Preparation and Instability of Nanocrystalline Cuprous Nitride

Article abstract of DOI:10.1021/acs.inorgchem.5b00679

Low-dimensional cuprous nitride (Cu₃N) was synthesized by nitridation (ammonolysis) of cuprous oxide (Cu₂O) nanocrystals using either ammonia (NH₃) or urea (H₂NCONH₂) as the nitrogen source. The resulting nanocrystalline Cu₃N spontaneously decomposes to nanocrystalline CuO in the presence of both water and oxygen from air at room temperature. Ammonia was produced in 60% chemical yield during Cu₃N decomposition, as measured using the colorimetric indophenol method. Because Cu₃N decomposition requires H₂O and produces substoichiometric amounts of NH₃, we conclude that this reaction proceeds through a complex stoichiometry that involves the concomitant release of both N₂ and NH₃. This is a thermodynamically unfavorable outcome, strongly indicating that H₂O (and thus NH₃ production) facilitate the kinetics of the reaction by lowering the energy barrier for Cu₃N decomposition. The three different Cu₂O, Cu₃N, and CuO nanocrystalline phases were characterized by a combination of optical absorption, powder X-ray diffraction, transmission electron microscopy, and electronic density of states obtained from electronic structure calculations on the bulk solids. The relative ease of interconversion between these interesting and inexpensive materials bears possible implications for catalytic and optoelectronic applications. (Figure Presented).

Full text of DOI:10.1021/acs.inorgchem.5b00679

Article
pubs.acs.org/IC
Preparation and Instability of Nanocrystalline Cuprous Nitride
Malinda D. Reichert, Miles A. White, Michelle J. Thompson, Gordon J. Miller, and Javier Vela*
Department of Chemistry, Iowa State University and Ames Laboratory, Ames, Iowa 50011, United States
S
* Supporting Information
ABSTRACT: Low-dimensional cuprous nitride (Cu3N) was
synthesized by nitridation (ammonolysis) of cuprous oxide
(Cu2O) nanocrystals using either ammonia (NH3) or urea
(H2NCONH2) as the nitrogen source. The resulting nano-
crystalline Cu3N spontaneously decomposes to nanocrystalline
CuO in the presence of both water and oxygen from air at
room temperature. Ammonia was produced in 60% chemical
yield during Cu3N decomposition, as measured using the
colorimetric indophenol method. Because Cu3N decomposition requires H2O and produces substoichiometric amounts of NH3,
we conclude that this reaction proceeds through a complex stoichiometry that involves the concomitant release of both N2 and
NH3. This is a thermodynamically unfavorable outcome, strongly indicating that H2O (and thus NH3 production) facilitate the
kinetics of the reaction by lowering the energy barrier for Cu3N decomposition. The three different Cu2O, Cu3N, and CuO
nanocrystalline phases were characterized by a combination of optical absorption, powder X-ray diffraction, transmission electron
microscopy, and electronic density of states obtained from electronic structure calculations on the bulk solids. The relative ease of
interconversion between these interesting and inexpensive materials bears possible implications for catalytic and optoelectronic
applications.
nanocrystals could also lead to Cu3N. Controlled syntheses
(monodispersity, size, morphology) of Cu2O nanocrystals by
electrodeposition, wet chemical, and solvothermal methods
exist, along with reports on their surface, catalytic, and electrical
properties.20,22,23
An interesting feature of Cu3N is its documented
decomposition upon heating or electron bombardment. For
example, Cu3N thin films convert to CuO nanowire arrays by
annealing in air.24 Cu3N deposited by magnetic sputtering
converts to Cu and N2 under electron bombardment or by
annealing in a vacuum.25,26 The optical properties of Cu3N thin
films made by ion-assisted deposition remained unchanged
upon conducting an aging test at 60 °C and 95% humidity for
15 months, but degraded to copper metal when heated to 300
°C.27 Cu3N nanoparticles were reported to degrade thermally,
forming CuO at temperatures ranging between 200 and 330 °C
under an argon atmosphere.3
In this work, we synthesize Cu3N nanocrystals by nitridation
of Cu2O nanocrystals with either ammonia or urea. We
characterize the structure, optical properties, and morphology
of both oxide and nitride phases using structural, optical, and
computational methods. We also describe the spontaneous
room-temperature decomposition of Cu3N nanocrystals upon
exposure to moisture and air, and used simple thermodynamics
calculations to explain that the observed concomitant release of
NH3 and N2 is due to kinetic factors.
INTRODUCTION
■
Copper(I) nitride (Cu3N) is a metastable semiconductor that
has a low reflectivity and high electrical resistivity.1 It has been
reported to have a narrow experimental band gap of 1.2−1.9
eV2 and features a cubic anti-ReO3 structure.3 Because of these
properties, Cu3N nanocrystals (NCs) are promising materials
for optical storage devices, random access memory chips,
conductive inks, high-speed integrated circuits, microscopic
metal links, photocatalysts, and electrocatalytic hydrogen
reduction devices.1−3
Cu3N NCs can be synthesized by several methods. Cu3N
nanocubes were synthesized by the one-pot decomposition of
copper(II) nitrate in octadecene, octadecylamine, and oleyl-
amine.1,2,4,5 Copper nitride nanocrystals were also produced by
decomposing Cu(OAc)2 with ammonia,3,6−9 a metastable
CuCl2-sodium azide precursor,10 or {[Cu5(CN)6-
(NH3)6(N3)2]}n.11 Cu3N nanopowders were produced by
ammonolysis of CuF2.12 Cu4N and Cu3N2 nanostructures
have been prepared as thin films.13−16
A versatile route for making nitrides and oxynitrides is the
nitridation (ammonolysis) of metal oxides. CuO has been used
as a precursor to Cu3N. Silica-supported CuO nanoparticles
were converted to Cu3N by nitridation with ammonia.17
Nitridation of sol−gel CuO with ammonia produced silica-
supported Cu3N nanocrystals.18 Ammonia is believed to reduce
CuO to a zerovalent (metallic) copper intermediate during
nitride formation.19
Cupric (Cu2+) and cuprous (Cu+) oxide nanocrystals are well
described in the literature. The synthesis, characterization, and
optoelectronic properties of CuO nanostructures have been
reviewed.20,21 Like CuO nanocrystals, nitridation of Cu2O
Received: March 26, 2015
Published: June 19, 2015
© 2015 American Chemical Society
6356
DOI: 10.1021/acs.inorgchem.5b00679
Inorg. Chem. 2015, 54, 6356−6362

