{rnsolat}))
549: $$
550: \beta_{hydr}=\frac{1}{2\xi}\ln(\epsilon+\gamma)
551: $$
552: and for the case of harmonic oscillator we have (see (\ref{rnOSCsol}))
553: $$
554: \beta_{osc}=\frac{\epsilon - \frac{1}{2}\alpha}{\xi}
555: $$
556:
557: \section{Energy Spectrum in Various Potentials}
558:
559: As we already seen, for a given potential it is straightforward to
560: compute corresponding energy levels. Indeed, from (\ref{rn}) we find
561: $r=r_n$ and upon substitution to (\ref{En}) we get $E_n$. Like in
562: quantum mechanics, potential $U=U(r)$ should be attractive and
563: strong enough to result in finite motion. Below we consider several
564: common central potentials which result in rather simple expressions
565: for energy spectrum. In what follows the constant $\xi$ is given by
566: (\ref{xidef}), note that it depends on time discreteness parameter
567: as $1/\tau^2$.
568:
569: \noindent a). Coulomb potential
570: $$
571: U(r)=-\frac{\alpha}{r},~~ r_n=n^{2/3}\left( \frac{\alpha}{\xi}
572: \right)^{1/3},~~
573: E_n=-\frac{1}{2n^{2/3}}\left(\alpha^2\xi\right)^{1/3}
574: $$
575: Note that energy spectrum is different from the $-\gamma/n^2$
576: spectrum obtained in quantum mechanics for this potential (see
577: section on energy levels of hydrogen atom for detailed discussion).
578:
579: \vspace{0.1cm} \noindent b). Linear potential
580: $$
581: U(r)=\alpha r,~~ r_n=\frac{n^2\alpha}{\xi},~~
582: E_n=\frac{3n^2\alpha^2}{2\xi}
583: $$
584:
585: \vspace{0.1cm} \noindent c). Logarithmic potential
586: $$
587: U(r)=\alpha\ln r,~~ r_n=n\sqrt{\frac{\alpha}{\xi}},~~
588: E_n=\alpha\left[\frac{1}{2}+\ln\left(n\sqrt{\frac{\alpha}{\xi}}\right)\right]
589: $$
590: For potentials (a), (b), and (c) $r_n>0$ if $\alpha>0$ in all three
591: cases it corresponds to the attraction field.
592:
593: \vspace{0.1cm} \noindent d). Polynomial potential
594: \begin{equation}
595: \label{polyU}
596: U(r)=\alpha r^\sigma,~~ r_n=\left(\frac{n^2
597: \alpha\sigma}{\xi}\right)^{\frac{1}{2-\sigma}},~~
598: E_n=\frac{1}{2}\alpha (2+\sigma)
599: \left(\frac{n^2\alpha\sigma}{\xi}\right)^{\frac{\sigma}{2-\sigma}}
600: \end{equation}
601: This case generalizes cases (a) and (b) described above, although
602: because of importance of these potentials we provided corresponding
603: expressions explicitly. Note that if we make $\sigma$ in (\ref{polyU}) satisfy the following equation
604: $$
605: \frac{2\sigma}{2-\sigma}=1,
606: $$
607: i.e. $\sigma=2/3$ then we get the linear energy spectrum $E_n\sim n$
608: for the 2D quantum harmonic oscillator (see previous section for
609: detailed discussion).
610:
611: \section{Discussion and Conclusion}
612:
613: In this paper we have shown that discrete time formalism leads to
614: some distinguishable properties of micro-observables that are used
615: to be described with quantum mechanics. In particular it was shown
616: that finite motion results in discrete energy spectrum. Of the main
617: interest in this paper is discrete energy levels of hydrogen atom.
618: We have shown that for unperturbed hydrogen atom the discrete time
619: formalism is able to give correct energy spectrum, more precisely we
620: have reconstructed the corresponding ``micro''-potential. Here we
621: did not consider Stark or Zeeman effects for hydrogen atom, it would
622: be interesting to study them from the point of view of discrete time
623: formalism.
