letters to nature
High critical current density and
enhanced irreversibility ®eld in
superconducting MgB2 thin ®lms
C. B. Eom*², M. K. Lee*, J. H. Choi*², L. J. Belenky*, X. Song²,
L. D. Cooley², M. T. Naus², S. Patnaik², J. Jiang², M. Rikel²,
A. Polyanskii², A. Gurevich², X. Y. Cai², S. D. Bu*, S. E. Babcock*²,
E. E. Hellstrom*², D. C. Larbalestier*², N. Rogado³, K. A. Regan³,
M. A. Hayward³, T. He³, J. S. Slusky³, K. Inumaru³, M. K. Haas³
& R. J. Cava³
* Department of Materials Science and Engineering, 1509 University Avenue; and
² Applied Superconductivity Center, 1500 Engineering Drive, University of
Wisconsin, Madison, Wisconsin 53706, USA
³ Department of Chemistry and Princeton Materials Institute,
Princeton University, Princeton, New Jersey 08544, USA
..............................................................................................................................................
The discovery of superconductivity at 39 K in magnesium
diboride1 offers the possibility of a new class of low-cost,
high-performance superconducting materials for magnets and
electronic applications. This compound has twice the transition
temperature of Nb3Sn and four times that of Nb-Ti alloy, and the
vital prerequisite of strongly linked current ¯ow has already been
demonstrated2±5. One possible drawback, however, is that the
magnetic ®eld at which superconductivity is destroyed is modest.
Furthermore, the ®eld which limits the range of practical applicationsÐthe irreversibility ®eld H*(T)Ðis approximately 7 T at
liquid helium temperature (4.2 K), signi®cantly lower than about
10 T for Nb-Ti (ref. 6) and ,20 T for Nb3Sn (ref. 7). Here we show
that MgB2 thin ®lms that are alloyed with oxygen can exhibit a much
steeper temperature dependence of H*(T) than is observed in bulk
materials, yielding an H* value at 4.2 K greater than 14 T. In
addition, very high critical current densities at 4.2 K are achieved:
1 MA cm-2 at 1 T and 105 A cm-2 at 10 T. These results demonstrate
that MgB2 has potential for high-®eld superconducting applications.
Three MgB2 ®lms were prepared by pulsed laser deposition at
room temperature onto (111) oriented single crystal SrTiO3 substrates. The base pressure before deposition was 3 3 10 2 6 Pa, and
deposition took place under 0.3 Pa of argon. The MgB2 target was
prepared by pressing and sintering, as described previously2,8. After
deposition, ®lm 1 was annealed in an evacuated niobium tube at
850 8C for 15 minutes. Film 2 was annealed in an evacuated quartz
tube at 750 8C for 30 minutes and quenched to room temperature.
Film 3 was annealed in a tantalum envelope inside an evacuated
niobium tube at 750 8C for 30 minutes. Magnesium pellets were
included in each tube to prevent magnesium loss. Film 1 showed
visible evidence of inhomogeneity in the centre of the ®lm, which
could not be excluded from samples used for electromagnetic
measurements. The thickness of each ®lm was approximately 500 nm.
The chemical compositions of the ®lms were determined by
wavelength dispersive X-ray spectroscopy using an electron microprobe at 3 keV to avoid X-ray generation in the substrate. Before
annealing, the ®lms had a Mg:B:O atomic ratio of about
1.0:1.0:0.17, the oxygen probably coming from the target or background gas during deposition. The primary differences in the ®lms
are the measured ratios of Mg:B:O obtained after annealing. Film 2
had a ratio of ,1.0:0.9:0.7, which corresponds to a relative oxygen
concentration approximately twice that of ®lm 1 and 3, as well as
higher Mg content. By contrast, ®lm 1 and ®lm 3 show 1.0:1.0:0.4
and 1.0:1.2:0.3 ratios, respectively. We believe that the higher oxygen
of post-annealed ®lm 2 came from the relatively poor vacuum
(approximately 150 Pa) and the larger volume of the tube used for
annealing. X-ray diffraction also revealed that the MgB2 lattice
parameter c is larger in ®lm 2 than in ®lm 1, ®lm 3, or bulk samples.
558
Figure 1 presents v±2v scans for ®lms 1, 2 and 3, which clearly show
that the (002) MgB2 peak of ®lm 2 is shifted to lower angle. We
believe the expansion of the c-lattice parameter in MgB2 is due to
oxygen alloying of the boron layer.
