Metallicity of Ca2Cu6P5 with single and double copper-pnictide layers

Article abstract of DOI:10.1016/j.jallcom.2016.02.084

We report thermodynamic and transport properties, and also theoretical calculations, for Cu-based compound Ca₂Cu₆P₅ and compare with CaCu_2-δP₂. Both materials have layers of edge-sharing copper pnictide tetrahedral CuP₄, similar to Fe-As and Fe-Se layers (with FeAs₄, FeSe₄) in the iron-based superconductors. Despite the presence of this similar transition-metal pnictide layer, we find that both Ca₂Cu₆P₅ and CaCu_2-δP₂ have temperature-independent magnetic susceptibility and show metallic behavior with no evidence of either magnetic ordering or superconductivity down to 1.8 K CaCu_2-δP₂ is slightly off-stoichiometric, with δ = 0.14. Theoretical calculations suggest that unlike Fe 3d-based magnetic materials with a large density of states (DOS) at the Fermi surface, Cu have comparatively low DOS, with the majority of the 3d spectral weight located well below Fermi level. The room-temperature resistivity value of Ca₂Cu₆P₅ is only 9 μΩ-cm, due to a substantial plasma frequency and an inferred electron-phonon coupling λ of 0.073 (significantly smaller than that of metallic Cu). Also, microscopy result shows that Cu-Cu distance along the c-axis within the double layers can be very short (2.5 ?), even shorter than metallic elemental copper bond (2.56 ?). The value of dρ/dT for CaCu_2-δP₂ at 300 K is approximately three times larger than in Ca₂Cu₆P₅, which suggests the likelihood of stronger electron-phonon coupling. This study shows that the details of Cu-P layers and bonding are important for their transport characteristics. In addition, it emphasizes the remarkable character of the DOS of '122' iron-based materials, despite much structural similarities.

Full text of DOI:10.1016/j.jallcom.2016.02.084

Accepted Manuscript
Metallicity of Ca Cu P with Single and Double Copper-Pnictide Layers
6 5
2
Li Li, David Parker, Miaofang Chi, Georgiy M. Tsoi, Yogesh K. Vohra, Athena S. Sefat
PII:
S0925-8388(16)30353-X
10.1016/j.jallcom.2016.02.084
JALCOM 36696
DOI:
Reference:
To appear in: Journal of Alloys and Compounds
Received Date: 2 December 2015
Revised Date: 5 February 2016
Accepted Date: 9 February 2016
Please cite this article as: L. Li, D. Parker, M. Chi, G.M. Tsoi, Y.K. Vohra, A.S. Sefat, Metallicity of
Ca Cu P with Single and Double Copper-Pnictide Layers, Journal of Alloys and Compounds (2016),
2
6 5
doi: 10.1016/j.jallcom.2016.02.084.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Metallicity of Ca2Cu6P5 with Single and Double Copper-Pnictide Layers
Li Li1, David Parker1, Miaofang Chi,2 Georgiy M. Tsoi,3 Yogesh K. Vohra,3 and Athena S.
Sefat1
1 Materials Science & Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN
37831, USA
2 Center for Nanophase Materials Sciences, Oak Ridge National Laboratory, Oak Ridge, TN,
37831, USA
3 Department of Physics, University of Alabama at Birmingham, Birmingham, AL 35294, USA
Corresponding author emails: lil2@ornl.gov; sefata@ornl.gov
Abstract
We report thermodynamic and transport properties, and also theoretical calculations, for
Cu-based compound Ca2Cu6P5 and compare with CaCu2-δP2. Both materials have layers of edge-
sharing copper pnictide tetrahedral CuP4, similar to Fe-As and Fe-Se layers (with FeAs4, FeSe4)
in the iron-based superconductors. Despite the presence of this similar transition-metal pnictide
layer, we find that both Ca2Cu6P5 and CaCu2-δP2 have temperature-independent magnetic
susceptibility and show metallic behavior with no evidence of either magnetic ordering or
superconductivity down to 1.8 K. CaCu2-δP2 is slightly off-stoichiometric, with δ = 0.14.
Theoretical calculations suggest that unlike Fe 3d-based magnetic materials with a large density
of states (DOS) at the Fermi surface, Cu have comparatively low DOS, with the majority of the
3d spectral weight located well below Fermi level. The room-temperature resistivity value of
Ca2Cu6P5 is only 9 µΩ-cm, due to a substantial plasma frequency and an inferred electron-
phonon coupling λ of 0.073 (significantly smaller than that of metallic Cu). Also, microscopy
result shows that Cu-Cu distance along the c-axis within the double layers can be very short (2.5
Å), even shorter than metallic elemental copper bond (2.56 Å). The value of dρ/dT for CaCu2-δP2
at 300 K is approximately three times larger than in Ca2Cu6P5, which suggests the likelihood of
stronger electron-phonon coupling. This study shows that the details of Cu-P layers and bonding
are important for their transport characteristics. In addition, it emphasizes the remarkable
character of the DOS of ‘122’ iron-based materials, despite much structural similarities.

