We demonstrate an integration of tunneling magnetoresistance and the Josephson effects within one tunneling junction. Several sets of Nb-Fe-Al-Al2O3-Fe-Nb wafers with varying Al and Fe layers thickness were prepared to systematically explore the competition of TMR and Josephson effects. A coexistence of the critical current IC(dFe) and the tunneling magnetoresistance ratio T M R(dFe) is observed for iron layer dFe thickness range 1.9 and 2.9 nm. Further optimization such as thinner Al2O3 layer leads to an enhancement of the critical current and thus to an extension of the coexistence regime up to dFe3.9 nm Fe.

Singlet superconductivity is based upon the formation of electron pairs with opposite spins and momenta, whereas the magnetization of ferromagnets is based on spontaneous parallel spin alignment. The competition of these antagonistic phenomena in thin films leads to interesting effects.1,2 For instance a ferromagnetic layer (F) sandwiched between two superconductors (S) gives rise to Josephson junctions with unusual π-shifted current phase relations.3 Other ferromagnets such as conical magnets,4 half-metallic ferromagnets,5 pseudospin valves6 and their stacks7 were also integrated in Josephson junctions and can lead to triplet superconductivity. The control of supercurrents by metallic pseudospin valves with the layer sequence SFNFS was attempted too, but its applicability was limited due to very low resistances.8 Recently, Josephson junctions with embedded pseudospin valves reveal that the Josephson coupling can be controlled in both magnitude9,10 and phase.11 

Initial experiments on Josephson junctions with spin valves as weak links based on the tunneling magnetoresistance (TMR) were made based on Nb and Ni.12 The TMR is the result of spin-dependent tunneling through an insulating barrier (I) of FIF structure and depends on mutual magnetization orientation of the ferromagnetic electrodes.13,14 In contrast if the F-layers are magnetized parallel the majority/minority spins tunnel into the same majority and minority bands on the other side of tunnel junction. A parallel magnetization usually results in a low resistance RP. If the magnetization is antiparallel the majority spins tunnel into the minority states of the other electrode and vice versa. This gives rise to a higher resistance RAP.

In this paper we integrate a spin-valve into a Josephson junction, using a SFIFS layer sequence. In contrast to all-metallic Josephson junctions (SFNFS) the use of an insulating aluminum oxide barrier Al2O3 gives raise to much larger resistance and hence, easily detectable voltages.2,15

Our Nb-Fe-Al2O3-Fe-Nb junctions (SFIFS) were prepared by magnetron sputtering technique. First a bottom stack Nb-Fe-Al is deposited. Oxidation of the Al-film forms insulating Al2O3 barrier and subsequent deposition of the double layer Fe-Nb creates complete SFIFS stack. In order to obtain the highest possible tunneling magnetoresistance ratio (T M R) one has to find the thickness of Al which can be fully oxidized at certain pressure without oxidizing the Fe. Thus Nb-Fe-Alwedge-Al2O3-Fe-Nb samples with 5 nm thick Fe layers were made. The Al thickness varies wedge-like across the wafer (Fig. 1(a)).15 The Al2O3 barrier was formed by exposing the Al layer into O2 atmosphere for 30 min under the pressure of 0.1 mbar. As seen in Fig. 1(b), the maximal value of T M R achievable under these conditions is 5.5 % at the aluminum thickness dAl = 1.8 nm. For dAl> 1.8 nm the Al layer is not completely oxidized, resulting in suppressed T M R. If the Al thickness is smaller than 1.8 nm the FeOx is formed at the bottom Fe-Al2O3 interface which suppresses T M R too. Both T M R suppression effects can be attributed to the spin-flip scattering at imperfect interfaces.16 Thus the proper thickness of the Al layer to form Al2O3 barrier is an essential point of sample preparation. Note that all measurements in this paper are performed at a temperature 300 mK.

FIG. 1.

(a) Sketch of Fe-Alwedge-Al2O3-Fe stack with Al wedged layer. The dashed arrow points to the wafer position with optimal Al thickness. The blue and orange layers depict parasitic Al metallic residuum and oxidized Fe layer, respectively. (b) Al layer thickness dependencies of normal conductance Gn(dAl) and T M R(dAl) of the Nb-Fe-Alwedge-Al2O3-Fe-Nb junctions. (c) Normal conductance Gn of Nb-Fe-Al-Al2O3-Fe-Nb junctions as a function of Fe layer thickness dFe for different dAl reveals a thickness inhomogeneity of Al layer over the 4 inch wafer. The junction areas was 50 μm2.

