Synthesis and characterization of overdoped Hg-1234 and Hg-1245 phases; the universal behavior of Tc variation in the HgBa2Can-1CunO2n+2+δ series

Article abstract of DOI:10.1016/S0921-4534(98)00066-5

HgBa₂Ca₃Cu₄O_10+δ and HgBa₂Ca₄Cu₅O_12+δ were prepared by high pressure/high temperature technique. The phase composition in the Hg-Ba-Ca-Cu-O system was shown to be very sensitive to synthesis conditions. A variation of the oxygen content in the starting oxide mixture by changing the BaO/BaO₂ ratio allowed the acquisition of Hg-1234 and Hg-1245 in the overdoped state. The heat treatments (nitrogen and oxygen flows, oxygen pressure 100 bar and 2000 bar) resulted in a change of T_cs from initial overdoped (116 and 100 K for Hg-1234 and Hg-1245, respectively) to underdoped (95 and 85 K) and, finally, to optimally-doped states (125 and 111 K). The dependencies of T_c vs. a-parameter (and, consequently, oxygen content) for these phases exhibit a cupola-shaped behavior, as it was previously found for the lower (n ≤ 3) members of the HgBa₂Ca_n-1Cu_nO_2n+2+δ series.

Full text of DOI:10.1016/S0921-4534(98)00066-5

Ž
.
Physica C 300 1998 71–76
Synthesis and characterization of overdoped Hg-1234 and
Hg-1245 phases; the universal behavior of Tc variation in the
HgBa2Cany1CunO2nq2qd series
a
a
a
b
K.A. Lokshin , D.A. Pavlov , M.L. Kovba , E.V. Antipov a,), I.G. Kuzemskaya ,
b
b
b
b,1
L.F. Kulikova , V.V. Davydov , I.V. Morozov , E.S. Itskevich
a
Department of Chemistry, Moscow State UniÕersity, 119899, Moscow, Russian Federation
Institute of High Pressure Physics RAS, 142092, Troitsk, Moscow, Russian Federation
b
Received 10 December 1997; accepted 14 January 1998
Abstract
HgBa2Ca3Cu4O10
and HgBa2Ca4Cu5O12qd were prepared by high pressurerhigh temperature technique. The phase
qd
composition in the Hg–Ba–Ca–Cu–O system was shown to be very sensitive to synthesis conditions. A variation of the
oxygen content in the starting oxide mixture by changing the BaOrBaO2 ratio allowed the acquisition of Hg-1234 and
Ž
.
Hg-1245 in the overdoped state. The heat treatments nitrogen and oxygen flows, oxygen pressure 100 bar and 2000 bar
Ž
.
resulted in a change of Tcs from initial overdoped 116 and 100 K for Hg-1234 and Hg-1245, respectively to underdoped
Ž
.
Ž
.
Ž
95 and 85 K and, finally, to optimally-doped states 125 and 111 K . The dependencies of Tc vs. a-parameter and,
.
consequently, oxygen content for these phases exhibit a cupola-shaped behavior, as it was previously found for the lower
Ž
.
nF3 members of the HgBa2Cany1CunO2 nq2qd series. q 1998 Elsevier Science B.V.
Keywords: Hg-1234; Hg-1245; High pressure synthesis; Heat treatments
Ž .
Ž .
1. Introduction
perovskite slab n and a concentration d of the
extra oxygen located in the Hg layers.
Hg-based superconducting Cu mixed oxides fam-
ily, HgBa2Cany1CunO2 nq2qd with ns1–6, is a
very attractive subject in the study of the fundamen-
tal aspects of high Tc superconductivity in layered
cuprates. It is well-known that there are two main
parameters influencing Tc in this family: a width of a
In the latter case, the dependence of Tc vs. extra
Ž .
oxygen content d exhibits a cupola-shaped behav-
ior as it was found for Hg-1201, Hg-1212, and
w
x
recently for Hg-1223 1–3 . The origin of this phe-
nomenon is a change of a carrier concentration in a
conducting band from underdoped to overdoped
states, followed by a Tc variation with a maximum
corresponding to an optimal hole doping. It should
be mentioned that the larger width of the perovskite
slab in the structure, the more difficult to obtain the
overdoped Hg-based superconductor. For instance, a
heat treatment under 100 bar oxygen pressure de-
)
Corresponding author. Fax: q7-095-9394788; e-mail:
antipov@inorg.chem.msu.su.