ACCEPTED MANUSCRIPT
1 Introduction
The discovery of high-TC superconductivity in LaFeAsO1 has generated extensive interest
in iron-based superconductors and other layered transition metal based materials, whose
structures contain MX (M= transition metal, X = pnicogen or chalcogen) layers composed of
edge-sharing MX4 tetrahedra2. Among the vast number of compounds, the ThCr2Si2-type
AM2Pn2 pnictides (A = Ca, Sr, Ba, Pn= P, As, Sb) have attracted much attention due to their
wide range of doping possibilities either by holes or electrons, as well as the recipe for single
crystals with sufficient sizes for various experiments. It is suggested that this type of material is
located at the borderline of an interlayer Pn-Pn bonding instability, and the forming and breaking
of a Pn-Pn bond will lead to a shape change of the tetrahedron as well as the charge
redistribution within the Pn-M-Pn bonded layers3. Different choices of A or M element may
affect the formation of Pn-Pn bond in these phases, and further influence the crystal and
electronic structures, magnetism, and presence of superconductivity in the ternary pnictides.
Following this assumption, many efforts have been made to revisited the 122-type pnictides and
varieties of interesting behaviors have been found other than the high-temperature
4
superconductivity, e.g., itinerant antiferromagnetism in BaCr2As2
and CaCo2P2,5
antiferromagnetic insulator in BaMn2As2,6 Pauli paramagnetic metals in CaFe2P2 7 and CaNi2P2,7
metal with nearly T-independent diamagnetism in ACu2Pn2 (A = Sr, Ba; Pn = As, Sb)8–10, and
superconductivity in BaNi2P2 and SrNi2P2 (TC < 4 K)12. By analogy with the cuprates, it is
11
interesting to consider whether high superconductivity can be observed in material with copper
phosphide layers, in a similar way to iron-based superconductors.
Our work here focuses on two calcium copper-pnictide layered compounds of Ca2Cu6P5
(“265”) and CaCu2-δP2 (“122”). Previous reports on the existence of the two compounds are
limited to crystallographic studies13–15, and there’s no following work on property measurements.
Both 265 and 122 have layers of edge-sharing copper pnictide tetrahedral (CuP4), similar to iron
arsenides in iron-based superconductors (FeSC). 265 has both single and double P-Cu-P layers,
while 122 is only made up of single layers. X-ray refinement result shows there are ~ 7%
vacancies on the Cu-sites in CaCu2-δP2, which is similarly found in arsenide version of
16
CaCu1.7As2 . Through extensive properties measurements, we find no evidence of either
magnetic ordering or superconductivity in these two copper pnictides, despite the resemblance of
crystal structure building units to the iron-based materials. Both Ca2Cu6P5 and CaCu2-δP2 show
temperature-independent weak magnetic susceptibility (χ≈10-5 emu/mol.Oe) and CaCu2-δP2
forms with Cu vacancies that lower the Fermi-level DOS and T-linear specific heat coefficient.
Our structural details and theoretical calculations explained the metallic behavior in these two
compounds and suggest the importance of the details of Cu-P layers on the transport
characteristics of these materials.
2 Experimental
The elemental reactants, dendritic Ca pieces (99.98%), Cu powders (99.999%), P lumps
(99.999%) were all from Alfa Aesar. Prior to its use, copper powder was reduced in a diluted H2
stream (5/96% H2/Ar) for 12 h, whereas other elements were used as received. All operations
were handled in a helium-filled glovebox with the moisture and oxygen below 0.1 ppm. CaP
binary was prepared by placing mixtures of Ca pieces and P lumps in a silica tube, which were

