Nano granular metallic Fe - oxygen deficient TiO$_{2-δ}$ composite films: A room temperature, highly carrier polarized magnetic semiconductor

0.47 MB·English

by  S. D. Yoon

Nano granular metallic Fe - oxygen deficient TiO composite films: A room
2−δ
temperature, highly carrier polarized magnetic semiconductor
S. D. Yoon,∗ C. Vittoria,† and V. G. Harris‡
Center for Microwave Magnetic Materials and Integrated Circuits,
Department of Electrical and Computer Engineering,
Northeastern University, Boston, MA. 02115 USA
A. Widom§
Department of Physics, Northeastern University, Boston, MA. 02115 USA.
8
0
K. E. Miller and M. E. McHenry
0
2 Department of Material Science and Engineering,
Carnegie Mellon University, Pittsburg, PA. 15213, USA.
n
a Nano granular metallic iron (Fe) and titanium dioxide (TiO2−δ) were co-deposited on (100) lan-
J
thanum aluminate (LaAlO3) substrates in a low oxygen chamber pressure using a pulsed laser
8 ablationdeposition(PLD)technique. Theco-depositionofFeandTiO2 resultedin≈10nmmetal-
lic Fe spherical grains suspended within a TiO2−δ matrix. The films show ferromagnetic behavior
] withasaturationmagnetizationof3100Gaussatroomtemperature. Ourestimateofthesaturation
i
c magnetization based on the size and distribution of the Fe spheres agreed well with the measured
s value. The film composite structure was characterized as p-type magnetic semiconductor at 300
l- K with a carrier density of the order of 1022/cm3. The hole carriers were excited at the interface
r
t betweenthenanogranularFeandTiO2−δ matrixsimilartoholesexcitedinthemetal/n-typesemi-
m conductor interface commonly observed in Metal-Oxide-Semiconductor (MOS) devices. From the
large anomalous Hall effectdirectly observed in thesefilmsit follows that theholes at theinterface
.
t werestronglyspinpolarized. StructureandmagnetotransportpropertiessuggestedthatthesePLD
a
m films havepotential nano spintronics applications.
-
PACSnumbers: 75.50.Pp,71.30.+h,72.20.-i,71.70.Gm,72.25.-b,73.40.Qv
d
n
o
I. INTRODUCTION samplepreparation11,12,13,14,15. Films ofTiO onsub-
c 2−δ
[ stratesof(100)lanthanumaluminate (i.e. LaAlO3)were

deposited by a pulsed laser ablation deposition (PLD)
1 The search for semiconductors exhibiting magnetism
technique at various oxygen chamber pressures ranging
v atroomtemperaturehasbeenlongandunyielding. How-
from0.3to400mtorr. TheoriginofthepresenceofTi2+
5 ever, recently much progress has been made toward this
8 goal1,2,3,4,5,6,7. By doping a host semiconductor mate- and Ti3+ (as well as Ti4+) ions was postulated as a re-
2 sultofthe lowoxygenchamberpressureduringthe films
rial with transition metal ferromagnetic atoms, dilute
1 growth9. Oxygen defects gave rise to valence states of
ferromagnetic semiconductors have been produced with
1. Curie temperatures (T ) as high as 100 K1. Hall effect Ti2+ andTi3+ (inthebackgroundofTi4+)wherebydou-
c
0 ble exchange between these sites dominated. The same
measurements below T showed evidence for carriers be-
c
8 carriersinvolvedindouble exchangealsogaverisetoim-
ing spin polarized raising hopes for spintronics applica-
0 purity donor levels accounting for the transport proper-
tions. Specifically, metallic manganese (Mn) was doped
:
v into gallium arsenide (GaAs) whereby the carriers were tiesofthefilm. Thedilutenumberofcarrierswerepolar-
Xi strongly spin polarized1. ized in external applied magnetic field, yet the magnetic
moment was still rather small10. For example, normal
r We have previously reported the magnetic and dc- Hall resistivity was measured to be much bigger than
a transport properties of magnetic semiconductor films of
theanomalousHallresistivity10. Inordertoincreasethe
TiO , where δ indicates the degree of oxygen defi-
2−δ anomalous contribution to the Hall effect and thereby
ciency or defects in the film8,9,10. The Curie tempera-
increase the number of polarized carriers, we have fabri-
ture,T ≈880K,waswellaboveroomtemperaturewith
c cated a nano-granular (NG) metallic iron (Fe) in semi-
a saturation magnetization, 4πM ≈ 400 Gauss. Tita-
S conducting TiO matrix. The intent was to introduce
nium dioxide,TiO , is a wellknownwidebandgapoxide 2−δ
2 a substantial magnetization component internally to the
semiconductor, belonging to the group IV-VI semicon-
semiconductor TiO .
