From:
c.h.thompson <c.h.thompson@newscientist.net>
To: David
Kielpinski <davidk@boulder.nist.gov>
Subject:
quant-ph/0102086: Recent Results in Trapped-Ion Quantum Computing
Date:
Wednesday, February 21, 2001 5:15 PM
Dear Dr
Kielpinski
These are
most impressive and fascinating results.
I am concerned, however, that the abstract of your paper could easily be
misconstrued as implying that you have demonstrated a "nonlocal"
effect. This would be the usual
implication of your statement that you have violated Bell's inequality without
using results from quantum mechanics and with the detector loophole
closed. I have spent the past day
trying to discover the true situation.
Provisionally,
I must assume that the fact that there is only one actual "detection"
event covering the two ions is the significant factor that enables the high
correlations. In other words, as you
say, the locality loophole is open -- indeed, it could not be much more open
than this!
On the
other hand, I see various difficulties and reasons why you might feel that
nevertheless you do have evidence of phenomena that are not covered by
"realist" theories. Having
studied in detail many of the supposed refutations of "local realism"
and found none that stood up to scrutiny, I am reluctant to admit that there is
any phenomenon that does not have a realist explanation. (For information about the optical Bell test
experiments please see my web site -- there are various papers in
http://www.aber.ac.uk/~cat/bibliography.htm .)
To return
to the current paper:
1. The
derivation of the Bell test presupposes that the probability of detection of
each particle individually is a well-defined quantity. Is this the case here? With only one measurement covering the two
particles, is it likely that the assumption of independence, necessary in order
to multiply probabilities, is correct?
2. Can you
tell me what the quantum mechanical prediction is? In most Bell tests, it is a function of the difference in
detectors settings. In your case,
therefore, I would expect to see it as a function of the difference of the two
phases. Your results, though, conflict
with this hypothesis. The value of q
for equal values of the two phases reverses sign as we change from two at -pi/8
to two at 3pi/8. Is there perhaps an
error here?
3. In most Bell tests there is some obvious
candidate for the hidden variable -- the factor that the two particles have in
common that varies randomly from one instance to the next. Can you tell me what it is that you consider
to vary randomly? The duration of the
Raman pulse, maybe? The phase
difference between this pulse and that of the circularly polarized laser beam?
4. At first
glance, there would seem to be a discrepancy between the range of parity values
shown in Fig. 3 (a) and the claimed accuracy of 98%. Why, when only two ions are present, is the maximum parity not
observed to be nearly 1?
As I said,
these are impressive results. You
clearly have control over some very delicate systems. You might be interested, however, in my paper
quant-ph/9912082. In this I suggest
that systems that are almost deterministic might be more useful than ones that
genuinely obey quantum mechanics when it comes to applications.