The transition temperature Tc was determined by both inductive
and resistive measurements. The resistive transitions, shown in
Fig. 2, have onset values of 36, 31 and 35 K for ®lms 1, 2 and 3
respectively, and fall to zero resistance just above the principal
inductive transitions. The room temperature resistivity r(300 K) of
®lm 2, 390 mQ cm, is clearly much higher than that of ®lm 1
(125 mQ cm) or ®lm 3 (60 mQ cm). High-resistivity ®lm 2 also
exhibits a slight resistivity increase with decreasing temperature,
indicating a resistivity ratio, RR ˆ r…300 K†=r…40 K†, less than 1. By
contrast, the RR of ®lm 1 is about 1.5, whereas that of ®lm 3 is 1.3.
For comparison, the RR of Nb-Ti alloys (Ginzburg±Landau parameter k < 40) is seldom higher than 1.4 (ref. 6), whereas that of
Nb3Sn (k < 30) is typically between 2 and 5 (ref. 7). Good bulk
samples of MgB2 can attain RR values of up to 25 and have k < 25
(refs 4 and 5).
The magnetization of 3 mm 3 3 mm pieces of each ®lm was
measured from 4.2 to 30 K and in perpendicular ®elds up to 14 T
using a vibrating sample magnetometer (VSM). Hysteresis curves at
4.2 K (Fig. 3a) close at ®elds very much higher for ®lm 2, at about
14 T, than for ®lm 1 (,6 T) or ®lm 3 (,6.5 T). We were not able to
unambiguously determine Hc2 owing to the smooth blending of the
superconducting magnetization into the background magnetization of substrate and sample holder. This is in marked contrast to
the behaviour seen for bulk samples; these have a signi®cant
reversible contribution to the remaining hysteretic magnetic
moment above the primary loop closure ®eld2,4,9±11. The oxygenrich ®lm 2 has a much higher irreversibility ®eld than ®lms 1 and 3
up to 25 K.
Magneto-optical examination of the ®lms showed strong ¯ux
shielding with only slightly distorted roof patterns, indicative of
macroscopic supercurrents. This validates use of the standard
expression12 for the magnetization of a thin ®lm, J c ˆ
30DmV 2 1 r 2 1 , to extract critical current densities Jc(H,T), where
Dm is the width of the hysteresis loop, V < 4:5 3 10 2 6 cm3 is the
®lm volume, and r ˆ 0:15 cm is the sample half-width. Figure 3b
shows that the highest Jc value at 1 T, 4.2 K exceeds 3 MA cm-2 for the
low-resistivity ®lm 3. This result is comparable to that of ref. 13.
160
MgB2(002)
120
MgO (200)
Intensity
.................................................................
Film 2
80
Film 3
40
Film 1
0
40
42
44
46
48
50
52
54
56
2θ (degrees)
Figure 1 X-ray diffraction v±2v scans of ®lms 1 (black curve), 2 (red curve) and 3 (blue
curve), showing both (002) MgB2 and (002) MgO peaks. The 2v position of the MgB2 (002)
peak for ®lm 2 (51.58) is smaller than for ®lms 1 and 3 (51.98). The lattice parameter c of
MgB2 in ®lm 2 is 0:3547 6 0:0007 nm, substantially larger than the 0.3521 nm values
for ®lms 1 and 3 and also for bulk material1,3. The sample was tilted at about 0.58 to avoid
the strong (111) SrTiO3 substrate diffraction.
© 2001 Macmillan Magazines Ltd
NATURE | VOL 411 | 31 MAY 2001 | www.nature.com
letters to nature
a
30
3
2
Moment (10–6 A m2)
20
1
0
–1
10
0
5
10
15
–2
–3
0
0
5
10
15
–10
4.2 K
Film 1
Film 2
Film 3
–20
–30
Field (T)
b
10 7
Film 3
4.2 K
Film 1
4.2 K
Jc (A m–2)
10 6
10 5
10
Film 2
4.2 K
4
25 K
10 3
20 K
0
10 K
15 K
5
10
15
Field (T)
c
20
Film 1
Film 2
Film 3
15
Field (T)
However, the much larger H*(4.2 K) of ®lm 2 enables Jc to achieve
1 MA cm-2 at 1 T and to remain above 105 A cm-2 up to 10 T, well
above H* for ®lms 2 and 3. Film 2 also exhibits substantial high-®eld
current densities up to 20 K. Jc values of ®lm 1 are well below those
of ®lm 3 at 4.2 and 20 K, which may be owing to its inhomogeneity.
