Silver delafossite oxides

Article abstract of DOI:10.1021/ic702197h

A single-step, low-temperature (<210°C) and -pressure (<20 atm) hydrothermal method has been developed to synthesize a series of silver delafossites, AgBO₂ (B = Al, Ga, Sc, and In). Experimental and computational studies were performed to understand the optical and electric properties of these silver delafossites, including the first in-depth study of AgAlO₂ and AgScO₂. Their properties were examined as a function of the trivalent cation radius and compared to those of copper delafossites to elucidate the role of both the A- and B-site cations. While optical band gaps for silver delafossites were larger and visible light absorption was lower than values previously reported for polycrystalline powder samples of copper delafossites, the conductivities of silver delafossites are similar or lower. Electronic structure calculations indicate that these properties are due to the scarcity of silver 4d states just below the valence band maximum.

Full text of DOI:10.1021/ic702197h

Inorg. Chem. 2008, 47, 2696-2705
Silver Delafossite Oxides
William C. Sheets,† Evan S. Stampler,† Mariana I. Bertoni,‡ Makoto Sasaki,§ Tobin J. Marks,†,‡
Thomas O. Mason,‡ and Kenneth R. Poeppelmeier*,†
Departments of Chemistry and of Materials Science and Engineering, Northwestern UniVersity,
EVanston, Illinois 60208, and, Department of Materials Science and Engineering, Muroran
Institute of Technology, 27-1 Mizumoto, Muroran, Hokkaido 050-8585, Japan
Received November 7, 2007
A single-step, low-temperature (<210 °C) and -pressure (<20 atm) hydrothermal method has been developed to
synthesize a series of silver delafossites, AgBO2 (B ) Al, Ga, Sc, and In). Experimental and computational studies
were performed to understand the optical and electric properties of these silver delafossites, including the first
in-depth study of AgAlO2 and AgScO2. Their properties were examined as a function of the trivalent cation radius
and compared to those of copper delafossites to elucidate the role of both the A- and B-site cations. While optical
band gaps for silver delafossites were larger and visible light absorption was lower than values previously reported
for polycrystalline powder samples of copper delafossites, the conductivities of silver delafossites are similar or
lower. Electronic structure calculations indicate that these properties are due to the scarcity of silver 4d states just
below the valence band maximum.
Introduction
conductivity of 0.95 S cm-1.8 Since then, delafossites,
especially the copper and silver members, have received
considerable attention owing to their application as transpar-
ent conducting oxides.9,10 Particular emphasis has been
placed on optimizing their optical and electrical properties
through proper selection of the parent delafossite, aliovalent
dopant, and/or control of the oxygen stoichiometry. Because
of their compositional versatility, delafossites also have been
studied intensely as catalysts,11–14 luminescent materials,15,16
batteries,17 and thermoelectrics.18–21 Recent photocatalytic
applications have exploited the sensitivity of delafossites to
The physical properties of numerous delafossites were first
reported as part of a large study on the synthesis, structure,
and electrical transport properties of delafossites by Rogers,
Shannon, and Prewitt.1–3 Later, Benko and Koffyberg
reported the electrical and optical properties of certain copper
delafossites in a series of papers.4–7 In 1997, interest in
delafossites was rekindled when a seminal paper by Kawazoe
and co-workers revealed that a 500 nm thick film of CuAlO2
transmitted 70% of visible light and exhibited p-type
(8) Kawazoe, H.; Yasukawa, M.; Hyodo, H.; Kurita, M.; Yanagi, H.;
Hosono, H. Nature (London) 1997, 389(6654), 939–942.
(9) Ginley, D. S.; Bright, C. MRS Bull. 2000, 25(8), 15–21.
(10) Kawazoe, H.; Yanagi, H.; Ueda, K.; Hosono, H. MRS Bull. 2000, 25(8),
28–36.
* To whom correspondence should be addressed. E-mail: krp@
northwestern.edu.
†
Department of Chemistry.
Department of Materials Science and Engineering.
Muroran Institute of Technology.
‡
(11) Carcia, P. F.; Shannon, R. D.; Bierstedt, P. E.; Flippen, R. B. J.
Electrochem. Soc. 1980, 127(9), 1974–1978.
§
(1) Shannon, R. D.; Rogers, D. B.; Prewitt, C. T. Inorg. Chem. 1971,
(12) Monnier, J. R.; Hanrahan, M. J.; Apai, G. J. Catal. 1985, 92(1), 119–
126.
10(4), 713–718.
(2) Prewitt, C. T.; Shannon, R. D.; Rogers, D. B. Inorg. Chem. 1971,
(13) Christopher, J.; Swamy, C. S. J. Mater. Sci. 1992, 27(5), 1353–1356.
(14) Domen, K.; Ikeda, S.; Takata, T.; Tanaka, A.; Hara, M.; Kondo, J. N.
Appl. Energy 2000, 67(1–2), 159–179.
10(4), 719–723.
(3) Rogers, D. B.; Shannon, R. D.; Prewitt, C. T.; Gillson, J. L. Inorg.
Chem. 1971, 10(4), 723–727.
(15) Doumerc, J. P.; Parent, C.; Chao, Z. J.; Le Flem, G.; Ammar, A. J.
Less-Common Met. 1989, 148, 333–337.
(4) Benko, F. A.; Koffyberg, F. P. J. Phys. Chem. Solids 1984, 45(1),
57–59.
(16) Jacob, A.; Parent, C.; Boutinaud, P.; LeFlem, G.; Doumerc, J. P.;
Ammar, A.; Elazhari, M.; Elaatmani, M. Solid State Commun. 1997,
103(9), 529–532.
(5) Benko, F. A.; Koffyberg, F. P. Can. J. Phys. 1985, 63(10), 1306–
1308.
(6) Benko, F. A.; Koffyberg, F. P. Phys. Status Solidi A 1986, 94(1), 231–
(17) Nagaura, T. Prog. Batteries Solar Cells 1982, 4, 105–107.
(18) Isawa, K.; Yaegashi, Y.; Ogota, S.; Nagano, M.; Sudo, S.; Yamada,
K.; Yamauchi, H. Phys. ReV. B: Condens. Matter 1998, 57(13), 7950–
7954.
234.
(7) Benko, F. A.; Koffyberg, F. P. J. Phys. Chem. Solids 1987, 48(5),
431–434.