ductors, described in terms of an ionic model of Ti4+ 2−δ
and O2−11,12,13,14,15. Its intriguing dielectric properties The basic difference between our film composite and
allow its use as a gate insulator material in the Field- thepreviouslyreportedmagneticsemiconductorsbyoth-
Effect-Transistor (FET)16. Also, TiO is characterized ers is that in our films the NG of metallic Fe co-exist in
2
to be an n-type semiconductor with an energy gap vary- a metallic state, whereas magnetic semiconductors pre-
ing in the range 3 volt < ∆ < 9 volt depending on pared by others the transition metal co-exists as metal
e
2

oxides and often oxide clusters4,5,6,7. This major differ- 0) 0) 0) 0)
ence is important in terms of the magnetic and trans- 0 0 0 0
1 2 3 4
pproortdupcreodpebrytieussoafntdhethmaatgnofetoicthseerms4i,c5o,6n,7d.ucFtoorr pexreasmenptlley, S ( 004) S ( S ( 008) S (
NGmetallicFecontainsasignificanthighermomentand nts) (O2 10) O (2 20)
Ciduers.ie tAelmsop,ertahteurperetsheannceanofyNotGhemr etrtaanllsicitiFoen amlleotwasl ofoxr- Cou Ti c (1 Ti c (2
c c
the creation of a reservoir of conduction electrons in the g( b b
o
conductionbandand,therefore,holesintheTiO2−δ ma- L
trix. As electrons from the conduction band of TiO
2−δ
arethermallyintroducedintothe metallic Fe conduction
band, holes arecreatedin TiO much like injunctions
2−δ
ofNGmetal/semiconductorinterfacesorinMetalOxide
Semiconductor (MOS) devices17. Conduction of holes
occurs in the TiO2−δ host. This mechanism gives rise to 20 40 60 80 100 120
lower resistivity at high temperature in contrast to the
pure TiO reported earlier9,10 where the carriers were 2 (degree)
2−δ
only electrons. No such mechanism is possible in mag- FIG. 1: X-ray diffraction pattern shows anatase of (001)
neticsemiconductorsdopedwithtransitionmetaloxides. TiO2−δ and (110) bccphase.
We reportherethatinour compositefilms ofNG metal-
lic Fe in anatase TiO the magnetization, 4πM ≈
2−δ S
3,100Gauss(also≈0.4µ /Fe),atroomtemperature,T peared at 2θ = 32.20 o could not be readily indexed.
B c
above800K,the carriersarestronglyspinpolarizedand Fromourspeculation,thepeakmaybeoriginatedbypos-
theroomtemperatureresistivityisloweredbyafactorof sible presence of iron oxide (FeO or Fe O ), or ilmenite
2 3
≈10−3 relativetofilmsofthe undopedTiO semicon- (FeTiO ) phase.