Inorganic Chemistry
Article
applied for the structural optimization with a 15 × 15 × 15
Monkhorst-pack k-points grid.34 During structure optimization, atomic
coordinates and cell volumes were allowed to optimize. Total energies
EXPERIMENTAL SECTION
■
Materials. Copper(II) chloride dihydrate (CuCl2·2H2O, 99.0%)
and sodium nitroprusside dihydrate (Na2[Fe(CN)5NO]·2H2O,
99.0%) were purchased from Sigma-Aldrich; ammonium carbonate
((NH4)2CO3, tech.), ascorbic acid (C6H8O6, tech.), and sodium
hydroxide (NaOH, tech.) from Fisher; urea (H2NCONH2, tech.) and
phenol (C6H5OH, tech.) from Alfa Aesar; ammonia gas (NH3,
99.99%) from Airgas; and sodium hypochlorite (NaOCl, 8.25%) from
Clorox. All chemicals were used as received.
were calculated using the tetrahedron method with Blochl35
̈
corrections applied. All VASP calculations treated exchange and
correlation by the Perdew-Burke-Enzerhoff (PBE) method.36 Because
of the highly correlated nature of these compounds, on-site Coulomb
interactions were used (LDA + U). The U value applied was 5.0 eV for
the Cu d states due to this value being successful for previous studies
on strongly correlated Cu compounds.37,38
Synthesis. Cu2O Nanocrystals. Cu2O nanocrystals were made by a
modified literature procedure:28 Briefly, 100 mL of a 0.4 M NaOH
solution was added to 100 mL of a 0.2 M CuCl2 solution. After the
solution was stirred for 5 min, 100 mL of a 0.2 M C6H8O6 was added.
The mixture was stirred for 25 min and centrifuged. The precipitate
was collected, washed three times with distilled water and once with
ethanol, and dried at room temperature (R.T., ca. 21−24 °C) for 12 h.
Cu3N nanocrystals from ammonia. Cu2O nanocrystals (0.15 g) were
placed in an alumina combustion boat and put in a fused silica tube
inside of a tube furnace (Lindberg 55035). The tube was purged with
NH3 (60 mL/min) for 30 min and then heated to 250 °C for 21 h.
Cu3N nanocrystals from urea. Cu2O nanocrystals (0.15 g) were weighed
along with H2NCONH2 (0.126 g) into a Teflon liner inside a 23 mL
steel autoclave. The autoclave was placed in a Thomas Scientific
(5300A25) muffle furnace and heated to 190 °C (10 °C/min ramp
rate, 6 h dwell time). CuO from Cu3N decomposition. Cu3N (∼100 mg,
synthesized using either NH3 or H2NCONH2) was exposed to
deionized water (10 mL) for 15 days. The product was isolated by
centrifugation (4500 rpm, 10 min). To determine the amount of
ammonia released during the Cu3N decomposition reaction, we used
the indophenol method.29,30 Solution A: C6H5OH (5 g, 53 mmol) and
Na2[Fe(CN)5NO]·2H2O (0.025 g, 0.084 mmol) were dissolved in
deionized water (500 mL) using a volumetric flask. Solution B: NaOH
(2.5 g, 62.5 mmol) was dissolved in deionized water, NaOCl (8.25%,
4.2 mL) added, and solution diluted with deionized water to 500 mL
in a volumetric flask. The two solutions were stored in amber bottles at
ca. 8−10 °C using a refrigerator and used within 2 weeks. Seven
different calibration solutions were made using solution A (5 mL) and
a 0.5 mM solution of (NH4)2CO3 in deionized water (0 to 350 μL in
50 μL increments). Solution B (5 mL) was added and the mixture was
stirred at 37 °C for 20 min. Absorbance values at λ = 623 nm were
used to construct the calibration curve. After centrifugation to isolate
the CuO during Cu3N decomposition, the same indophenol method
was used to analyze the supernatant solutions (100 μL) for NH3
content. Values reported are three-run averages.
Structural Characterization. X-ray Diffraction. Powder X-ray
diffraction (XRD) data were measured using Cu Kα radiation on a
Rigaku Ultima U4 difractometer. Sample percent composition was
