- Email: [email protected]
SIS mixers
2010
Physica 1()9 & I li)B 11982) 2i110-2(11'~ Norlh-Holland Publishing ('ompany
SIS M I X E R S T.G. PHILLIPS ('alifornia Institute of Technology. Pasadena. (~.4 ~)1125. I/.gA
and G.J. D O L A N Bell Labs, Murray Hill. N.J. 07974, /J.S'A
Small area superconducting thin tilm tunnel junctions have properties which make them suitable for high frequency (~ I00 GHz) heterodyne receivers. Both pair and single quasiparticle tunneling is present m the devices, but pair tunneling is found to be excessively noisy, whereas quasiparticle tunneling apparently gives hope of near quantum noise limited performance. The physical effect involved is photon assisted quasiparticle tunneling first observed by Dayem and Martin. A recent theory by Tucker has allowed a good understanding of many aspects of the phenomenon so that considerable hope exists for successful application of SIS mixer devices in fields such as millimeter and submillimeterwave astrononly. Some of the laboratory experiments and receiver construction efforts are reviewed in this article.
1. Introduction In the past few y e a r s s e v e r a l a d v a n c e s have b e e n m a d e in the use of l o w - t e m p e r a t u r e techn o l o g y to a c h i e v e i m p r o v e m e n t s in m i c r o w a v e a n d m i l l i m e t e r w a v e d e t e c t o r s . In p a r t i c u l a r , the field of a s t r o n o m i c a l m i l l i m e t e r w a v e s p e c t r o s c o p y using h e t e r o d y n e r e c e i v e r t e c h n i q u e s has b e n e f i t e d f r o m t h e d e v e l o p m e n t of s u p e r c o n d u c tor-insulator-superconductor (SIS) thin film junction mixer elements. S u p e r c o n d u c t i n g t u n n e l i n g effects h a v e b e e n in use for m a n y y e a r s for m i c r o w a v e d e t e c t i o n , b u t t h e m o s t p o p u l a r effect has b e e n that of J o s e p h s o n p a i r t u n n e l i n g . T r a d i t i o n a l l y this was developed with point contact (whisker) g e o m e t r i e s in o r d e r to k e e p the d e v i c e c a p a c i t a n c e small to allow h i g h - f r e q u e n c y o p e r a t i o n . H o w e v e r , it s e e m s that such J o s e p h s o n d e v i c e s are intrinsically noisy (see t h e r e v i e w article by R i c h a r d s [3D, a l t h o u g h t h e physical origin for t h e excess noise is difficult to d e s c r i b e . W i t h t h e m o r e r e c e n t a d v e n t of high r e s o l u t i o n p h o t o l i t h o g r a p h y a n d e l e c t r o n b e a m l i t h o g r a p h y tech0378-4363/82/0000-0000/$02.75
(T') 1982 N o r t h - H o l l a n d
nologies, it b e c a m e c l e a r that the w e l l - s t u d i e d q u a s i p a r t i c l e t u n n e l i n g effects c o u l d be useful for d e t e c t i o n , since the thin fihn j u n c t i o n s could be m a d e small e n o u g h to r e d u c e the u n w a n t e d j u n c t i o n p a r a l l e l c a p a c i t a n c e to negligible values, at least for m i c r o w a v e f r e q u e n c i e s . This line of r e a s o n i n g led to the use of SIS a n d SIN j u n c t i o n s for r e c e i v e r s [4,5]. H o w e v e r , i n d e p e n d e n t l y w o r k e r s using the f a m i l i a r S c h o t t k y b a r r i e r d i o d e r e c e i v e r s c o n c l u d e d that g r e a t e r sensitivity c o u l d be a c h i e v e d with d i o d e s by r e p l a c i n g the n o r m a l m e t a l e l e c t r o d e with a s u p e r c o n d u c t i n g elect r o d e , in effect using the n o n l i n e a r i t y of the q u a s i p a r t i c l e t u n n e l i n g curve in w h a t is k n o w n as t h e s u p e r - S c h o t t k y d e t e c t o r [6]. O f these v a r i o u s t y p e s of q u a s i p a r t i c l e t u n n e l i n g d e t e c t o r s the most p r o m i s i n g s e e m s to be the SIS. A l t h o u g h p a i r t u n n e l i n g is a l l o w e d in this device, thc excess noise a s s o c i a t e d with J o s e p h s o n d e t e c t o r s can be a v o i d e d by c o n s t r a i n i n g o p e r a t i o n to the quasiparticle current dominated regime. The I-V curve n o n l i n e a r i t y is s t r o n g e r for t h e SIS t h a n for e i t h e r t h e SIN o r s u p e r - S c h o t t k y . A s is c l e a r f r o m t h e e a r l y p a p e r s [4, 51, t h e full
T.G. Phillips and G.J. Dolan / SIS mixers
importance of the photon assisted tunneling process was not immediately grasped. The theoretical discussion of Tucker [2] primarily concerned the super-Schottky diode, which, like the SIN, does not have a very obvious manifestation of the photon assisted tunneling process in the 1 - V characteristics. However, as soon as clear photon assisted tunneling steps were observed in the SIS I - V characteristics [7] it was obvious that the detection process was the same as for the D a y e m - M a r t i n effect [l] and the quantum nature of the devices was established. It should be pointed out that the requirement of thermal cycling and chemical stability, which is present for computer applications of superconducting circuits, is the same as that for SIS mixers. Thus, another reason for attempting to use SIS mixers for receiver applications is the recent advance in metallurgy for the thin film junctions. A successful high-frequency SIS mixer junction must be of very small area, probably considerably less than !/xm:, and be stable and cyclable. With these constraints it is hard to produce a structure with the same sharpness in the I - V characteristics as can be obtained using traditional metal (e.g. Sn) large area junctions. This is the primary reason that the early attempts [4, 5] did not definitively reveal the photon assisted effects.
2. Heterodyne receivers Before discussing the SIS mixer receivers specifically, some attempt will be made to put them in perspective by briefly describing the current status for high-frequency mixer receivers in general. To achieve high spectral resolution it is usual in the radio, millimeter and submillimeter wave bands to employ heterodyne receivers in which a local oscillator wave is coherently mixed with the signal. The mixer element is usually a diode-type device providing nonlinear response. The device must be capable of carrying currents at the signal
2011
frequency. The difference frequency between the signal and the local oscillator is available as an intermediate frequency (IF) over a somewhat restricted range (usually less than a GHz) depending on the capability of the low noise IF amplifier. Such a receiver is characterized by a noise temperature (TR) which is contributed to by the noise generated in the mixer, which is equal to or greater than the quantum noise due to fluctuations in the local oscillator power, plus the noise of the IF amplification chain which appears as effectively multiplied by the power conversion loss factor (L) of the mixer. (The case of gain is properly handled by this expression in that L is less than 1 and TIF becomes less important.) T~= TM+ L T w .