2696 Inorganic Chemistry, Vol. 47, No. 7, 2008
10.1021/ic702197h CCC: $40.75  2008 American Chemical Society
Published on Web 02/13/2008

SilWer Delafossite Oxides
light in the visible region, and the variability in their optical
band gaps for different A- and B-site cation combinations
offers many further applications.22–27 For copper delafossites,
the effect of altering the B-site cation on both the optical
and electrical properties has been investigated exten-
sively.7,28–30 In comparison, the optical and electrical proper-
ties of silver delafossites are not well understood, and some
disagreement exists on the reported optical properties. For
example, Vanaja et al. reported an optically measured band
gap of 4.12 eV for a thin film sample of AgGaO2,31 whereas
Maruyama and co-workers reported an optically measured band
gap of 2.4 eV for a polycrystalline powder sample.24 Similar
optically measured band-gap anomalies have been observed in
copper delafossites and have been studied by absorption
spectroscopy and electronic structure calculations.32–34
Figure 1. Schematic representation of the delafossite structure (ABO2).
The gray polyhedra and black spheres represent edge-shared B3+O6 distorted
octahedra and linearly coordinated A+ cations, respectively.
The layered delafossite A+B3+O2 structure, as exemplified
by the parent delafossite mineral, CuFeO2,35,36 maintains the
expected valences for the monovalent A-site cations (Ag+
and Cu+) and trivalent B-site cations [0.535 < r(B3+) < 1.03
Å].37,38 This structure consists of alternate layers of two-
dimensional close-packed A cations with linear O–A+–O
bonds and slightly distorted edge-shared B3+O6 octahedra
(Figure 1).2 Furthermore, each oxygen is coordinated by four
cations (one A+ and three B3+) in a pseudotetrahedral
arrangement. Depending on the stacking of the double layers
(close-packed A cations and BO6 octahedra), two polytypes
are possible. The 3R polytype consists of “AaBbCcAaBbCc”
stacking along the c axis and has rhombohedral symmetry
j
with the space group R3m, whereas the 2H polytype consists
of an alternate “AaBbAaBb” stacking sequence in the P63/
mmc space group.
The synthesis of silver delafossites by high-temperature
ceramic methods never occurs in open systems because Ag2O
decomposes, owing to its low free energy of formation,
before any appreciable reaction can occur. This low value
(∆Ff ) –2.6 kcal/mol) results in decomposition to silver
metal and oxygen at a temperature of 300 °C.1 Therefore,
alternative synthetic methods, including high-oxygen-pres-
sure solid-state, metathetical (cation-exchange), oxidizing
flux, and hydrothermal reactions, have been used to generate
silver delafossites. For example, silver ferrate (AgFeO2), the
first reported silver delafossite, was prepared at low tem-
peratures (∼100 °C) by Krause and Gawryck from the
combination of a “meta-ferric hydroxide gel” (γ-FeOOH)
and Ag2O in a boiling NaOH solution.39 To prepare small
single crystals, Croft et al. employed thin-walled platinum
tubes at higher temperatures (400 °C) and pressures (2700
atm) to synthesize AgFeO2 from Ag2O and Fe2O3 and
therefore took advantage of a closed system (hydrothermal
conditions) to prevent decomposition of Ag2O.40 Later,
Shannon et al. reported the first comprehensive study of three
closed-system syntheses (metathetical, oxidizing flux, and
hydrothermal) to generate numerous single-crystal and
powder samples of delafossites, including silver delafossites.1
In general, the techniques involved low temperatures (e.g.,
metathesis reactions with the formation of a fused salt
byproduct), oxidizing conditions (e.g., solid-state reactions
at high pressures of internally generated oxygen), or both
(e.g., hydrothermal or oxidizing flux reactions). There was
no single technique reported, however, that could be used
to obtain the various silver delafossites and the authors noted
limitations for each synthetic method. First, the metathesis
reactions always led to the formation of highly stable silver
halides with little to no delafossite present. Oxidizing flux
reactions, where the flux was removed with a postsynthetic
leaching step, resulted in higher yields of the delafossite
(19) Yagi, H.; Seo, W.-S.; Koumoto, K. Key Eng. Mater. 2000, 181–182,
63–66.
(20) Koumoto, K.; Koduka, H.; Seo, W.-S. J. Mater. Chem. 2001, 11(2),
251–252.
(21) Takahashi, Y.; Matsushita, H.; Katsui, A. Mater. Sci. Forum 2007,
534–536(2), 1081–1084(Progress in Powder Metallurgy).
(22) Bessekhouad, Y.; Trari, M.; Doumerc, J. P. Int. J. Hydrogen Energy
2003, 28(1), 43–48.
(23) Koriche, N.; Bouguelia, A.; Aider, A.; Trari, M. Int. J. Hydrogen
Energy 2005, 30, 693.
(24) Maruyama, Y.; Irie, H.; Hashimoto, K. J. Phys. Chem. B 2006, 110(46),
23274–23278.
(25) Trari, M.; Bouguelia, A.; Bessekhouad, Y. Sol. Energy Mater. Sol.
Cells 2006, 90(2), 190–202.
(26) Saadi, S.; Bouguelia, A.; Derbal, A.; Trari, M. J. Photochem.
Photobiol., A 2007, 187(1), 97–104.
(27) Brahimi, R.; Bessekhouad, Y.; Bouguelia, A.; Trari, M. J. Photochem.
Photobiol., A 2007, 186(2–3), 242–247.
(28) Nagarajan, R.; Duan, N.; Jayaraj, M. K.; Li, J.; Vanaja, K. A.; Yokochi,
A.; Draeseke, A.; Tate, J.; Sleight, A. W. Int. J. Inorg. Mater. 2001,
3(3), 265–270.
(29) Ingram, B. J.; Harder, B. J.; Hrabe, N. W.; Mason, T. O.; Poeppelmeier,
K. R. Chem. Mater. 2004, 16(26), 5623–5629.
(30) Marquardt, M. A.; Ashmore, N. A.; Cann, D. P. Thin Solid Films
2006, 496(1), 146–156.
(31) Vanaja, K. A.; Ajimsha, R. S.; Asha, A. S.; Jayaraj, M. K. Appl. Phys.
Lett. 2006, 88(21), 212103/1–212103/3.
(32) Pellicer-Porres, J.; Segura, A.; Gilliland, A. S.; Munoz, A.; Rodriguez-
Hernandez, P.; Kim, D.; Lee, M. S.; Kim, T. Y. Appl. Phys. Lett.
2006, 88(18), 181904/1–181904/3.
(33) Makhova, L.; Wett, D.; Lorenz, M.; Konovalov, I. Phys. Status Solidi
A 2006, 203(11), 2861–2866.
(34) Nie, X.; Wei, S.-H.; Zhang, S. B. Phys. ReV. Lett. 2002, 88(6), 066405/
1–066405/4.