2−δ 3
ductors previously produced8,9,10. Experimental results, In FIG.2, reflection electron diffraction (see upper in-
discussions and concluding remarks follow. set) supports the existence of the epitaxial TiO film
2−δ
host while the TEM image (and lower inset) reveal the
presenceofnanogranular(NG)metallicFeparticlessus-
II. EXPERIMENTAL RESULTS AND ANALYSIS pended within the TiO host. A JOEL 2000EX high
2−δ
resolutiontransmissionelectronmicroscope operatingat
200keVand400keVwasusedintheanalysis. AHRTEM
Thin films consisting of nano granular (NG) metallic
image shown in FIG.2 illustrates that NG metallic Fe
Fe and TiO were deposited by a pulsed laser abla-
2−δ
spheres were formed at the interface between TiO
tion deposition (PLD) technique from binary targets of 2−δ
layers and LaAlO substrate. Average diameter of NG
TiO and metallic Fe on (100) LaAlO substrates. Tar- 3
2 3 sphereswasmeasuredtobeintheorderof≈10to15nm
gets of TiO and Fe were mounted on a target rotator
2
as shown in the TEM image. The image suggests that
driven by a servomotor and synchronized with the trig-
theAT-PLDprocessmayleadimplantationofNGmetal-
ger of the pulsed excimer laser λ = 248 nm. In each
deposition cycle, the ratio of laser pulses incident upon
the TiO targettothoseuponthe Fe targetwas6:1,this
2
technique denotedas alternatingtargets-pulsedlaserab-
lation deposition (AT-PLD) technique. The substrate
temperature, laser energy density, and pulse repetition
rate were maintained at 700 oC,≈8.9 J/cm2, and 1 Hz,
respectively. The deposition was carried out in a pure
oxygen background of around 10−5 torr in order to in-
duce defects in the TiO host. There were a total of
2
4,206 laser pulses (3,606 pulses on TiO target and 600
2
pulses onFe target)for eachfilm resulting ina thickness
ofapproximately200nmasmeasuredbyaDek-Tekstep
profilometer.
Crystallographic and micrographic properties of the
AT-PLD films were measured using x-ray diffractome-
try(XRD)andtransmissionelectronmicroscopy(TEM).
Results indicate phases of anatase TiO , metallic body
2
centered cubic (bcc) Fe. (001) plane of anatase TiO
2−δ
and (110) plane of bcc Fe phase are clearly exhibited in FIG.2: TEMimagesforarepresentativefilmofnanogranular
thexrdpattern18,19,20 showninFIG.1,whereaspeakap- Fein TiO2−δ.
3

0
10
pure TiO2-
m)] -1 (Fe)TiO2-
10
c

(
T) 10-2
(
g[
o
L -3
10
-4
10
0 50 100 150 200 250 300 350 400
T (K)
FIG.3: Resistivity,ρasafunctionoftemperature,T,forthe
nano granular Fe in TiO2−δ film (dashed line) and the pure
TiO2−δ (solid line) from thereference10.
lic Fe spheres into host TiO . Average atomic ratio
2−δ
of Fe/Ti ions of the AT-PLD films was obtained by en-
ergy dispersivex-rayspectroscopy(EDXS) within anul-
tra high resolution scanning electron microscope (UHR-
SEM) column of Hitachi S-4800 that ≈ 6 ± 0.5% of Fe
arepossiblypresentedinthe1 × 1cm2 areaofthefilms.
ForsteadycurrentsandwiththemagneticintensityH
FIG. 4: (a) Energy band structure of nano granular metallic
directed normal to the film, the resistance matrix R in
the plane of the film may be written as21 Fe and semiconducting TiO2−δ, where it is similar to metal
oxidesemiconductor. (b) Contact potential model.