(1)
For the purposes of this paper all numbers are quoted for single side band operation (SSB), which is the usual spectroscopic mode. Noise temperature measurements are usually made with white noise, fixed temperature loads applied to the front end of the receiver, which affect both side bands equally. For small values of the IF it is often the case that the double side band values are one-half the SSB values, since both side bands convert power equally. In principle power loss to the unwanted side band can be avoided (image suppression) by differential matching if the IF is large enough, or by the use of a two-element front end. At higher frequencies these techniques are not usually practicable. Spectroscopy is performed by dividing the IF band into channels of width Au and rectifying and integrating the noise in each channel separately. All mixer receivers operating in the Rayleigh-Jeans limit (hu < kTR) provide a signal to noise ratio given by the Dicke radiometer equation: s _ T~ ~ / ~
N
(2)
T.
where Ts is the equivalent black body temperature of the signal and At is the integration
2012
T.G. Phillips a n d G.J. D o l a n / SIS mixers
time. Clearly it is very important to reduce TR to a minimum. All radiometric devices are governed by eq. (2); however, not all are classical diode mixers. Quantum mixers, discussed below, and bolometer mixers, not discussed in this article, have properties which make them suitable for certain purposes. Figure 1 shows a graph for measured mixer noise temperatures (TM, SSB) equation (1), for various devices at frequencies up to ~ 8 0 0 GHz. The lowest noise temperatures at high frequencies are found for InSb bolometer mixers. These are bulk, single crystal devices. which do not suffer from embedding network difficulties (i.e. they have easily computed resistive and reactive impedance c o m p o n e n t s in a waveguide mount), however they have very limited IF bandwidths due to the intrinsic electronic relaxation time of the material.
~4 L/
4
BOOK
O
SUPER
0
£nSb B J : 2 , Z E T E R MIXEF~
0
JOSE PH S:l": SIS
effect
Photon assisted tunneling effects in SIS june* tions were first observed by D a y e m and Martin [1]. The basic effect is that the tunneling current is modified at bias voltages nhHe away from the gap. This is distinct from pair tunneling photon modified effects (Josephson tunneling) which occur at nhH2e f r o m V :: II. Fig. 2(a) shows this effect for a modern junction with an incident millimeterwave field a! 115 GHz. An initial explanation was given by Tien and Gordon [221 in terms of a photon modified density of states.
i\y
15
(d)
%'-H:Z,TTK',
IFAwTIL
~E
+
q~
t
/
\J
I
II
i°il
SCH')TTk~
QLJA
3. 1. The Dayem-Martin
(c)
l
1,05
3. SIS mixer elements
I
i I
1
/
i'
(b)
295K :
\,,
LCAD
~'7K LOAD
[1_
÷ ! 'L ,a j
L, X
i:} 2
~
( I
40
ILl
i,} I
44 47 50 53
:L
0
~?,,
~
20
dbm dbm .lbm dbm
•,
.z
~}~
j..dS iO o 10
j,y
V
L o OFF
1
I?
I( ~'
z (GHz)
0 0
I
2
V
Fig. 1. Single side band mixer noise temperatures for various millimeter and submillimeter heterodyne detectors. Data for 3I)() K Schottky mixers are from [8,-11]; 15 K Schottky mixers from [12 141; super Schottkys from [51 and [15]: lnSb bolometer mixers from [161: Josephsnn mixers from [1"7, b,~l: quasiparticle mixers from [l~q and [20, 2 ! ]
3
( rr, Y '.,
Fig. 2. (a) Current ",s. voltage characteristics for an SIS junction with various 115GH;, local oscillator power levels applied. (b) IF output power with various temperature signal loads applied. (c) Single side hand mixer noise temperature (d) Single side band mixer conversion loss.
T.G. Phillips and G.J. Dolan / SIS mixers
The recent detailed theory of Tucker [2] is based on the same explanation and seems to be entirely appropriate for the Pb and Sn junctions currently in use. However, it has been suggested by EntinWohlman [23] that effects of the microwave field on the quasiparticle distribution function will be important for A1 junctions. No detailed discussion of the theory will be attempted in this article, but it must be emphasized that one of the major advantages that the SIS detector has is the fairly complicated quantum structure imposed on the device characteristics, which can be accurately modelled using Tucker's theory. As will be shown in the following sections, this capability should lead to the achievement of near quantum noise limited performance because the effects of the microwave environment on the device can be deduced and unwanted imbedding network effects minimized. Fig. 2(b) shows the mixer effect of the SIS detector. When both local oscillator and signal power are applied the power output is of the form shown, as a function of junction bias voltage. The two curves represent the effects of white noise inputs due to two different temperature sources. Mixing is strongest on the fiat current steps of the Dayem-Martin effect, with the first step below the gap voltage (2A - 2.6 mV for the Pb-ln alloy of this junction) being dominant. From the curves of fig. 2(b) it is possible to deduce the parameters of eq. (1), provided TIE is known independently. T~ and L are displayed in fig. 2(c and d). The mixer noise temperature and conversion loss are both minimum in the region of the first photon assisted tunneling step. A major advantage of the SIS mixer, as compared to semiconductor diode devices, is the very low level of local oscillator power required, only 10-sW in the example of fig. 2. It may be noticed in fig. 2 that at bias voltages below 1.5mV the detector performance is strongly degraded. The noise power increases rapidly as the bias voltage decreases, with an equivalent increase in conversion loss and mixer noise temperature. This bias region is that in
2013
Fig. 3. An SIS junction manufactured at Bell Labs. by photoresist bridge techniques [4]. which Josephson current effects are noticeable, particularly the well-known hysteretic behavior of the I - V characteristics. These effects can be suppressed by application of a magnetic field, or
Fig. 4. A series array of 20 SIS junctions.
21114
7\G. Phillips, and G.J. l)olan / ,S'l,gmixers
simply ignored by confining operation to larger bias values. The reader may recognize that this latter p r o c e d u r e will fail at high frequencies where tip approaches 2A. A discussion of the ditticulties and possible solutions may be found m refs. 1~+ and 21. 3.2.
Device
structure
Among the advantages of the SIS devices is the planar nature of the thin fihn structure which is suited to production by m o d e r n p h o t o or electron b e a m lithography. O n e technique for small junction (<1 # m e) fabrication is that of Dolan el al. [4]. Electron beam lithography is used to generate masks from which photoresist bridge structures arc made. By evaporating past the bridge at different angles, overlapping structures are formed. Fig. 3 shows an example of a small area (0.1 # m : ) P b - i n junction made in this way. Very importantly, arrays of junctions can also be
formed. Fig. 4 shows a series array of 2(I SIS junctions nlade at Bell l+abs. The ( ' h a h n c r s I_Jniversiiv g r o u p [21,24] has been active in pursuing the series array technology. Example of superc~mducting thin film ~,trtictures m o u n t e d in waveguides are given in ligs 5 and n. Fig. 5 shows a Bell Labs single junction nlounted across a 1 1 5 G H i waveguide. Integrated to the UllCtiOll (which is at the center ~ll the waveguide) is tl ,,uperconducling choke ,,co lion which prevents the signal and local oscillator radiation fl-tml propagating in the striplinc 11: section. F'ig. (~ indicates the lawmt for the Sw'cdish array 12@ which makes use of ',in a n t e n n a structure within the wavcguidc to couple the currents to the junctitm array slightly outside the waveguide itself. Many other structures are possible, and as progress is made in understanding the simple structures it should bc possible It> design more complex structures to optimize the w a v e g u i d e - j u n c t i o n thatch.