1 Also corresponding author. Fax: q7-095-9394788; e-mail:
itskev@ns.hppi.troitsk.ru.
0921-4534r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.
Ž
.
PII S0921-4534 98 00066-5

(
)
72
K.A. Lokshin et al.rPhysica C 300 1998 71–76
creases Tc for the first two members, while no
significant changes in Tc has been found for the third
ters might allow the synthesis of the high members
in overdoped states.
w
x
member 2,4,5 . Recently, we have shown that the
overdoped Hg-1223 samples with the much smaller
The goal of the present work was to obtain over-
doped Hg-1234 and Hg-1245 phases and to clarify
the relationship between Tc and extra oxygen content
variation for these compounds.
Ž
.
Tc s118 K than the optimal value 135 K could be
prepared under high pressure using the initial oxide
w x
mixture with a high oxygen content 3 . In contrast to
Ž
.
these phases the lower members , the overdoped
state has not yet been reported for the higher mem-
2. Experimental
w x
6 and
bers with nG4. Earlier, Loureiro et al.
w x
Capponi et al. 7 showed that the Tc s for the under-
doped phases with ns4 and 5 could be enhanced
by a heat treatment under 1300 bar oxygen pressure
up to 125 and 110 K, respectively, but a decrease of
Tc due to overdoping was not observed.
The high pressurerhigh temperature synthesis
used in the present study was described in detail in
w x
Ref. 3 . The stoichiometric mixtures of HgO
Ž
.
Ž
.
99.9% , BaO, BaO2 , CaO and CuO 99.9% were
used as the starting materials. BaO was prepared by
Ž
.
The second structural parameter influencing Tc in
this series is a width of the perovskite-like slab. The
highest reported value of Tc for each member in-
a decomposition of BaCO3 99.9% in vacuum at
11008C for 72 h. CaO was synthesized by a similar
Ž
.
way from CaCO3 99.9% annealed at 9008C. BaO2
was obtained by heating BaO at 450–5008C in puri-
fied oxygen. The initial mixtures ‘HgBa2Ca3Cu4Ox’
and ‘HgBa2Ca4Cu5Ox’ were prepared from the start-
ing reagents with a different oxygen concentration
Ž
.
creases from the first HgBa2CuO4qd , Tc s97 K up
Ž
to the third member HgBa2Ca2Cu3O8qd , Tc s135
.
K
and then decreases to 107 K for HgBa2Ca5-
w
x
Cu6O14qd 7,8 . Thus, a dependence of Tc vs. n
exhibits a cupola-shaped behavior too. The reason
for this phenomenon is still unknown. The probable
explanation might be an insufficient carrier concen-
tration for the higher members of the series with
nG4 or specific structural features of these phases.
Therefore, it would be interesting to prepare Hg-1234
and Hg-1245 samples in overdoped states to accept
or reject the first assumption.
Ž .
x . The variation of oxygen content was made by
changing the BaOrBaO2 ratio. All preliminary oper-
ations were carried out in a glove box filled with
nitrogen, and NaOH and P2O5 were used for a
purification from CO2 and water vapor. The synthe-
ses were performed under high pressure 2–4 GPa
in gold capsules volume of a capsule was about 0.04
Ž
.
Ž
.
ml at the temperature range 900–11008C for 2 h.
Earlier, we showed that a variation of the
BaO2rBaO ratio in the initial oxide mixture was an
important parameter, which influenced not only the
extra oxygen content in the phase, but a phase
Subsequent heat treatments of the obtained sam-
ples were performed at 3008C in purified oxygen and
nitrogen flows. Several annealings were carried out
under high oxygen pressure of 100 and 2000 bar at
3008C.