(35) Soller, W.; Thompson, A. J. Phys. ReV. 1935, 47, 644.
(36) Pabst, A. Am. Mineral. 1946, 31, 539–546.
(37) Shannon, R. D.; Prewitt, C. T. Acta Crystallogr., Sect. B 1969, 25(Pt.
5), 925–946.
(39) Krause, A.; Gawryck, S. Z. Anorg. Allg. Chem. 1938, 238, 406–412.
(40) Croft, W. J.; Tombs, N. C.; England, R. E. Acta Crystallogr., Sect. A
1964, 17, 313–314.
(38) Shannon, R. D. Acta Crystallogr., Sect. A 1976, A32(5), 751–767.
Inorganic Chemistry, Vol. 47, No. 7, 2008 2697

Sheets et al.
aluminum foil and kept away from light. Reference copper
delafossites (CuAlO2, CuGaO2, CuScO2, and CuInO2) were syn-
thesized by literature methods.45–47
phase, but the presence of a second phase was also reported
for certain compositions. Hydrothermal reactions, carried out
at 500-700 °C with 3000 atm of externally applied pressure,
produced small amounts of the target delafossite in a mixture
with Ag2O, which was easily removed by leaching the
mixture in dilute nitric acid. As with most high-pressure and
-temperature hydrothermal techniques, the use of sealed, thin-
walled gold or platinum tubes limited the mass yield of the
reaction.
Small amounts of AgAlO2 and AgScO2 have been
synthesized previously by high-temperature and -pressure
reactions.1,41 However, the optical and electrical properties
of these silver delafossites have not been studied in-depth.
Here, we demonstrate that low-temperature hydrothermal
techniques originally developed for the synthesis of copper
delafossites can be exploited to access numerous silver
delafossites. In this approach, the hydrothermal synthesis of
the desired product was enabled by controlling the pH using
a NaOH mineralizer. As a result, we report here an
experimental and computational analysis of the optical
and electrical properties of silver delafossites. The optical
and electrical properties of these delafossites are presented
and compared with each other and copper delafossites, where
A ) Ag and Cu and B ) Al, Ga, Sc, and In, which has
permitted a discussion on the influence that both the A- and
B-site cations has on their properties.
Powder X-ray Analysis. Powder X-ray diffraction (PXD) data
were recorded on a Rigaku XDS 2000 diffractometer with Ni-
filtered Cu KR radiation operating at 40 kV and 20 mA. Data were
collected in the 2θ range of 10-80° by scanning every 0.05° for
1 s. Where available, the PXD patterns were matched to their
respective JCPDS files using the Jade software suite.48 Samples
not having a JCPDS file were matched to calculated PXD patterns
generated from published crystallographic data by the PowderCell
software suite.49
Composition Analysis. Cation ratios were determined by induc-
tively coupled plasma atomic emission spectroscopy (ICP-AES). The
polycrystalline powder samples were dissolved in HNO3 at 80 °C
for 3 h and then diluted to measurable concentrations (ppm) with
deionized water. A Thermo Jarrell Ash Atomscan model 25
Sequential ICP spectrometer equipped with vacuum optics covering
the spectral range from 160 to 850 nm was used in the current
work. Three readings were taken per sample and averaged.
Thermogravimetric analysis (TGA) was performed to determine
the oxygen content of selected samples. Approximately 10–30 mg
of material was reduced under flowing N2 gas in a thermogravi-
metric analyzer model Q50 (TA Instruments), and the weight loss
was determined precisely. The resultant products were analyzed
by X-ray diffraction (XRD) to verify that the following general
reaction occurred:
3+
1+y
4Ag1+xM
O
2+δ 9∆8 4(1 + x)Agmetal + 2(1 + y)M32+O3 +
Experimental Section
(1 + 2δ- 6y)O2 (1)
Synthesis. Polycrystalline silver delafossites were prepared from
a direct hydrothermal reaction between silver oxide (99.99%, Alfa
Aesar) and the appropriate B-site cation sesquioxide or trihydroxide
(Supporting Information).42–44 Stoichiometric amounts (0.8 g of
total mass) of these reactants were placed along with fused NaOH
into individual fluoro(ethylene-propylene) Teflon pouches. The
effect of pH and temperature on product formation is discussed
elsewhere.44 For any individual preparation, multiple pouches were
placed in each 125-mL poly(tetrafluoroethylene) (PTFE) Teflon-
lined pressure vessel (Parr Instruments), which was then backfilled
with 50 mL of distilled water. The pressure vessel was then sealed
and heated to 150 °C, followed by a controlled incremental increase
over 5 h to the desired reaction temperature, from 180 to 210 °C
depending on the delafossite. This incremental temperature increase
is necessary to prevent a rapid influx of water from rupturing the
pouch.44 The maximum temperature was held constant for 60 h
with subsequent cooling to room temperature at 6 °C/h. After the
latter step, the pouch was opened and polycrystalline products were
recovered by filtration, followed by a deionized water rinse to
remove any NaOH. To minimize any photochromic decomposition
of the silver delafossites after drying, samples were wrapped in
Electron Microscope Analysis. Electron microscopy studies of
delafossites were carried out using a Hitachi S-4500 scanning
electron microscope (SEM). To prepare SEM samples, polycrys-
talline samples were attached to an aluminum mount using carbon
tape and coated with 5 nm of gold.
Optical Analysis. Optical data were obtained via diffuse
reflectance. In a diffuse-reflectance experiment, the nonspecular
component of reflection (i.e., radiation that has undergone multiple
scattering events on different particles and returned to the sample
surface) is measured relative to a standard.50 The transition edge
onset was used to estimate an optical gap. Diffuse-reflectance data
were collected over the spectral range 200–800 nm using a Cary
500 scanning double-beam spectrophotometer equipped with an
integrating sphere (110 mm in diameter; Cary 1E with Cary 1/3
attachment, Varian, Walnut Creek, CA). Baseline spectra were
collected using pressed PTFE placed in the sample and reference
beams. Data were collected with a scan rate of 300 nm/min, a data
interval of 1 nm, an average time of 0.2 s, and a signal bandwidth
of 3 nm. Pellets were mounted on a blackened sample mask. The
reflectance data were translated into absorption data (R/S) by using
an equation developed by Kubelka and Munk, R/S ) (1–R)2/2R,
where R is the absorption coefficient, S is the scattering coefficient,
and R is the diffuse reflectance at a certain energy.51
(41) Brachtel, G.; Jansen, M. Cryst. Struct. Commun. 1981, 10(1), 173–
174.