Zig. 5. A substrate supporting an SIS junction with d.c. and IF connections and r.f. choke structure, mounted across a me-quarter height WR-8 waveguide [16].
T.G. Phillips and G.J. Dolan / SIS mixers
(a)
2015
£ -,
/
~";~
\ ........
~
(b)
•
//
<.- ~. - ~ > "
~,,~ ,, ground
/
Backshort
-lo~m
,if" i
f
Fig. 6. A monopole waveguide a n t e n n a coupled to a series array of SIS junctions and an IF/bias line with R.F. choke. After R u d n e r et al. [24].
i
S
i
4. SIS mixer performance A major hope for the SIS mixer is that the great deal of quantum structure will provide the receiver engineer with an excellent tool for analyzing his device. That this same philosophy has failed for the Josephson effect is surprising. In what follows the Josephson current is ignored, since we confine our attention to the structure near the gap voltage. Tucker's theory [2] has been computationally encoded by Wengler and Woody [25] so that a plot can be made of any one of the device parameters as a function of d.c. or local oscillator bias for variable values of the parameters. This formalism is exact in so far as the theory is exact, including, for instance, the quantum mechanically induced reactance term which appears in the device impedance parameters [2]. Previous formalisms have been approximate in some senses [26, 27]. The object of the exercise is to deduce the environment that the tunnel junction sees in terms of the real and imaginary parts of an admittance function. Ys. Changes to the device structure or to the waveguide environment can then be contemplated to improve the device performance. Fig. 7 shows a set of computer generated I - V characteristics for various values of the amplitude and phase of Ys. The input to the theory is the digitized I - V characteristic with local
j V -'~
ARG (YS) = - l . S
V --~
0 0
I V-'-~ l.S
Fig. 7. Plots of I - V characteristics for a P b - I n junction at 115 GHz. T h e full line without local oscillator power is from experiment and is input data for the computer program, The crosses represent the experimental characteristics with local oscillator power applied, and the full lines are the computer generated local oscillator applied curves for various values of the amplitude and phase of Ys (see text).
oscillator power applied. The local oscillator amplitude has been fixed by requiring that the computer generated and observed I - V characteristic (L.O. power applied) agree at the center of the first photon step. Clearly, an excellent fit is obtained (for all the quasiparticle structure) for only one value of Ys. The device under investigation here is a Bell Labs. junction at 97 GHz, in a waveguide mount. Apparently the junction sees an admittance of about three times the waveguide admittance. An example of the power of the theoretical expression is shown in fig. 8, where the best fit Ys p a r a m e t e r found for one value of the local oscillator power is held constant, while the local oscillator power is varied in 3 d b steps, both
T.G. Phillips and G.J. Dolan / SIS mixers"
2/11~
'
'
I
I "
//~
.~'/*
i
{
4.-. t. ,
al ,
O
.~.
0.5
••
,- / / y7 M ~ /
v/vG
t
10
15
•
<
'2 ~
~ •
Fig. ~, P l o t s of r e d u c e d c u r r e ~ t vs. ~ o [ t a g e ft~r the s a m e i u n c t i o n as in fig. 7, f o r v a r i o u s v a l u e s of t h e local o s c i l l a t o r power. A-zero L . O . : ( ' - I . . ( ) . as f o r fig. 7: B - 3 d B less t h a n (7: l ) - 3 d B m o r e t h a n ('. T h e c r o s s e s r e f e r to e x p e r i -
i ¢:
.
,,r , ,
Fig. ll). ! V c h a r a c t e r i s t i c s fl~l a tin i u n c t i o n at 3 ( ~ G H / , s h o w i n g h y s t e r e t i c b e h a v i o r at t h e first s t e p . A f t e r lVlcGrath e! al. [27].
mental data.
experimentally and theoretically. A good tit is obtained for all the steps. The shape of the local oscillator induced steps in fig. 8 is not very remarkable, the slope at the step center being similar to that of the zero 1..O. curve at the same bias voltage. However. under different conditions of imbedding admittance (Y,) other workers have observed negative differential resistance steps, implying a device with arbitrarily large gain [281. An example is the work of Kerr el al. [2q] who observed gain at 1 1 5 G H z with an array of Pb alloy junctions. Their I - V characteristics are shown in tig. ~). A
60
second example is that observed by McGrath et al. [27], shown in fig. It) for a 36 G H z experiment with a tin junction. In this case the junction is current biased so that hysteresis is observed m the negative resistance region. It is not obvious that a device will have its lowest noise temperature when the conversion loss is a minimum. (Operation under conditions of large gain would be dilBcult in an engineering sense and might lead to considerable
>10(2
O C2~
Ii 20 _ _ i
Ci
0 -
I
o
IO
0 5
I
3
IO
30
I¥%t 20
5o
40
5o
BI~S VOLTAGE (MV)
Fig. ~. I - V c h a r a c t e r i s t i c s f o r a Pb alloy series a r r a y at 115 GDIz, s h o w i n g n e g a t i v e d i f f e r e n t i a l r e s i s t a n c e at the firsl s t e p . A f t e r K e r r et al_ {2g{.
Fig. 11. C o n t o u r s of c o m p u t e d ~,ystem t e m p e r a l u r c " ;t ~, ;l f u n c t i o n of a m p l i t u d e a n d p h a s e of Y,. T h e c r o s s r e p r e s e n t s t h e p o s i t i o n of t h e j u n c t i o n of fig. 7. T h e r e g i o n to the left of t h e d a s h e d line is w h e r e n e g a t i v e d i f f e r e n t i a l r e s i s t a n c e o c c u r s o n t h e lirst s t e p .
T.G. Phillips and G.J. Dolan / SIS mixers 0
....
, ....
, ....
, ....
2017
HOT AND COLD LOADS I% DIELECTRIC BEAMSPLITTER
.~ -tO
/ / / / /
/ / / / /
.
-20
.
.
.
.
.
.
.
/
) LENS
/
I
V/VG Fig. 12. C o n v e r s i o n loss as a function of r e d u c e d bias v o l t a g e for the j u n c t i o n of fig. 7. T h e c o m p u t e d p e r f o r m a n c e is the full line, t h e e x p e r i m e n t a l v a l u e s are the crosses.
excess noise.) For the junction of fig. 7, a computed plot of system noise (TM), arising from current shot noise plus a 10 K IF amplifier contribution, is shown in fig. 11 as a function of the amplitude and phase of Y~. The region of negative differential resistance steps and arbitrarily large gain is also shown. It is interesting to note that the position of the actual junction is well away from the proper value of Y~ to achieve minimum noise performance. Fig. 12 shows the computed conversion loss for that junction at its present Ys value as compared with the measured loss. The agreement is excellent. The best reported values of mixer noise temperatures are 9 K at 3 6 G H z [27], 6 0 K at 115 G H z [16, 29] and 300 K at 230 GHz [16].