ACCEPTED MANUSCRIPT
evacuated to a pressure of 10-2 Torr and placed in a horizon tube furnace. The mixtures were
slowly heated up to 400 °C (annealed for 1 day), then heated up to 750 °C (annealed for 3 days),
then slowly cooled to room temperature. Cu2P binary was prepared from the mixtures of Cu
powders and P lumps by using the same method. Ca2Cu6P5 and CaCu2-δP2 were grown out of
CaP and Cu2P binaries, with the ratio of 2:1 and 1:1, respectively. The reaction mixtures were
sealed inside evacuated silica tubes, annealed at 900 °C (for 265) and 950 °C (for 122) for 12
hours, and then slowly cooled to room temperature over 10 hours. The samples were ground and
pressed into pellets, sealed under vacuum inside silica tubes, heated to 1000 °C (for 265) and
950 °C (for 122), and annealed for 24 hours. After another regrinding and pelletizing, the pellets
were sealed in evacuated silica tubes and annealed at 900 °C (for 265) and 950 °C (for 122) for
48 hours before turning off the furnace. The obtained samples were black and stable in air.
The powder x-ray diffraction (XRD) patterns were collected at room temperature by a
PANalytical X’Pert PRO MPD x-ray powder diffractometer with Cu Kα radiation. The Rietveld
refinements of the XRD patterns were done with FullProf software suite.17 The electrical and
thermal transport measurements were performed in a Quantum Design (QD) Physical Property
Measurement System (PPMS), using a standard four-probe method. Magnetization
measurements were performed in a QD Magnetic Property Measurement System (MPMS).
Scanning Transmission Electron Microscopy (STEM) images on 265 were acquired using
High Angle Annular Dark Field (HAADF) and Annular Bright Field (ABF) imaging methods in
STEM mode with an FEI Titan 60-300S microscope equipped with a Gatan Imaging Filter
Quantum 865. The electron probe used in the imaging was ~0.7A at a full-width half-maximum
with a convergence semi-angle of 23mrad at 300kV. A collection semi-angle of 55-300mrad for
HAADF imaging, and of 11.5- 23mrad for ABF imaging was implemented in this experiment.
The TEM specimens were prepared by focused ion beam sectioning with a low ion-beam current
followed by nano-milling on a cryo-stage for the final thinning.
High pressure electrical resistance measurements at cryogenic temperatures down to 10 K
were performed in a six-tungsten microprobe designer diamond anvil cell (DAC) with 280 ꢀm
size culet18,19. The tungsten microprobes are encapsulated in a homoepitaxial diamond film and
are exposed only on the culet surface to make contact with the sample. The sample was loaded
into an 80 ꢀm hole of a 230 ꢀm thick spring-steel gasket that was first pre-indented to a ∼80 ꢀm
thickness and mounted between the designer diamond anvil and 300 ꢀm culet size matching
diamond. Two electrical leads pass constant 2 mA current through the sample and two additional
leads measure voltage across the sample. The sample pressure was monitored continuously by
the in situ ruby fluorescent technique20 and care was taken to carefully calibrate the ruby R1
emission to the low temperature of 10 K.18 In this study no pressure medium was used in order to
ensure good electrical contacts between the probes and the sample.
First principles density functional theory calculations were performed using the linearized
augmented plane wave code WIEN2K21, within the Generalized Gradient Approximation of
Perdew, Burke and Ernzerhof22. Internal coordinates were relaxed until forces were less than
2mRyd/Bohr, and sphere radii of 2.1 Bohr for P and 2.3 Bohr for Ca and Cu were used for
Ca2Cu6P5, and 1.96 and 2.5, respectively, for CaCu2P2. An RKmax of 9.0 was used, where R is
the smallest LAPW sphere radius and Kmax the largest plane wave vector. For the purposes of
computing the plasma frequencies used to understand the resistivity, we used the Boltzmann
transport code BoltzTraP23. Approximately 1000 k-points in the full Brillouin zone were used