The bulk pinning-force, F p …H† ˆ m0 HJ c …H†, exceeds 40 GN m-3 at
4.2 K for ®lm 3, whereas Fp approaches 20 GN m-3 at 5 T for ®lm 2
and remains above 10 GN m-3 between 1 and 10 T. These data are
comparable to or exceed the pinning force of the established
technological superconductors Nb47wt%Ti (ref. 6) and Nb3Sn
(ref. 7), which lie in the range 15±30 GN m-3.
Figure 3c summarizes the irreversibility ®elds for bulk and thin®lm MgB2 derived from VSM data. The extrapolation to zero of the
0.25
Kramer function (J0.5
) was used to de®ne the irreversibility ®eld
c H
for the thin ®lms. These extrapolations were nearly straight for ®lms
1 and 3, just as in bulk samples2, whereas the Kramer plots for ®lm 2
were curved because of stronger pinning at higher ®elds. These
analyses produce a H*(T) line for ®lm 2 that has a much steeper
slope, m0 dH*=dT < 0:7 T K 2 1 , than the H*(T) data for bulk samples, which have a slope of about 0.3 T K-1. This steeper H*(T)
characteristic allows H*(T) for ®lm 2 to intersect and exceed bulk
sample Hc2(T) values below about 10 K. In strong contrast to the
®lm 2 data, the H*(T) analyses for lower resistivity ®lms 1 and 3 lie
along much lower lines, and do not exceed bulk H*(T) values at low
temperature.
Transmission electron microscopy (TEM) was used to characterize the phase constitution and nanostructure of ®lm 2. The MgB2
grain size was remarkably small, approximately 10 nm, as shown in
Fig. 4. Selected-area diffraction patterns were all ring patterns (Fig. 4
inset) and indicated the presence of signi®cant volume fractions of
both MgB2 and MgO. As in the wavelength-dispersive microprobe
results discussed earlier, energy-dispersive X-ray microanalysis in
the TEM also suggests a considerably higher concentration of
oxygen compared to what we have observed in TEM specimens
from bulk MgB2 (X.S. et al., unpublished work).
The TEM analyses of ®lm 2 and X-ray diffractometry of all three
®lms indicate that all ®lms have a [00l] texture with a large mosaic
spread normal to the substrate and random in-plane orientation.
The rocking curve widths of the (002) re¯ection of MgB2 are 4.88,
7.38 and 8.28 for ®lms 1, 2 and 3, respectively. The broadening of the
(002) X-ray peaks is consistent with the ®ne grain size observed by
TEM. Using the Sherrer formula d ˆ 0:9 l=B cosv, where B is the full
width at half maximum of the 2v diffraction peaks and l is the X-ray
wavelength used, grain dimensions along the c axis are 17 nm, 7 nm
and 8 nm for ®lms 1, 2 and 3, respectively. The MgO grains appeared
10
Hc2(T) bulk
H*(T) bulk
5
500
0
0
Resistivity (µΩ cm)
400
80
Film 1
Film 2
Film 3
60
40
200
20
0
20
25
30
35
40
45
50
100
0
0
50
20
30
40
Temperature (K)
100
300
10
100
150
200
250
300
Temperature (K)
Figure 2 Resistivity as a function of temperature. Film 1 is indicated by the red curve with
circles, ®lm 2 by the black curve with squares, and ®lm 3 by the blue curve with crosses.
The inset shows a magni®ed view of the resistive transitions near the critical temperature.
The uncertainty of the resistivity value is about 620% owing to the large voltage contacts
relative to sample dimensions.
NATURE | VOL 411 | 31 MAY 2001 | www.nature.com
Figure 3 Magnetization measurements for ®lms 1 (red curves), 2 (black curves), and 3
(blue curves). a, Raw magnetization data, obtained using a vibrating sample
magnetometer, is compared at 4.2 K. The inset shows a magni®ed view of the closure of
hysteresis loops, after subtracting a linear background that is due to the paramagnetic
moment of the samples and the sample holder. b, The derived values of the critical current
density are plotted as a function of ®eld at 4.2 and 20 K for ®lms 1 and 3, and at 4.2, 10,
15, 20, and 25 K for ®lm 2. We note that the data for ®lm 2 at 4.2 K are above 105 A cm-2
(a common benchmark for superconducting magnet applications) at nearly 10 T. c, The
magnetization data are used to estimate an H±T phase diagram for MgB2 thin ®lms in
comparison to bulk samples. The grey bands indicate established H c2(T ) and H*(T ) lines
for bulk MgB2, based on inductive2,4,5,9±11 and resistive4,5,9,11 measurements. The
irreversibility ®elds for ®lms 1 (red) and 3 (blue) lie close to the H*(T ) line for bulk
materials, even though their Tc values are lower than the approximate 39 K bulk value. The
irreversibility ®eld for ®lm 2 increases at a faster rate of approximately 0.7 T K-1 below the
critical temperature of 28.5 K, crossing the bulk H c2(T ) line below about 10 K.