R R 1 ρ −ρ
R= xx xy = H (1)
(cid:18)Ryx Ryy(cid:19) t (cid:18)ρH ρ (cid:19) band. Since the Fermi energy in metallic Fe is about
4.6 eV above the conduction energy level it implies that
wherein ρ and ρ represent, respectively, the normal-
H
the energy band gap (3 eV) of TiO is degenerate
resistivity and the Hall resistivity. A ρ as a function 2−δ
with the iron conduction band. This means that the
of temperature for the film was measured in an applied
field of H = 0 Oe and shown in FIG.3. We note that donor levels in TiO2−δ are also degenerate, and thereby,
electrons hopping between Ti2+ and Ti3+ sites would
the value of ρ(T) was quite small at high temperature.
The ρ(T) behavior for pure TiO film9 (solid line) is find a conduit into the metallic conduction band leav-
2−δ
ing behind holes in TiO . Therefore,a smallpotential
also shown in FIG.3 exhibiting a typical metal-insulator 2−δ
barrier must exist at the interface. This is illustrated
transitioninthe temperaturerangeof4Kand300K.In
schematically in FIG.4a. As noted in FIG.2, the NG
contrast to the ρ for pure TiO , ρ for the composite
2−δ
metallic Fe sphere particles are isolated or disconnected,
NGmetallic Fe andTiO filmis afactorof1000lower
2−δ
implying that conduction in the TiO is via hole con-
andis constantfortemperatures between225K and400 2−δ
duction. This can be also explained by contact poten-
K. The temperature variations are of the characteristic
tial at the interface between NG metallic Fe sphere and
formexpectedfrommetalsemiconductorinterfaces. The
ρvalueat300Kwasmeasuredtobe183µΩ−cm,andis anataseTiO2−δ asdescribedinLandauLifshitz21. Inor-
about a factor of 20 largerthan resistivity of pure Fe22. der to remove electron trough the surface of a metallic
Fe, work must be done thermodynamically. According
Presence of these unique transport properties in the
to the reference21, Poisson’s equation for potential Φ(x)
NG metallic Fe spheres embedded in TiO films have
2−δ along the x−axis normal to the interface between NG
been modeled as coexisting electronic structures of both
metallic Fe / TiO implies
semiconducting TiO and NG metallic Fe. We have 2−δ
2−δ
modeled the mechanism for transport in this compos- −Φ′′(x)=4πρ(x),
ite film in a sketch shown in FIG.(4)a17,23,24. A com-
∞ 1 ∞
mon chemical potential energy or Fermi energy in both ̟ = xρ(x)dx=− xΦ′′(x)dx,
TiO2−δ andmetallicFeimpliesthattheconductionband Z−∞ 4π Z−∞
of TiO is degenerate with the metallic Fe conduction Φ(∞)−Φ(−∞)=V =4π̟, (2)
2−δ c
4

)
m
0.0
c
40
3 3 cm)30
- x10 -6.0 21 x10 /20
) ( p (10
H
(H 0
-12.0 100 150 200 250 300
T (K)
H (H=0 kOe)
H (H = 50 kOe)
-18.0
0 50 100 150 200 250 300 350 400
T (K)
FIG. 5: (Hall resistivity, ρH(H), versus temperature for H
= 0 (circle symbol) and H = 50.0 kOe (dashed line) in the
nanogranularFeinTiO2−δ film. Insetshowsrelation carrier
density in function of temperature..
wherein ̟ is defined to be dipole moment per unit con-
tact area and V is contact potential. The schematic of
c
contact potential at the interface is shown in FIG.4b.
Electronconductionbetweenmetallicspheresmaynot
be possible, see FIG.2. In FIG.5, Hall resistivity is plot-
ted as a function of temperature that at high tempera-
tures carriers are holes which also confirmed by Seebeck
measurements. Furthermore, the number of carriers rel-
ative to pure TiO has increased by factor 1,000 and
2−δ
the mass of holes is approximatelyabout 10 times larger
than the electronmass10. Also,in FIG.5 the carrierhole FIG. 6: (a) Magneto resistivity, ρ(H) versus H in different
densitypisplottedasafunctionoftemperature. Inorder temperatures. (b)Hallresistivity,ρH(H)versusHindifferent
temperatures.
to calculate p, we employ
1
ρ = R B+4πMR and R = , (3)
H 0 S 0
pec normal to the film plane (out-of-plane measurement).