ACCEPTED MANUSCRIPT
for the self-consistent calculations, and approximately 10,000 points were used for the transport
calculations.
3 Results and discussion
Fig. 1 (a) shows the XRD patterns and Rietveld refinements for Ca2Cu6P5 and CaCu2-δP2
structures respectively. The reflections can be well indexed with the space group of I4/mmm (No.
139) for both. The schematic images of the crystal structure are shown in the inset. Only trace
amount of impurity phases are present in the XRD pattern, which have been taken into account
for a refinement that improved the fitting. The lattice parameters are a = 4.0169(3) Å, c =
24.6634(2) Å for Ca2Cu6P5, and a = 4.0223(2) Å, c = 9.6333(1) Å for CaCu2-δP2. Impurity phase
was estimated using the Hill and Howard method24. For 265, the only impurity detected was
CuP2, refined to less than 3% by mass; for 122, the total content of impurities were less than 6%.
For CaCu2-δP2, electron density of Cu-site occupancy according to diffraction data is about 0.93,
i.e., the vacancy concentration is ≈7% (= δ/2), corresponding to an approximate composition of
16
CaCu1.86P2. Vacancies were similarly reported for CaCu1.7As2 .
The STEM results of HAADF image of Ca2Cu6P5 and its simultaneously acquired ABF
are shown Fig.1(b), which are sensitive to heavy and light elements, respectively. By combining
these two imaging techniques, the atomic columns of all three elements, Ca, Cu and P, are
clearly observed. An alternating single and double Cu-P tetrahedron layers are resolved. Here,
the lattice distortions or bonding length modification from the incorporation of different layered
configurations are detected. In fact, the interlayer Cu-Cu bond length in the double P-Cu2-P
layers is between 0.25 nm and 0.29 nm, well consistent with the x-ray refinement result (≈ 2.65
Å). In fact, the Cu-Cu distances in the double layer seem to alternate from the detailed image
analyses; the atomic position are identified using Gaussian fitting (see Fig. 1b). The Cu-Cu
distance within ab-plane is 2.84 Å (=a/√2), slightly larger than reported for dFe-Fe in BaFe2As2
(2.80 Å).
Fig. 2 shows the temperature dependence of electrical resistivity ρ(T) for Ca2Cu6P5 and
CaCu2-δP2. Both of them exhibit almost similar metallic behavior over the whole measured
temperature range. At room temperature, the electrical resistivity is about 9 µΩ-cm for Ca2Cu6P5
and 47 µΩ-cm for CaCu2-δP2. As the temperature decreases, the resistivity decreases
monotonously down to 1.5 µΩ-cm for Ca2Cu6P5 and 19 µΩ-cm for CaCu2-δP2 around 20 K, and
then keeps almost constant below 20 K. The ρ(T) of CaCu2-δP2 does not show any feature down
to 2 K, in disagreement with CaCu1.7As2 exhibiting an anomaly (at ~ 45 K) assumed to be
associated with ordering of the Cu vacancies16. For the single- and double-layer compound of
Ca2Cu6P5, we performed a pressure study to see if some correlated behavior can be induced
within the double layers. The temperature dependence of resistivity below 100 K with applying
hydrostatic pressure of 1.4 GPa, 3.7 GPa, 8.9 GPa, and 11.7 GPa, is shown in Fig. 2 inset.
Unfortunately, the sample remains metallic without anomalies. Magnetic susceptibility
measurements show that Ca2Cu6P5 is Pauli paramagnetic, while CaCu2-δP2 is diamagnetic with
signals of χ ~ 10-5 emu/mol.Oe (not shown), and a Curie tail below due to paramagnetic
impurities as observed in diffraction.
Fig. 3(a) illustrates the temperature dependence of the specific heat for Ca2Cu6P5 and
CaCu2-δP2. There is no anomaly in the whole temperature range, confirming no structural
transition or magnetic ordering in either system. The plots of Cp/T versus T2 for Ca2Cu6P5 and