w x
formation as well 3 . For instance, the variation of x
from 10.2 to 10.6 in the initial oxide mixture
‘HgBa2Ca3Cu4Ox’ resulted in a drastic change of
the phase composition of the as-prepared samples. A
high fraction of Hg-1234 was obtained when the
synthesis was performed under 2 GPa at 9008C for 2
h and x was equal to 10.2, while no traces of this
phase were detected in the sample prepared under
the same conditions with xs10.6. The latter sample
contained mainly overdoped Hg-1223. The formation
of the HgBa2Cany1CunO2 nq2qd compounds in the
Hg–Ba–Ca–Cu–O system is influenced by other
parameters like temperature, pressure, and cation
composition. Therefore, a variation of these parame-
X-ray powder diffraction study of the prepared
samples was performed using a Guinier focusing
Ž
.
monochromatic camera FR-552 Cu–K a1 radiation
with Ge as an internal standard. The lattice parame-
ters of the Hg-based superconductors were deter-
mined using the same set of non-overlapping strong
reflections for all samples. Fractions of different
phases were estimated visually using intensities of
the main reflections.
The temperature dependence of AC susceptibility
was measured by an induction method with an exter-
nal field amplitude of 1 Oe at a modulation fre-
quency of 27 Hz.

(
)
K.A. Lokshin et al.rPhysica C 300 1998 71–76
73
Table 2
3. Results and discussion
Characteristics of the samples prepared from the starting mixture
‘HgBa2Ca4Cu5Ox’ under 2 GPa at T s9508C for 2 h
3.1. High pressure synthesis
N
x
Phase composition, % Lattice parameters
Tc , K
˚
˚
c, A
For the preparation of overdoped Hg-1234 and
Hg-1245 superconductors, we decided to use two
a, A
Ž
.
.
Ž .
Ž .
Ž .
Ž .
5
12.6 Hg-1234 80
3.8485 6 19.001 7 116
3.849 1 22.18 1
Ž .
Ž
different ways: i to vary temperature and pressure
Hg-1245 10
101
Ž .
Ž .
CaO 5 , CuO 5
for annealing the samples with the fixed initial stoi-
Ž
.
.
Ž .
Ž .
Ž .
Ž .
6
12.8 Hg-1245 80
3.8486 4 22.151 5 100
3.848 2 19.01 2
Ž .
chiometry ‘HgBa2Ca3Cu4O10.5’ and ii to vary oxy-
Ž
Hg-1234 10
Ž .
114
gen content in the starting mixture ‘HgBa2Ca4-
Ž .
CuO 5 , CaO 5
Ž
Cu5Ox’ annealed under fixed conditions 2 GPa,
Ž
.
.
Ž .
Ž .
Ž .
Ž .
7
13.0 Hg-1245 85
3.8482 5 22.147 2 100
3.850 2 18.97 4
.
Ž
9508C, 2 h .
Hg-1234 10
114
Ž .
Ž .
CuO 5 , CaO 5
Earlier, we have obtained optimally doped Hg-
1234 with Tc s125 K from the initial mixture
‘HgBa2Ca3Cu4Ox’ with xF10.4, while the samples
with xs10.6 contained only overdoped Hg-1223
purity phase instead of Hg-1234 in sample a2. The
cell constants of the Hg-based superconductors and
their Tc s are in agreement with those reported earlier
6,7 . Thus, we could not obtain the overdoped sam-
ples of Hg-1234 or Hg-1245 phases using starting
composition ‘HgBa2Ca3Cu4O10.5’.
The second way, which was based on a variation
of an oxygen content, was more successful. Earlier,
we have prepared overdoped Hg-1223 from the mix-
ture with a starting composition ‘HgBa2Ca3Cu4Ox’
where x was equal to 10.4–10.6. Now we decided
to use the starting mixture ‘HgBa2Ca4Cu5Ox’ with
x lying in the range 12.6–13.0. The results of these
experiments are summarized in Table 2. The fraction
of Hg-1245 increases with the increase of x. Sam-
ples a5 and a7 contain predominantly Hg-1234 or
w x
superconductor 3 . So, it was logical to use the
initial oxide mixture with an intermediate oxygen
content: HgBa2Ca3Cu4O10.5. The results of the high
pressure synthesis are presented in Table 1. The
phase formation in the Hg–Ba–Ca–Cu–O system
was found to be very sensitive to a temperature
variation. An increasing of synthesis temperature
w
x
Ž
.
under the same pressure 2 GPa results in a prepara-
tion of the higher members of the homologous series.
