Professional Documents
Culture Documents
Measuring The Spin Polarization of A Metal With A Superconducting Point Contact
Measuring The Spin Polarization of A Metal With A Superconducting Point Contact
generate magnetic fields and coronal x-ray References and Notes 16. J. D. Kirkpatrick, T. J. Henry, D. W . McCarthy, Astro-
phys. J. Suppl. 77,417 (1991); J. D. Kirkpatrick, T. J.
emission. A rotation-induced dynamo is needed 1. A. Burrows and J. Liebert, Rev. Mod. Phys. 65,301
Henry, D. A. Simon, Astron. J. 109,797 (1995).
(1993); S. R. Kulkarni, Science 276,1350 (1997).
( 6 ) .The star VB 8 wit11 spectral type M7 is the 2. A. Burrows e t a/., Nature 375,299 (1995). 17. C. F. Prosser, J. R. Stauffer, R. P. Kraft, Astron. J. 101,
star with the latest spectral type that shows 3. J. R. Stauffer, D. Hamilton, R. Probst, Astron. J. 108, 1361 (1991); E. L. Martin, R. Rebolo, M. R. Zapatero-
quiescent x-ray emission, namely; log (LulLbol) 155 (1994); R. Rebolo, M. R. Zapatero-Osorio, E. L. Osorio, Astrophys. J. 469,706 (1996).
Martin, Nature 377, 129 (1995); C. Basri, C. W. 18. K. L. Luhman and C. H. Rieke, Astrophys. J. 497,354
= -2.8 (where Lx is the x-ray luminosity and
Marcy, J. Craham,Astrophys. J. 458,600 (1996); M. R. (1998).
Lbolis the bolometric luminosity) (19). Also, Zapatero-Osorio, R. Rebolo, E. L. Martin, Astron. As- 19. M. S. Ciampapa e t a/., ibid. 463,707 (1996).
the Iv18 star \'B 10 was detected at log (L,ILbol) trophys. 317,164 (1997). 20. T. Fleming, J. H. M. M. Schmitt, M. S. Ciampapa, in
= -3, but only dusing a flare. The quiescent 4. F. Comeron, C. H. Rieke, P. Claes, J. Torra, R. J. preparation.
Laureijs, Astron. Astrophys. 335,522 (1998). 21. K. M. Strom and S. E. Strom, Astrophys. J. 424,237
upper limit is log (L,IL,,,) 5 -4.5 before and 5. K. L. Luhman, J. Liebert, C. H. Rieke, Astrophys. J. 489, (1994).
after the flare (20). 111addition, the M6 to M7 T 1165 (1998). 22. K. L. Luhman e t a/., ibid. 493,909 (1998).
Tauli star V410 x-ray 3 was also detected as an 6. B. Dorman, L. Nelson, W. Chau, ibid. 342, 1003 23. A slightly larger x-ray upper limit was reported in (5)
(1989); F. D'Antona and I. Mazzitelli, ibid. 296,502 obtained from the l a noise level i n the hard band
x-ray source (21). With a mass of 0.08 to 0.15 (1985). ROSAT map published by S. Casanova e t al. [Astro-
1li(,; an age of lo6 years, and log (LxiL,bol)= 7. F. Comeron, C. H. Rieke, R. Neuhauser, in preparation. phys. J. 439,752 (1995)l.
-2.8 (22). it is similar to Cha Ha 1. but slightly 8. S. K. Leggett, F. Allard, C. Berriman, C. C. Dahn, P. H. 24. None o f t h e 1 1 bona fide BDs published so far,
Hauschildt, Astrophys. J. Suppl. 104,117 (1996). all w i t h ages 2 3 x l o 6 years and rotational
more massive. The object 1623-2426, a young
9. F. Allard, P. H. Hauschildt, D. R. Alexander, S. Starr- velocities 2 2 0 k m l s (as far as known), were
BD in p Oph (5).was not detected in the 33-ks field, Annu. Rev. Astron. Astrophys. 35,137 (1997). detected in x-ray observations (R. Neuhauser e t al.,
PSPC pointed obsen-ation 200045. newly re- 10. A. Burrows e t al., Astrophys. J. 491,856 (1997); F. in preparation).
duced by us (23), with the upper limit being log D'Antona and I. Mazzitelli, in Proceedings of the
25. W. A. Lawson, E. D. Feigelson, D. P. Huenemoerder,
Meeting on Cool Stars i n Clusters and Associations, C. Mon. Not. R. Astron. Soc. 280,1071 (1996).