ACCEPTED MANUSCRIPT
CaCu2-δP2 in the temperature range from 1.8 to 10 K are shown in the inset of Fig. 3Fig. 3(b).
The experimental data points were fitted using the formula Cp = γT + βT3, in which the first and
the second terms are the electronic and lattice contributions to the heat capacity, respectively.
The obtained results from the fitting are γ = 5.64(2) mJ/mol-K2, β = 1.34(1) mJ/mol-K4 for
Ca2Cu6P5, and γ = 1.66(3) mJ/mol-K2, β = 0.21(2) mJ/mol-K4 for CaCu2-δP2. The Debye
temperatures are estimated to be 265 K for Ca2Cu6P5 and 359 K for CaCu2-δP2, from the ΘD =
(12π4NR / 5β) 1/3, where N is the number of atoms per formula.
The Seebeck coefficients of Ca2Cu6P5 and CaCu2-δP2 are shown in Fig. 4(a). The
thermopower values increase monotonically within the whole temperature range, achieving
about 9.67 ꢀV K-1 for Ca2Cu6P5 and 4.84 ꢀV K-1 for CaCu2-δP2, which are both relatively low for
metallic compounds. Their thermopowers are positive, indicating that holes are the major charge
carriers in the system. The inset of Fig. 4 Fig. 4(b) shows the temperature dependence of the
thermal conductivity κ for Ca2Cu6P5 and CaCu2-δP2. The thermal conductivity of Ca2Cu6P5
rapidly increases to ~100 WK-1m-1 with a linear behavior from 2 K to 30 K, and after that it is
temperature independent up to 300 K. For comparison, the κ value of Ca2Cu6P5 at room
temperature is close to 109 WK-1m-1 of brass (~70%Cu-30%Zn)25, but is only one fourth of 401
WK-1m-1 for the value of pure copper25. For CaCu2-δP2, the thermal conductivity value (19 WK-
1m-1) at 300 K is much lower than Ca2Cu6P5. The thermal conductivity also shows a rapid
increase from 2 K to 30 K (κ30K ≈ 11 WK-1m-1), and after that it slowly increases with warming
up to 300 K.
As mentioned above, the room temperature resistivity value of Ca2Cu6P5 is 9 µΩ-cm,
which is exceptionally low and may be a result for having small Cu-Cu intraplanar distance of
2.5 Å. By way of comparison, the review of resistivity of cubic metallic elements by Allen26
shows values at 295 K ranging from 1.61 µΩ-cm for Ag to 39 µΩ-cm for Ba, and fully half these
elements exceed the 9 µΩ-cm value. In order to understand the high level of metallicity for
Ca2Cu6P5, we performed first principles DOS calculations. Fig. 5 shows a DOS comparison of
265 to 122. For 265, the greatest DOS is located between 3 and 5 eV below the Fermi level. This
is unlike 3d-based magnetic materials in which there is typically a large DOS at the Fermi
surface and in fact bears some resemblance to metallic Cu itself. The Fermi-level DOS is just
2.38 states/eV-f.u., and much of the spectral weight within the spheres is Cu, providing the
beginnings of an explanation for the low resistivity. For Ca2Cu6P5, when we convert the Fermi-
level DOS to a T-linear electronic specific heat coefficient γ, we find a γ of 5.62 mJ mol-1 K-2.
This compares very well to the experimental value of 5.64 mJ mol-1 K-2 and this agreement is in
fact suggestive of very weak electron-phonon coupling.
The standard theory of electron-phonon coupling26 tells us that electron-phonon coupling
enhances “bare” band structure values of γ by a factor (1 + λ), where λ is the dimensionless
electron-phonon coupling constant. As pointed out by Allen, typically (1 + λ) is uncertain by
~10%, so that the theory and experiment together here should not be taken to imply a λ of
5.64/5.62 – 1 = 0.0036, but are suggestive of a λ of order 0.1 or less, which is comparable to that
of Cu itself. To put this on still more solid computational foundation, we consider that the
resistivity ρ in this regime is given by the following expression: ꢀ = 8ꢁꢂꢃꢄꢅꢆ/ℏꢇꢈꢂ, where ωp
2
is the plasma frequency and kB is the Boltzmann’s constant. Now, ωp is directly calculable from
the electronic structure results (it is effectively identified with σ / τ output from the BoltzTraP
results; here τ is an average scattering time). There is one slight complication in that in Ca2Cu6P5

