On The 0 Limit of The Schrödinger Equation

On the →0 limit of the Schrödinger equation
Song Ling
Citation: The Journal of Chemical Physics 96, 7869 (1992); doi: 10.1063/1.462388
View online: http://dx.doi.org/10.1063/1.462388
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/96/10?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in

A discussion of the → 0 limit of the Schrödinger equation
Am. J. Phys. 75, 718 (2007); 10.1119/1.2737474
The → 0 limit of the Schrödinger equation

Am. J. Phys. 74, 557 (2006); 10.1119/1.2186338
Some extensions of semiclassical limit →0 for Wigner functions on phase space

J. Math. Phys. 36, 622 (1995); 10.1063/1.531145
The Hamiltonian structures of the nonlinear Schrödinger equation in the classical limit
J. Math. Phys. 28, 1094 (1987); 10.1063/1.527553
Numerical Solution of the SLimit Schrödinger Equation

J. Chem. Phys. 48, 1879 (1968); 10.1063/1.1668984
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.247.166.234 On: Sat, 22 Nov 2014 22:57:51
On the Ii-O limit of the SchrOdinger equation
Song Ling a)
Department of Pharmaceutical Chemistry, University of California, San Francisco, California 94143-0446
(Received 17 December 1991; accepted 7 February 1992)
It is believed that the time dependent Schrodinger correspondence principle. However, we have found a
equation contains the classical Hamilton-Jacobi equation counter example (more examples can be given for qua-
in the Ii-O limit. This is done in the following way:1 writ- dratic potentials with Gaussian initial wave functions), i.e.,
ing down the time dependent Schrodinger equation the quantum potential of a system turns out to be indepen-
dent of Ii, which we will show in the next paragraph. The
a1/J(r,t) fil 2
iii at = - 2m V tf;(r,t) + V(r)tf;(r,t), (1) failure of the previous inference lies in the fact that besides
the prefactor - fil/2m in Q(r,t), R(r,t) can also have
[where VCr) is a real potential] and casting the complex factors of Ii hidden in it. The case here very much depends
wave function tf;(r,t) in "polar coordinates" on what kind of wave function (Le., the initial wave func-
tion) is being studied.
1/J(r,t) = R(r,t)exp[iS(r,t)/Ii], (2) We choose to deal with the coherent state wave func-
where R(r,t) and S(r,t) are real, and are the amplitude tion of a one dimensional harmonic oscillator, since har-
and phase of 1/J(r,t), respectively. [To be mathematically monic oscillators not only well serve as a paradigm for
rigorous, the phase is S(r,t) divided by Ii.] Substituting (2) atomic, molecular and solid state bound potentials near the
into (1), we obtain, by equating the real and imaginary minima, but also still draw the attention of the present day
parts of (1), respectively, researchers. 3 The time dependent Schr6dinger equation in
this case reads
aR(r,t) 1
at = - 2m [R(r,t)V 2S(r,t) + 2VR(r,t)'VS(r,t)],
(3)
as(r,t) _ [[Vs(r,t) f ~ V 2R(r,t)] The coherent state wave function is a known analytic
at - - 2m + VCr) - 2m R(r,t) . (4) solution4,5 of (7),
Following the conventional interpretation, the probability mw) 1/4 { mw }