© 2001 Macmillan Magazines Ltd
559
letters to nature
to be randomly oriented with a grain size of about 10 nm in all three
®lms.
Taken as a whole, the experiment shows two primary results that
have important implications for the potential of MgB2 in superconducting magnets. First, unlike bulk samples, which appear to
have a common Tc, Hc2(T) and H*(T), here we see a strong
enhancement of H*(T) that correlates well to the strongly enhanced
resistivity and concomitant reduction of the electron mean free path
and superconducting coherence length of ®lm 2. Films 1 and 3,
which were annealed in metal tubes and have lower oxygen
concentrations, have metallic resistivity and smaller H* values and
H*(T) slopes. Film 2, which was annealed in quartz and has higher
measured oxygen concentration and larger c axis parameter
(0.3547 nm versus 0.3521 nm for bulk) has a weak increase of
resistivity with decreasing temperature and a much steeper H*(T)
curve, suggesting that oxygen can substitute for boron and expand
the c axis of MgB2. In both alloy and intermetallic-compound
superconductors, substitutional alloying to decrease the electron
mean free path is very important for enhancing high-®eld superconductivity, as indicated by the direct dependence of Hc2(0) on the
normal-state resistivity through m0 H c2 …0† ˆ 3:110rgT c (ref. 14).
Taking the coef®cient of electronic speci®c heat g ˆ 153 J m 2 3 K 2 2
for MgB2 (ref. 15), the resistivity values predict m0Hc2(0) values of
12, 57 and 7.6 T for ®lms 1, 2 and 3 respectively, assuming that the
dirty limit of the Ginsburg±Landau theory applies. Although we
were not able to deduce Hc2 values from the magnetization measurements, these predictions are at least qualitatively reasonable
given the data in Fig. 3, which suggest extrapolated values of
m0H*(0) of 10, 22 and 7 T for ®lms 1, 2 and 3, respectively.
Second, the thin-®lm process produces textured MgB2 ®lms with
extremely small grains and a substantial fraction of extremely small
a
20 nm
b
c
Figure 4 Transmission electron microscopy. a, Dark-®eld, plan-view image of ®lm 2 in
which only a fraction of the grains show bright intensity, clearly revealing the 10 nm grain
size of the material. b, c, Selected area diffraction patterns from about 1 mm square areas
of ®lm 2. All of the rings can be attributed to either MgB2 or MgO. The pattern in b was
collected with the incident electron beam aligned approximately parallel to the ®lm
normal. It shows only complete rings and those rings that correspond to the MgB2 phase
are all of the (hk0) type. The pattern in c was collected with the incident beam oriented at
an angle to the ®lm normal. It shows an uneven distribution of intensity along all MgB2
rings. Considered together, these diffraction patterns indicate that the ®lm possesses a
(00l ) (that is, c axis) ®bre texture oriented parallel to the ®lm normal with little or no texture
in the plane of the ®lm.
560
(,10 nm) MgO particles. A high number density of ¯ux pinning
centres is crucial for obtaining useful (.105 A cm-2) current densities at high ®elds. The exceptionally ®ne 10-nm scale of the grain
size would be consistent with strong grain boundary pinning, as in
Nb3Sn, for which F p ~ d 2 1 (refs 16 and 17). Additional corepinning by the non-superconducting MgO particles can also contribute to the overall pinning force17. Moreover, the very high Jc
values con®rm the absence of electromagnetic granularity2. Evidently the threshold at which grain boundaries block current lies
well above Jc(T,H) and closer to the depairing current density of
about 108 A cm-2, an important requisite for attaining strong
pinning in polycrystalline superconductors18.
In summary, it appears that MgB2 can be alloyed, probably with
oxygen, so as to increase the irreversibility ®eld to levels where it
may compete with today's materials of choice, Nb46wt%Ti and
Nb3Sn. In addition to removing this fundamental limitation, the
®ne-grained ®lms have technologically important current densities
and strong ¯ux pinning in perpendicular ®elds, which is the weaker
orientation for layered superconductors. These results demonstrate
that the principles that have guided the development of established
high-®eld superconductors can also be applied successfully to
MgB2. Improved high-®eld properties may thus be obtained in
alloyed magnesium boride wires. We do not yet know how the
development of texture, the extremely small grain size, the alloying
of the MgB2, and the substantial presence of MgO are related, if
at all. Further study of epitaxial MgB2 and alloyed magnesium
boride ®lms on different substrate materials will be helpful in this
M
regard.