ACCEPTED MANUSCRIPT
the plasma frequency is anisotropic due to the tetragonal symmetry; we address this by
modification ω2p,total=(2/3) ω2
+ (1/3) ω2 (this is equivalent to assuming an isotropic
p,planar
p,c
scattering time). This yields an ωp,total of 3.14 eV. Again, this is a comparatively large value; it
is comparable to the smallest plasma frequency in the cubic metals compiled by Allen26, thus
forming another piece of the explanation of the low resistivity of Ca2Cu6P5. We note that the
Fermi surface of this material (depicted in Fig. 5a) shows a large area, consistent with the large
plasma frequencies. The resistivity value can be fitted using the first principles calculated plasma
frequencies, finding an electron-phonon coupling λ in Ca2Cu6P5 of 0.073. This is significantly
smaller than even the value of 0.13 estimated for metallic Cu, and it is this very small value that
is mainly responsible for the surprisingly small resistivity.
Note that we consider the resistivity as being due to electron-phonon coupling, as is
usually the case in metals. It is conceivable, though rather unlikely, that there is a contribution
from spin-fluctuation scattering arising from magnetic fluctuations in the Cu atoms. No evidence
of a magnetic phase transition was seen, either in the experiment (ambient and even pressure
work) or in our theoretical calculations (we checked both ferromagnetic and a nearest-neighbor
antiferromagnetic state). It is of interest to consider the theoretical results on this layered
structure in light of the literature work on the superconducting iron-based materials. Unlike in
these materials, here we find no evidence of either magnetic ordering, despite the presence of a
Cu-P layer that bears a certain resemblance to the Fe-As or Fe-Se layers in the iron-based
materials. These iron-based materials are near to itinerant antiferromagnetism27, due largely to
the presence of the Fe 3d states at the Fermi level, along with the favorable Fermi surface
topology. In the 265 compound, however, as shown in Fig. 5a, the highest density of Cu states
are generally far from the Fermi level, as in Cu itself; the Cu Fermi level density of states here is
approximately 0.2/Cu-eV, while the calculated Fe Fermi-level DOS in LaOFeAs is
approximately 2/Fe-eV, or ten times higher. Hence there is not sufficient spectral weight near EF
in Ca2Cu6P5 to drive a magnetic instability.
To elaborate on this a little further, we note that Stoner theory predicts a magnetic
instability when the Stoner parameter IN(EF) > 1, where N(EF) is the transition metal Fermi-level
density-of-states (DOS) and I is the exchange correlation integral, as defined in Ref. [28]. For
metallic Cu I is approximately 0.4 eV [28], so that with a Cu Fermi level DOS of 0.2/Cu-eV one
finds a Stoner parameter of just 0.08, so that we are far away from a magnetic instability.
Explicit spin-polarized calculations find neither a ferromagnetic nor nearest-neighbor
antiferromagnetic ground state.
The comparatively low DOS, but substantial plasma frequency, in this material indicates
large Fermi velocities. It is interesting that the Fermi surface structure of Ca2Cu6P5 contains
nearly one dimensional sheets, running vertically and suggestive of nesting, although we find no
indication of a magnetic or structural instability in this compound. It is likely the lack of spectral
weight at EF, as manifested by the low Fermi-level DOS, precludes such a possibility.
Depicted in Fig. 5b is the electronic structure of CaCu2-δP2 that bears some similarity to
that of Ca2Cu6P5; in both cases there is a comparatively low Fermi-level DOS and the majority of
the Cu spectral weight is located well below the Fermi level. The Fermi level DOS in the 122
compound is calculated to be 1.39 states/eV.formula-unit, which translates to a T-linear specific
heat coefficient γ = 3.26 mJ/mol-K2. This is nearly double the experimental value of 1.66