FIG. 1.

(a) Sketch of Fe-Alwedge-Al2O3-Fe stack with Al wedged layer. The dashed arrow points to the wafer position with optimal Al thickness. The blue and orange layers depict parasitic Al metallic residuum and oxidized Fe layer, respectively. (b) Al layer thickness dependencies of normal conductance Gn(dAl) and T M R(dAl) of the Nb-Fe-Alwedge-Al2O3-Fe-Nb junctions. (c) Normal conductance Gn of Nb-Fe-Al-Al2O3-Fe-Nb junctions as a function of Fe layer thickness dFe for different dAl reveals a thickness inhomogeneity of Al layer over the 4 inch wafer. The junction areas was 50 μm2.

Close modal

Having determined the thickness of Al for optimal oxidation and a maximal T M R value, several wafers of the Nb-Fe-Al2O3-Fe-Nb junctions with four different dAl and with varying iron thickness dFe were prepared. To achieve a better homogeneity of Al layer over 4 inch wafer it was deposited during rotation of the sample holder. Both, top and bottom iron wedged layers have the same thickness. Despite the rotation the Al film has a slightly convex thickness profile over 4 inch wafer. Due to this inhomogeneity the shape of Gn(dFe) differs for each Al mean thickness (Fig. 1(c)). For an average Al thickness dAl=2.45 nm, Gn(dFe) has a deep minimum near dFe 3 nm. When dAl is further decreased (2.1 and 1.75 nm) Gn(dFe) goes over a maximum due to a slight oxidation of the Fe layer in middle of the wafer. This effect does not occur in usual Nb-Al-Al2O3-X-Nb junctions, where Al is thick enough to avoid oxidation of adjacent Nb layer. In our SFIFS junctions one cannot completely avoid these variations. Nevertheless, the inhomogeneity of Al layer over 4 inch wafer does not impede the barrier homogeneity of the individual junctions with a lateral size of 10 μm.

A few junctions from all four wafers with thick Fe layers were measured first. As seen in Fig. 4, the T M R has a maximum value of 5.2% similar to above mentioned experiment at ca. 1.75 nm of Al thickness. For both thinner and thicker Al layers the T M R drops.

Now we turn towards the Josephson effect in SFIFS junctions for dAl = 1.75 nm, which display the best TMR ratio. As seen in Fig. 2 the current-voltage characteristics for thin Fe layers have the typical hysteretic behavior of underdamped Josephson junctions. For larger Fe layer thickness the hysteretic behavior vanishes, i.e. junctions become overdamped due to low critical current. The variation of the slope of the IV-characteristics at high voltage V > 2.2 mV, i.e. the normal state conductance Gn, is caused by the above mentioned slight inhomogeneity of Al layer shown in Fig. 1(c).

FIG. 2.

Current-voltage curves of SFIFS Josephson junctions measured for different Fe thickness at 300 mK. Inset: a perfect Fraunhofer dependence of IC(Φ) demonstrates the homogeneity of the tunnel barrier.

FIG. 2.

Current-voltage curves of SFIFS Josephson junctions measured for different Fe thickness at 300 mK. Inset: a perfect Fraunhofer dependence of IC(Φ) demonstrates the homogeneity of the tunnel barrier.

Close modal

Typical tunneling magnetoresistance curves are shown on Fig. 3. To determine the T M R ratio in the SFIFS devices these curves were measured around V 4 mV, i.e. well above 2ΔNb, with constant bias current. Two clearly distinguishable parallel and antiparallel states were observed. The difference of substrate, i.e., Nb for the first and Al2O3 for the second Fe-film, respectively, result in a difference of switching fields. The T M R ratio is calculated as T M R = (RAPRP)/RP and plotted vs. dFe (triangles in Fig. 4). For thick dFe> 4.7 nm the T M R is independent of dFe. As the ferromagnetic layer thickness decreases below 4.7 nm, the T M R ratio also decreases. This is probably caused by the gradual suppression of the magnetic order in the thinnest Fe films.