1/J(x,t) = ( 1rli exp - ~[x - (X(t»]2
density p(r,t) is the square of the absolute value of 1/J(r,t) ,
Le., [R(r,t)f, substituting p(r,t) for R(r,t) in (3), we ob-
tain, xexp{~[ (p(t»x - ~ t - (x(t) )2(P(t»]l,
ap(r,t) ( Vs(r,t»)
- a - + V· p(r,t) = O. (5a) (8)
t m
where (x(t» and (P(t» are the expectation values of the
It can be proved (with just a few analytic operations) that
position and momentum at time t, respectively. It is
this is exactly the equation of the conservation of proba-
known4,5 that they follow classical trajectories
bility flux 2
ap(r,t) . (P(O» .
~+ V'J(r,t) =0, (5b) (x(t» = (x(O) )cos«())() + -mw- sm(wt), (9)
where by definition 2 j(r,t) = ili/2m[1/J(r,t)Vtf;*(r,t) (p(t» = (P(O) )cos(wt) - mw(x(O) )sin(wt), (10)
- tf;*(r,t)Vtf;(r,t)]. On the other hand, (4) has the same
form as the Hamilton-Jacobi equation in classical mechan- where (x(O» and (P(O» are the initial expectation values
ics with an "effective" time dependent potential as the sum that must be specified (chosen) before their time evolution
of the classical potential V(r) plus the "quantum begins. It is clear from (8) that R(x,t) and S(x,t) are
potential" 1(a)
2
fil [V R(r,t) ]
mw) 1/4
R(x,t) = ( 1rli
{ mw
exp - 21i [x- (X(t»]2 ,
I (11 )
Q(r,t) = - 2m R(r,t) . (6)
Iiw (x(t» (p(t»
It was plausibly suggested 1 that in the limit of Ii going to S(x,t) = (p(t»x - 2" t - 2 . (12)
zero, the quantum potential drops out regardless of what-
ever R(r,t) and then (4) is reduced to the classical The quantum potential is readily calculated by using (6)
Hamilton-Jacobi equation, as a manifestation of the Bohr and (11),
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chern. Phys. 96 (10), 15 May 1992 On: Sat, 22 Nov 2014 22:57:51 © 199.2 American Institute of Physics
0021-9606/92/107869-02$006.00
155.247.166.234 7869
7870 Letters to the Editor
Kouri, Craig Martens, and Michael Kellman for their en-

(13) couragements. Special acknowledgment goes to Professor
Robert Langridge and the Computer Graphics Laboratory
We immediately see that the first term in (13) is indepen-
dent of Ii! Only the second term (which but lifts the energy ofUCSF.
zero level) drops as Ii-.O. This contradicts the previous
a) Address after May 1, 1992: Department of Chemistry/Physics/
beliefl that (4) would be reduced to the exact classical Astronomy, Indiana University Northwest, Gary, IN 46408.
Hamilton-Jacobi equation in the Ii-.O limit. IThe examples are (a) D. Bohm, Phys. Rev. 85, 166 (1952); (b) A.
The conclusion is that, generally speaking, the time Messiah, Quantum Mechanics (North-Holland, Amsterdam, 1961);
dependent Schrodinger equation does not contain the clas- (c) L.1. Schiff, Quantum Mechanics (McGraw-HilI, New York, 1968).
sical Hamilton-Jacobi equation in the Ii-.O limit, and ge- 2See, for example, S. Gasiorowicz, Quantum Physics (Wiley, New York,
1974).
nerically the meaning of S(r,t) does not correspond to that
3G. Torres-Vega and J. H. Frederick, J. Chern. Phys. 93, 8862 (1990).
of the Hamilton's principal function in classical mechanics.
4C. Cohen-Tannoudji, B. Diu, and F. Laloe, Quantum Mechanics, Vol. 1
Finally, we comment that since the example given in this (Wiley, New York, 1977).
note is derived from a well-known classically integrable S The general solutions of the time dependent Schrodinger equation with
system, the interesting question remains whether analo- quadratic potentials and initial Gaussian wave functions are given in E.
gous cases can be found in classically chaotic systems. 6 J. Heller, J. Chern. Phys. 62, 1544 (1975).
I thank Colin Wright for introducing Ref. 1(a) to me 6 For an interesting discussion on topics concerning "quantum chaol-
and for helpful discussions, Professors Eric Heller, Donald ogy," see, e.g., M. Berry, Physica Scr. 40, 335 (1989).
J. of
This article is copyrighted as indicated in the article. Reuse Chern. Phys., Vol.
AIP content 96, No.
is subject to 10,
the 15 Mayat:1992
terms http://scitation.aip.org/termsconditions. Downloaded to IP:
155.247.166.234 On: Sat, 22 Nov 2014 22:57:51

On The 0 Limit of The Schrödinger Equation

Uploaded by

Copyright:

Available Formats

On The 0 Limit of The Schrödinger Equation

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

On The 0 Limit of The Schrödinger Equation

Uploaded by

Copyright:

Available Formats

On the →0 limit of the Schrödinger equation

Articles you may be interested in

The → 0 limit of the Schrödinger equation

Some extensions of semiclassical limit →0 for Wigner functions on phase space

Numerical Solution of the SLimit Schrödinger Equation

Following the conventional interpretation, the probability mw) 1/4 { mw }

Kouri, Craig Martens, and Michael Kellman for their en-

You might also like