Received 20 March; accepted 27 April 2001.
1. Nagamatsu, J., Nakagawa, N., Muranaka, T., Zenitani, Y. & Akimitsu, J. Superconductivity at 39 K in
magnesium diboride. Nature 410, 63±64 (2001).
2. Larbalestier, D. C. et al. Strongly linked current ¯ow in polycrystalline forms of the superconductor
MgB2. Nature 410, 186±189 (2001).
3. Bud'ko, S. L. et al. Boron isotope effect in superconducting MgB2. Phys. Rev. Lett. 86, 1877±1880 (2001).
4. Finnemore, D. K., Ostenson, J. E., Bud'ko, S. L., Lapertot, G. & Can®eld, P. C. Thermodynamic and
transport properties of superconducting Mg10B2. Phys. Rev. Lett. 86, 2420±2422 (2001).
5. Can®eld, P. C. et al. Superconductivity in dense MgB2 wires. Phys. Rev. Lett. 86, 2423±2426 (2001).
6. Lee, P. J. in Wiley Encyclopedia of Electrical and Electronics Engineering (ed. Webster, J. G.) Vol. 21, 75±
87 (Wiley, New York, 1999).
7. Smathers, D. B. in Metals Handbook 10th edn Vol. 2, Properties and Selection: Nonferrous Alloys and
Special-Purpose Materials 1060±1076 (ASM Intl, 1990).
8. Slusky, J. S. et al. Loss of superconductivity with the addition of Al to MgB2 and a structural transition
in Mg1-xAlxB2. Nature 410, 343±345 (2001).
9. MuÈller, K.-H. et al. The upper critical ®eld in superconducting MgB2. Preprint cond-mat/0102517 at
hxxx.lanl.govi (2001).
10. Takano, Y. et al. Superconducting properties of MgB2 bulk materials prepared by high pressure
sintering. Appl. Phys. Lett. 78, 2914±2916 (2001).
11. Bugoslavsky, Y., Perkins, G. K., Qi, X., Cohen, L. F. & Caplin, A. D. Critical currents and vortex
dynamics in superconducting MgB2. Nature 410, 563±565 (2001).
12. Gyorgy, E. M., van Dover, R. B., Jackson, K. A., Schneemeyer, L. F. & Waszzak, J. V. Anisotropic critical
currents in Ba2YCu3O7 analyzed using an extended Bean model. Appl. Phys. Lett. 55, 263±265 (1989).
13. Kang, W. N., Kim, H.-J., Choi, E.-M., Jung, C. U. & Lee, S.-I. Epitaxial MgB2 superconducting thin
®lms with a transition temperature of 39 kelvin. Science (in the press).
14. Orlando, T. P., McNiff, E. J. Jr, Foner, S. & Beasley, M. R. Critical ®elds, Pauli paramagnetic limiting,
and material parameters of Nb3Sn and V3Si. Phys. Rev. B 19, 4545±4561 (1979).
15. Wang, Y., Plackowski, T. & Junod, A. Speci®c heat in the superconducting and normal state (2±300 K,
0±16 T), and magnetic susceptibility of the 38-K superconductor MgB2: Evidence for a multicomponent gap. Preprint cond-mat/0103181 at hxxx.lanl.govi (2001).
16. Dew-Hughes, D. The role of grain boundaries in determining Jc in high-®eld, high-current superconductors. Phil. Mag. 55, 459±479 (1987).
17. Campbell, A. M. & Evetts, J. E. Flux vortices and transport currents in type II superconductors. Adv.
Phys. 21, 199±428 (1972).
18. Gurevich, A. & Cooley, L. D. Anisotropic ¯ux pinning in a network of planar defects. Phys. Rev. B 50,
13563±13576 (1994).
Acknowledgements
The work at the University of Wisconsin was supported by funding from the US
Department of Energy, the Air Force Of®ce of Scienti®c Research, the National Science
Foundation through the Materials Research Science and Education Center for
Nanostructured Materials and a David Lucile Packard Fellowship (C.B.E.). The work at
Princeton University was supported by the National Science Foundation and the US
Department of Energy.
Correspondence and requests for materials should be addressed to C.B.E.
(e-mail: eom@engr.wisc.edu).
© 2001 Macmillan Magazines Ltd
NATURE | VOL 411 | 31 MAY 2001 | www.nature.com