ACCEPTED MANUSCRIPT
mJ/mol-K2, but we note that the electronic DOS in Fig. 5b varies quite rapidly around EF. It
should be noted that our XRD refinements place the occupancy of the Cu site as 0.93, and these
vacancies would tend to move the Fermi level leftward on the graph, to regions of lower DOS
and hence T-linear specific heat coefficient. To check this we constructed a 2x1x1 supercell of
the 122 structure, with one Cu removed, for a total of 4 units of CaCu1.75P2, and calculated its
DOS (black dotted line in Fig.5b DOS plot, inset). It is evident that Cu vacancies hole dope the
system and lower the Fermi-level DOS; one calculates a γ of 1.76 mJ/mol-K2, in much better
agreement with the experimental value. We have also calculated the Fermi surface of CaCu1.75P2
using the virtual crystal approximation, assuming Cu to be divalent. We depict this in Figure 5b
– it has a cage-like structure, along with pockets near the top and bottom one faces. The Cu
vacancies found in the 122 structure may play a role in higher resistivity, with the residual
resistivity being an order of magnitude larger. Nevertheless, the slope dρ/dT of the resistivity at
300 K is approximately three times larger than in 265 compound, which given the similarity in
plasma frequencies suggests the likelihood of stronger electron-phonon coupling. This illustrates
the importance of the number of Cu-P layers on the detailed transport characteristics of these
materials.
4 Conclusions
In summary, we have synthesized two calcium copper-pnictide layered compounds,
Ca2Cu6P5 and CaCu2-δP2 and systematically studied their thermodynamic and transport properties.
Through properties measurement, we find metallicity for both Ca2Cu6P5 and CaCu2-δP2 with no
evidence of either magnetic ordering or superconductivity, despite the presence of a Cu-P layer
that bears a certain resemblance to the Fe-As or Fe-Se layers in the iron-based materials.
Theoretical calculations suggest that, unlike other 3d-based magnetic materials (with a large
DOS at the Fermi surface), Cu-based 122 and 265 have comparatively low Fermi-level density-
of-states with the majority of the Cu spectral weight locating well below. The analyses of the
structure, the DOS, the plasma frequency, and the electron-phonon coupling λ explain the
metallic behavior in the two compounds. Our results indicate the details of structures dictate
characteristics of materials.
Acknowledgement
This work was supported by the U. S. Department of Energy (DOE), Office of Science, Basic
Energy Sciences, Materials Science and Engineering Division (A.S.S.). This study was partially
funded (L.L., D.P.) by ORNL’s Lab-directed Research & Development (LDRD). The electron
microscopy work was performed at the ORNL’s Center for Nanophase Materials Sciences
(CNMS), which is an Office of Science User Facility. We finally acknowledge support from the
Department of Energy - National Nuclear Security Administration under Grant No. DE-
NA0002014 (Y.V., G.M.T.) for pressure measurements

ACCEPTED MANUSCRIPT
Reference
1 Y. Kamihara, T. Watanabe, M. Hirano, and H. Hosono, J. Am. Chem. Soc. 130, 3296 (2008).
2 A.S. Sefat and D.J. Singh, MRS Bull. 36, 614 (2011).
3 D. Johrendt, C. Felser, O. Jepsen, O.K. Andersen, A. Mewis, and J. Rouxel, J. Solid State Chem. 130,
254 (1997).
4 D.J. Singh, A.S. Sefat, M.A. McGuire, B.C. Sales, D. Mandrus, L.H. VanBebber, and V. Keppens, Phys.
Rev. B 79, 094429 (2009).
5 R.E. Baumbach, V.A. Sidorov, X. Lu, N.J. Ghimire, F. Ronning, B.L. Scott, D.J. Williams, E.D. Bauer,
and J.D. Thompson, Phys. Rev. B 89, (2014).
6 Y. Singh, A. Ellern, and D.C. Johnston, Phys. Rev. B 79, 094519 (2009).
7 S. Jia, S. Chi, J.W. Lynn, and R.J. Cava, Phys. Rev. B 81, (2010).
8 V.K. Anand, P.K. Perera, A. Pandey, R.J. Goetsch, A. Kreyssig, and D.C. Johnston, Phys. Rev. B 85,
214523 (2012).
9 B. Saparov and A.S. Sefat, J. Solid State Chem. 191, 213 (2012).
10 S.F. Wu, P. Richard, A. van Roekeghem, S.M. Nie, H. Miao, N. Xu, T. Qian, B. Saparov, Z. Fang, S.
Biermann, A.S. Sefat, and H. Ding, Phys. Rev. B 91, 235109 (2015).
11 T. Mine, H. Yanagi, T. Kamiya, Y. Kamihara, M. Hirano, and H. Hosono, Solid State Commun. 147,
111 (2008).
12 F. Ronning, E.D. Bauer, T. Park, S.-H. Baek, H. Sakai, and J.D. Thompson, Phys. Rev. B 79, (2009).
13 I. Pilchowski and A. Mewis, Z. Für Anorg. Allg. Chem. 581, 173 (1990).
14 Albrecht Mewis und Peter Blumstengel, Z Naturforsch 83b 671-672 1978 (n.d.).
15 Albrecht Mewis, Z Naturforsch 35 B 141-145 1980 (n.d.).
16 V.K. Anand and D.C. Johnston, Phys. Rev. B 86, 214501 (2012).
17 J. Rodríguez-Carvajal, Phys. B Condens. Matter 192, 55 (1993).
18 G. Tsoi, A.K. Stemshorn, Y.K. Vohra, P.M. Wu, F.C. Hsu, Y.L. Huang, M.K. Wu, K.W. Yeh, and S.T.
Weir, J. Phys. Condens. Matter 21, 232201 (2009).
19 S.T. Weir, J. Akella, C. Aracne-Ruddle, Y.K. Vohra, and S.A. Catledge, Appl. Phys. Lett. 77, 3400
(2000).
20 J. Yen and M. Nicol, J. Appl. Phys. 72, 5535 (1992).
2143--P. Blaha, K. Schwarz, G. Madsen, D. Kvasnicka and J. Luitz, WIEN2k, An Augmented Plane Wave
+ Local Orbitals Program for Calculating Crystal Properties (Karlheinz Schwarz, Techn. Universität
Wien, Austria), 2001. ISBN 3-9501031-1-2 (n.d.).
22 J.P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 77, 3865 (1996).
23 G.K.H. Madsen, K. Schwarz, P. Blaha, and D.J. Singh, Phys. Rev. B 68, 125212 (2003).
24 R.J. Hill and C.J. Howard, J. Appl. Crystallogr. 20, 467 (1987).
25 D.R. Lide, CRC Handbook of Chemistry and Physics, 85th Edition (CRC Press, 2004).
26 P.B. Allen, Phys. Rev. B 36, 2920 (1987).
27 D.J. Singh and M.-H. Du, Phys. Rev. Lett. 100, 237003 (2008).
28 J.F. Janak. Phys. Rev. B 16, 255 (1977).