_•FEEDHORN
IM_'xERL_..__.JGaAs FET L -- 1.0-15GHz IF. BACKSHIBLOCKoRT, [ J Ii AMpLIF'E~I DC 8~AS
Fig. 13. S c h e m a t i c d i a g r a m of the 115 G H z receiver.
signal to pass freely and couples in only - 1 % of the L.O. power. This is possible because of the low L.O. power requirement for the SIS mixer. Fig. 14 is a photograph of the receiver dewar with radiation shields removed. A prototype version of this receiver, built at
5. Receiver performance A receiver to mount at the focus of a 10m telescope at Caltech's Owens Valley Radio Observatory, using an SIS mixer element, has been completed. A simplified block diagram of the receiver is shown in fig. 13. This figure also demonstrates the method of measurement of the receiver and mixer performance parameters. The parts within the box are in vacuum, bolted to the cold plate of a liquid helium cryostat. Some measurements have been made at 4.2 K, but the best results are generally found at lower temperatures. The dielectric beam splitter allows the
Fig. 14. Picture of the I 1 5 G H z r e c e i v e r to b e used for a s t r o n o m i c a l o b s e r v a t i o n s . T h e principal e l e m e n t s are: A, L.O. f e e d h o r n ; B, 1% dielectric b e a m s p l i t t e r ; C, m i x e r f e e d h o r n ; D, m i x e r block; E, b a c k s h o r t t u r n e r ; F, 3 0 d B d i r e c t i o n a l c o u p l e r for c o u p l i n g test signals into the IF line; G, 1.4 G H z c o o l e d G a A s F E T I F amplifier; H. IF o u t p u t port.
201;4
T.G. Phillips and G.J. Dolan / SIS mixers
Bell Labs., was used in June 1979 to determine its suitability for astronomical operation. No m a j o r problems were encountered. The current receiver was successfully used during the winter of 1980. Table I indicates p e r f o r m a n c e figures for the prototype and current receiver. The current receiver has a system t e m p e r a t u r e of 23(1 K at l l 5 G H z , which is comparable with the best c o n t e m p o r a r y cooled Schottky diode receivers. Also shown is a very recent report from Linke for a similar prototype receiver at Bell Labs, [29]. It seems that SIS receivers are already competitive with the best available alternatives in the millimeterwave regime and should eventually
,i
o
dA£: i76~,
/
J
) I
3
0
I
()VR()
h m k e (BTI .)
95 K 50 K ~ dB 4110 K
t 41 ) K II K u dB 22711K
$1) K s K S dB 131) K
prove superior due to the very great developmerit potential. As a final note, it is probable that considerable i m p r o v e m e n t s in the metallurgical aspects of junction fabrication will aid developments. An example is the Nb-A1 oxide-PbBi system of Gurvitch and Rowell [31]. These workers have produced very high quality junctions using a Nb base electrode coated with a thin htyer of /\1 which is then oxidized. As shown in lig. 15, the I - V characteristic of such junctions can bc sharper in curvature than Pb alloys over a wide range of AI thicknesses. Also, Nb-based junctions are known to be very durable.
The authors gratefully acknowledge the many contributions of their colleagues M. Wengler, D. Woody, R. Miller, E. Sutton and J. Rowell. Work at Caltech is supported by National Science Foundation grant no. AST-8007645.
References
dA2 : 364~
2
FM Tw L 1'~
Prototype (BTI.)
Acknowledgements
i
=
Table I
~
3
4-
VOUTAGE (MI~LIVOLTS~ Fig. 15. I - V characteristics of large area Nb-A1 oxide-PbBi junctions for different thicknesses of AI (dAI). Very sharp characteristics are obtained in this way over a wide range of da~, suggesting that the small high current density devices required for the present application could be m a d e using this metallurgical system. After Gurvitch and Rowell [311.
[11 A.H. D a y e m and R.,I. Martin, Phys. Re~. I,ctt. S (I~,~(Q) 24~. [21 J.R. Tucker, I E E E J. ()uantunl Electron. ()E-15 (197t, ~) 1234. 131 P.l,. Richards, in: Semiconductors and Semimclals, xol 12 Willardson and Beer. eds. (Academic Press. New York, 1977) p. 395. [4] G.J. Dolan, T.G. Phillips and D.P. Woody, Appl. Ph,.s. Lett. 34 (1979) 347. [5] P . L Richards. T.M. Shen, R.E. Harris and F.l.. Lhwd, Appl, Phys. Len. 34 11970) 345.
T.G. Phillips and G.J. Dolan / SIS mixers
[6] M. McColl, M.F. Bottjer, A.B. Chase, R.J. Pedersen, A.H. Silver and J.R. Tucker, IEEE Trans. Magn. MAG15 (1979) 468. [7] D.P. Woody, T.G. Phillips and G.J. Dolan, Proc. Nat. Radio Sci. Meeting, Nat. Acad. Sci, USNC/URSI (1979) p. 43. [8] A.R. Kerr, IEEE Trans. Microwave Theory Tech. MTF-23 (1975) 781. [9] E.R. Carlsen, M.V. Schneider and T.F. McMaster, IEEE Trans. Microwave Theory Tech. MTT-26 (1978) 71/6. [111] N.R. Erickson, 1EEE Int. Microwave Syrup. MT'f-S (1978) 438. [11] M.R. Fetterman, P.E. Tannenwald, B.J. Clifton, C.D. Parker, W.D. Fitzgerald and N.R. Erickson, Appl. Phys. Lett. 33 (1978) 151. [12] E. Kollberg, URSI Symp. Millimeter Tech. in Radioastronomy, Grenoble (1980). [13] R.A. Linke, M.V. Schneider and A.Y. Cho, IEEE Trans. Microwave Theory Tech. MTlr-26 11978) 935. [14] S. Weinreb and J. Payne, private communication. [15] W. Wilson, private communication. [16] T.G. Phillips, D.P. Woody, G.J. Dolan, R.E. Miller and R.A. Linke, IEEE Trans. Magn. MAG-17 (1981) 684. [17] Y. Taur, J.H. Classen and P.L. Richards, Trans.
2019
Microwave Theory Tech. MTIr-22 (1976) 10115. [18] Y. Taur and A.R. Kerr, Appl. Phys. Lett. 32 (1978) 775. [19] Th. de Graauw, private communication. [20] T.M. Shen, P.L. Richards, R.E. Harris and F.L. Lloyd, Appl. Phys. Lett. 36 (1980) 777. [21] S. Rudner, M.J. Feldman, E. Kollberg and T. Claeson, in: SQUID '80' ed. Halbohm and Lubbig, eds. (W. de Gruyter, Berlin, 1981) p. 91/1. [22] P.K. Tien and J.P. Gordon, Phys. Rev. 129 (1963) 647. [23] O. Entin-Wohlman, Phys. Rev. B22 (1980) 5225. [24] S. Rudner, M.J. Feldman, E. Kollberg and T. Claeson, J. Appl. Phys. (1981) to be published. [25] M.J. Wengler and D.P. Woody, in preparation. [26] G J . Dolan, R.A. Linke, T.G.L.C. Sollner, D.P. Woody and T.G. Phillips, IEEE Trans. Microwave Theory Tech. M'VF-29 (1981) 87. [27] W.R. McGrath, P.L. Richards, A.D. Smith, H. van Kemper and R.A. Batchelor, Appl. Phys. Lett. (1981), to be published. [28] J.R. Tucker, Appl. Phys. Lett. 36 (1980) 477. [29] A.R, Kerr, S.K. Pan, M.J. Feldman and A. Davidson, Physica, this volume. [30] R.A. Linke, private communication. [31] M. Gurvitch and J.M. Rowell, to be published.
