15.08.2013 Views

shot noise in mesoscopic conductors - Low Temperature Laboratory

shot noise in mesoscopic conductors - Low Temperature Laboratory

shot noise in mesoscopic conductors - Low Temperature Laboratory

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

and the <strong>in</strong>itial state of our two-particle scatter<strong>in</strong>g experiment is thus AK (x , t) AK (x , t)0.<br />

<br />

We are aga<strong>in</strong> <strong>in</strong>terested <strong>in</strong> determ<strong>in</strong><strong>in</strong>g the probabilities that two particles appear <strong>in</strong> an output<br />

branch or that one particle appears <strong>in</strong> each output branch, P(2, 0)"dk dk n( (k )n( (k ),<br />

<br />

P(1, 1)"dk dk n( (k )n( (k ), and P(0, 2)"dk dk n( (k )n( (k ). Let us aga<strong>in</strong> consider<br />

<br />

P(1, 1). Its evaluation proceeds <strong>in</strong> much the same way as <strong>in</strong> the case of pure scatter<strong>in</strong>g states. We<br />

re-write P(1, 1) <strong>in</strong> terms of the absolute square of an amplitude,<br />

<br />

P(1, 1)" dk dk 0bK (k )bK (k ) . (25)<br />

<br />

<br />

We then write the a( operators <strong>in</strong> the AK <strong>in</strong> terms of the output operators bK . Instead of Eq. (20), we<br />

obta<strong>in</strong><br />

P(1, 1)"¹#R$2¹RJ , (26)<br />

where<br />

<br />

J" dk H(k) (k) exp(ik(x !x )) (27)<br />

<br />

<br />

Ya.M. Blanter, M. Bu( ttiker / Physics Reports 336 (2000) 1}166 17<br />

is the overlap <strong>in</strong>tegral of the two particles. For the case of complete overlap J"1 we obta<strong>in</strong><br />

Eq. (20). For the case that we have no overlap we obta<strong>in</strong> the classical result P(1, 1)"¹#R<br />

which is <strong>in</strong>dependent of whether a boson or fermion is <strong>in</strong>cident on the scatterer. In the general case,<br />

the overlap depends on the form of the wave packet. If two Gaussian wave packets of spatial width<br />

and central velocity v are timed to arrive at time and at the scatterer, the overlap <strong>in</strong>tegral is<br />

<br />

J"exp[!v( ! )/2] . (28)<br />

<br />

A signi"cant overlap occurs only dur<strong>in</strong>g the time /v. For wave packets separated <strong>in</strong> time by more<br />

than this time <strong>in</strong>terval the Pauli pr<strong>in</strong>ciple is not e!ective.<br />

Complete overlap occurs <strong>in</strong> two simple cases. We can assume that the two wave packets are<br />

identical and are timed to arrive exactly at the same <strong>in</strong>stant at the scatterer. Another case, <strong>in</strong> which<br />

we have complete overlap, is <strong>in</strong> the basis of scatter<strong>in</strong>g states. In this case (k)"(k!k ) for<br />

<br />

a scatter<strong>in</strong>g state with wave vector k and consequently for the two particles with k and k we have<br />

<br />

J" .The"rst option of timed wave packets seems arti"cial for the thermal sources which we<br />

<br />

want to describe. Thus <strong>in</strong> the follow<strong>in</strong>g we will work with scatter<strong>in</strong>g states.<br />

The probabilities P(2, 0), P(1, 1) and P(0, 2) for these scatter<strong>in</strong>g experiments are shown <strong>in</strong> Table 2.<br />

The considerations which lead to these results now should be extended to take the polarization of<br />

photons or sp<strong>in</strong> of electrons <strong>in</strong>to account.<br />

2.2.4. Two-particle scatter<strong>in</strong>g: sp<strong>in</strong><br />

Consider the case of fermions and let us <strong>in</strong>vestigate a sequence of experiments <strong>in</strong> each of which<br />

we have an equal probability of hav<strong>in</strong>g electrons with sp<strong>in</strong> up or down <strong>in</strong> an <strong>in</strong>cident state.<br />

Assum<strong>in</strong>g that the scatter<strong>in</strong>g matrix is <strong>in</strong>dependent of the sp<strong>in</strong> state of the electrons, the results<br />

discussed above describe the two cases when both sp<strong>in</strong>s po<strong>in</strong>t <strong>in</strong> the same direction (to be denoted<br />

as P(1,1) and P(1,1)). Thus what rema<strong>in</strong>s is to consider the case <strong>in</strong> which one <strong>in</strong>cident particle<br />

has sp<strong>in</strong> up and one <strong>in</strong>cident particle has sp<strong>in</strong> down. If the detection is also sp<strong>in</strong> sensitive, the<br />

probability which we determ<strong>in</strong>e is P(1,1). But <strong>in</strong> such an experiment we can tell which of the two

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!