FIG. 3.

(a-e) Magnetoresistance curves measured at 300 mK for different Fe layer thickness. The resistance is always measured around a voltage 4 mV. The field range ± 300 mT is the same for all curves. (f) Magnetoresistance curve measured up 2 T shows the transition of Nb electrodes into the normal state.

FIG. 3.

(a-e) Magnetoresistance curves measured at 300 mK for different Fe layer thickness. The resistance is always measured around a voltage 4 mV. The field range ± 300 mT is the same for all curves. (f) Magnetoresistance curve measured up 2 T shows the transition of Nb electrodes into the normal state.

Close modal
FIG. 4.

The critical current IC and T M R as a function of Fe layer thickness for Nb-Fe-Al2O3-Fe-Nb junctions at 300 mK. The dashed arrow on IC(dFe) curve points out the transition to a continuous magnetic layer. Dashed IC(dFe) line shows an increase of critical current for lower Al-oxidation.

FIG. 4.

The critical current IC and T M R as a function of Fe layer thickness for Nb-Fe-Al2O3-Fe-Nb junctions at 300 mK. The dashed arrow on IC(dFe) curve points out the transition to a continuous magnetic layer. Dashed IC(dFe) line shows an increase of critical current for lower Al-oxidation.

Close modal

The critical current as function of dFe is depicted as circles in Fig. 4. The behavior of the IC(dFe) dependence can be divided into two regimes. For dFe 1.9 nm the exponential variation of IC(dFe) can be extrapolated to dFe = 0 where one obtains a critical current IC=10 mA. From the junction area 5x10 μm2 one determines a critical current density jC = 2 kA/cm2. This value is typical for a reference Nb-Al2O3-Nb junction with the same oxidation parameters as used for our SFIFS junctions. In this thickness range both Fe layers behave as if the magnetic order is weakened but still with strong spin scattering. For values dFe> 1.9 nm IC(dFe) is much more rapidly suppressed due to the fully developed magnetic order in both iron layers. Hence, a substantial overlap of the IC(dFe) and T M R(dFe) curves is observed for dFe range in between 2.4 and 2.9 nm.

A shorter time of oxidation (3 min) and lower oxidation pressure (0.01 mbar) leads to only 10 % drop of the T M R value from 5.5 % to of 5 %, at an optimal thickness of Al layer dAl = 1 nm with a much narrower T M R peak compared to that for thicker Al2O3 barrier (see dashed line in Fig. 5).

FIG. 5.

T M R(dAl) dependence for different oxidation parameters. The shorter Al oxidation time results in a lower optimal Al thickness and a small drop of T M R.

FIG. 5.

T M R(dAl) dependence for different oxidation parameters. The shorter Al oxidation time results in a lower optimal Al thickness and a small drop of T M R.

Close modal

To show the impact of the weaker tunneling barrier on the critical current we prepared a Nb-Fe-Al-Al2O3-Fe-Nb junction set with both Fe-wedged layers and with 1 nm thick Al. This results in significant increase of the critical current by two orders of magnitude (see dashed IC(dFe) in Fig. 4) compared to Josephson junctions with thicker Al2O3 barrier. For high critical currents IC the IV-characteristics are unstable because of a parasitic weak link at the interface between the Nb wiring and upper Nb-layer of the stack. Thus in this case one can measure the TMR for high dFe (negligible IC) only. As seen in Fig. 5 the T M R value drops only by 0.5 %, therefore we expect that the T M R(dFe) dependence is similar for both cases of oxidation. Our experiment then shows that the upper limit for the observation of Josephson effect is shifted to dFe 3.9 nm. Hence, a substantial increase of the maximal T M R(dFe) and IC(dFe) can be achieved, i.e. IC 20 μA and TMR 0.75 % are found at dFe 3.1 nm, as compared to IC 2.5 μA and TMR 0.15 % at dFe 2.6 nm for the thicker Al2O3 barrier.

In conclusion we have prepared superconducting spinvalves that display a substantial supercurrent in the presence of the tunneling magnetoresistance effect. Optimization of Al layer thickness and the use of lower oxidation pressure to form the Al2O3 barrier leads to significant enhancement of the critical current. The interplay of TMR and Josephson effects, such as F-layer magnetization orientation influence on IC(B), i.e. controllability of the Josephson junctions, will be described elsewhere.