Ž
For instance, Hg-1212 was obtained at 9008C sam-
.
ple a1 as a main phase, while at 11008C, sample
a4 contained mainly Hg-1245 and a minority of
Hg-1234. An increase of pressure from 2 GPa to 4
GPa did not allow the preparation of Hg-1234, and,
moreover, sample a3 contained Hg-1212 as an im-
Table 1
Characteristics of the samples prepared from the starting mixture HgBa2Ca3Cu4O10.5
N
1
2
3
4
P, GPa
T,8C
Phase composition, %
Lattice parameters
Tc , K
˚
˚
c, A
a, A
Ž
.
Ž .
3.8587 5
y
Ž .
12.687 5
y
2
2
4
2
900
Hg-1212 80
124
132
Ž .
Hg-1223 5
Ž .
Ž .
Ž .
5
CuO 5 , CaO 5 , BaCuO2
Ž
.
Ž .
3.8490 4
y
Ž .
15.779 2
y
1000
1000
1100
Hg-1223 90
128
y
Ž .
Hg-1234 5
Ž .
CaO 5
Ž
.
.
Ž .
3.8546 4
y
Ž .
15.855 4
y
Hg-1223 70
133
95
Ž
Hg-1212 10
Ž .
Ž
.
Ž .
CuO 5 , CaO 10 , BaCuO2 5
Ž
.
.
Ž .
3.8490 4
y
Ž .
22.157 5
y
Hg-1245 70
107
122
Ž
Hg-1234 20
Ž .
Ž
.
Ž
.
CaO 5 , CuO 2–5 , Hg-1223 2–5

(
)
74
K.A. Lokshin et al.rPhysica C 300 1998 71–76
reported earlier for these optimally doped com-
pounds. Taking into consideration the small values
Ž
of their a-parameter and, consequently, the in-plane
Cu–O bond distances which are very sensitive to the
.
hole concentration variation , we suggested that these
superconductors were indeed overdoped. However,
the latter observation gave us an indirect support of
this supposition only. Therefore, to clarify this prob-
lem the different subsequent heat treatments of sam-
ples a5 and 7 were performed.
3.2. Tc Õariation in Hg-1234 and Hg-1245 caused by
heat treatments
Fig. 1. Temperature dependence of the AC susceptibility of
Ž
.
as-prepared Hg-1234 solid square , and after heat treatments: in
Samples a5 and 7 were subsequently heat treated
Ž
.
Ž
.
nitrogen flow triangle , in oxygen flow circle and under 2000
bar oxygen pressure open square .
Ž
.
Ž .
under the following conditions: 1 3008C, oxygen
Ž .
Ž .
3
flow, 50 h;
2
3008C, nitrogen flow, 20 h;
Ž .
Ž
.
3008C, oxygen flow, 50 h; 4 3008C, P O2 s100
Hg-1245, respectively, and their fractions in the
samples are not smaller than 80%, while the amount
of superconducting impurities Hg-1245 and Hg-
Ž .
Ž
.
bar, 30 h; 5 3008C, P O2 s2000 bar, 10 h.
The lattice constants of the Hg-1234 and Hg-1245
phases and their Tc s are presented in Table 3, and the
Ž
.
1234, respectively is not higher than 10%. The
Ž .
Ž .
x T curves are shown in Figs. 1 and 2. Both
x T measurements of these as-prepared samples
Ž
compounds exhibit a similar behavior Tc and a-
parameter variations caused by heat treatments. The
Ž
.
Figs. 1 and 2 revealed two well-defined transitions
.
on each curve, and the main slopes on these curves
corresponded to the dominating superconductor in
the sample.
It should be noted that the as-prepared Hg-1234
and Hg-1245 phases exhibit a relatively low Tc 116
first annealing in an oxygen flow did not change the
characteristics of the compounds in comparison with
those of the as-prepared samples. However, a reduc-
tion in nitrogen flow led to a noticeable increase of
the a-parameters and a change of Tc s. It should be
mentioned that the presence of a small amount of
Ž
.
and 100 K, respectively in comparison with those
Table 3
Ž
.