ACCEPTED MANUSCRIPT
Figures & Captions:
Figure 1. (a) Refined powder X-ray diffraction pattern for Ca2Cu6P5 and CaCu2-δP2. Red circles
represent observed data; black and green solid lines represent the calculated intensity and
difference between the observed and calculated intensity. The peak of trace impurity phase are
indicated by the symbols of arrow (CuP2) in Ca2Cu6P5, “*”(CaCu4P2) and “#” (Ca2Cu6P5) in
CaCu2-δP2. (b) STEM images reveal the structure of Ca2Cu6P5 with n = 1 and 2 layers, and
interlayer atoms, by imaging light and heavy elements simultaneously. Also, the alternating
short-long Cu-Cu distances are clearly revealed.

ACCEPTED MANUSCRIPT
Figure 2. Temperature-dependence of resistivity for Ca2Cu6P5 and CaCu2-δP2 up to room
temperature. The inset is the temperature dependence of resistivity for Ca2Cu6P5 measured below
100 K with the application of approximately hydrostatic pressure of 1.4, 3.7, 8.9, and 11.7 Gpa .

ACCEPTED MANUSCRIPT
Figure 3. Temperature dependence of heat capacity for Ca2Cu6P5 and CaCu2-δP2; C/T versus T2
in the low temperatures, along with linear fits are in inset.

ACCEPTED MANUSCRIPT
Figure 4. Temperature dependence of the Seebeck coefficient for Ca2Cu6P5 and CaCu2-δP2 up to
room temperature; thermal conductivity is shown in the inset.

ACCEPTED MANUSCRIPT
Figure 5. The calculated density-of-states of Ca2Cu6P5 (a, left) and CaCu2P2 (b,left), and Fermi
surface of Ca2Cu6P5 (a, right) and CaCu1.75P2 (b, right). Inset of (b) shows the DOS difference
between 122 (solid line) and slightly Cu-deficient cell (dotted line).
(a)
(b)

ACCEPTED MANUSCRIPT
1. A comprehensive study on Cu-based compound Ca2Cu6P5 and compare with CaCu2-δP2.
2. Both materials have layers of edge-sharing CuP4 tetrahedral.
3. Ca2Cu6P5 and CaCu2-δP2 have comparatively low Fermi-level density-of-states.
4. Our results indicate the details of structures dictate characteristics of materials.