determined using PowderCell 2.4 (PCW) refined against standard
XRD patterns for Cu2O, Cu3N, and CuO. Single-crystalline domain
sizes were calculated using the Scherrer equation and the typical
uncertainty estimated to be 0.2 nm. Transmission Electron Microscopy.
Transmission electron microscopy (TEM) was conducted on carbon-
coated copper grids using a FEI Tecnai G2 F20 field emission
scanning transmission electron microscope (STEM) at 200 kV (point-
to-point resolution <0.25 nm, line-to-line resolution <0.10 nm).
Particle dimensions were measured manually and/or by using ImageJ.
Size measurements and particle statistics were obtained for at least
>100 particles. Average sizes standard deviations are reported.
Optical Characterization. Absorption spectra were measured
using a photodiode array Agilent 8453 UV−vis spectrophotometer.
Solvent absorption was subtracted from all spectra. Diffuse reflectance
measurements were made using a StellarNet Inc. Black-Comet-SR
spectrometer (200−1080 nm).
RESULTS AND DISCUSSION
■
Synthesis of Cu2O Nanocrystals. On the basis of previous
reports of successful nitridation of CuO, we attempted to
perform a similar reaction using nanocrystalline Cu2O as the
starting material. We first synthesized Cu2O nanocrystals by
reacting copper(II) chloride and sodium hydroxide using a
modified literature procedure.28 This reaction proceeds through
a blue copper-hydroxide intermediate, which upon treatment
with ascorbic acid produces an orange solution of Cu2O. X-ray
diffraction confirmed the formation of cubic copper(I) oxide
(Cu2O) nanocrystals (Figures 1 and 2a). The unit cell of Cu2O
Figure 1. Experimental powder XRD patterns before and after treating
Cu2O nanocrystals with ammonia (NH3) or urea (H2NCONH2).
Reported powder XRD patterns for bulk cubic Cu2O and cubic (anti-
ReO3) Cu3N are shown for comparison.
consists of a body-centered cubic (bcc) arrangement of oxide
ions (O2−) with two-coordinate cuprous ions (Cu+) linking
every second corner with the central oxide (Figure 2a). The
size of the Cu2O nanocrystals, as estimated from XRD peak
widths using the Scherrer equation, is approximately 25 nm.
Figure 3 shows the general appearance and optical absorption
of the Cu2O nanocrystals. Absorption starts at 600 nm for both
the diffuse reflectance of a solid film and the solution phase
absorption spectrum of the Cu2O nanocrystals, which roughly
agree with a reported experimental Cu2O band gap of 2.2 eV
(see calculated vs experimental band gap discussion below).28
TEM shows that the Cu2O nanocrystals have a cubic
morphology, with an average size of 43 10 nm (Figure 4).
Table 1 summarizes these observations.
Nitridation of Cu2O Nanocrystals. We used two different
reagents to nitride nanocrystals of Cu2O to produce Cu3N,
namely, ammonia (NH3) and urea (H2NCONH2) (eqs 1 and
2, respectively, in Scheme 1). XRD shows that, in both cases,
the main product is made of anti-ReO3 Cu3N nanocrystals with
Scherrer sizes of 15 nm (NH3) and 22 nm (urea) (Figure 1 and
Table 1). The cubic, anti-ReO3-type unit cell of Cu3N consists
of a primitive cubic arrangement of nitride ions with two-
coordinate cuprous ions linking every two neighboring nitrides
Density of States (DOS) Calculations. Electronic structure
calculations of CuO, Cu2O, and Cu3N were performed using the
Vienna Ab initio Simulation Package (VASP).31,32 All VASP
calculations were performed using projected augmented-wave
(PAW) pseudopotentials with a cutoff energy of 500 eV and a
convergence energy of 1 × 10−6 eV.33 A conjugated algorithm was
6357
DOI: 10.1021/acs.inorgchem.5b00679
Inorg. Chem. 2015, 54, 6356−6362