Physica 1()9 & I li)B 11982) 2i110-2(11'~ Norlh-Holland Publishing ('ompany
SIS M I X E R S T.G. PHILLIPS ('alifornia Institute of Technology. Pasadena. (~.4 ~)1125. I/.gA
and G.J. D O L A N Bell Labs, Murray Hill. N.J. 07974, /J.S'A
Small area superconducting thin tilm tunnel junctions have properties which make them suitable for high frequency (~ I00 GHz) heterodyne receivers. Both pair and single quasiparticle tunneling is present m the devices, but pair tunneling is found to be excessively noisy, whereas quasiparticle tunneling apparently gives hope of near quantum noise limited performance. The physical effect involved is photon assisted quasiparticle tunneling first observed by Dayem and Martin. A recent theory by Tucker has allowed a good understanding of many aspects of the phenomenon so that considerable hope exists for successful application of SIS mixer devices in fields such as millimeter and submillimeterwave astrononly. Some of the laboratory experiments and receiver construction efforts are reviewed in this article.
1. Introduction In the past few y e a r s s e v e r a l a d v a n c e s have b e e n m a d e in the use of l o w - t e m p e r a t u r e techn o l o g y to a c h i e v e i m p r o v e m e n t s in m i c r o w a v e a n d m i l l i m e t e r w a v e d e t e c t o r s . In p a r t i c u l a r , the field of a s t r o n o m i c a l m i l l i m e t e r w a v e s p e c t r o s c o p y using h e t e r o d y n e r e c e i v e r t e c h n i q u e s has b e n e f i t e d f r o m t h e d e v e l o p m e n t of s u p e r c o n d u c tor-insulator-superconductor (SIS) thin film junction mixer elements. S u p e r c o n d u c t i n g t u n n e l i n g effects h a v e b e e n in use for m a n y y e a r s for m i c r o w a v e d e t e c t i o n , b u t t h e m o s t p o p u l a r effect has b e e n that of J o s e p h s o n p a i r t u n n e l i n g . T r a d i t i o n a l l y this was developed with point contact (whisker) g e o m e t r i e s in o r d e r to k e e p the d e v i c e c a p a c i t a n c e small to allow h i g h - f r e q u e n c y o p e r a t i o n . H o w e v e r , it s e e m s that such J o s e p h s o n d e v i c e s are intrinsically noisy (see t h e r e v i e w article by R i c h a r d s [3D, a l t h o u g h t h e physical origin for t h e excess noise is difficult to d e s c r i b e . W i t h t h e m o r e r e c e n t a d v e n t of high r e s o l u t i o n p h o t o l i t h o g r a p h y a n d e l e c t r o n b e a m l i t h o g r a p h y tech0378-4363/82/0000-0000/$02.75
(T') 1982 N o r t h - H o l l a n d
nologies, it b e c a m e c l e a r that the w e l l - s t u d i e d q u a s i p a r t i c l e t u n n e l i n g effects c o u l d be useful for d e t e c t i o n , since the thin fihn j u n c t i o n s could be m a d e small e n o u g h to r e d u c e the u n w a n t e d j u n c t i o n p a r a l l e l c a p a c i t a n c e to negligible values, at least for m i c r o w a v e f r e q u e n c i e s . This line of r e a s o n i n g led to the use of SIS a n d SIN j u n c t i o n s for r e c e i v e r s [4,5]. H o w e v e r , i n d e p e n d e n t l y w o r k e r s using the f a m i l i a r S c h o t t k y b a r r i e r d i o d e r e c e i v e r s c o n c l u d e d that g r e a t e r sensitivity c o u l d be a c h i e v e d with d i o d e s by r e p l a c i n g the n o r m a l m e t a l e l e c t r o d e with a s u p e r c o n d u c t i n g elect r o d e , in effect using the n o n l i n e a r i t y of the q u a s i p a r t i c l e t u n n e l i n g curve in w h a t is k n o w n as t h e s u p e r - S c h o t t k y d e t e c t o r [6]. O f these v a r i o u s t y p e s of q u a s i p a r t i c l e t u n n e l i n g d e t e c t o r s the most p r o m i s i n g s e e m s to be the SIS. A l t h o u g h p a i r t u n n e l i n g is a l l o w e d in this device, thc excess noise a s s o c i a t e d with J o s e p h s o n d e t e c t o r s can be a v o i d e d by c o n s t r a i n i n g o p e r a t i o n to the quasiparticle current dominated regime. The I-V curve n o n l i n e a r i t y is s t r o n g e r for t h e SIS t h a n for e i t h e r t h e SIN o r s u p e r - S c h o t t k y . A s is c l e a r f r o m t h e e a r l y p a p e r s [4, 51, t h e full
T.G. Phillips and G.J. Dolan / SIS mixers
importance of the photon assisted tunneling process was not immediately grasped. The theoretical discussion of Tucker [2] primarily concerned the super-Schottky diode, which, like the SIN, does not have a very obvious manifestation of the photon assisted tunneling process in the 1 - V characteristics. However, as soon as clear photon assisted tunneling steps were observed in the SIS I - V characteristics [7] it was obvious that the detection process was the same as for the D a y e m - M a r t i n effect [l] and the quantum nature of the devices was established. It should be pointed out that the requirement of thermal cycling and chemical stability, which is present for computer applications of superconducting circuits, is the same as that for SIS mixers. Thus, another reason for attempting to use SIS mixers for receiver applications is the recent advance in metallurgy for the thin film junctions. A successful high-frequency SIS mixer junction must be of very small area, probably considerably less than !/xm:, and be stable and cyclable. With these constraints it is hard to produce a structure with the same sharpness in the I - V characteristics as can be obtained using traditional metal (e.g. Sn) large area junctions. This is the primary reason that the early attempts [4, 5] did not definitively reveal the photon assisted effects.
2. Heterodyne receivers Before discussing the SIS mixer receivers specifically, some attempt will be made to put them in perspective by briefly describing the current status for high-frequency mixer receivers in general. To achieve high spectral resolution it is usual in the radio, millimeter and submillimeter wave bands to employ heterodyne receivers in which a local oscillator wave is coherently mixed with the signal. The mixer element is usually a diode-type device providing nonlinear response. The device must be capable of carrying currents at the signal
2011
frequency. The difference frequency between the signal and the local oscillator is available as an intermediate frequency (IF) over a somewhat restricted range (usually less than a GHz) depending on the capability of the low noise IF amplifier. Such a receiver is characterized by a noise temperature (TR) which is contributed to by the noise generated in the mixer, which is equal to or greater than the quantum noise due to fluctuations in the local oscillator power, plus the noise of the IF amplification chain which appears as effectively multiplied by the power conversion loss factor (L) of the mixer. (The case of gain is properly handled by this expression in that L is less than 1 and TIF becomes less important.) T~= TM+ L T w .