624:
625: As we have seen in above the discrete time model requires potentials
626: which are different from QM potentials. One may argue this as a
627: disadvantage of the model. We pay attention that there are no
628: reasons to expect to reproduce QM by using the standard classical
629: potentials; Bohr, Zommerfeld, Heisenberg and many others tried to do
630: this, but they did not succeed. D. Bohm developed a new model
631: \cite{BH} in which quantum mechanics can be reproduced on the
632: classical basis, but, of course, the classical potential could not
633: be preserved -- it is perturbed by the quantum potential. And the
634: latter looks not so natural from the classcial viewpoint, see e.g.
635: the quantum potential for the two slit experiment in \cite{BH}. The
636: discrete time model has an analogy with Bohmian mechanics -- it
637: tries to reproduce QM by changing potentials. But there is of course
638: the fundamental difference: the only postulate that is used in the
639: proposed approach is that there exists a quant of time $\tau$.
640:
641: Another interesting point is that one might expect that our
642: dynamical equations are essentially difference equations which might
643: produce discrete spectrum. This is not correct -- we recall that in
644: our model only time is discrete, but space is still continuous.
645:
646: There is still an open problem of the quantative value of the
647: discreteness parameter $\tau$. One might speculate its relation with
648: Plank time constant \cite{Plank,QuantGrav} -- the smallest
649: measurable time interval in ordinary QM and gravity -- which is
650: quantatively given by
651: $$
652: t_{Pl}=\sqrt{\frac{\hbar G}{c^5}}\approx 5.39 10^{-44}~(sec.)
653: $$
654: The detailed analysis of this issue is out of the scope of the
655: present article.
656:
657: There might be an interesting interconnection on how the discrete time
658: is used in information dynamics theory\cite{IKO,Ohya} and
659: the discrete time dynamics as it appears in presented study.
660: In particular, it would be interesting to consider the
661: equations for information dynamics with discrete time.
662:
663: Finally, we would like to comment also that there might be a deep
664: interrelation between the energy-time uncertainty relations
665: \cite{MT,Rue,PfeFro} and Bohr-Somerfeld quantization rules
666: \cite{MasFed} in quantum mechanics and our discrete time model. In
667: particular one can try to write the Bohr-Somerfeld semi-classical
668: quantization rules for energy and time as canonical variables. For a
669: system with conserved energy one might get $E_n T_n \sim n\hbar$,
670: this relation holds for example for energy levels $E_n$ and
671: classical periods $T_n$ of Hydrogen atom. One the other hand the
672: relation (\ref{percd}) in discrete mechanics could be treated as the
673: condition for the period to take only discrete values $T_n\sim
674: n\tau$. We can see that although relations are similar there is an
675: extra factor $E_n$ in the quantum-mechanical relation. In fact one
676: may argue that if we make the $\tau$ in equations of motion depend
677: on the energy of the system as
678: $$
679: \tau=\tau_0 \frac{\varepsilon}{E},
680: $$
681: where $\tau_0$ is the ``fundamental'' time quantum and $\varepsilon$
682: a ``fundamental'' energy quantum, we get precisely the semiclassical
683: quantization rules. The question arise how to treat the energy $E$
684: here and what will happen with the dynamics. Further investigation
685: of this interrelation will be discussed elsewhere.
686:
687: \section{Acknowledgments}
688:
689: The authors would like to thank B.~Hiley, A.~Plotnitsky, G.~`t
690: Hooft, H.~Gus\-tafson, and K.~Gustafson for discussions on
691: quantum-like models with discrete time and M.~Ohya and D.~Petz for
692: discussions on classical and quantum information theory and information
693: dynamics.
694: Y.V. would like to thank L.~Joukovskaya for fruitful
695: discussions on atomic spectra measurements.
696:
697: \begin{thebibliography}{99}
698:
699: \bibitem{Heis} W.~Heisinberg,
700: \emph{\"{U}ber Quantentheoretische Undeutung kinematischer und
701: mechanischer Beziehungen}, Zs. Phys., \textbf{33}, pp. 879-883,
702: (1925).
703:
704: \bibitem{Heis2} W. Heisenberg, {\it Physical principles of quantum theory,}
705: Chicago Univ. Press, 1930.
706:
707: \bibitem{2} D. Hilbert, J. von Neumann, L. Nordheim, {\it Math. Ann.}, {\bf
708: 98}, 1-30 (1927).
