Moosey - reading the articles on QC it would appear that the two approaches you describe are basically different algorithms - one approach sequentially destroys the qubit state on read out whilst the other (more difficult approach) leaves the qubit state in place. I have not fully got my mind around the issue (it's pretty involved) but it doesn't appear to be involved with the "engineering" and is applicable to many different approaches (nitrogen vacancies, trapped ions etc) and not particular to Silicon based methods. Of course aviodance of heat generation (particularly for QC) is front and centre as an issue for QC (as it is for conventional computing) and it would be of considerable benefit fo get a x10 improvement in conductivity. but this is not the reason for using ZSSi - that reason is to provide a vacuum type environment for a phosphorus atom qubit to preserve the combined 0/1 decoherence state for as long as possible (during which time the QC computation is proceeding). ZSSi removes the
29Si isotope so as to provide this environment. Whats not clear to me is that the SLX/SQC literature says ZSSi is "
Natural silicon (Si) consists of 3 isotopes: 92.2% Si-28, 3.1% Si-30 (each with zero electron spin) and 4.7% Si-29 (with a spin state of ½
. The presence of Si-29 in concentrations above 100ppm (0.01%) prevents effective QC performance, so ZS-Si must be produced by isotopic elimination of Si-29" - it makes no mention of removing the
30Si. Literature that refers to the conductivity benefits of Si refers to using pure
28Si and attributes the improved conductivity to a regualr crystal matrix since all the Si atoms are identical in size and other properties. To me presence of
30S would likely destroy this "perfect" crystal structure since
30Si has 2 more neutrons than
28Si and would seem to distort this "perfect" crystal structure and hence not provide the beneficial conductivity of pure
28Si.
I am NOT disagreeing with you over the desirability of much improved conductivity - it would be great. What I am asking you about is to point me to literature which either says:
1. That SLX ZSSi does NOT contain
30Si
OR
2. That
30Si in a Si crystal structure does NOT inhibit the improved conductivity characteristics of pure
28Si crystal
It would be great to claim the conductivity improvements for the SLX product - my problem is that I don't know if that's correct as SLX/SQC literature does not appear to claim to have removed the
30Si. Removal of the
29Si appears to be ESSENTIAL for QC but the conductivity benefits only appear desirable (as would be the case for all other Si semiconductor substrates. As you will be aware SLX ran experiments with Suminito in around 2005 on pure (or near pure)
28Si and Suminito concluded (then) that there was not a business case for
28Si for normal Si circuits but if
30Si is unimportant to this improvement then the business case may now be different due greater needs to squash more circuitry into much smaller areas thus exasperating the heat generation issue and reducing the amount of Si required.
(20min delay)