(1)
For the purposes of this paper all numbers are quoted for single side band operation (SSB), which is the usual spectroscopic mode. Noise temperature measurements are usually made with white noise, fixed temperature loads applied to the front end of the receiver, which affect both side bands equally. For small values of the IF it is often the case that the double side band values are one-half the SSB values, since both side bands convert power equally. In principle power loss to the unwanted side band can be avoided (image suppression) by differential matching if the IF is large enough, or by the use of a two-element front end. At higher frequencies these techniques are not usually practicable. Spectroscopy is performed by dividing the IF band into channels of width Au and rectifying and integrating the noise in each channel separately. All mixer receivers operating in the Rayleigh-Jeans limit (hu < kTR) provide a signal to noise ratio given by the Dicke radiometer equation: s _ T~ ~ / ~
N
(2)
T.
where Ts is the equivalent black body temperature of the signal and At is the integration
2012
T.G. Phillips a n d G.J. D o l a n / SIS mixers
time. Clearly it is very important to reduce TR to a minimum. All radiometric devices are governed by eq. (2); however, not all are classical diode mixers. Quantum mixers, discussed below, and bolometer mixers, not discussed in this article, have properties which make them suitable for certain purposes. Figure 1 shows a graph for measured mixer noise temperatures (TM, SSB) equation (1), for various devices at frequencies up to ~ 8 0 0 GHz. The lowest noise temperatures at high frequencies are found for InSb bolometer mixers. These are bulk, single crystal devices. which do not suffer from embedding network difficulties (i.e. they have easily computed resistive and reactive impedance c o m p o n e n t s in a waveguide mount), however they have very limited IF bandwidths due to the intrinsic electronic relaxation time of the material.
~4 L/
4
BOOK
O
SUPER
0
£nSb B J : 2 , Z E T E R MIXEF~
0
JOSE PH S:l": SIS
effect
Photon assisted tunneling effects in SIS june* tions were first observed by D a y e m and Martin [1]. The basic effect is that the tunneling current is modified at bias voltages nhHe away from the gap. This is distinct from pair tunneling photon modified effects (Josephson tunneling) which occur at nhH2e f r o m V :: II. Fig. 2(a) shows this effect for a modern junction with an incident millimeterwave field a! 115 GHz. An initial explanation was given by Tien and Gordon [221 in terms of a photon modified density of states.
i\y
15
(d)
%'-H:Z,TTK',
IFAwTIL
~E
+
q~
t
/
\J
I
II
i°il
SCH')TTk~
QLJA
3. 1. The Dayem-Martin
(c)
l
1,05
3. SIS mixer elements
I
i I
1
/
i'
(b)
295K :
\,,
LCAD
~'7K LOAD
[1_
÷ ! 'L ,a j
L, X
i:} 2
~
( I
40
ILl
i,} I
44 47 50 53
:L
0
~?,,
~
20
dbm dbm .lbm dbm
•,
.z
~}~
j..dS iO o 10
j,y
V
L o OFF
1
I?
I( ~'
z (GHz)
0 0
I
2
V
Fig. 1. Single side band mixer noise temperatures for various millimeter and submillimeter heterodyne detectors. Data for 3I)() K Schottky mixers are from [8,-11]; 15 K Schottky mixers from [12 141; super Schottkys from [51 and [15]: lnSb bolometer mixers from [161: Josephsnn mixers from [1"7, b,~l: quasiparticle mixers from [l~q and [20, 2 ! ]
3
( rr, Y '.,
Fig. 2. (a) Current ",s. voltage characteristics for an SIS junction with various 115GH;, local oscillator power levels applied. (b) IF output power with various temperature signal loads applied. (c) Single side hand mixer noise temperature (d) Single side band mixer conversion loss.
T.G. Phillips and G.J. Dolan / SIS mixers
The recent detailed theory of Tucker [2] is based on the same explanation and seems to be entirely appropriate for the Pb and Sn junctions currently in use. However, it has been suggested by EntinWohlman [23] that effects of the microwave field on the quasiparticle distribution function will be important for A1 junctions. No detailed discussion of the theory will be attempted in this article, but it must be emphasized that one of the major advantages that the SIS detector has is the fairly complicated quantum structure imposed on the device characteristics, which can be accurately modelled using Tucker's theory. As will be shown in the following sections, this capability should lead to the achievement of near quantum noise limited performance because the effects of the microwave environment on the device can be deduced and unwanted imbedding network effects minimized. Fig. 2(b) shows the mixer effect of the SIS detector. When both local oscillator and signal power are applied the power output is of the form shown, as a function of junction bias voltage. The two curves represent the effects of white noise inputs due to two different temperature sources. Mixing is strongest on the fiat current steps of the Dayem-Martin effect, with the first step below the gap voltage (2A - 2.6 mV for the Pb-ln alloy of this junction) being dominant. From the curves of fig. 2(b) it is possible to deduce the parameters of eq. (1), provided TIE is known independently. T~ and L are displayed in fig. 2(c and d). The mixer noise temperature and conversion loss are both minimum in the region of the first photon assisted tunneling step. A major advantage of the SIS mixer, as compared to semiconductor diode devices, is the very low level of local oscillator power required, only 10-sW in the example of fig. 2. It may be noticed in fig. 2 that at bias voltages below 1.5mV the detector performance is strongly degraded. The noise power increases rapidly as the bias voltage decreases, with an equivalent increase in conversion loss and mixer noise temperature. This bias region is that in
2013
Fig. 3. An SIS junction manufactured at Bell Labs. by photoresist bridge techniques [4]. which Josephson current effects are noticeable, particularly the well-known hysteretic behavior of the I - V characteristics. These effects can be suppressed by application of a magnetic field, or
Fig. 4. A series array of 20 SIS junctions.
21114
7\G. Phillips, and G.J. l)olan / ,S'l,gmixers
simply ignored by confining operation to larger bias values. The reader may recognize that this latter p r o c e d u r e will fail at high frequencies where tip approaches 2A. A discussion of the ditticulties and possible solutions may be found m refs. 1~+ and 21. 3.2.
Device
structure
Among the advantages of the SIS devices is the planar nature of the thin fihn structure which is suited to production by m o d e r n p h o t o or electron b e a m lithography. O n e technique for small junction (<1 # m e) fabrication is that of Dolan el al. [4]. Electron beam lithography is used to generate masks from which photoresist bridge structures arc made. By evaporating past the bridge at different angles, overlapping structures are formed. Fig. 3 shows an example of a small area (0.1 # m : ) P b - i n junction made in this way. Very importantly, arrays of junctions can also be
formed. Fig. 4 shows a series array of 2(I SIS junctions nlade at Bell l+abs. The ( ' h a h n c r s I_Jniversiiv g r o u p [21,24] has been active in pursuing the series array technology. Example of superc~mducting thin film ~,trtictures m o u n t e d in waveguides are given in ligs 5 and n. Fig. 5 shows a Bell Labs single junction nlounted across a 1 1 5 G H i waveguide. Integrated to the UllCtiOll (which is at the center ~ll the waveguide) is tl ,,uperconducling choke ,,co lion which prevents the signal and local oscillator radiation fl-tml propagating in the striplinc 11: section. F'ig. (~ indicates the lawmt for the Sw'cdish array 12@ which makes use of ',in a n t e n n a structure within the wavcguidc to couple the currents to the junctitm array slightly outside the waveguide itself. Many other structures are possible, and as progress is made in understanding the simple structures it should bc possible It> design more complex structures to optimize the w a v e g u i d e - j u n c t i o n thatch.