The authors would like to thank M. Weides, M. Aprili, Š. Gaži and Ch. Back for fruitful discussions. This work is supported by the German Science Foundation via SFB 689 and KO1953/11-1, and by Emmy Noether project Hu 1808/1.

1.
V. V.
Ryazanov
,
V. A.
Oboznov
,
A. Y.
Rusanov
,
A. V.
Veretennikov
,
A. A.
Golubov
, and
J.
Aarts
,
Phys. Rev. Lett.
86
,
2427
(
2001
).
2.
T.
Kontos
,
M.
Aprili
,
J.
Lesueur
,
F.
Genet
,
B.
Stephanidis
, and
R.
Boursier
,
Phys. Rev. Lett.
89
,
137007
(
2002
).
3.
A.
Bauer
,
J.
Bentner
,
M.
Aprili
,
M. L.
Della Rocca
,
M.
Reinwald
,
W.
Wegscheider
, and
C.
Strunk
,
Phys. Rev. Lett.
92
,
217001
(
2004
).
4.
J. W. A.
Robinson
,
J. D. S.
Witt
, and
M. G.
Blamire
,
Science
329
,
59
(
2010
).
5.
R. S.
Keizer
,
S. T. B.
Goennenwein
,
T. M.
Klapwijk
,
G. X.
Miao
,
G.
Xiao
, and
A.
Gupta
,
Nature
439
,
825
(
2006
).
6.
J. W. A.
Robinson
,
G. B.
Halasz
,
A. I.
Buzdin
, and
M. G.
Blamire
,
Phys. Rev. Lett.
104
,
207001
(
2010
).
7.
C.
Klose
,
T. S.
Khaire
,
Y.
Wang
,
W. P.
Pratt
, Jr.
,
N. O.
Birge
,
B. J.
McMorran
,
T. P.
Ginley
,
J. A.
Borchers
,
B. J.
Kirby
,
B. B.
Maranville
, and
J.
Unguris
,
Phys. Rev. Lett.
108
,
127002
(
2012
).
8.
C.
Bell
,
G.
Burnell
,
C. W.
Leung
,
E. J.
Tarte
,
D. J.
Kang
, and
M. G.
Blamire
,
Appl. Phys. Lett.
84
,
1153
(
2004
).
9.
B.
Baek
,
W. H.
Rippard
,
S. P.
Benz
,
S. E.
Russek
, and
P. D.
Dresselhaus
,
Nature Comm.
5
,
3888
(
2014
).
10.
M.
Abd El Qader
,
R. K.
Singh
,
S. N.
Galvin
,
L.
Yu
,
J. M.
Rowell
, and
N.
Newman
,
Appl. Phys. Lett.
104
,
022602
(
2014
).
11.
E. C.
Gingrich
,
B. M.
Niedzielski
,
J. A.
Glick
,
Y.
Wang
,
D. L.
Miller
,
R.
Loloee
,
W. P.
Pratt
, Jr.
, and
N. O.
Birge
,
Nat. Phys.
12
,
564
(
2016
).
12.
I. P.
Nevirkovets
,
O.
Chernyashevskyy
,
G. V.
Prokopenko
,
O. A.
Mukhanov
, and
J. B.
Ketterson
,
IEEE Appl. Supercond.
25
,
1800705
(
2015
).
13.
J. S.
Moodera
,
L. R.
Kinder
,
T. M.
Wong
, and
R.
Meservey
,
Phys. Rev. Lett.
74
,
3273
(
1995
).
14.
T.
Miyazaki
and
N.
Tezuka
,
J. Magn. Magn. Mater.
139
,
L231
(
1995
).
15.
M.
Weides
,
M.
Kemmler
,
E.
Goldobin
,
D.
Koelle
,
R.
Kleiner
,
H.
Kohlstedt
, and
A.
Buzdin
,
Appl. Phys. Lett.
89
,
122511
(
2006
).
16.
S. A.
Wolf
,
D. D.
Awschalom
,
R. A.
Buhrman
,
J. M.
Daughton
,
S.
von Molnar
,
M. L.
Roukes
,
A. Y.
Chtchelkanova
, and
D. M.
Treger
,
Science
294
,
1488
(
2001
).