The film can be fully saturated with a field of ≈ 6.6
wherein R and R are the normal and anomalous Hall
0 S kOe,sincethecoercivefieldwasmeasuredtobe2.0kOe.
coefficients,respectively. InFIG.6,ρ (H)scaleslinearly
H Field cooled (FC) and zero-field-cooled (ZFC) M(H,T)
with magnetic field for temperatures below 300 K. The
data in FIG.7 shows no difference which implies that no
slopeofρ (H)vs.H givesrisetothenormalHallcoeffi-
H spin glass effects are present in the films. There are
cientwhichcorrespondstothefirstterminEq.(3),where
two sources for magnetism in this composite structure:
the hole carrierdensity, p, may be deduced. Fortemper-
(1)ferromagnetisminTiO and(2)ferromagnetismin
atures well above 250 K, p is governed by the second 2−δ
metallic Fe spheres. The ferromagnetism in TiO is
term in Eq.(3), the anomalous Hall coefficient. Eq.(3) 2−δ
due to double exchangeas calculatedby us ina previous
may re-written as25
calculation and it gives rise to a relatively small satura-
tion magnetization at room temperature (≈ 400 Gauss).
H 1
ρH = + +RS (4πM) where RS ≫R0. (4) Based on the data presented in FIG.2, we estimate a
pec (cid:18)pec (cid:19) sphere density of 0.125 to 0.25 × 1018 spheres/cm3 and
spherediameterbetween10and15nm. Thisimpliesthat
The anomalous Hall effect is dominant at high temper-
the loading factor of metallic Fe in the composite film is
atures (T > 250 K) and the second term in Eq.(4) is
in the order of 0.05 to 0.25 which corresponds
constant for saturation magnetization M =M .
S
The spontaneous magnetization, M(H,T), was mea- to 1,100 to 5,100 Gauss for the saturation magneti-
suredto be nearly constantas a function of temperature zation of the composite. This compares with a mea-
as shown in FIG.7. The measurement was performed sured value of 3,100 Gauss. We have measured the
with an external dc-magnetic field of 10.0 kOe applied transverse magnetoresistivity, ρ(H), and Hall resistiv-
5

magnetic field of 6.8 kOe in Fe spheres at (say) room
0.35
temperature. These data also show relatively smaller
coercive field for the (M/M ) hysteresis loop than for
0.30 S
the ρ (H) hysteresis loop. The difference in hysteresis
H
)
G0.25 loopsisduetothefactthattheNGFesphericalsamples
3 0 give rise to dipole internal magnetic fields. In order to
1
x0.20 reverse the magnetization in each sphere it requires an
) ( external field to overcome this interactive internal field.
H
0.15
M( Hence, the coercive field is approximately constant with
temperature as it scales with magnetization. In the Hall
0.10
resistivitymeasurementthecoercivefieldisstronglytem-
0.05 H= 10 kOe FC perature dependent, since there are two contributions to
H = 10 kOe ZFC the resistivity measurement: (1) Normal Hall resistivity
0.00 contribution which is not hysteretic with respect to H
0 50 100 150 200 250 300
and (2) the anomalous Hall effects which is hysteretic
T (K)
with H, since this effect is proportional to the magne-
FIG. 7: Static magnetization, M(H,T), versus temperature tization. At high temperatures the AHE contribution
with H = 10.0 kOenormal to thefilm plane. dominatestheresistivitymeasurementincontrasttolow
temperatures where the normal Hall contribution domi-
nates. Thus, spin polarizationas reflectedin ρ (H), see
H
ity, ρH(H), as a function of magnetic field sweeps be- FIG.8, correlates extremely well with the magnetic hys-
tween - 90 and + 90 kOe as shown in FIGS.6a and teresis loopimplying that the magnetizedFe spheres are
6b, and at different temperatures. In FIG.6a, the nor- polarizing the carriers.