Zig. 5. A substrate supporting an SIS junction with d.c. and IF connections and r.f. choke structure, mounted across a me-quarter height WR-8 waveguide [16].
T.G. Phillips and G.J. Dolan / SIS mixers
(a)
2015
£ -,
/
~";~
\ ........
~
(b)
•
//
<.- ~. - ~ > "
~,,~ ,, ground
/
Backshort
-lo~m
,if" i
f
Fig. 6. A monopole waveguide a n t e n n a coupled to a series array of SIS junctions and an IF/bias line with R.F. choke. After R u d n e r et al. [24].
i
S
i
4. SIS mixer performance A major hope for the SIS mixer is that the great deal of quantum structure will provide the receiver engineer with an excellent tool for analyzing his device. That this same philosophy has failed for the Josephson effect is surprising. In what follows the Josephson current is ignored, since we confine our attention to the structure near the gap voltage. Tucker's theory [2] has been computationally encoded by Wengler and Woody [25] so that a plot can be made of any one of the device parameters as a function of d.c. or local oscillator bias for variable values of the parameters. This formalism is exact in so far as the theory is exact, including, for instance, the quantum mechanically induced reactance term which appears in the device impedance parameters [2]. Previous formalisms have been approximate in some senses [26, 27]. The object of the exercise is to deduce the environment that the tunnel junction sees in terms of the real and imaginary parts of an admittance function. Ys. Changes to the device structure or to the waveguide environment can then be contemplated to improve the device performance. Fig. 7 shows a set of computer generated I - V characteristics for various values of the amplitude and phase of Ys. The input to the theory is the digitized I - V characteristic with local
j V -'~
ARG (YS) = - l . S
V --~
0 0
I V-'-~ l.S
Fig. 7. Plots of I - V characteristics for a P b - I n junction at 115 GHz. T h e full line without local oscillator power is from experiment and is input data for the computer program, The crosses represent the experimental characteristics with local oscillator power applied, and the full lines are the computer generated local oscillator applied curves for various values of the amplitude and phase of Ys (see text).
oscillator power applied. The local oscillator amplitude has been fixed by requiring that the computer generated and observed I - V characteristic (L.O. power applied) agree at the center of the first photon step. Clearly, an excellent fit is obtained (for all the quasiparticle structure) for only one value of Ys. The device under investigation here is a Bell Labs. junction at 97 GHz, in a waveguide mount. Apparently the junction sees an admittance of about three times the waveguide admittance. An example of the power of the theoretical expression is shown in fig. 8, where the best fit Ys p a r a m e t e r found for one value of the local oscillator power is held constant, while the local oscillator power is varied in 3 d b steps, both
T.G. Phillips and G.J. Dolan / SIS mixers"
2/11~
'
'
I
I "
//~
.~'/*
i
{
4.-. t. ,
al ,
O
.~.
0.5
••
,- / / y7 M ~ /
v/vG
t
10
15
•
<
'2 ~
~ •
Fig. ~, P l o t s of r e d u c e d c u r r e ~ t vs. ~ o [ t a g e ft~r the s a m e i u n c t i o n as in fig. 7, f o r v a r i o u s v a l u e s of t h e local o s c i l l a t o r power. A-zero L . O . : ( ' - I . . ( ) . as f o r fig. 7: B - 3 d B less t h a n (7: l ) - 3 d B m o r e t h a n ('. T h e c r o s s e s r e f e r to e x p e r i -
i ¢:
.
,,r , ,
Fig. ll). ! V c h a r a c t e r i s t i c s fl~l a tin i u n c t i o n at 3 ( ~ G H / , s h o w i n g h y s t e r e t i c b e h a v i o r at t h e first s t e p . A f t e r lVlcGrath e! al. [27].
mental data.
experimentally and theoretically. A good tit is obtained for all the steps. The shape of the local oscillator induced steps in fig. 8 is not very remarkable, the slope at the step center being similar to that of the zero 1..O. curve at the same bias voltage. However. under different conditions of imbedding admittance (Y,) other workers have observed negative differential resistance steps, implying a device with arbitrarily large gain [281. An example is the work of Kerr el al. [2q] who observed gain at 1 1 5 G H z with an array of Pb alloy junctions. Their I - V characteristics are shown in tig. ~). A
60
second example is that observed by McGrath et al. [27], shown in fig. It) for a 36 G H z experiment with a tin junction. In this case the junction is current biased so that hysteresis is observed m the negative resistance region. It is not obvious that a device will have its lowest noise temperature when the conversion loss is a minimum. (Operation under conditions of large gain would be dilBcult in an engineering sense and might lead to considerable
>10(2
O C2~
Ii 20 _ _ i
Ci
0 -
I
o
IO
0 5
I
3
IO
30
I¥%t 20
5o
40
5o
BI~S VOLTAGE (MV)
Fig. ~. I - V c h a r a c t e r i s t i c s f o r a Pb alloy series a r r a y at 115 GDIz, s h o w i n g n e g a t i v e d i f f e r e n t i a l r e s i s t a n c e at the firsl s t e p . A f t e r K e r r et al_ {2g{.
Fig. 11. C o n t o u r s of c o m p u t e d ~,ystem t e m p e r a l u r c " ;t ~, ;l f u n c t i o n of a m p l i t u d e a n d p h a s e of Y,. T h e c r o s s r e p r e s e n t s t h e p o s i t i o n of t h e j u n c t i o n of fig. 7. T h e r e g i o n to the left of t h e d a s h e d line is w h e r e n e g a t i v e d i f f e r e n t i a l r e s i s t a n c e o c c u r s o n t h e lirst s t e p .
T.G. Phillips and G.J. Dolan / SIS mixers 0
....
, ....
, ....
, ....
2017
HOT AND COLD LOADS I% DIELECTRIC BEAMSPLITTER
.~ -tO
/ / / / /
/ / / / /
.
-20
.
.
.
.
.
.
.
/
) LENS
/
I
V/VG Fig. 12. C o n v e r s i o n loss as a function of r e d u c e d bias v o l t a g e for the j u n c t i o n of fig. 7. T h e c o m p u t e d p e r f o r m a n c e is the full line, t h e e x p e r i m e n t a l v a l u e s are the crosses.
excess noise.) For the junction of fig. 7, a computed plot of system noise (TM), arising from current shot noise plus a 10 K IF amplifier contribution, is shown in fig. 11 as a function of the amplitude and phase of Y~. The region of negative differential resistance steps and arbitrarily large gain is also shown. It is interesting to note that the position of the actual junction is well away from the proper value of Y~ to achieve minimum noise performance. Fig. 12 shows the computed conversion loss for that junction at its present Ys value as compared with the measured loss. The agreement is excellent. The best reported values of mixer noise temperatures are 9 K at 3 6 G H z [27], 6 0 K at 115 G H z [16, 29] and 300 K at 230 GHz [16].