(42) Shahriari, D. Y.; Barnabe, A.; Mason, T. O.; Poeppelmeier, K. R.
Inorg. Chem. 2001, 40(23), 5734–5735.
Conductivity Measurements. The electrical conductivity of the
as-prepared polycrystalline silver delafossites was obtained by the
(43) Shahriari, D. Y.; Erdman, N.; Haug, U. T. M.; Zarzyczny, M. C.;
Marks, L. D.; Poeppelmeier, K. R. J. Phys. Chem. Solids 2003,
64(9–10), 1437–1441.
(44) Sheets, W. C.; Mugnier, E.; Barnabe, A.; Marks, T. J.; Poeppelmeier,
K. R. Chem. Mater. 2006, 18(1), 7–20.
(48) Jade, 5.0; Materials Data Inc.: Livermore, CA, 1999.
(49) PowderCell, 2.4; Federal Institute for Materials Research and Testing:
Berlin, Germany, 2000.
(50) Hecht, H. G. Proceedings of the Symposium on Modern Aspects of
Reflectance Spectroscopy; Plenum Press New York, 1968; pp 1–26.
(51) Kubelka, P.; Munk, F. Z. Tech. Phys. (Leipzig) 1931, 12, 593.
(45) Ingram, B. J.; Mason, T. O.; Asahi, R.; Park, K. T.; Freeman, A. J.
Phys. ReV. B: Condens. Matter 2001, 64(15), 155114/1–155114/7.
(46) Park, S.; Keszler, D. A. J. Solid State Chem. 2003, 173(2), 355–358.
(47) Koehler, B. U.; Jansen, M. Z. Anorg. Allg. Chem. 1986, 543, 73–80.
2698 Inorganic Chemistry, Vol. 47, No. 7, 2008

SilWer Delafossite Oxides
powder-solution-composite (PSC) technique.52 In this technique,
the bulk conductivity of a ceramic powder is obtained by measuring
the impedance spectra of composite slurries of the powder with
electrolytic solutions of NaCl. Powder conductivity is obtained from
a series of experiments in which the conductivity of the aqueous
electrolytic solution phase is varied from 2.6 × 10-4 S cm-1 (0.002
M) to 0.25 S cm-1 (4.8 M). Each of these solutions was mixed
with 100 mg of a silver delafossite powder to form a slurry, which
was placed in the measuring apparatus (described elsewhere in
detail),52 and the AC-impedance spectrum was acquired. The
application of this technique to materials with known conductivities
produces results that are in good agreement with expected values.52
Electronic Structure Calculations. Linear muffin-tin orbital
(LMTO) calculations53,54 on the delafossites were performed within
the atomic-sphere approximation using version 47C of the Stuttgart
TB-LMTO-ASA program.55 Scalar-relativistic Kohn–Sham equations
within the local density approximation (LDA)56 were solved by
taking all relativistic effects into account except for the spin–orbit
coupling. The calculations were performed on 132 irreducible k
points within the primitive wedge of the Brillouin zone. Calculations
were performed using an ideal crystal structure of the 2H delafos-
sites (space group P63/mmc) reported by Kandpal and Seshadri.57
It has been shown previously that the precise stacking sequence of
the delafossites (i.e., 2H or 3R polytype) has little effect on their
electronic structure.58,59
Figure 2. SEM images for (a) a mixture of AgScO2 crystallites ranging in
size from a few tenths of one micron up to ∼5 µm and (b) a close-up view
of some micron-sized hexagonal plates of AgGaO2.
Results and Discussion
The difficulties normally encountered in the synthesis of
silver delafossites were overcome by the Teflon pouch
hydrothermal technique.44 Upon optimization of the reaction
temperature and mineralizer concentration, each silver
delafossite formed in >90% mass yield in the direct reaction
between Ag2O and the B-site cation sesquioxide or trihy-
droxide. This method avoids the high temperature and
oxygen pressures required previously to synthesize small
amounts of both AgAlO2 (crystal)41 and AgScO2 (powder)1
delafossites and, indeed, represents a direct method to
produce several grams of polycrystalline powder per reaction.
Increasing the product yield is necessary for the potential
application of silver delafossites as transparent electrodes
(laser ablation or sputtering targets) and photocatalysts where
large amounts of powder are required. The PXD patterns of
the pure-phase silver delafossite samples indexed as either
Figure 3. Plot of the TGA of AgAlO2 reduced in a N2 atmosphere. The
final products (verified by XRD) were Al2O3 and silver metal.
common products for cation-exchange reactions involving
the smaller B-site cations, were not observed.
The SEM images in Figure 2 illustrate that the particle
sizes for silver delafossite crystallites prepared range from
500 nm to a few microns. The cation ratio of each silver
delafossite was verified with ICP analysis. For each silver
delafossite, the average [Ag]/[B] ratio was 1.02(1). The
presence of trace amounts of metallic silver in some product
mixtures is not surprising because it has been evidenced in
previous attempts at silver delafossite synthesis.61–66 Prior
transmission electron microscopy studies from several com-
mercial vendors reveal Ag2O to be the source of the silver
metal.43 The oxygen composition of the silver delafossites
was verified experimentally by TGA as per eq 1, and the
results for AgAlO2 are shown in Figure 3. A 4.71% weight
loss is observed beginning at 700 °C. Based on the final
products, as characterized by XRD, and the level of weight
loss, an initial oxygen content of 2.00(1) per formula unit
j
the 3R (R3m) polytype or a mixture of both the 3R and 2H
(P63/mmc) polytypes (Supporting Information), and the unit
cell parameters for each silver delafossite are in agreement
with those of previous reports.1 It should be noted that the
orthorhombic ꢀ-AgAlO2 and ꢀ-AgGaO2 phases,60 which are
(52) Ingram, B. J.; Mason, T. O. J. Electrochem. Soc. 2003, 150(8), E396–
E402.
(61) Wichainchai, A.; Dordor, P.; Doumerc, J. P.; Marquestaut, E.;
Pouchard, M.; Hagenmuller, P.; Ammar, A. J. Solid State Chem. 1988,
74(1), 126–131.
(53) Andersen, O. K. Phys. ReV. B: Solid State 1975, 12(8), 3060–3083.
(54) Jepsen, O.; Andersen, O. K. Z. Phys. B: Condens. Matter 1995, 97(1),
35–47.
(62) Shin, Y. J.; Doumerc, J. P.; Dordor, P.; Pouchard, M.; Hagenmuller,
P. J. Solid State Chem. 1993, 107(1), 194–200.
(63) Attili, R. N.; Uhrmacher, M.; Lieb, K. P.; Ziegeler, L.; Mekata, M.;
Schwarzmann, E. Phys. ReV. B: Condens. Matter 1996, 53(2), 600–
608.