709:
710: \bibitem{Dir} P. A. M. Dirac, {\it The Principles of Quantum Mechanics,}
711: Oxford Univ. Press, 1930.
712:
713: \bibitem{5} J. von Neumann, {\it Mathematical foundations
714: of quantum mechanics,} Princeton Univ. Press, Princeton, N.J., 1955.
715:
716: \bibitem{6} A. Einstein, B. Podolsky, N. Rosen, Phys. Rev., {\bf 47}, 777--780
717: (1935).
718:
719: \bibitem{Hooft1}
720: G. 't Hooft, \textit{Determinism in Free Bosons}, Int.J.Theor.Phys.
721: 42 (2003) 355-361
722:
723: \bibitem{Hooft2}
724: G. 't Hooft, \textit{Determinism beneath Quantum Mechanics},
725: Proceedin\-gs of ``Quo Vadis Quantum Mechanics'', Philadelphia,
726: 2002, quant-ph/0212095
727:
728: \bibitem{Hooft3}
729: G. 't Hooft, \textit{Quantum Mechanics and Determinism},
730: hep-th/0105105
731:
732: \bibitem{IKO}
733: R.S.~Ingarden, A.~Kossakowski, and M.~Ohya,
734: \textit{Information Dynamics and Open Systems: Classical and Quantum Approach},
735: Kluwer Acad. Publishers, 1997
736:
737: \bibitem{OP}
738: M.~Ohya, D.~Petz,
739: \textit{Quantum entropy and its use},
740: Springer-Verlag, Heidelberg 1993
741:
742: \bibitem{Ohya}
743: M.~Ohya,
744: \textit{Information Dynamics and its Application to Recognition Process},
745: in ``A Garden of Quanta: Essays in Honor of Hiroshi Ezawa'', World Scientific Pub Co Inc,
746: pp. 445-462, 2003, quant-ph/0406229
747:
748: \bibitem{Bohm} D. Bohm, {\it Quantum theory,} Prentice-Hall,
749: Englewood Cliffs, New-Jersey, 1951.
750:
751: \bibitem{9} A. Lande, {\it Foundations of quantum theory,} Yale Univ. Press,
752: 1955.
753:
754: \bibitem{11} A. S. Wightman, Hilbert's sixth problem: mathematical treatment of
755: the axioms of physics, {\it Proc. Symposia in Pure Math.,} {\bf 28},
756: 147-233 (1976).
757:
758: \bibitem{12} L. De Broglie, {\it The current interpretation of wave mechanics,
759: critical study.} Elsevier Publ., Amsterdam-London-New York, 1964.
760:
761: \bibitem{Bell} J. S. Bell, {\it Speakable and unspeakable in quantum mechanics,}
762: Cambridge Univ. Press, 1987.
763:
764: \bibitem{Acc}
765: L.~Accardi,
766: \textit{Urne e camaleonti. Dialogo sulla realt`a, le leggi del caso e la teoria quantistica},
767: Il Saggiatore (1997)
768:
769: \bibitem{AIR}
770: L.~Accardi, K.~Imafuku, M.~Regoli,
771: \textit{On the physical meaning of the EPR--chameleon experiment},
772: quant-ph/0112067
773:
774: \bibitem{BH}
775: D. Bohm, B.J. Hiley, \textit{The Undivided Universe: An Ontological
776: Interpretation of Quantum Theory}, London: Routledge \& Kegan Paul.
777:
778: \bibitem{Hol}
779: P.R.~Holland, \textit{The Quantum Theory of Motion}, Cambridge Univ.
780: Press, 1993.
781:
782: \bibitem{MasFed}
783: V.P.~Maslov, M.V~Fedoriuk, \textit{Semi-Classical Approximations in
784: Quantum Mechanics}, Boston: Reidel, 1981
785:
786: \bibitem{14} G. W. Mackey, {\it Mathematical foundations of quantum mechanics,}
787: W. A. Benjamin INc, New York, 1963.
788:
789: \bibitem{16} L. E. Ballentine, {\it Rev. Mod. Phys.}, {\bf 42}, 358--381 (1970).
790:
791: \bibitem{22} S. P. Gudder, {\it Axiomatic quantum mechanics and generalized
792: probability theory,} Academic Press, New York, 1970.