_•FEEDHORN
IM_'xERL_..__.JGaAs FET L -- 1.0-15GHz IF. BACKSHIBLOCKoRT, [ J Ii AMpLIF'E~I DC 8~AS
Fig. 13. S c h e m a t i c d i a g r a m of the 115 G H z receiver.
signal to pass freely and couples in only - 1 % of the L.O. power. This is possible because of the low L.O. power requirement for the SIS mixer. Fig. 14 is a photograph of the receiver dewar with radiation shields removed. A prototype version of this receiver, built at
5. Receiver performance A receiver to mount at the focus of a 10m telescope at Caltech's Owens Valley Radio Observatory, using an SIS mixer element, has been completed. A simplified block diagram of the receiver is shown in fig. 13. This figure also demonstrates the method of measurement of the receiver and mixer performance parameters. The parts within the box are in vacuum, bolted to the cold plate of a liquid helium cryostat. Some measurements have been made at 4.2 K, but the best results are generally found at lower temperatures. The dielectric beam splitter allows the
Fig. 14. Picture of the I 1 5 G H z r e c e i v e r to b e used for a s t r o n o m i c a l o b s e r v a t i o n s . T h e principal e l e m e n t s are: A, L.O. f e e d h o r n ; B, 1% dielectric b e a m s p l i t t e r ; C, m i x e r f e e d h o r n ; D, m i x e r block; E, b a c k s h o r t t u r n e r ; F, 3 0 d B d i r e c t i o n a l c o u p l e r for c o u p l i n g test signals into the IF line; G, 1.4 G H z c o o l e d G a A s F E T I F amplifier; H. IF o u t p u t port.
201;4
T.G. Phillips and G.J. Dolan / SIS mixers
Bell Labs., was used in June 1979 to determine its suitability for astronomical operation. No m a j o r problems were encountered. The current receiver was successfully used during the winter of 1980. Table I indicates p e r f o r m a n c e figures for the prototype and current receiver. The current receiver has a system t e m p e r a t u r e of 23(1 K at l l 5 G H z , which is comparable with the best c o n t e m p o r a r y cooled Schottky diode receivers. Also shown is a very recent report from Linke for a similar prototype receiver at Bell Labs, [29]. It seems that SIS receivers are already competitive with the best available alternatives in the millimeterwave regime and should eventually
,i
o
dA£: i76~,
/
J
) I
3
0
I
()VR()
h m k e (BTI .)
95 K 50 K ~ dB 4110 K
t 41 ) K II K u dB 22711K
$1) K s K S dB 131) K
prove superior due to the very great developmerit potential. As a final note, it is probable that considerable i m p r o v e m e n t s in the metallurgical aspects of junction fabrication will aid developments. An example is the Nb-A1 oxide-PbBi system of Gurvitch and Rowell [31]. These workers have produced very high quality junctions using a Nb base electrode coated with a thin htyer of /\1 which is then oxidized. As shown in lig. 15, the I - V characteristic of such junctions can bc sharper in curvature than Pb alloys over a wide range of AI thicknesses. Also, Nb-based junctions are known to be very durable.
The authors gratefully acknowledge the many contributions of their colleagues M. Wengler, D. Woody, R. Miller, E. Sutton and J. Rowell. Work at Caltech is supported by National Science Foundation grant no. AST-8007645.
References
dA2 : 364~
2
FM Tw L 1'~
Prototype (BTI.)
Acknowledgements
i
=
Table I
~
3
4-
VOUTAGE (MI~LIVOLTS~ Fig. 15. I - V characteristics of large area Nb-A1 oxide-PbBi junctions for different thicknesses of AI (dAI). Very sharp characteristics are obtained in this way over a wide range of da~, suggesting that the small high current density devices required for the present application could be m a d e using this metallurgical system. After Gurvitch and Rowell [311.
[11 A.H. D a y e m and R.,I. Martin, Phys. Re~. I,ctt. S (I~,~(Q) 24~. [21 J.R. Tucker, I E E E J. ()uantunl Electron. ()E-15 (197t, ~) 1234. 131 P.l,. Richards, in: Semiconductors and Semimclals, xol 12 Willardson and Beer. eds. (Academic Press. New York, 1977) p. 395. [4] G.J. Dolan, T.G. Phillips and D.P. Woody, Appl. Ph,.s. Lett. 34 (1979) 347. [5] P . L Richards. T.M. Shen, R.E. Harris and F.l.. Lhwd, Appl, Phys. Len. 34 11970) 345.
T.G. Phillips and G.J. Dolan / SIS mixers
[6] M. McColl, M.F. Bottjer, A.B. Chase, R.J. Pedersen, A.H. Silver and J.R. Tucker, IEEE Trans. Magn. MAG15 (1979) 468. [7] D.P. Woody, T.G. Phillips and G.J. Dolan, Proc. Nat. Radio Sci. Meeting, Nat. Acad. Sci, USNC/URSI (1979) p. 43. [8] A.R. Kerr, IEEE Trans. Microwave Theory Tech. MTF-23 (1975) 781. [9] E.R. Carlsen, M.V. Schneider and T.F. McMaster, IEEE Trans. Microwave Theory Tech. MTT-26 (1978) 71/6. [111] N.R. Erickson, 1EEE Int. Microwave Syrup. MT'f-S (1978) 438. [11] M.R. Fetterman, P.E. Tannenwald, B.J. Clifton, C.D. Parker, W.D. Fitzgerald and N.R. Erickson, Appl. Phys. Lett. 33 (1978) 151. [12] E. Kollberg, URSI Symp. Millimeter Tech. in Radioastronomy, Grenoble (1980). [13] R.A. Linke, M.V. Schneider and A.Y. Cho, IEEE Trans. Microwave Theory Tech. MTlr-26 11978) 935. [14] S. Weinreb and J. Payne, private communication. [15] W. Wilson, private communication. [16] T.G. Phillips, D.P. Woody, G.J. Dolan, R.E. Miller and R.A. Linke, IEEE Trans. Magn. MAG-17 (1981) 684. [17] Y. Taur, J.H. Classen and P.L. Richards, Trans.
2019
Microwave Theory Tech. MTIr-22 (1976) 10115. [18] Y. Taur and A.R. Kerr, Appl. Phys. Lett. 32 (1978) 775. [19] Th. de Graauw, private communication. [20] T.M. Shen, P.L. Richards, R.E. Harris and F.L. Lloyd, Appl. Phys. Lett. 36 (1980) 777. [21] S. Rudner, M.J. Feldman, E. Kollberg and T. Claeson, in: SQUID '80' ed. Halbohm and Lubbig, eds. (W. de Gruyter, Berlin, 1981) p. 91/1. [22] P.K. Tien and J.P. Gordon, Phys. Rev. 129 (1963) 647. [23] O. Entin-Wohlman, Phys. Rev. B22 (1980) 5225. [24] S. Rudner, M.J. Feldman, E. Kollberg and T. Claeson, J. Appl. Phys. (1981) to be published. [25] M.J. Wengler and D.P. Woody, in preparation. [26] G J . Dolan, R.A. Linke, T.G.L.C. Sollner, D.P. Woody and T.G. Phillips, IEEE Trans. Microwave Theory Tech. M'VF-29 (1981) 87. [27] W.R. McGrath, P.L. Richards, A.D. Smith, H. van Kemper and R.A. Batchelor, Appl. Phys. Lett. (1981), to be published. [28] J.R. Tucker, Appl. Phys. Lett. 36 (1980) 477. [29] A.R, Kerr, S.K. Pan, M.J. Feldman and A. Davidson, Physica, this volume. [30] R.A. Linke, private communication. [31] M. Gurvitch and J.M. Rowell, to be published.