(LyiLb0,)5 -3.26, above the value measured Micela, R. Pallavicini, S. Sciortino, Eds. (Astronomical
26. W e are grateful t o H. Zinnecker, J. Schmitt, E.
for Cha Ha 1. Society of Italy, Florence, 1997), pp. 807-822; F.
Martin, and C. Rieke for useful discussion about
With an optical magnitude in the V band of D'Antona e t a/., in preparation; A. Burrows e t a/., in
BDs, ESO for allocation o f telescope time, C. Lid-
preparation.
21 magnitudes (7). Cha Ha 1 is the optically 11. J. Trumper, Adv. Space Res. 2 (no. 4). 241 (1983); E.
man and C . Carraro for allowing the use o f their
faintest low-mass object we observed. Yet, it is telescope time, A. Burrows and F. D'Antona for
Pfeffermann e t ai., Proc. SPIE 733,519 (1988).
providing their new pre-main-sequence tracks, E.
the x-ray brightest object: and the x-ray to 12. E. D. Feigelson and C. A. Kriss, Astrophys. J. 338,262
Feigelson for allowing t h e use o f his ROSAT obser-
(1989).
bolo~netricluminosity relation cannot explain 13. E. D. Feigelson e t al., ibid. 416,623 (1993).
vation, t h e second processing o f which is n o t y e t
kvhy only Cha Ha 1 is detected in x-rays. It is publicly available, and B. Stelzer and T. Hearty for
14. M. Braun, thesis, Friedrich-Schiller Universitat, Jena,
part o f t h e specific software development. ROSAT
possible that Cha Ha 1 rotates fast to suppost a Germany (1992).
is supported b y the German government (BMBFI
strong dynamo. The spectra! resolutions of our 15. J. Liebert e t al., in 7th Cambridge Workshop on Cool
DLR) and the Max-Planck-Society.
Stars, Stellar Systems, and the Sun, vol. 26 of ASP
observations are too low to detelmine the rota- Conference Series, M. S. Ciampapa and J. A. Bookbinder,
tional velocity. Because p Oph 1623-2426 has a Eds. (ASP, San Francisco, 1992), pp. 282-283. 6 July 1998; accepted 2 September 1998
mass similar to Cha Ha 1 and is 3 to 10 times
older, but it is not detected as an x-ray somce.
only the co~nbinationof a young age ( 5 3 X
lo6 vears) and fast rotation ( 2 20 luds) may Measuring the Spin Polarization
allow us to detect x-ray emission from a BD
(24). Alternative models for BD x-ray emission of a Metal with a
appear less likely: (i) Flare activity without (or
with faint) q~~iescent emission: for example, due
to magnetic field recolu~ections,as in the late-
Superconducting Point Contact
type star \IB 10 (20). is not supported by our R. J. Soulen Jr., J. M. Byers,* M. S. Osofsky, B. Nadgorny,
observations. because we find no evidence for T. Ambrose, S. F. Cheng, P. R. Broussard, C. T. Tanaka, J. Nowak,
variability in the x-ray emission. (ii) If Cha Ha J. S. Moodera, A. Barry, J. M. D. Coey
1 were a close binary with magnetic field con-
figurations si~nilarto those in x-ray bright, in- A superconducting point contact is used to determine the spin polarization at the
teracting low-mass binalies; it should be bright- Fermi energy of several metals. Because the process of supercurrent conversion at
er in the optical than obsened (7). (iii) The a superconductor-metal interface (Andreev reflection) is limited by the minority
x-ray e~nissioncantlot be linked with any cir- spin population near the Fermi surface, the differential conductance of the point
cumstellar material. because we do not see any contact can reveal the spin polarization of the metal. This technique has been
S I R excess (7). Hence, coro~lalact~vityappears applied to a variety of metals where the spin polarization ranges from 35 to 90
to be the most plausible explanation for the percent: Nio,,Feo,,, Ni, Co, Fe, NiMnSb, Lao,,Sro,,MnO,, and CrO,.