mal resistivity exhibits negative magnetoresistivity de-
fined as, ∆f = (ρ(H) − ρ(0))/ρ(0) for temperatures
100 K < T < 200 K. ∆f is negligible for temperatures III. DISCUSSION AND CONCLUSIONS
above300K.However,plots ofρ (H) showninFIG.6b,
H
clearly demonstrate that saturation effects at tempera-
The plots of ρ (H) at T < 200 K exhibit paramag-
turesnearroomtemperatureasaresultoftheholecarrier H
netic hysteresis behavior, see FIG.6b. This transition
density increasing toward saturation levels, see FIG.3.
between paramagnetic and ferromagnetic ρ (H) hys-
Notice that plots of ρ (H) at T ≥ 200 K exhibit clear H
H
teresis behavior was also reported and correlated with
hysteresisloopbehaviorsimilartoferromagnetichystere-
hole carrier density related with RKKY theory in pre-
sis loops. Interestingly, saturation for both ρ (H) and
H vious research1,26. Their estimate of the threshold hole
(M/M ), normalizedmagnetization,occurredatexactly
S
density for the carrier spin polarization transition was
the same external magnetic field value of ≈ 6.6 kOe, see
p = 3×1020/cm3 from paramagnetic p < 3×1020/cm3
FIG.8. This is no coincidence, since one requires a de-
to ferromagnetic p ≥ 3×1020/cm3 behavior1,25,26. Our
magnetization factor of of (4π/3) to magnetically satu-
estimate of hole carrier density of the films was p ≈
rate a sample of spherical shape in order to overcome a
2.1×1021/cm3 and p ≈ ×1022/cm3 at T = 100 K and
200 K, respectively. The film exhibited strong polariza-
tionandhysteresisbehaviorasafunctionofappliedfield
1.5 at 300 K. Since anomalous Hall effects may be observed
0.70
inthepresenceofspontaneousmagnetization27,28,wein-
1.0 fer that about ≈ 1022 /cm3 hole carriers are polarized
m) 0.35 at 300K at 300 K by an applied field H. This indicates that nearly all of
c 0.5 the carriers were polarized by the spontaneous magneti-
- M S zationat T = 300 K.According to FIG.8, carrierpolar-
- 6 0 0.00 0.0 M/ ization correlated very well with the magnetic hysteresis
x1 loop of NG Fe spheres. Also, FIG.8 indicates that the
(
H -0.5 carrier polarization is not affected by external fields up
-0.35
to 3.0 kOe, which can be an advantage for memory de-
-1.0 vice applications29,30 However, if the shape of the par-
ticles embedded in the composite could be shaped into
-0.70
-1.5 needles, it would imply spin polarization of carriers in
-20 -15 -10 -5 0 5 10 15 20
external magnetic fields below 0.1 kOe which is ideally
H (kOe)
suited for spintronics applications.
FIG.8: Symbol(X)showsatypicalanomalousHallhysteresis Insummary,magneticandmagneto-transportdatafor
loop, ρH(H), at T = 300 K with magnetic hysteresis loop films of nano granular metallic Fe and oxygen defected
(symbol (o)) at 300 K.
TiO are reported. The essence of this paper showed
2−δ
6
that conduction carriers of the films were strongly cou- was measured to be near ≥3.0 ×1022 /cm3. Therefore,
pled to residual magnetic moments of metallic Fe grains spintronicsandspindependentmemoryapplicationscan
inthenanocompositestructure. Thedramaticreduction be based upon the results presented here.
of normal resistivity (ρ(T)) of the films is a consequence
of two factors: (1) oxygendefects in the TiO hostin-
2−δ
duced electron hopping; (2) electrons from the TiO
2−δ
wereintroducedintothe conductionbandofFetocreate
IV. ACKNOWLEDGMENT
holes in TiO similar to a Metal Oxide Semiconduc-
2−δ
tor (MOS) structure. As a result the number of carriers
increased at room temperature. The holes in TiO This research was supported by the National Science
2−δ
werepolarizedduetothepresenceofferromagneticnano Foundation (DMR 0400676)and the Office of Naval Re-
granular metallic Fe, where the carrier polarized density search (N00014-07-1-0701).