(55) Jepsen, O.; Andersen, O. K. The Stuttgart TB-LMTO-ASA Program,
version 47;MPI fur Festkorperforschung: Stuttgart, Germany, 2000.
(56) Von Barth, U.; Hedin, L. J. Phys. C: Solid State Phys. 1972, 5(13),
1629–1642.
(64) Otabe, T.; Ueda, K.; Kudoh, A.; Hosono, H.; Kawazoe, H. Appl. Phys.
Lett. 1998, 72(9), 1036–1038.
(57) Kandpal, H. C.; Seshadri, R. Solid State Sci. 2002, 4(8), 1045–1052.
(58) Seshadri, R.; Felser, C.; Thieme, K.; Tremel, W. Chem. Mater. 1998,
10(8), 2189–2196.
(65) Clayton, J. E.; Cann, D. P.; Ashmore, N. Thin Solid Films 2002, 411(1),
140–146.
(59) Ong, K. P.; Bai, K.; Blaha, P.; Wu, P. Chem. Mater. 2007, 19(3),
634–640.
(66) Tate, J.; Jayaraj, M. K.; Draeseke, A. D.; Ulbrich, T.; Sleight, A. W.;
Vanaja, K. A.; Nagarajan, R.; Wager, J. F.; Hoffman, R. L. Thin Solid
Films 2002, 411(1), 119–124.
(60) Gessner, W. Silikattechnik 1970, 21(2), 45–46.
Inorganic Chemistry, Vol. 47, No. 7, 2008 2699

Sheets et al.
Figure 4. Diffuse-reflectance spectra of as-prepared silver delafossites.
Figure 6. Plot of (Rhν)2 versus hν for AgAlO2 (black) and AgScO2 (gray).
The absorption coefficient (R) is estimated by the Kubelka–Munk conversion
of diffuse reflectance.
Figure 5. Kubelka–Munk absorption spectra of as-prepared silver delafos-
sites.
was measured, consistent with the composition AgAlO2. The
calculated oxygen compositions for the other silver delafos-
sites, AgGaO2, AgScO2, and AgInO2, were 2.03, 2.04, and
2.02, respectively (Supporting Information). For the samples
of AgGaO2 and AgInO2, a low-temperature weight loss step
commencing at 200 °C was observed and can be attributed
to the loss of surface water, as determined by IR spectro-
scopy.67
The color of the as-prepared silver delafossites was as
follows: white (AgAlO2), light gray (AgScO2), olive green
(AgGaO2), and bright orange (AgInO2). To quantify their
optical properties, diffuse-reflectance spectra were collected
on uniaxially cold-pressed, polycrystalline specimens (150
MPa). The diffuse reflectance (normalized versus a PTFE
Halon standard) and converted Kubelka–Munk optical ab-
sorbance data are shown in Figures 4 and 5, respectively.
Precise determination of the optically measured band gap at
Γ, from these plots was complicated because diffuse-
reflectance spectroscopy has several inherent limitations
when applied to powder samples. First, for crystallite sizes
smaller than ∼10 µm, Kubelka–Munk theory systematically
underestimates the optically measured band gap, with an error
Figure 7. Kubelka–Munk absorption spectra of as-prepared silver delafos-
sites (black lines) compared to their corresponding copper delafossites (gray
lines) shown for (a) AAlO2, (b) AScO2, (c) AGaO2, and (d) AInO2.
that increases with decreasing crystallite size.68 Second, while
a plot of (Rhν)2 versus hν (Figure 6) can be used to determine
the optical band gaps of 3.6 and 3.8 eV for AgAlO2 and
AgScO2 samples, respectively, additional light absorption in
the visible region for AgGaO2 and AgInO2 samples makes
the exact determination of their optical band gaps more
challenging. Nonetheless, the optical properties of these silver
delafossite are understood by comparing them with data
obtained from experimental and computational studies on
the copper delafossites, whose optically measured band gaps
have been determined accurately from thin film samples.
Figure 7 illustrates that there is a high degree of qualitative
similarity between the optical absorption of silver and copper
delafossite samples with the same B-site cation. Electronic
structure calculations indicate that this similarity results from
(67) Shahriari, D. Y. Ph.D. Thesis, Northwestern University, Evanston, IL,
2004.
(68) Delgass, W. N.; Haller, G. L.; Kellerman, R.; Lunsford, J. H.
Spectroscopy in Heterogeneous Catalysis; Academic Press Inc. 1979.
2700 Inorganic Chemistry, Vol. 47, No. 7, 2008

SilWer Delafossite Oxides
as MAs and CuMS2 (M ) Al, Ga, and In), whose
fundamental direct band gap decreases when the radius of
the group 13 element increases.72 To better understand why
the optically measured band gap increases for copper
delafossites, Nie et al. applied first-principles methods’ LDAs
using the full-potential linear augmented plane-wave method
to calculate the electronic and optical properties of CuBO2
(B ) Al, Ga, and In).34 On the basis of their calculations,
the authors concluded that the fundamental direct band gap
at Γ decreases from 2.68 eV (CuAlO2) f 1.64 eV (CuGaO2)
f 0.73 eV (CuInO2) and thus does follow the trend observed
for conventional group 13 containing semiconductors. Sub-
sequent UV–vis-NIR and X-ray absorption spectroscopic
measurements have confirmed that the energies of the
forbidden fundamental direct and indirect band gaps decrease
down the group.32,33 As illustrated in Figure 9, our density
functional electronic structure calculations using the LMTO
method indicate that the fundamental direct band gap of silver
delafossites also decreases at Γ from 1.7 eV (AgAlO2) f
1.2 eV (AgGaO2) f 0.6 eV (AgInO2). The decrease in the
direct band gap at Γ occurs mainly because the unit cell
volume of the delafossites expands with an increase in the
size of the B-site cation (Figure 8a), which lowers the energy
of the s antibonding states of the group 13 element, thereby
decreasing the energy of the conduction band minimum
(CBM).45,73 It should be noted that the LDA-calculated band
gaps are consistently underestimated by approximately
0.5–2.0 eV, a level of discrepancy that is not unusual for
oxides.74 The systematic error is expected to cancel, however,
because we focus on the difference between band gaps within
a group of similar compounds.