Lattice constants and Tc of Hg-1234 first line and Hg-1245
Ž
.
second line after subsequent heat treatments
Tconset, K a, A
˚
˚
c, A
Treatment
As-prepareda
116
100
95
85
116
106
3.8485 6 19.001 7
Ž .
Ž .
Ž .
Ž .
3.8482 5 22.149 5
Ž . Ž .
Nitrogen flow, 3008C, 20 h
Oxygen flow, 3008C, 50 hb
3.8532 6 19.003 4
Ž . Ž .
3.8531 6 22.181 4
Ž . Ž .
3.8522 7 19.009 8
Ž . Ž .
3.8518 5 22.161 6
Ž . Ž .
Ž
.
P O2 s2000 bar, 3008C, 10 h 125
3.8506 5 19.011 5
Ž . Ž .
3.8503 5 22.157 4
111
a The heat treatment in oxygen flow at 3008C for 50 h did not
Fig. 2. Temperature dependence of the AC susceptibility of
change Tc s and lattice parameters.
b
Ž
.
Ž
.
as-prepared Hg-1245 solid square , and after heat treatments: in
The subsequent heat treatment under P O2 s100 bar at 3008C
for 30 h did not change neither Tc s nor lattice parameters of the
phases.
Ž
.
Ž
.
nitrogen flow triangle , in oxygen flow circle , and under 2000
Ž
.
bar oxygen pressure open square .

(
)
K.A. Lokshin et al.rPhysica C 300 1998 71–76
75
impurities in the studied samples did not allow the
performance of a chemical analysis, but a variation
of the a-parameter gives us an indication of an
alteration of the extra oxygen content in the Hg-based
superconductor. The subsequent oxidation in an oxy-
gen flow resulted in both a-parameter decreasing
Ž
and Tc s increasing up to 116 and 106 K for Hg-1234
.
and Hg-1245, respectively . In contrast to the lower
Ž
.
members of the series nF3 , annealing in an oxy-
gen flow does not allow the acquisition of optimal
Tc s for Hg-1234 and Hg-1245. Moreover, annealing
under 100 bar oxygen pressure did not change lattice
constants and Tc s. Therefore, a heat treatment under
Ž
.
the higher P O2 was necessary to achieve an opti-
mal hole concentration, like what was done in Refs.
Fig. 3. The dependencies of Tc s vs. a-parameter for the Hg-1212,
w
x
6,7 . Indeed, the oxidation under 2000 bar oxygen
pressure resulted in a significant increase of Tc s for
Ž
w
x.
Hg-1223 data taken from Refs. 1,3 , Hg-1234 and Hg-1245
Ž
.
present work . The dotted lines are guides to the eye.
Ž
.
both phases 125 and 110 K, respectively , while the
a-parameters decreased. These values are close to
those reported earlier for these optimally-doped com-
the similar structure arrangements of these phases.
However, to achieve the optimal hole concentration
w
x
pounds 6,7 . It should be mentioned that these treat-
ments resulted in changing Tc of impurity phases
also, and moreover, their values agree very well with
those described before for the corresponding com-
pounds.
Ž
.
in the higher members, a treatment under high P O2
is necessary. From a chemical point of view, this fact
looks quite surprising because it means that the
stabilization of the similar copper valence in the
closely related compounds require different oxygen
pressures. We think that the origin of this phe-
nomenon has a mainly structural reason. An interca-
lation of an extra oxygen in the Hg-layers causes a
4. Conclusion
Ž
Summarizing these experiments, we can conclude
that the heat treatments changed a carrier concentra-
tion in the same Hg-1234 and Hg-1245 samples from
the overdoped state for the initial as-prepared com-
static displacement of the surrounding cations 4 Hg
.
cations and 2 Ba ones toward the inserted anion due
to a chemical interaction between them. However, an
insertion of the larger amount of an extra oxygen in
Ž
Ž
pounds to the underdoped phases after treatments in
the structure of the higher members which is neces-
.
nitrogen and then in oxygen flows and, finally, to
the optimally-doped phases after a treatment under
2000 bar oxygen pressure . In Fig. 3 we plotted the
sary to achieve the same hole concentration as in the
lower members could be less energetically prof-
Ž
.