∗ Corresponding author e-mail: [email protected] 17 B.G. Streetman, “Solid State Electronic Devices,” ch. 8,
† Electronic address: [email protected] 301, Prentice Hall, Inc., Englewood Cliffs, New Jersey
‡ Electronic address: [email protected] (1990).
§ Electronic address: [email protected] 18 Natl. Bur. Stand. (U.S.) Monogr. 25 7, 82 (1969).
1 H. Ohno,Science 281, 951 (1998). 19 C.J. Howard, T.M. Sabine, and F. Dickson, Acta Crystal-
2 K. Ueda, H. Tahata, and T. Kawai, Appl. Phys. Lett. 79, lographica, B: Structural Science B47, 462 (1991).
988 (2001). 20 H.E. Swanson, J.C. Doukhan, and G.M. Ugrinic, Natl.
3 Y. Masumoto et al., Science 291, 854 (2001). Bur. Stand. (U.S.), Circ. 5, 539 (1955).
4 S.J. Pearton et al., Mat. Sci. and Engn. R40, 137 (2003). 21 L.D. Landau and E.M. Lifshitz, “Electrodynamics of con-
5 S.A.ChambersandR.F.C.Farrow, MRS Bulletin 28,729 tinuous media,” 1984 ch. 4, Pregamonn Press, Oxford
(2003). (1984).
6 I.Z˘uti´c,J.Fabian,andD.Sarma,Rev. Mod. Phys.76.32 22 R.C. Weast andM.J. Astle, “CRC Handbook of chemistry
(2004). and physics,” section E, 81, CRC press, Inc. Boca Raton,
7 J.M.D. Coey, M. Venkatesan, and C.B. Fitzgerald, Nat. Florida, 63rd edition (1982).
Maters. 4, 173 (2005). 23 M.L.KnotekandP.J.Feibelman,Phys. Rev. Lett.40,964
8 N. H.Hong et al., Phys. Rev., B 73, 132404 (2006). (1978).
9 S.D. Yoon et al., J. Phys.; Condens. Matter. 18, L355 24 Z. Zhang, S.P. Jeng, and V. Henrich, Phys. Rev. B 43,
(2006). 12004 (1991).
10 S.D. Yoon et al., J. Phys.: Condens. Matter. 19, 326202 25 C.M.Hurd,“The Hall Effect in Metals and Alloys,”1972
(2007). Ch. 5, Plenum Press, New YorkLondon.
11 M. Earle, Phys. Rev. 61, 56 (1942). 26 T. Story,et al., Phys. Rev. Lett. 56, 777 (1986).
12 R.G. Breckenridge and W.R. Hosler, Phys. Rev. 91, 793 27 C.L.ChienandC.R.Westgate,“TheHalleffectanditsap-
(1953). plications,” Plenum Press, NewYork and London (1980).
13 N. Daude, C. Gout, and C. Jouanin, Phys. Rev. B 15, 28 R. Karplus and J. M. Luttinger, Phys. Rev. 95, 1154
3229 (1977). (1954).
14 J.Pascual,J.Camassel,andH.Mathieu,Phys.Rev.B18, 29 G. Prinz, Science 282, 1660 (1998).
5606 (1978). 30 S.A. Wolf et al., Science 294, 1488 (2001).
15 H. Tang et al., Solid State Commun.87, 847 (1993).
16 S.A.Campbelletal.,IBMJ.Res.Develop.43,383(1999).