x-ray emission that is cons~stentwith all the
other obsemational data. A new class of electronics is emerging ferromagnet (FM) is not easy. A typical
The x-ray detection of Cha Ha 1 sug- based on the ability of ferromagnetic met- transition-metal FM has two components to
gests that a young BD can support a mag- als to conduct spin-polarized currents ( I ) . its electronic structure: narrow d bands that
netic corona. Therefore, it may be possible The effectiveness of magnetoelectronics may be fully or partially spin-polarized
to find more young BDs in star-forming depends on the extent to which a current is (due to the on-site exchange energy) and
-
regions as counterparts to faint x-ray sourc- spin-polarized. All device designs improve broad s bands with a lesser degree of spin
es in x-ray observations with long exposure their performance as the spin polarization P polarization (due to hybridization with the
times. Establishing the BD x-ray luminos- 100%. For both scientific and techno- LJ bands). The quantity P can be defined as
ity function and estimating the integrated logical reasons it is important to be able to
iVT(EF) - iVL(E,.)
x-ray elnission are important for assessing directly and easily measure the electronic P = (1
the BD contribution to the diffuse galactic spin polarization at the Fermi energy, E,, iVT(EF) + :V-(EF)
x-ray emission and the baryonic dark mat- of a candidate material. where ,Vu (E) is the spin-dependent density of
ter in the galactic halo Unfortunately, determining P at E, of a states. The value of P is controlled by the
extent to which these s and d bands cross the we form a metallic point contact between the into a Cu foil at a temperature of 1.6 K. At
Fermi surface. If the orbital character at the sample and a superconductor using a simple low voltage the normalized conductance is
Fermi surface of a FM is primarily d-like, mechanical adjustment. Unlike a tunnel junc- indeed twice that of the normal state, and an
then P will be high. If, however, the orbital tion, a metallic contact allows coherent two- excess current of -0.2 rnA is present.
character is s-like or s-d-hybridized, then P particle transfer at the interface between the The probes for this study were fabricated
can be low or high depending on the details normal metal and the SC. The electronic by mechanically polishing SC rods of super-
of the electronic structure. The magnetization transport properties at the point contact mea- conducting material [Nb and tantalum (Ta)]
of a material may show that all of the elec- sures the conversion between superconduct- to a sharp point with progressively finer sand-
tronic spins associated with the d orbitals are ing pairs and the single-particle charge carri- paper. Examination of the sharpened points
aligned but that P at E, can be depressed (2). ers of the metal. with a scanning electron microscope indicat-
However, metallic oxide FMs, for example, The conversion of normal current to su- ed that all were roughly cone shaped and
have a greater opportunity for high values of percurrent at a metallic interface is called tapered to a rounded end with an approximate
P because of the predominance of d-orbital Andreev reflection (5) and is a well-known radius of 100 pm. However, the extreme
character at E,. phenomenon in superconductivity. To under- portion of the tips was studded with several
Measuring P requires a spectroscopic stand this process, consider Fig. 1A showing protrusions that were 1 pm or smaller and
technique that can discriminate between the an electron in a metal with P = 0 propagating likely formed the actual point contact. Posi-
spin-up and spin-down electrons near E,. toward the interface. For the electron to enter tioning and adjustment of the point contact
Spin-polarized photoemission spectroscopy the superconducting condensate and proceed was achieved by simple mechanical means.
is technically capable of providing the most as part of the supercurrent, it must be a The tip was attached to a drive shaft vertical-
direct measurement of P, but lacks the nec- member of a pair. The other electron required ly positioned above the sample material. The
essary energy resolution (-1 meV) (3). An for the formation of the pair is obtained from shaft was driven by a micrometer mechanism
effective alternative to photoemission is the the metal, thus leaving behind a hole at the capable of moving the point linearly by 100
use of spin-polarized tunneling in a planar interface. This hole has the opposite momen- pm per revolution. All of the transport mea-
junction geometry that does allow the elec- tum of the incident electron and propagates surements were made with a conventional
tronic spectrum near E, to be probed with away from the interface. The Andreev reflect- four-terminal arrangement while the point
submillielectron volt energy resolution. Ted- ed holes act as a parallel conduction channel contact and sample were immersed in a liquid
row and Meservey (4) pioneered this tech- to the initial electron current, doubling the He bath at either 4.2 or 1.6 K. The dIldV data
nique by making FM-superconductor (SC) normal-state conductance G, (where G = in this study were obtained by standard ac
tunnel junctions and Zeeman splitting the dIldV and V is the voltage) of the point lock-in techniques at a fGuency of 2 kHz.