The discrepancy between the aforementioned trends in the
fundamental direct band gaps and the optically measured
band gaps for copper and silver delafossites results from
transitions across the fundamental direct band gap being
dipole forbidden because both the conduction and valence
band states at Γ have the same parity.34,45 Instead, the
optically measured band-gap absorption of a delafossite is
determined by the direct band-gap transition at the next
lowest energy where the dipolar optical transition matrix
element is high. Hence, the optically measured band-gap
transition occurs at M for the 2H polytype of delafossite and
at L for the 3R polytype, which are assigned to transitions
from the valence hybridized copper 3d/silver 4d and oxygen
2p states to copper 4p/silver 5p states (with a lesser
contribution from the B-cation ns states) in the conduction
band.45 Unlike the fundamental direct band gap at Γ, the
calculated direct band gap at M (Figure 9) widens with an
increase in the size of the B-site cation, 3.3 eV (AgAlO2) f
3.2 eV (AgGaO2) f 3.5 eV (AgInO2). The slight decrease
at M from AgAlO2 to AgGaO2 is an exception that results
from the Ga 4s orbital having much lower energy than that
Figure 8. Graph illustrating the relationship between the ionic radius of
the B-site cation and (a) the unit cell volume, (b) the a-axis length, and (c)
the c-axis length for delafossites.
a parallel expansion of the unit cell volume for copper and
silver delafossites (Figure 8a) when the ionic radius of the
group 13 B-site cation increases, 0.535 Å (Al3+) f 0.62 Å
(Ga3+) f 0.80 Å (In3+).38 This parallel unit cell expansion
with increasing group 13 B-site cation size decreases the
onset energy of a parity-forbidden band-gap transition,
resulting in noticeable visible light absorption, and increases
the energy of the optically measured band gap. Indeed, the
optically measured band gaps for copper delafossites, 3.5
eV (CuAlO2)69 f 3.6 eV (CuGaO2)70 f 3.9 eV (CuInO2),71
and silver delafossites, 3.6 eV (AgAlO2) f 4.1 eV (Ag-
GaO2)31 f 4.2 eV (AgInO2),64 expand when the radius of
their B-site cation increases. As first noted by Nie et al.,
however, there is a discrepancy between the above-noted
trend in the optically measured band gap of delafossites and
most conventional group 13 containing semiconductors, such
(72) Madelung, O.; Schulz, M., Landolt-Boernstein. Numerical Data and
Functional Relationships in Science and Technology; Springer-Verlag:
Berlin, 1987; Vol. 22a.
(69) Yanagi, H.; Inoue, S.-i.; Ueda, K.; Kawazoe, H.; Hosono, H.; Hamada,
N. J. Appl. Phys. 2000, 88(7), 4159–4163.
(70) Ueda, K.; Hase, T.; Yanagi, H.; Kawazoe, H.; Hosono, H.; Ohta, H.;
Orita, M.; Hirano, M. J. Appl. Phys. 2001, 89(3), 1790–1793.
(71) Yanagi, H.; Hase, T.; Ibuki, S.; Ueda, K.; Hosono, H. Appl. Phys.
Lett. 2001, 78(11), 1583–1585.
(73) Robertson, J.; Peacock, P. W.; Towler, M. D.; Needs, R. Thin Solid
Films 2002, 411(1), 96–100.
(74) Eng, H. W.; Barnes, P. W.; Auer, B. M.; Woodward, P. M. J. Solid
State Chem. 2003, 175, 94–109.
Inorganic Chemistry, Vol. 47, No. 7, 2008 2701

Sheets et al.
Figure 9. Calculated band structure for (a) AgAlO2, (b) AgGaO2, and (c) AgInO2. Energy zero is at the highest valence band at M.
of the Al 3s orbital.34 The s orbital contributes to these
particular conduction band states and therefore modulates
the energy of the CBM. This trend is also observed in the
calculated direct band-gap trend at L for 3R polytype copper
delafossites, 2.68 eV (CuAlO2) f 2.54 eV (CuGaO2) f 3.08
eV (CuInO2).34 The energies of these transitions increase
with the radius of the B-site cation because the a axis
lengthens (Figure 8b) and reduces the A-A interaction in
the close-packed A-site cation layer, elevating the conduction
band energy of the states with respect to the top of the
valence band. Based on the aforementioned band-gap trends
of both copper and silver delafossites, it is not surprising
that their optically measured band gap expands while their
direct band gap decreases when the radius of the B-site cation
increases; because their unit cell volumes increase in a similar
fashion (Figure 8a).
Despite this increase in the optically measured band gap
of delafossites from Al or Ga to In, the previously mentioned
low-energy “forbidden” direct band gap at Γ (e.g., 1.2 eV
for AgGaO2 and 0.6 eV for AgInO2) results in some
measurable absorption of photons in the visible light range,
as observed in Figures 4 and 5. Indeed, delafossites, such as
CuInO2 and AgInO2 (orange) and CuGaO2 and AgGaO2
(green), are distinctly colored despite having reported band
gaps greater than 3.1 eV.31,64,69,70 Specifically, while the
optically measured band gap for a thin film sample of
AgGaO2 was reported to be 4.1 eV, there is noticeable light
absorption in the transmission spectra of the visible range
with a maximum at ∼500 nm (2.5 eV), an energy that
corresponds to the onset of light absorption for powder
samples of AgGaO2 (Figures 4 and 5).31 The absorption is
significant enough that Maruyama et al. recently reported
an optically measured band gap of 2.4 eV for a polycrys-
talline powder sample of AgGaO2.24 This reduced band gap
and green coloration of AgGaO2 are the result of the
“forbidden” direct band-gap transition at Γ and the indirect
band gap between Γ and M being partially allowed.
Compared to direct and “allowed” band-gap transitions, light
absorption owing to a “forbidden” transition increases more
Figure 10. Calculated band structure for AgScO2. Energy zero is at the
highest valence band at M.
gradually with energy in the vicinity of the band gap,75 and
significant light absorption can occur when the energy of
visible photons is much greater than the energy of a
“forbidden” band-gap transition. As expected, the smaller
direct band gap at Γ for AgInO2 (0.6 eV) when compared to
that for AgGaO2 (1.2 eV), shown in Figure 9, results in the
gradual absorption of visible light at a lower energy (2.0
eV) and orange coloration for a powder sample of AgInO2
(Figure 5).
