Supercurrents in lead-copper-lead sandwiches

TL;DR: In this article, the de Gennes theory of the proximity effect has been used to account quantitatively for the behaviour of thin-film lead-copper-lead junctions with the lead in the superconducting state.

Abstract: The resistance of thin-film lead-copper-lead junctions has been studied with the lead in the superconducting state. The junctions will sustain a supercurrent up to a certain critical value above which a voltage appears, rising smoothly from zero as the current is increased. The effect of a magnetic field upon the critical current has demonstrated that the sandwiches behave phenomenologically as Josephson junctions. The critical current rises rapidly as the temperature is lowered, decreases exponentially with increasing thickness of copper and mcreases with increase of the mean free path of the copper. A simplified version of the de Gennes theory of the proximity effect has been used to account quantitatively for this behaviour. The experiments show that the coherence length of the paired electrons in the copper increases as the temperature decreases, implying that thermal fluctuations govern the decay of the pairs. From the value of the decay length, the interaction parameter in copper is estimated to lie between +0$\cdot$06 and +0$\cdot$14. The properties of these junctions are compared with those of junctions with insulating barriers.

Summary

UCRL-18l59

• This effect demonstrated the quenching of the superconductivity by the adjacent normal metal.
• This paper also contains a review of the proximity effect experiments mentioned above.
• The lead was evaporated from an electrically heated molybdenum boat.
• The authors shall therefore assume that critical current for which A J > w/4 are not significantly self-field limited.

Choice of materials

• The choice of the superconductor and normal metal from which specimens are to be made is of the highest importance.
• It is essential to select two metals which do not form intermetallic compounds (Chiou and Klockholm 1964, 1966) and whose mutual solubility is very low so that the diffusion of one into the other is negligible (Rose-Innes and Serin .1961 ). chosen as the superconductor and normal metal, respectively.
• It was intended to reduce the effects of boundary disorder and interdiffusion in the proximity effect specimens by using materials of very short mean free path.

UCRL-18159

• Bassewitz and Minnigerode (1964) have demonstrated that copper deposited on a substrate at 7tK has a density of about two-thirds the bulk.
• The results of experiments on films prepared and kept at /' low temperatures, although they may provide excellent evidence for the existence of the proximity-effect, cannot be used to deduce information on the bulk.
• The critical cUrrent of each junction was determined by passing a current between the appropriate pair of lead strips and noting the value at which a voltage first appeared; for reasons which will emerge later, the current was divided equally between the ends of each lead strip, as is indicated in Fig.
• The authors shail then describe the theory which relates the critical current of the junctions to their various parameters and finally analyse their data in the light of this theory.
• A correction has been applied for those specimens in which the mean free path differed markedly from the average value (different specimens are represented at different temperatures so that the average mean free path is not necessarily the same).

Determination of mean tree path

• For each film evaporated, a strip was deposited simultaneously on the same glass slide.
• The lead had a mean free path of about 10,000~,much greater than its coherence length, .~S' so that it was in the clean limit.
• This method of determining the mean free path in the Cu/Al films has some disadvantages.
• In addition, the resistance was measured longitudinally rather than transversely; if the films were at all non-uniform, as seems likely in practice, this measurement-gives an incorrect value.
• The uncertainties in the mean free path, perhaps as high as ±2o%, were the largest in the whole experiment.

Asymmetry of specimens

• The method of specimen preparation just described may give rise to some degree of asynnnetry (Hauser et al.1964:'.19 66 ; Hilsch andHilsch 1964) .
• Acertainamoimt of Qxidewill inevitably develop',on a freshly-deposited metal before the next layer isevaporated~, The oxid.e' groWth-rate on le~d is greater than on ,copper, so'that there maybe different amounts of.oxide at each SN'interface~:.
• It is hoped that,the>,high -vacua usedrena.erea this effect relatively unimportant.
• The second cause-,6fasymmetry is the dependence of the morphology of a thin film cmthe nature of its substrate: lead deposited on copper may nucleate ina.different~wayfromleadon glass.
• H"owever, si~ce the lead was in the clean limit:, slight differences in its structure should not introduce serious discrepancies.

'rhe cryostat

• The cryostat was designed to operate betw'een L2°K and 8°K and a longitudinal section of the lower part is shown in figure 2~ A thinwalled 3 em diameter copper-nickel tube, sealed at the low'er end, supported a vacuum can in the hellum bath.
• The specimen slide, mounted on a copper plate, was lO'W'ered into the tube on a thin stainless steel pipe attached to,a gas-tight plug at the top of the 3 cm tube.
• An expanded pOlystyrene plug separated the slide from the voltmeter which was maintained at the temperature of the helium bath by the exchange gas (helium at a pressure of about 5 torr) in the 3 cm tube.
• The plug was not gas-tight but 'inhibited convection of the exchange gas.

The voltmeter

• When the current through a junction exceeded the critical value, the voltage developed was detected by means of a superconducting galvanometer.
• The writer gratefully acknowledges the loan of a calibrated germanium thermometer by Dr. C. R. Barber of the National Physical Laboratory, Teddington, Middlesex. -10-.

i-v characteristics

• The voltage developed when the critical current was exceeded was small; this is to be expected because of the very low resistance of the metal barrier.
• Figure 4 shows on a larger scale the region around the critical current; there was no voltage step but rather a continuous rise in voltage as the current was increased from the critical value.
• The estimated resistance of the barrier w'as 7XIO-7 D and the measured differential resistance at currentsw'ell above the critical current approximately 5><10:-7 D.,.