SC's strongly peaked single-particle excita- contact for applied voltages eV < A, where A To understand the effect of P on the An-
tion spectrum by the application of a rnagnet- is the superconducting gap at the interface. In dreev reflection process, consider Fig. 1A
ic field. The resulting spectrum of the SC an I-Vmeasurement, the supercurrent conver- again. Because a superconducting pair is
roughly corresponds to two fully spin-polar- sion appears as an excess current added to the composed of a spin-up and spin-down elec-
ized peaks (neglecting spin-orbit coupling ef- ohmic response at the interface. We illustrate tron, an incident spin-up electron in the metal
fects) that can be used to detect P of a current the effect experimentally in Fig. 1B for a requires a spin-down electron to be removed
I from the FM film. The tunnel junction superconducting niobium (Nb) point pressed from the metal as well for conversion to
technique has been successfully used to find
P for a number of magnetic metals. The
drawback of the technique is the constraint of
fabricating a layered device consisting of a
thin-film FM on top of a uniform oxide layer
10 to 20 A thick that is formed on top of the
SC base. The need for a uniform oxide layer
is a severe limitation of the technique because
many interesting materials cannot be made
within this stringent constraint.
Accordingly, we have developed an ap-
proach to measuring P of a metal that re-
quires no magnetic field and places no special
constraints on a sample; thin films, single
crystals, and foils of several metals have been
successfully measured. In contrast to the tun-
nel junctions used by Tedrow and Meservey,
0 *S
Fig. 3. The differential conductance for a Fe-Ta
configuration where Fe is the sample and Ta
the point and vice versa. The spin polarization
a.0 for the Fe (PC = 43%) is nearly the same in
-4 -2 0 2 4 either configuration. Note that A(T = 0) = 0.7
Mmv) meV for Ta.
our technique. When these conditions are (14). The measured spin polarization of Ni tained for the gap A and PC. The slight
relaxed (Z + 0 and finite T), a numerical and Ni-rich alloys has been shown to be veiy difference near zero bias is due to varying
fitting procedure over the entire voltage range sensitive to impurities, possibly accounting amounts of interfacial scattering Z.
with our modified form of the BTK model for the discrepancies obsei~red.Another pos- Our results are summarized in Table 1 with
can be used to obtain PC. sible source of disagreement is that the dif- the materials arranged in ascending values of
In Fig. 2 we present a series of samples ferences in the definition of spin polarization PC. In our teclmique, the deviation from the
that show vaiying degrees of PC to demon- may be more pronounced for Ni con~pounds. mean value of polarization PC(Table 1) is not
strate the applicability of the superconducting The three lowest cllidVcurves of Fig. 2 are controlled solely by experimental error because
point contact technique. As prescribed by for compounds that have not been examined it may be due to the actual distribution of the
Blonder and Tinkham (13). only data from by the tunnel junction technique. These ma- spin polarization within the sample as well as
those point-contact configurations with an terials are easily measured with our point surface-scattering effects. Nevertheless, the sta-
ohmic contact resistance l/G,, between 1 and
100 ohms were analyzed. This criterion also
contact technique. A Heusler alloy film.
NiMnSb, grown by three-source co-evapora-
tistical variation is com~arableto the tunneling
results. The sinlplicity of our method allows
-
helm remove consideration of contacts with tion (description of growth technique to be materials not previously measured for spin po-
too much plastic deformation that causes low published), shows a polarization of PC = larization at E, (NiMnSb, La,,,Sr, ,MnO,, and
contact resistance and heating due to high 58% in contrast to expectations that it should CrOJ z,
to be studied. The ~ o i n tcontact tech-
current densities. For a Sharvin point contact be fully spin-polarized (16). Examination of nique has the added advantage of providing
(10) between typical metals (Cu and Nb), this the structure by extended x-ray absorption rapid feedback in the development of spin-
range implies contact areas of -10"'. Ad- fine structure spectroscopy showed some seg- polarized materials for magnetoelectronics by
ditionally, we include in Fig. 2 only those regation of the Sb near the surface of Heusler making the determination of electronic spin
contacts where the interfacial scattering, Z, films grown by this technique (17). A thin polarization no more difficult than a low-tem-
appears to be minimal, allowing a direct es- film of the colossal magnetoresistance mate- perature inagnetization measurement.