Although the ionic radius of Sc3+ (0.745 Å) is between
that of Ga3+ and In3+,38 the optical properties of AgScO2
are different from those of AgGaO2 and AgInO2. Indeed,
the optical absorption data of AgScO2 are more similar to
those of AgAlO2, where no significant light absorption is
observed in the visible range for powder samples at photon
energies less than those for the onset of the optically
measured band-gap transition (Figure 5). To better understand
the optical properties of AgScO2, LMTO band structure
calculations were used to calculate the electronic structure
of 2H-AgScO2, which is shown in Figure 10. The funda-
mental direct band gap at Γ for AgScO2 (2.4 eV) is larger
than that of AgGaO2 (1.2 eV) and AgInO2 (0.6 eV) because
the CBM of AgScO2 is at higher energy, owing to the lack
of 3s electrons for the group 3 B-site cation Sc3+. The 3s
orbital contributes to the conduction band states at Γ and
(75) Yu, P. Y., Cardona, M., Eds. Fundamentals of Semiconductors: Physics
and Materials Properties; Springer, 1995.
2702 Inorganic Chemistry, Vol. 47, No. 7, 2008

SilWer Delafossite Oxides
Figure 12. Kubelka–Munk absorption spectra of as-prepared AgScO2
before and after exposure to visible light for 72 h.
4d and oxygen 2p states to silver 5p states in the conduction
band thus requires more energy owing to the scarcity of silver
4d states just below the valence band maximum (VBM) of
silver delafossites. As a result, silver delafossites have a wider
optical band gap compared to the copper delafossites with
the same B-site cation (Figure 7). In addition to a larger
optical band gap, the as-prepared silver delafossites have
lower absorption coefficients than copper delafossites across
the visible range. For example, Figure 7a shows that the
absorption coefficient of AgAlO2 is one order of magnitude
lower than that of CuAlO2. All things considered, the larger
band gap and lower absorption coefficients of silver delafos-
sites would translate into more transparent thin films.
The synthesized samples, noticeably AgScO2, exhibit
photosensitivity to visible light. This was not surprising
because the photochromic behavior of certain silver complex
oxides, including AgInO2, has been reported previously.76
Figure 12 shows the optical absorption data of AgScO2 as-
prepared and after continuous exposure to laboratory lighting
for 72 h. After light exposure, absorption increases in the
visible range at energies less than those of the absorption
edge; however, the absorption edge remained constant. The
mechanism of the photosensitive properties is analogous to
those reported for other complex silver metal oxides where
the Ag+ cations reduce to silver metal with the release of
oxygen (see eq 1).76–78 Electron microscopy has revealed
previously the presence of globules, which by electron
diffraction are shown to be metallic silver, attached to the
surface of larger silver oxide particles.43
Figure 11. Projected densities of A d and O p states for (a) AAlO2, (b)
AScO2, (c) AGaO2, and (d) AInO2, where A ) Cu (gray lines) or Ag (black
lines). In these plots, the top of the valence band is taken as zero on the
energy axis.
therefore modulates the energy of the CBM.34,45 Therefore,
no significant light absorption, resulting from “forbidden”
direct or indirect band-gap transitions, occurs in the visible
range before the onset of the optically measured band-gap
transition of 3.8 eV, which corresponds to the direct band-
gap transition at M (Figure 10).
If we now compare copper and silver delafossites with
the same B-site cation (Figures 7), one observes that the onset
of the optically measured band gaps of the silver delafossites
is a few tenths of an electronvolt greater than that of the
corresponding copper delafossites. The larger optical band
gaps of silver delafossites do not result from the unit cell
volume expansion upon replacement of copper with silver
on the A site. Because the O–A–O bonds are parallel to the
c axis, the length of the a axis does not change significantly
when copper is replaced by silver in a delafossite with a
constant B-site cation (Figure 8b). For example, the a-axis
lengths of CuAlO2 (2.858 Å) and AgAlO2 (2.890 Å) differ
by only 0.032 Å, a 1.1% increase. Instead, the unit cell
volume increase for silver over copper delafossites with the
same B-site cation (Figure 8a) results primarily from an
expansion of the c axis (Figure 8c). With a minimal change
in the a axis, the A-A interaction between the closed packed
A-site cations should remain constant for copper and silver
delafossites with the same B-site cation. If the A-A
interaction does not change, the CBM states do not change,
leading to the conclusion that the optical gap increases for
another reason. Indeed, density functional electronic structure
calculations using the LDA LMTO method indicate that the
increase of the optical band gap from copper to silver
delafossites is due to a change in energy valence band of
the optical band gap that includes the A-site cation d states.57
Figure 11 illustrates that while the partial density of states
(PDOS) of copper 3d states is sharply peaked near the top
of the valence band, the PDOS for silver 4d states is broader,
tails to lower energies, and traces the oxygen p character.
The transition of a valence electron from hybridized silver
Because the silver delafossite powders could not be
sintered into dense ceramics without partial or complete
thermal decomposition, a PSC technique52 was used to
measure their electrical properties. The results were analyzed
with effective medium theory and equivalent circuit model-
ing, and the results for AgAlO2 are shown in Figure 13 (see
the Supporting Information for PSC data of all silver
delafossites). SEM images (Figure 2) showed a similar
(76) Hirono, T.; Yamada, T. J. Appl. Phys. 1984, 55(3), 781–785.
(77) Gurney, R. W.; Mott, N. F. Proc. R. Soc. London, Ser. A 1938, 164,
151–167.
(78) L’Vov, B. V. Thermochim. Acta 1999, 333(1), 13–19.
Inorganic Chemistry, Vol. 47, No. 7, 2008 2703

Sheets et al.
much lower than typically required for application as a
transparent conductor, the conductivities of the as-prepared
silver delafossites are comparable to or lower than values
previously reported for polycrystalline powder samples of
copper delafossites with the same B-site cation (Table 1).
Based on the argument of Nagarajan et al. that holes are
more mobile in the d states than in the p states,28 it has been
proposed previously that the admixture of oxygen 2p states
with the silver 4d states at the VBM (Figure 11) limits the
mobility of holes and thus the conductivity of silver
delafossites.57 Indeed, copper delafossites are more conduct-
ing because their VBM comprises isolated copper 3d states.
Nonetheless, the conductivities of both copper and silver
delafossites containing main-group B-site cations are low
owing to a diffusion-limited polaron conduction mechanism
in which trapped holes and resultant lattice distortions “hop”
between sites, which limits room temperature mobility (<1
cm2 V-1 s-1) and therefore conductivity.45 Similar to copper
delafossites, acceptor doping via aliovalent substitution or
oxygen intercalation could be used to enhance the conductiv-
ity of silver delafossites by increasing the carrier concentra-
tion. For example, the conductivities of both CuYO281,82 and
CuCrO283 increase significantly upon aliovalent doping with
Ca and Mg, respectively.
Figure 13. Plot that summarizes the PSC conductivity measurements for
AgAlO2. The gray data points represent the measured conductivity of the
PSC, while the black line represents the conductivity of the electrolyte
solution. A conductivity estimate is given where the composite and
electrolyte conductivities are equivalent. A Bruggeman asymmetric model
considering randomly arranged spherical particles and a broad range of sizes
(volume fraction ∼ 0.2–0.3) was used.
Table 1. Measured and Literature Room Temperature Conductivities of
Certain Delafossites
compound
form
σ (S cm-1
)
ref
CuAlO2
CuAlO2
CuAlO2
AgAlO2
CuGaO2
CuGaO2
AgGaO2
AgGaO2
crystal
powder
film
powder
powder
film
1 × 10-3
84
4, 52, 85
30
this work
6, 86
30
3
31
this work
29
this work
87
88
3
65
64
1.7 × 10-3–3.6 × 10-2
8 × 10-2–2
9 × 10-3
5.6 × 10-3–1.1 × 10-2
6.3 × 10-2–0.56
2 × 10-8
crystal
film
3.2 × 10-4
e10-4
Conclusions
a
AgGaO2
powder
powder
powder
powder
film
crystal
powder
film
CuScO2
AgScO2
CuInO2
CuInO2
AgInO2
AgInO2
AgInO2
AgInO2
2 × 10-2
The silver delafossites AgAlO2, AgScO2, AgGaO2, and
AgInO2 have been synthesized on a preparative scale by a
low-temperature (<210 °C) and -pressure (<20 atm) Teflon
pouch hydrothermal technique. As a result, the optical and
electrical properties of a series of these silver delafossites
have been examined for the first time. Electronic structure
calculations and analysis of the Kubelka–Munk absorption
data reveal that, similar to copper delafossites, silver delafos-
sites have a disparity in energy between the “forbidden”
fundamental direct and indirect band gaps and optically
measured band gap. While their optically measured band
gaps widen with an increase in the radius of the B-site cation,
the decreased fundamental band gaps for larger B-site cations
result in some absorption of photons in the visible light range
for AgGaO2 and AgInO2, which reduces their optical
transparency and thereby colors these delafossites. When
corresponding copper and silver delafossites with the same
B-site cation are compared, however, silver delafossites have
larger band gaps and lower visible light absorption, owing
4 × 10-2
10-4
5.6 × 10-2
1 × 10-4
10-7
1 × 10-5
powder
8 × 10-3
this work
a The conductivity of the sample is below the measurable range of
conductivity (e10-4).
platelike morphology for all silver delafossites, and the best
fit to the PSC composite data for these particles was given
by the model proposed by Fricke in which the Maxwell–
Wagner dispersion was extended to a suspension of ellipsoids
(platelike).79 Although some samples were a mixture of two
polytypes (2H and 3R), it has been reported previously that
the existence of two polytypes, which alters stacking along
the c axis, has little effect on the measured polycrystalline
conductivity values.80
The electrical conductivities of the polycrystalline silver
delafossites and other reference delafossites are presented
in Table 1. Comparisons of the electrical properties for
copper and silver delafossites are complicated by the fact
that the carrier type of the silver delafossite powders could
not be established reliably. Prior studies of AgGaO2 and
AgInO2 have reported thin film samples to be p and n type,
respectively, and the small transference number of Ag+ in
such samples, as determined by the direct current polarization
method, indicates that ionic contribution to conductivity is
negligible.31,64 While the measured conductivity values are
(81) Cava, R. J.; Peck, W. F.; Krajewski, J. J.; Cheong, S. W.; Hwang,
H. Y. J. Mater. Res. 1994, 9(2), 314–317.
(82) Jayaraj, M. K.; Draeseke, A.; Tate, J.; Sleight, A. W. Thin Solid Films
2001, 397, 244–248.
(83) Nagarajan, R.; Draeseke, A. D.; Sleight, A. W.; Tate, J. J. Appl. Phys.
2001, 89(12), 8022–8025.
(84) Lee, M. S.; Kim, T. Y.; Kim, D. Appl. Phys. Lett. 2001, 79(13), 2028–
2030.
(85) Ingram, B. J.; Gonzalez, G. B.; Mason, T. O.; Shahriari, D. Y.;
Barnabe, A.; Ko, D.; Poeppelmeier, K. R. Chem. Mater. 2004, 16(26),
5616–5622.
(86) Ashmore, N. A.; Cann, D. P. J. Mater. Sci. 2005, 40, 3891–3896.
(87) Sasaki, M.; Shimode, M. J. Phys. Chem. Solids 2003, 64(9–10), 1675–
1679.
(79) Fricke, H. J. Phys. Chem. 1953, 57, 934–937.
(80) Routbort, J. L.; Rothman, S. J. J. Appl. Phys. 1994, 76(10), 5615–
5628.
(88) Yang, B.; He, Y.; Polity, A.; Meyer, B. K. MRS Proc. 2005, 865,
399–403.
2704 Inorganic Chemistry, Vol. 47, No. 7, 2008

SilWer Delafossite Oxides
to a shift in the valence band states to lower energy upon
replacement of copper 3d states with silver 4d states. Thus,
we have demonstrated that the electronic structure prediction
is general (i.e., any silver delafossite has a larger band gap
than the corresponding copper delafossite). Moreover, while
silver delafossites have conductivities lower than those of
polycrystalline powders of copper delafossites, this study
provides a starting point for the difficult process of improving
the conductivity of delafossites through extrinsic doping
without significantly compromising their optical properties.
program of the National Science Foundation (Grant DMR-
0520513) at the Materials Research Center of Northwestern
University. The SEM work was performed by M. Russell
and S. DiBenedetto in the EPIC facility of NUANCE Center
at Northwestern University. NUANCE Center is supported
by NSF-NSEC, NSF-MRSEC, Keck Foundation, the State
of Illinois, and Northwestern University. W.C.S was ad-
ditionally supported through a Natural Sciences and Engi-
neering Research Council of Canada (NSERC) Julie Payette
postgraduate doctoral scholarship.
Supporting Information Available: Commercial sources and
preparation methods for reagents used, PXD spectra matched to
calculated PXD patterns generated from published crystallographic
data, TGA data, and PSC conductivity data and references. This
materialisavailablefreeofchargeviatheInternetathttp://pubs.acs.org.
Acknowledgment. The authors thank R. Seshadri, who
contributed to the theoretical calculations on delafossites. The
authors gratefully acknowledge support from the Department
of Energy, National Renewable Energy Laboratory Subcon-
tract (Award XAT-5-33636-02/DE-AC36-98GO). The au-
thors made use of Central Facilities supported by the MRSEC
IC702197H
Inorganic Chemistry, Vol. 47, No. 7, 2008 2705