Self-field limiting

• It has been shown (Anderson 1964 , Ferrell and Prange 1963 , Josephson 1964 ) that if the junction area is large or the current density high, the I self-field geherat~d by the current , is not negligible and gives rise to a Meissner effect within the junction.
• Unfortunately, a quantitative description for the present specimen configuration is hardly tractable but the authors shall very briefly discuss the one-dimensional situation.
• The authors main purpose is to determine which of their re.sults are seriously self-field limited.

c -19-UCRL-18159

• According to the B.CoS. theory of superconductivity (Bardeen et al,. 1957i'), a metal may be character'ized by the parameter NY, where N is the density of states at the Fermi surface and V the' effective electronelectron interaction.
• It should be remarked that in the present experiments the superconductor was in the clean limit so that there is some doubt as to the applicabilityof (13).
• There seems no a priori justification for preferring the latter value of ~. Temperature dependence of IceT) at low temperatures Although ( 13) is strictly true only near TCS it might be expected to hold approximately at much lower temperatureso.

The value of NV in copper

• It is evident that the experimental errors associated with the value of TeN are enormous and that it should not be taken too seriously.
• Has pointed out that in the vicinity of the SN interface, where the kernel in (6) contains contributions from both materials, it is * See de Gennes ( 1964).
• -30-UCRL-18159 not clear as to how one should treat the cut-off of the electron-phonon interaction.
• He has showed further that different 'methods of performing the cut-off in the superconductor gave rise to markedly different values of NV-for copper; this is a consequence of the fact that lead, which was used as the superconductor in all the experiments, is strong-coupling.
• -31-UCRL-18159 ;N rises with decreasing temperature a's suggested by ( 7), indicating that thermal fluctuations determine the decay length of the electron pairs.

Figures

Fig. 8'

Fig. 1;5

Fig. 1

Fig. 11'

Fig. 1

Fig. 5

Table II compares the present value of NV in copper with those obtained

Table III: Variation of ~'N with T

Fig. 6

Full text

Lawrence Berkeley National Laboratory
Recent Work
Title
SUPERCURRENTS IN LEAD-COPPER-LEAD SANDWICHES
Permalink
https://escholarship.org/uc/item/3wb7h547
Author
Clarke, John.
Publication Date
1968-04-01
eScholarship.org Powered by the California Digital Library
University of California

TWO-WEEK
LOAN
COPY
This
.is
a
l'ibrary
C-irculating
Copy
whi~h';'ay
be,borrowed
for
two
weeks.
For'
a
personal
retention
copy.
call
Tech.
Info.
Division.
Ext.
5545

DISCLAIMER
This document was prepared
as
an account
of
work sponsored by the United States
Government. While this document is believed to contain correct information, neither the
United States Government nor any agency thereof, nor the Regents
of
the University
of
California, nor any
of
their employees, makes any warranty, express or implied, or
assumes any legal responsibility for the accuracy, completeness, or usefulness
of
any
information, apparatus, product, or process disclosed, or represents that its use would not
infringe privately owned rights. Reference herein to any specific commercial product,
process, or service by its trade name, trademark, manufacturer, or otherwise, does not
necessarily constitute or imply its endorsement, recommendation, or favoring by the
United States Government or any agency thereof, or the Regents
of
the University of
California. The views and opinions
of
authors expressed herein do not necessarily state or
reflect those
of
the United States Government or any agency thereof or the Regents
of
the
University
of
California.

Submitted
tot
Proc.
Roy. Soc.
UNIVERSrry
OF
CALD"ORNIA
Lawrence
Radiation
Laboratory
Berkeley,
California
AEC
Contract
No. W-7405-eng-48
SUPERCURRENTS
IN
LEAD-CaPPER-LEAD
SANDWICHES
John
Clarke
·April1968
UCRL-18159
Prep:dnt

SUPERCURRENTS
IN
LEAD-COPPER-LEAD
SANDWICHES
*
John
Clarke
Cavendish
Laboratory,
Cambridge;
England
ABSTRACT
UCRL-18l59
The
resistance
of
thin-film
lead-copper-lead
junctions
has
been
studied
with
the
lead
in
the
superconducting
state.
The
junctions
will
sustain
a
supercurrf'nt
up
to
a
certain
critical
value
above which a
voltage
appears,
rising
smoothly
from
zero
as
the
current
is
increased.
The
effect
of
a
magnetic
field
upon
the
critical
current
has
demonstrated
that
the
sandwiches
behave
phenomenologically
as
Josephson
junctions.
The
critical
current
rises
rapidly
as
the
temperature
is
lowered,
decreases
exponentially
with
increasing
thickness
of
copper
and
increases
with
in-
crease
of
the
mean
'free
path
of
the
copper.
A
simplified
version
of
the
de Gennes
theory
of
the
proximity
effect
has
been
used
to
account
quanti-
tatively
for
this
behaviour.
The
experiments
show
that
the
coherence
length
of
the
paired
electrons
in
the
copper
increases
as
the
temperature
decreases,
implying
that
thermal
fluctuations
govern
the
decay
of
the
pairs.
From
the
value
of
the
decay
length,
the
interaction
parameter
in
copper
is
estimated
to
lie
between
+0.06
and
+0.14.
The
properties
of
these
junctions
are
compared
with
those
of
junctions
with
insulating
barriers.
*
Present
address:
Department
of
Physics,
University
of
California
at
Berkeley,
and
Inorganic
Materials
Research
Division
of
the
Lawrence
Radiation
Laboratory,
Berkeley,
California.