.
Ž
.
itable. It requires the higher P O2 , because some of
this extra oxygen should be placed in the mesh
formed by the 4 Hg cations already shifted towards
the oxygen in another cell.
The higher members have the smaller a-parame-
ters compared with those for the lower members.
dependencies of Tc vs. the a-parameter for the
HgBa2Cany1CunO2 nq2qd superconductors with n
s2–5. We would like to note that the only experi-
mental data for the same as-prepared and heat-treated
Hg-1234 and Hg-1245 samples were used. It can be
clearly seen that all these dependencies demonstrate
a cupola-shaped behavior, as well as one for the
Hg-1201 phase 1,2 . Therefore, we can conclude
that all HgBaCany1CunO2 nq2qd superconductors
have the same type of Tc variation caused by a
Ž
.
The successive intercalation of additional CaCuO2
slabs into the structures of Hg-based superconductors
is accompanied by compressing the in-plane parame-
ters for the next members of this series, thus provid-
ing the stronger in-plane interactions. This can be a
reason why the Tc variation for the lower members
w
x
Ž .
change of the extra oxygen content d , because of

(
)
76
K.A. Lokshin et al.rPhysica C 300 1998 71–76
is accompanied by the larger variation of the a-
laev for the technical assistance. The research is
supported by the Russian Scientific Program on Su-
Ž
.
parameter, while the Tc a-parameter dependencies
become narrower for the higher members. Moreover,
the a-parameters have practically the same values
Ž
.
perconductivity Projects—‘Poisk’ and 96045 and
Ž
.
INTAS 93-2483ext . K.A.L. and D.A.V. are grate-
ful to the ISSEP for the financial support grants
Ž
Ž
for the higher members treated under the same
.
.
conditions in contrast to the first three members
a861 and s1648 .
where the difference in their values is significant. It
can be clearly seen from Fig. 3 that the Hg-1223,
Hg-1234 and Hg-1245 phases with the same a-
parameter values have significantly different Tc s,
which decrease from the first to the final phase. On
the left side of each dotted curve, the superconduc-
References
w x
1
E.V. Antipov, S.N. Putilin, E.M. Kopnin, J.J. Capponi, C.
Chaillout, S.M. Loureiro, M. Marezio, A. Santoro, Physica C
Ž
.
235-240 1994 21.
Ž
tors are in an overdoped state with the higher than
w x
2
A. Fukuoka, A. Tokiwa-Yamamoto, M. Itoh, R. Usami, S.
Adachi, K. Tanabe, Phys. Rev. B 55 1997 6612.
.
optimal carrier concentration . Therefore, we should
Ž
.
w x
reject the simple idea of decreasing Tc due to under-
doping. We can suggest that the origin of this phe-
nomenon indeed has structural reasons, which can be
found by a further precise structural study of these
materials.
3
K.A. Lokshin, I.G. Kuzemskaya, L.F. Kulikova, E.V. An-
Ž
.
tipov, E.S. Itskevich, Physica C 279 1997 11.
w x
4
B. Morosin, E.L. Venturini, J.E. Schirber, P.P. Newcomer,
Ž
.
Physica C 226 1994 175.
w x
5
S. Adachi, A. Tokiwa-Yamamoto, M. Itoh, K. Isawa, H.
Ž
.
Yamauchi, Physica C 214 1993 313.
w x
6
S.M. Loureiro, E.V. Antipov, E.M. Kopnin, M. Brunner, J.J.
Ž
.
Capponi, M. Marezio, Physica C 257 1996 117.
w x
7
J.J. Capponi, E.M. Kopnin, S.M. Loureiro, E.V. Antipov, E.
Gautier, C. Chaillout, B. Souletie, M. Brunner, J.L. Tholence,
Acknowledgements
Ž
.
M. Marezio, Physica C 256 1996 1.
The authors would like to thank Dr. P.E. Kazin
for the magnetic measurements and Dr. N.A. Niko-
w x
8
B.A. Scott, E.Y. Suard, C.C. Tsuei, D.B. Mitzi, T.R. McGuire,
B.-H. Chen, D. Walker, Physica C 230 1994 239.
Ž
.