793:
794: \bibitem{23} S. P. Gudder, {\it ``An approach to quantum probability''} in
795: {\it Foundations of Probability and Physics,} edited by A. Yu.
796: Khrennikov, Quantum Prob. White Noise Anal., 13, WSP, Singapore,
797: 2001, pp. 147-160.
798:
799: \bibitem{FeyGib} R. Feynman and A. Hibbs, {\it Quantum Mechanics and Path Integrals,}
800: McGraw-Hill, New-York, 1965.
801:
802: \bibitem{MT}
803: L.~Mandelstam, I.E.~Tamm, J. Phys. (Moscow) \textbf{9}, 249, (1945).
804:
805: \bibitem{Rue}
806: D.~Ruelle, Nuovo Cimento A \textbf{61}, 655, (1969).
807:
808: \bibitem{PfeFro}
809: P.~Pfeifer, J.~Fr\"ohlich, Rev. Mod. Phys., \textbf{67} 759, (1995).
810:
811:
812: \bibitem{26} A. Peres, {\em Quantum Theory: Concepts and Methods,} Dordrecht,
813: Kluwer Academic, 1994.
814:
815: \bibitem{27} L. Accardi, {\it ``The probabilistic roots of the quantum
816: mechanical paradoxes''} in {\em The wave--particle dualism. A
817: tribute to Louis de Broglie on his 90th Birthday,} edited by S.
818: Diner, D. Fargue, G. Lochak and F. Selleri, D. Reidel Publ. Company,
819: Dordrecht, 1984, pp. 297--330.
820:
821: \bibitem{33} P. Busch, M. Grabowski, P. Lahti, {\it Operational Quantum Physics,}
822: Springer Verlag,Berlin, 1995.
823:
824: \bibitem{Plank}
825: C.~Callender, N.~Huggett (edditors), \textit{Physics Meets
826: Philosophy at the Planck Scale: Contemporary Theories in Quantum
827: Gravity}, Cambridge Univ. Press, 2001.
828:
829: \bibitem{QuantGrav}
830: C.~Rovelli, \textit{Quantum Gravity}, Cambridge Univ. Press, 2004.
831:
832: \bibitem{Vax3} A. Yu. Khrennikov (editor), {\it Foundations of Probability and
833: Physics}-2, Ser. Math. Modeling, 5, V\"axj\"o Univ. Press, 2003.
834:
835: \bibitem{Vax4} A. Yu. Khrennikov (editor), {\it Quantum Theory: Reconsideration
836: of Foundations}-2, Ser. Math. Modeling, 10, V\"axj\"o Univ. Press,
837: 2004.
838:
839: \bibitem{Kh} A. Yu. Khrennikov, {\it J. Phys.A: Math. Gen.,} {\bf 34}, 9965-9981
840: (2001).
841:
842: \bibitem{Kh5} A. Yu. Khrennikov, {\it J. Math. Phys.}, {\bf 44}, 2471- 2478
843: (2003).
844:
845: \bibitem{Kh6} A. Yu. Khrennikov, {\it Phys. Lett. A}, {\bf 316}, 279-296 (2003).
846:
847: \bibitem{Kh8} A. Yu. Khrennikov, {\it Annalen der Physik,} {\bf 12}, 575-585 (2003).
848:
849: \bibitem{KV1} A.Yu. Khrennikov, Ya.I.~Volovich,
850: \textit{Discrete Time Leads to Quantum-Like Interference of
851: Deterministic Particles}, in: Quantum Theory: Reconsiderations of
852: Foundations, ed. A. Khrennikov, V\"axj\"o University Press, pp. 455, (2002)
853:
854: \bibitem{KV2} A.Yu. Khrennikov, Ya.I.~Volovich,
855: \textit{Interference as a statistical consequence of conjecture on
856: time quant}, quant-ph/0309012
857:
858: \bibitem{KV3} A.~Khrennikov, Ya.~Volovich,
859: \textit{Discrete Time Dynamical Models and Their Quantum Context
860: Dependant Properties}, J.Mod.Opt. Vol.51, No.6-7, pp. 1113, (2004).
861:
862: \end{thebibliography}
863:
864:
865: \end{document}
866: