High conductivity in hydrothermally grown AgCuO2 single crystals verified using focused-ion-beam-deposited nanocontacts

Article abstract of DOI:10.1021/ic101420c

The silver-copper mixed oxide AgCuO₂ (also formulated as Ag ₂Cu₂O₄) possesses a peculiar electronic structure in which both Ag and Cu are partially oxidized, with the charge being delocalized among the three elements in the oxide. Accordingly, a quasi-metallic behavior should be expected for this oxide, and indeed bulk transport measurements show conductivity values that are orders of magnitude higher than for other members of this novel oxide family. The presence of silver makes thermal sintering an inadequate method to evaluate true conductivity, and thus such measurements were performed on low density pellets, giving an underestimated value for the conductivity. In the present work we present a new synthetic route for AgCuO₂ based on mild hydrothermal reactions that has yielded unprecedented large AgCuO₂ single-crystals well over 1 μm in size using temperatures as low as 88 °C. We have used a dual beam instrument to apply nanocontacts to those crystals, allowing the in situ measurement of transport properties of AgCuO₂ single crystals. The results show a linear relationship between applied current and measured voltage. The conductivity values obtained are 50 to 300 times higher than those obtained for bulk low density AgCuO₂ pellets, thus confirming the high conductivity of this oxide and therefore supporting the delocalized charge observed by spectroscopic techniques.

Full text of DOI:10.1021/ic101420c

Inorg. Chem. 2010, 49, 10977–10983 10977
DOI: 10.1021/ic101420c
High Conductivity in Hydrothermally Grown AgCuO Single Crystals Verified
2
Using Focused-Ion-Beam-Deposited Nanocontacts
,
†,^
‡,§
‡,§,||
§,||
David Mu ~n oz-Rojas,* Rosa C ꢀo rdoba, Amalio Fern ꢀa ndez-Pacheco,
Jos ꢀe Mar ´ı a De Teresa,
^
^
#
†
Guillaume Sauthier, Jordi Fraxedas, Richard I. Walton, and Nieves Casa ~n -Pastor
†
Instituto de Ciencia de Materiales de Barcelona, ICMAB-CSIC, Campus de la UAB, Bellaterra 08193, Spain,
‡
§
Instituto de Nanociencia de Arag oꢀ n, Departamento de Fı ´s ica de la Materia Condensada, Facultad de Ciencias, and
|
|
Instituto de Ciencia de Materiales de Arag oꢀ n, CSIC, Facultad de Ciencias, Universidad de Zaragoza, Zaragoza 50009,
^
Spain, Centro de Investigaci oꢀ n en Nanociencia y Nanotecnologı ´a , CIN2 (CSIC-ICN), Campus de la UAB,
Bellaterra 08193, Spain, and Department of Chemistry, University of Warwick, Coventry CV4 7AL, U.K.
#
Received July 16, 2010
The silver-copper mixed oxide AgCuO (also formulated as Ag Cu O ) possesses a peculiar electronic structure in
2
2
2 4
which both Ag and Cu are partially oxidized, with the charge being delocalized among the three elements in the oxide.
Accordingly, a quasi-metallic behavior should be expected for this oxide, and indeed bulk transport measurements show
conductivity values that are orders of magnitude higher than for other members of this novel oxide family. The presence of
silver makes thermal sintering an inadequate method to evaluate true conductivity, and thus such measurements were
performed on low density pellets, giving an underestimated value for the conductivity. In the present work we present a new
synthetic route for AgCuO based on mild hydrothermal reactions that has yielded unprecedented large AgCuO single-
2
2
crystals well over 1 μm in size using temperatures as low as 88 ꢀC. We have used a dual beam instrument to apply
nanocontacts to those crystals, allowing the in situ measurement of transport properties of AgCuO single crystals. The
results show a linear relationship between applied current and measured voltage. The conductivity values obtained are
2
50 to 300 times higher than those obtained for bulk low density AgCuO pellets, thus confirming the high conductivity of this
oxide and therefore supporting the delocalized charge observed by spectroscopic techniques.
2
3
Introduction
Ag Cu O performed chemically, that is, with ozone, or elec-
2 2 3
4,5
trochemically yielded the third member in the family: Ag2-
Cu O . The same compound was also obtained from a AgNO
3
and Cu(NO ) solution using persulfate as the oxidizing
1
999 saw the birth of the new family of silver-copper
mixed oxides, with Ag Cu O being the first natural or syn-
2
4
2
2 3
1
thetic silver-copper oxide ever known. Research in this family
was prompted by the need to find alternatives to the highest Tc
Hg-Cu based superconductors, in which the Hg could be
replaced by a less hazardous element. Silver, presenting simila-
rities in terms of chemistry and coordination with Hg, stands up
as a good candidate to replace it. Although no superconductor
phase has so far been discovered within the Ag-Cu oxide
family, the new compounds found to date have shown very
interesting structures and properties so as to deserve a deeper
study in their own right. Thus, after the discovery of Ag Cu O ,
which was synthesized by a simple coprecipitation reaction,
other members were synthesized. Coprecipitation also yielded a
solid solution of formula Ag Pb Cu O , while oxidation of
3 2
6
agent. By a completely different approach, the azide/nitrate
route gave rise to Rb Ag Cu O , a further member in the
3
0.5
0.5 2
7
family, although air and moisture sensitive in this case. Finally,
a low-temperature hydrothermal reaction yielded the last
family member so far described: AgCu0.5Mn0.5
O , a mixed
2
8
B-site delafossite with significant magnetic properties.
(
3) Mu n~ oz-Rojas, D.; Fraxedas, J.; G oꢀ mez-Romero, P.; Casa n~ -Pastor,
2
2
3
1
N. J. Solid State Chem. 2005, 178, 295–305.
(4) Mu n~ oz-Rojas, D.; Or oꢀ , J.; G oꢀ mez-Romero, P.; Fraxedas, J.; Casa n~ -
Pastor, N. Electrochem. Commun. 2002, 4, 684–689.
(5) Mu
n~ oz-Rojas, D.; Tejada-Rosales, E. M.; G oꢀ mez-Romero, P.;
Casa n~ -Pastor, N. Bol. Soc. Esp. Ceram. Vidrio 2002, 41, 55–58.
6) (a) Curda, J.; Klein, W.; Jansen, M. J. Solid State Chem. 2001, 162,
2
5
2-x
x
6
(
*To whom correspondence should be addressed. Present address: Department
220–224. (b) Curda, J.; Klein, W.; Liu, H.; Jansen, M. J. Alloys Compd. 2002,
of Materials Science and Metallurgy, University of Cambridge, Pembroke Street,
Cambridge CB2 3QZ, U.K. Phone: þ441233334374. Fax: þ441223334375.
E-mail: davidmunozrojas@gmail.com.
338, 99–103.
(7) Sofin, M.; Peters, E.-M.; Jansen, M. Z. Kristallogr. NCS 2003, 218,
379–380.
(
1) G oꢀ mez-Romero, P.; Tejada-Rosales, E. M.; Palacı
Chem., Int. Ed. 1999, 38, 524–525.
2) Tejada-Rosales, E. M.; Or oꢀ -Sol ꢀe , J.; G oꢀ mez-Romero, P. J. Solid State
´
n, M. R. Angew.
(8) Mu n~ oz-Rojas, D.; Subı
´
as, G.; Or oꢀ -Sol ꢀe , J.; Fraxedas, J.; Martı
´
nez, B.;
Casas-Cabanas, M.; Canales-V ꢀa zquez, J.; Gonz ꢀa lez-Calbet, J.; Garcı
Gonz ꢀa lez, E.; Walton, R. I.; Casa n~ -Pastor, N. J. Solid State Chem. 2006,
179, 3883–3892.
´
a-
(
Chem. 2002, 163, 151–157.
r 2010 American Chemical Society
Published on Web 11/04/2010
pubs.acs.org/IC

1
0978 Inorganic Chemistry, Vol. 49, No. 23, 2010
Mu ~n oz-Rojas et al.
Among these novel oxides, AgCuO presents very inter-
2
esting crystallographic and electronic structures. This com-
pound was initially reported as Ag Cu O , as it was obtained
2
2 4
from the parent phase Ag Cu O (which implies an oxidation
2
2 3
of þ1 per unit formula of the latter). Although initial studies
claimed, on the one hand, that copper was being oxidized from
2
þ
3þ
Cu to Cu basing it on structural studies under different
6
conditions, simultaneous X-ray photoelectron spectroscopy
(
XPS) analyses seemed to support a mixed valence for silver
1þ
3þ
2þ
9
(i.e., Ag Ag Cu O ,). Later, a deeper X-ray Absorption
2
4
Spectroscopy (XAS) study showed that both metals are indeed
partly oxidized and that the charge is actually delocalized
1
0
among all the elements, including oxygen. Thus, the XPS
data was interpreted as an opening of the Ag 4d band, and a
more detailed study of the O 1s signal showed also a true
9
,10
oxidation of oxygen atoms.
scribe this oxide is as Ag
Therefore, the best way to de-
Cu
(
1þx)þ
(2þy)þ -(2-z)
O 2, where the
values of x and y depend on the synthetic method used, and in
1
0
which all elements share an intermediate oxidation state. More
recent XAS studies performed by other groups using highly
crystalline AgCuO samples have confirmed the charge de-
2
11
localization and the oxidation of silver in the compound.
On the crystallographic side, AgCuO is isostructural with
2
Figure 1. Crystallographic structure of AgCuO ; red, oxygen; blue, cop-
2
per; gray, silver. All atoms are drawn with the same radius for the sake of
4
crednerite (CuMnO ), having the space group C2/m. But
2
2
clarity. While structurally equivalent to crednerite (CuMnO ), the coordina-
2
tion of the metals in AgCuO and, therefore, the electronic structure and
oxidation states are completely different. As shown at the bottom, Ag is
while copper atoms are linearly coordinated by 2 oxygen atoms
1þ
(1þx)þ
in crednerite (O-Cu -O), the oxidized Ag
atoms in
AgCuO are coordinated by 6 oxygen atoms, 2 axial and four
2
coordinated by six oxygen atoms forming a deformed octahedron (right).
This octahedron is nevertheless more close to being regular than the one
formed by the six closest oxygen atoms around Cu cations (left), as illustrated
by the superimposed spheres; and indeed cooper has a 4 planar square co-
ordination only. The small distance of the 4 equatorial oxygen atoms around
equatorial forming an elongated octahedron, as deduced from
1
0
extended X-ray absorption fine structure (EXAFS), and
accordingly to the higher oxidation state of silver (see Figure 1).
The peculiar electronic structure described above implies
3þ
10
Ag (as compared for instance with Ag coordination in AgO) is what
accounts for the higher oxidation state observed by spectroscopic techniques.
that AgCuO should be a quasi-metal in which the delocalized
2
electrons would play the equivalent role of a partly filled
10
conduction band. A first hint on this high conductivity comes
In addition to these results concerning bulk measurements,
direct measurements on single crystals would be of most
interest to better characterize the transport properties of
from contrasting the darker color of AgCuO , as compared
2
with Ag Cu O or AgCu Mn O , in which the metals have a
2
2
3
0.5
0.5 2
fixed oxidation state and no charge delocalization takes place.
AgCuO , by eliminating any contribution from grain bound-
1þ
3þ
2
A second hint of the absence of localized Ag , Cu states, is
the absence of diamagnetism expected for those states, and the
observation of a small Pauli paramagnetism typical of many
aries. The presence of silver in this compound makes it very
difficult to grow a single macroscopic crystal to allow a stan-
dard electrical measurement. However, nanotechnology tools
can be brought into action to carry out electrical measure-
ments at the micro/nanoscale. In this sense, Focused-Ion-Beam
1
2
metals, checked for all pure samples in several measurement
holders. The actual evaluation of the transport properties in
these phases is not simple since the presence of silver does not
allow any sintering and thus, dense pellets are difficult to
obtain. In spite of such difficulties, bulk transport measure-
(FIB) or Focused-Electron-Beam (FEB) nanolithography can
be used to pattern metallic nanocontacts by introducing metal-
lorganic precursors in the process chamber, which become
1
3
ments have been reported for Ag Cu O , AgCu Mn O ,
and AgCuO2,
15
2
2
3
0.5
0.5 2
dissociated by the ion or electron beam, the resolution of this
12,14
the latter presenting conductivity values
process reaching down to a few nanometers. These techniques
have been applied to the transport characterization of indivi-
orders of magnitude higher than the former two, as expected
from their respective electronic structures, and thereby showing
16
17
dual nanowires and nanoparticles. In addition, the use of in
situ transport measurements inside a dual beam (FIB/FEB) is
very useful to quickly characterize the conductive properties of
the predicted enhanced conductivity of AgCuO as a result of
2
charge delocalization.
18
nanostructures.
(
9) Mu n~ oz-Rojas, D.; Or oꢀ , J.; Fraxedas, J.; G oꢀ mez-Romero, P.; Casa n~ -
Pastor, N. Cryst. Eng. 2002, 5, 459–467.
10) Mu n~ oz-Rojas, D.; Subıas, G.; Fraxedas, J.; G oꢀ mez-Romero, P.;
(
´
(15) Utke, I.; Hoffmann, P.; Melngailis, J. J. Vac. Sci. Technol. B 2008, 26,
Casa n~ -Pastor, N. J. Phys. Chem. B 2005, 109, 6193–6203.
1197–1276.
(
(
11) Dr. G. Waterhouse. Private Communication.
(16) Cronin, S. B.; Lin, Y. M.; Rabin, O.; Black, M. R.; Ying, J. Y.;
Dresselhaus, M. S.; Gai, P. L.; Minet, J. P.; Issi, J. P. Nanotechnology 2002,
13, 653.
12) Mu n~ oz-Rojas, D. New Ag-Cu Oxides by Electrochemical Inter-
calation and other soft methods. Ph.D. Thesis, Autonomous University of
Barcelona (UAB), Barcelona, Spain, May 2004.
(17) Martı
Appl. Phys. Lett. 2009, 94, 062507.
(18) (a) De Teresa, J. M.; C oꢀ rdoba, R.; Fern ꢀa ndez-Pacheco, A.; Montero,
O.; Strichovanec, P.; Ibarra, M. R. J. Nanomaterials 2009, 936863. (b)
Fern ꢀa ndez-Pacheco, A.; De Teresa, J. M.; C ꢀo rdoba, R.; Ibarra, M. R. Phys.
Rev. B 2009, 79, 174204.
nez-Boubeta, C.; Balcells, Ll.; Monty, C.; Ordejon, P.; Martınez, B.
´ ´
(
13) Tejada-Rosales, E. M.; Rodrı
´
guez-Carvajal, J.; Casa n~ -Pastor, N.;
ꢀ
Alemany, P.; Ruiz, E.; El-Fallah, M. S.; Alvarez, S.; G oꢀ mez-Romero, P.
Inorg. Chem. 2002, 41, 6604–6613.
(
14) Sauvage, F.; Mu n~ oz-Rojas, D.; Poeppelmeier, K. R.; Casa n~ -Pastor,
N. J. Solid State Chem. 2009, 182, 374–380.

Article
Inorganic Chemistry, Vol. 49, No. 23, 2010 10979
Table 1. Summary of Different Reactions Assayed for the Hydrothermal Synthesis of AgCuO
2
reaction no.
reactants/g
H
2
O/mL
T/ꢀC
t
oxidizer/g
products
O, CuO
1
2
3
4
5
6
7
8
9
0.2 AgO, 0.2 CuSO
0.2 AgO, 0.2 CuSO
0.13 AgO, 0.26 CuSO
0.13 AgO, 0.26 CuSO
0.13 AgO, 0.18 CuSO
4
5H
5H
2
O, 0.1 NaOH
O, 0.1 NaOH
12
12
12
12
12
36
36
30
36
240
160
120
r.t.
120
100
90
10 h
10 h
10 h
17 h
17 h
17 h
20 h
17 d
15.5 d
Ag
Ag
2
3
3
4
2
2
2
O, CuO, (AgCuO )
4
4
4
3
3
5H
5H
2
O, 0.3 KOH
O, 0.3 KOH
AgCuO
AgCuO
AgCuO
AgCuO
AgCuO
AgCuO
AgCuO
2
2
2
2
2
2
2
, Ag
2
O
2
, 1 KOH
0.3 K
2
S
2
O
8
0.2 AgNO
0.4 Ag Cu
0.13AgO, 0.18 CuSO
0.1 Ag Cu , 4 KOH
3
, 0.28 Cu(NO
, 6 KOH
, 8 KOH
3
)
2
3 3H2O, 2 KOH
1 K
2 K
2
S
2
O
O
8
, Ag
2
2
O, CuO
O, CuO
2
2
O
3
2
S
2
8
4
88
88
1.5 K
2 K
2
S
2
O
8
, Ag
2
2
O
3
2
S
2
O
8
, (CuO)
The largest particles of AgCuO , the ones obtained by elec-
electrical measurements, a drop of the microparticle-containing
suspension is first deposited onto a thermally oxidized silicon wafer
2
trochemical oxidation of Ag Cu O , reach a maximum of only
2
2 3
4
(
∼200 nm of SiO
2
on top of the Si substrate) where Al microelec-
∼
500 nm width, and several nanometers thickness. Despite the
trodes have been pre-patterned in a well-defined geometry to enable
transport measurements by means of conventional optical litho-
graphy. The final mask consist of large metallic pads (400 μm ꢀ 400
μm), which are connected via 4 μm wide electrodes. The target
microparticle was located by means of the electron microscope, and
Pt or W nanocontacts were patterned on the microparticle, con-
necting it to the Al-microelectrodes. The experimental setup to
measure in situ the resistance by four-point method without break-
ing vacuum has been previously used in our group in Pt-based
fine resolution of the current dual-beam technology, these
dimensions are still quite prohibitive and no reliable results
could be obtained. To overcome this limitation we have deve-
loped a new synthetic route for AgCuO based in low tem-
2
perature hydrothermal reactions that have allowed us to
synthesize crystals several micrometers in size for the first time.
Hydrothermal reactions have proven very effective in the
synthesis of complex oxides at relatively low temperatures in
18
1
9
nanodeposits and Co-based nanodeposits.
the past. In this work we describe the hydrothermal synthesis
of micrometric AgCuO single crystals and the study of their
transport properties by means of a Dual-Beam microscope.
2
Results and Discussion
Hydrothermal Synthesis of AgCuO . The hydrothermal
2
synthesis of AgCuO , was performed from starting products
2
Experimental Section
matching the formal oxidation state in the final product, as it
would be the case in an ordinary high temperature solid state
reaction. Following such approach, solid AgO (in which half
of the silver atoms have oxidation state of þ3, according to
Materials and Characterization. Reagent grade AgO (Aldrich),
AgNO (Panreac), CuSO (PROBUS), CuSO 5H O (Fluka), Cu-
3
4
4
3
2
(
NO3)2 3
3H O (Merk), Sodium Persulfate (Na S O , Aldrich),
2 2 2 8
2 2 3
KOH, and NaOH were purchased and used as received. Ag Cu O
was synthesized as previously reported.
The hydrothermal synthesis of AgCuO , was performed in either
1
þ
3þ
20
the formula Ag Ag O ) was suspended in an Cu(SO)
2
4
2
solution, according to a 1:1 metal molar ratio. NaOH or
KOH was then added to act as mineralizer, and the suspen-
sion was stirred for 5 min before the hydrothermal treatment
at different temperatures. (Hydrothermal heterogeneous
redox reactions, that is, using a solid insoluble reactant
instead of a soluble salt, have proven a very effective path to
a 23 mL Parr acid digestion bomb or a screw capped 60 mL plastic
bottle. The different conditions and precursors tried are summa-
rized in Table 1.
X-ray powder diffraction data were collected using a Rigaku
X-ray powder diffractometer RotaflexRu-200B, with Cu KR radia-
tion. Scanning electron microscopy (SEM) was performed with a
Hitachi S-570. XPS measurements were performed at room tem-
perature with a SPECS EA10P hemispherical analyzer using non-
monochromatic Mg KR radiation (1253.6 eV) as excitation source
8
the formation of novel oxide phases and even remarkable
2
1
hybrid nanostructures .) Table 1 shows a summary of dif-
ferent reactions performed and the respective products
obtained. The corresponding X-ray diffraction (XRD) patterns
are presented in Figure 2.
-
9
in a base pressure of about 10 mbar, on pressed pellets, avoiding
using silver epoxy.
In Situ Measurements of the Transport Properties. For the
present experiments, a commercial “dual beam” instrument (Nova
As shown in Table 1, the product of the reaction is very
sensitive to the temperature used and, thus, only Ag O and
2
200 NanoLab from FEI) was used. It integrates a 30 kV field-
CuO were obtained when the reaction was carried at 240 ꢀC
emission electron column and a Ga-based 30 kV ion column placed
forming 52ꢀ with coincidence point 5 mm away from the electron-
column pole. The contacts on the microparticles were performed by
means of Pt or W deposition using an automatized gas-injection
(reaction 1). On the contrary, when the temperature was
lowered to 160 ꢀC some tiny peaks corresponding to
AgCuO could be observed along with those of Ag O and
2
2
system (GIS), with (CH
platinum) and W(CO) (tungsten hexacarbonyl) as the precursors,
6
3
)
3
Pt(CpCH ) (trimethylcyclopentadienyl-
3
CuO (reaction 2, see small black arrows in Figure 2). In
reactions 1 and 2 in Table 1, an excess of AgO was used so that
the possible formation of AgCuO could be achieved in spite
respectively. The GIS tip was positioned about 150 μm away from
the region of interest in the z direction and about 50 μm away in the
x/y direction. The GIS was heated to about 40 ꢀC for operation and
a 10-min preheating period was performed before deposition. In the
case of Pt, contacts made with focused-ion-beam-induced deposi-
tion results in low contact resistance compared to focused-electron-
beam-induced deposition and were thus preferred for the present
2
of the expected high reduction rate of AgO to give Ag O. At
2
1
20 ꢀC the main product was already AgCuO , but the
2
presence of other products, especially Ag O (as in reaction
3 for instance), was very hard to avoid. Finally, and very
interestingly, an equivalent reaction carried out at room
2
18a
study. W contacts were also deposited by FIB. To perform the
(20) Fischer, P.; Jansen, M. J. Less-Common Met. 1988, 137, 123–131.
(
19) (a) Sheets, W. C.; Mugnier, E.; Barnabe, A.; Marks, T. J.; Poeppelmeier,
(21) (a) Mu n~ oz-Rojas, D.; Or oꢀ -Sol ꢀe , J.; Ayyad, O.; G oꢀ mez-Romero, P.
Small 2008, 4, 1301–1306. (b) Mu ~n oz-Rojas, D.; Or ꢀo -Sol ꢀe , J.; G ꢀo mez-Romero, P.
J. Phys. Chem. C 2008, 112, 20312–20318. (c) Mu ~n oz-Rojas, D.; Or ꢀo -Sol ꢀe , J.;
G ꢀo mez-Romero, P. Chem. Commun. 2009, 5913–5915. (d) Mu ~n oz-Rojas, D.; Or ꢀo -
Sol ꢀe , J.; Ayyad, O.; G ꢀo mez-Romero, P. J. Mater. Chem. DOI:10.1039/C0JM01449D.
K. R. Chem. Mater. 2006, 18, 7–20. (b) Sheets, W. C.; Stampler, E. S.; Bertoni,
M. I.; Sasaki, M.; Marks, T. J.; Mason, T. O.; Poeppelmeier, K. R. Inorg. Chem.
008, 47, 2696–2705. (c) Modeshia D. R.; Walton, R. I. Chem. Soc. Rev. 2010,
9, 4303-4325.
2
3

1
0980 Inorganic Chemistry, Vol. 49, No. 23, 2010
Mu ~n oz-Rojas et al.
Figure 2. XRD patterns of the products obtained for the different hydro-
thermal reactions summarized in Table 1. The red and black dotted lines point
to the main Ag O and CuO peaks that can be observed in some of the samples.
2
temperature for 17 h (see Table 1) yielded AgCuO as the
2
sole product. Solid AgO is therefore capable of oxidizing the
2þ
Cu ions in the suspension to form the mixed oxide even at
room temperature and ambient pressure. Under hydrother-
mal conditions the mechanism is equivalent to other meth-
ods, with AgO being the oxidizing agent to finally form
AgCuO which contains both oxidized silver and copper.
3
2
The mixed oxide is less stable as temperature increases and
tends to decompose to Ag O and CuO and/or other copper
2
species in solution (soluble complexes formed at high pH) in
those conditions. Higher temperatures also accelerate the
reduction of AgO to Ag O before AgCuO can even be
Figure 3. SEM images of (a) AgCuO
oxidation of Ag Cu
obtained using AgO and CuSO
2
obtained by electrochemical
2
2
O
3
; and via mild hydrothermal reactions: (b) product
as starting products (reaction 3 in Table 1);
Cu and persulphate as starting products
reaction 7 in Table 1); (d) product obtained using AgO and CuSO and a
2
2
4
formed, which would explain the absence of AgCuO in
2
(
c) product obtained using Ag
2
2 3
O
reaction 1 (at 240 ꢀC) and the low yield in reaction 2 (at
(
4
1
60 ꢀC). On the contrary, it is interesting to note that metallic
temperature of 88 ꢀC during 17 days (reaction 8 in Table 1).
silver is never obtained, not even at 240 ꢀC, the reduction of
AgO proceeding to form AgCuO (in which the silver ions
and a better mixing before the hydrothermal treatment
translated to higher yields. Having persulfate as oxidiz-
ing agent also allowed the use of soluble AgNO as silver
source (similarly to the method reported in ref 6), instead of
solid AgO, though in this case CuO and Ag O were hard to
avoid (see reaction 6 for instance). In a slightly different
approach, and to avoid diffusion (kinetic) limitations that
could be facilitating the precipitation of Ag O or CuO before
2
have an oxidation state of 1 þ x) or Ag O for the higher
2
temperatures. This is quite different to what happens at
ambient pressure, conditions in which silver is readily re-
duced to metallic Ag as the temperature is increased (around
3
2
2
2
1
00 ꢀC for AgO and 200 ꢀC for Ag O).
2
Figure 3 shows SEM images for different AgCuO
2
samples. In Figure 3a, crystals obtained by electrochemi-
cal oxidation of Ag Cu O can be observed. Figure 3b
shows some of the crystals obtained in reaction 3 (Table 1),
2
the mixed oxide could form, a suspension of Ag Cu O was
2
used, as opposed to using different precursors for Ag and Cu
2 3
2
2
3
(
see reaction 7 in Table 1). This again yielded pure AgCuO2,
in which AgCuO was the main product. Although the
2
sample presents a high dispersion in particle size, particles
of up to about 3 μm could be found, as opposed to what is
but even though the formation of Ag OorCuOwaseasierto
2
avoid at temperatures below 100 ꢀC in this case, the increase
in temperature translated to more byproduct being ob-
tained. Figure 3c shows an image of the particles obtained
observed for electrochemically synthesized AgCuO , which
2
presents much smaller particles.
In an effort to minimize the formation of Ag O and
for reaction 7 in Table 1 (using Ag Cu O and persulfate)
2
2 3
2
where crystals over 2 μm in size can be observed despite using
a temperature of only 90 ꢀC for the synthesis.
CuO at 120 ꢀC and higher temperatures, and favor the
formation of pure AgCuO with even higher particle
2
As it was stated in the introduction, AgCuO presents
a delocalized charge distribution with silver and copper
sizes, an additional oxidizing agent, namely, sodium
persulfate, Na S O , was added to the initial suspen-
sion. Under these conditions, the formation of AgCuO2
2
2
2
8
1
0
having oxidation states of 1 þ x and 2 þ y, respectively.
The values of x and y are different for different synthetic
methods and can also vary under strong radiation. Because
could be achieved at 120 ꢀC without any Ag O/CuO
2
9
impurities being obtained (reaction 5). Notwithstanding,
even when using persulfate, great care had to be taken to
avoid the formation of byproduct. From the output of
many other reactions carried out (not shown) we could
conclude that higher pHs, higher amounts of persulfate,
of this particular electronic configuration, AgCuO could be
2
considered a quasi-metal, and indeed bulk conductivity
measured in pressed pellets is several orders of magnitude
larger than that of AgCu Mn O or Ag Cu O . To
1
2
4
13
0.5
0.5
2
2 3
probe the valence of silver cations in the samples obtained
using the new hydrothermal approach, XPS was performed.
XPS has proven a very sensitive and effective tool to probe
(
22) CRC Handbook of Chemistry and Physics, 85th ed.; CRC Press:
Boca Raton, FL, 2004.

Article
Inorganic Chemistry, Vol. 49, No. 23, 2010 10981
As explained in the Experimental Section, the sample was
then scanned to find big isolated particles to be contacted.
Figure 5 shows an example of a particle of AgCuO from
2
reaction 9 contacted using Pt as described in the Experimental
Section. The particle has approximate dimensions of 1 ꢀ 1 ꢀ
3
μm, and does not show any observable damage caused by
the ion beam used to make the contacts. The particle seems to
be the result of twinning (as is the case for many of the big
particles obtained using the hydrothermal method, shown in
Figure 3), yet it can be considered a single crystal since both
twin crystals are clearly welded and aligned at the bottom face
(see insets in Figure 5). The EDX also shows that the particle
contains both Ag and Cu (1:1 ratio), thus confirming that it is
a crystal of AgCuO and not a Ag O or CuO impurity.
2
2
Contacting the particles was indeed highly challenging
because of their height (1 μ or over) and their low aspect
ratio. Most in situ transport measurements reported to date
have been performed by applying nanocontacts to nanowires
or nanotubes, which have a maximum diameter of some tens
of a nanometer. One-dimensional (1D) nanostructures are
easily contacted by laying the nanocontact over the wire/tube
and without risk of shortcircuit between the different con-
tacts because of their high aspect ratio. In this work, we have
successfully applied neat nanocontacts on 1 μm tall particles
with an aspect ratio as low as 3:1.
Once the particle was contacted, in situ four-probe current
versus voltage measurements were performed. To check the
quality of the contacts, initial 2-probe measurements were
made, using alternative couples of nanocontacts. In both
cases, a linear dependence between applied current and
voltage was obtained. Linear fitting of the lines obtained
for the different set of nanocontacts gave comparable resis-
tance values of 250.6 and 235.0 kΩ, respectively, thus
indicating the good quality of the nanocontacts on the
Figure 4. 3d region of the XPS spectra of Ag
synthesized under hydrothermal conditions (reaction 9 in Table 1). * The
XPS spectra of electrochemically and chemically synthesized AgCuO
2 2 3 2
Cu O and AgCuO
2
samples (from reference 10) are also included for comparison.
the electronic state of Ag in AgCuO , showing a clear shift to
2
lower binding energies of the Ag 3d band, as compared to
that of Ag Cu O or Ag O. In some cases the spectrum
2
2
3
2
presents two clearly separate peaks. Figure 4 shows the XPS
spectra of the Ag 3d region for Ag Cu O and for the
AgCuO , obtained in reaction 7. XPS spectra of AgCuO
2
2 3
2
2
prepared by electrochemical and chemical oxidation are also
included for comparison (the spectra are taken from
reference 10). A clear shift of both the Ag 3d3/2 and Ag
3d5/2 bands to lower energies for AgCuO , as compared
with Ag Cu O , is clearly observed being equivalent to that
2
2
2 3
10
observed for AgCuO samples prepared by other methods,
2
as shown in Figure 4; this represents unequivocal evidence
that silver has an oxidation state higher than 1þ in hydro-
AgCuO particle. 4-Probe in situ measurements were then
2
thermally synthesized AgCuO , as was indeed expected. The
2
made using the four nanocontacts simultaneously. To avoid
any risk of short-circuit between the nanocontacts, the two
intensity probes were connected to nanocontacts placed in
opposite corners of the crystal. The measurements again
showed a linear dependence between applied current and
voltage, as shown in Figure 6, which corresponds to the
contacted particle shown in Figure 5. Linear fitting of the
obtained points gave a resistance value of 1.75 kΩ. This value
is much smaller than the one obtained for 2-probe measure-
ments, and consistent with the high conductivity expected for
XPS spectrum for the O 1s region is also similar to the one
obtained for other AgCuO samples (not shown). Thus, it
10
2
can be concluded that the product obtained using the
hydrothermal approach has a similar electronic structure
to those obtained by other means, and thereby, high con-
ductivity values should be expected. On the other hand, the
influence of the synthetic method in the electronic struc-
ture can still be observed in the different shape of the XPS
spectrum for the hydrothermal AgCuO , as compared
2
to the ones obtained for AgCuO synthesized by other
2
methods.
Transport Properties of AgCuO Single Crystals. For
2
AgCuO . Taking into account the particle shape, and the
2
10
resistance obtained from the four-probe measurement, we
-
1
calculated a conductivity value of ∼17 S cm for AgCuO2.
Similar results were obtained for another particle contacted
with W instead of Pt (see Figure 6). In-situ four-probe current
versus voltage measurements again showed a linear depen-
dence when using W contacts. In this case, the resistance
values obtained for 2-probe measurements were 51.0 and
49.2 kΩ, while the 4-probe measurement showed a resis-
tance value of 0.2 kΩ. Taking into account the shape of the
particle, the conductivity value obtained was in this case ∼98
the in situ transport measurements, the reaction conditions
were adjusted to favor the formation of big particles. In
particular, mild hydrothermal reactions at 88 ꢀC were carried
during long periods, namely, 15.5 days using Ag Cu O and
2
2 3
persulphate (reaction 8 in Table 1). The XRD pattern of this
sample clearly indicates a significant increase in average
particle size, as deduced from the much better resolved peaks,
as compared with the electrochemical sample (see inset in
Figure 1). SEM confirmed the larger average particle size,
which is caused by the long reaction times used in this case.
Figure 3d shows a single crystal of over 5 μm in size obtained
for reaction 8, along with smaller particles. The mild tem-
perature used favored the minimal decomposition of Ag-
-
1
S cm . The difference in conductivity obtained for different
nanocontacts can be due to either the difference in conduc-
tivity between the nanocontacts themselves or the error
associated with the geometrical effects when measuring
2
3
resistance in non-symmetric and small objects. FIB-W
CuO , and only small amounts of CuO were present after
2
such long reaction times (see XRD pattern in Figure 2).
(23) der Pauw, V. Philips Tech. Rev. 1958, 20, 220.

1
0982 Inorganic Chemistry, Vol. 49, No. 23, 2010
Mu ~n oz-Rojas et al.
Figure 5. SEM image of a AgCuO
2
particle from reaction 8 contacted using ion assisted deposition of Pt. The insets show a close-up of the particle from
different perspectives, as well as an EDX spectrum of the contacted particle showing the corresponding peaks for Ag and Cu.
silver that occurs upon sintering. These measurements are
then expected to be affected by an important component
because of grain boundary resistance and, therefore, give
only an underestimated value of the actual conductivity
of AgCuO . The in situ measurements presented here thus
2
confirm the high conductivity of AgCuO , and yield a
2
closer estimation of the actual value. Although not a
direct evidence, the results presented here are a necessary
consequence of the charge delocalization among Ag, Cu
and O observed using spectroscopic techniques (XPS,
1
0
XANES, and EXAFS).
Conclusions
A new synthetic route for the synthesis of micrometer-sized
AgCuO twinned and single crystals has been developed
2
Figure 6. V versus I plots obtained for contacted particles using FIB-Pt
and FIB-W nanocontacts. The inset shows the 4 probes contacting the
patterned Al microelectrodes in the wafer, to which the metallic nano-
using mild hydrothermal reactions. Although the synthesis
is possible using several precursors, the best results are
obtained when Ag Cu O and persulfate are used at
wires, Pt in this case, are connected. The nanocontacted AgCuO
is in this case the one shown in Figure 5.
2
particle
2
2
3
temperatures below 100 ꢀC and for long reaction times.
Using this novel approach, single crystals of several micro-
meters have been obtained for the first time, using tempera-
tures as low as 88 ꢀC. The big crystals obtained have been
used to get a better estimation of the conductivity of
nanocontacts are indeed much more conducting than FIB-Pt
2
4
ones, as can be appreciated in the resistance values obtained
for 2-probe measurements.
The conductivity values obtained in situ for AgCuO2
single crystals were thus 50 to 300 times higher than the
one obtained for bulk measurements using nonsintered
pellets. Such a difference is not surprising given the fact
that the bulk measurements were done with low-density,
non-sintered pellets to avoid the formation of metallic
AgCuO . In situ 4 probe measurements of nanocontacted
2
AgCuO particles show a linear dependence between volt-
2
age and current, and the resistance values calculated are
between 50 and 300 times smaller than those obtained for
bulk measurements using low density pellets. Therefore, the
present results confirm the high conductivity expected for
1
4
AgCuO and, furthermore, they add to the consistency of
2
the set of previously reported observations that evidence
that the phase has an electronic structure similar to that
(
24) Marcano, N.; Sangiao, S.; Plaza, M.; P ꢀe rez, L.; Fern ꢀa ndez Pacheco,
A.; C oꢀ rdoba, R.; S ꢀa nchez, M. C.; Morell oꢀ n, L.; Ibarra, M. R.; De Teresa,
J. M. Appl. Phys. Lett. 2010, 96, 082110.
found in the field of high T superconducting cuprates; and
c

Article
Inorganic Chemistry, Vol. 49, No. 23, 2010 10983
that by itself constitutes an unique example from the more
fundamental point of view in the field of argentates, cupr-
ates, and mixed valence oxides.
and Science through Grants MAT2005-07683-C02-01
and MAT2008-06643-C02-01, and for a FPI fellowship
for D.M-R. D.M-R. is also grateful to the CSIC and the
European Social Fund for funding through the I3P
program. Dr. F. Sauvage is acknowledged for fruitful
discussions.
Acknowledgment. The authors thank financing from
Nanoaracat and from the Spanish Ministry of Education