Inorganic Chemistry
Article
Figure 2. Crystalline unit cells for cubic (a = 4.27 Å) (cuprite) Cu2O
(a), cubic (a = 3.81 Å) (anti-ReO3) Cu3N (b), and monoclinic (a =
4.68 Å, b = 3.47 Å, c = 5.12 Å, and β = 99.7°) (tenorite) CuO (c).
Figure 3. General appearance (suspended in methanol) (a), solid
phase diffuse reflectance (b), and solution phase electronic absorption
spectra (in methanol) (c) of nanocrystalline Cu2O, Cu3N, and CuO.
Scheme 1
(Figure 2b). The product of nitridation with ammonia contains
some unreacted Cu2O (6%), while the product of nitridation
with urea contains smaller amounts of Cu2O (1.6%) and CuO
(0.4%) (Table 1). Figure 3 shows diffuse reflectance (solid) and
optical absorption (solution) spectra of the Cu3N nanocrystals
made using ammonia as the nitridation agent as a
representative example. These nanocrystals show an absorption
band edge of around 700 nm, which very roughly matches
different reports of experimental Cu3N band gaps between 1.1
eV and 1.9 eV.39 The absorption spectra of the reddish-brown
Cu3N nanocrystals in both solid form and in solution
demonstrate this red shifting (Figure 3b,c). TEM measure-
ments indicate a reagent-dependent change in nanocrystal
morphology upon nitridation. Nitridation with ammonia leads
to aggregated, spheroidal Cu3N nanocrystals with a size of 22
7 nm (Figure 4b and Table 1). Nitridation with urea leads to
aggregated Cu3N nanocrystals with a wider range of
morphologies, including rods, sheets, and clusters with sizes
electronic structure calculations on bulk solids to obtain
electronic DOS curves (Figure 5). Cu2O and Cu3N both
display similar electronic structures. In both species, a band gap
is present at the Fermi level with a magnitude of 0.95 and 0.73
eV for Cu3N and Cu2O, respectively. This value agrees well
with previously reported calculations: Cu2O is similar to prior
calculations that reported a theoretical value of 0.78;40 Cu3N
calculations claimed band gaps that were generally smaller but
ranged in theoretical value from 0.23 to 0.9 eV.41 Our
calculated band gap for Cu3N agrees well with experimental
values, while the Cu2O band gap is much smaller than those
obtained experimentally. An explanation for this discrepancy is
the difference between electronic and optical band gaps in
Cu2O. Previous work has found that the direct transition at Γ is
forbidden, which leads to no appreciable optical absorption
until above 2 eV.42
of 370
160 nm, 310
160 nm, and 220
140 nm,
respectively (Figure 4d and Table 1).
Electronic Structure Calculations. To gain more insight
into the optical properties of the different nanocrystalline
copper phases presented here, we performed first-principles
6358
DOI: 10.1021/acs.inorgchem.5b00679
Inorg. Chem. 2015, 54, 6356−6362

Inorganic Chemistry
Article
Figure 5. Density of state (DOS) curves for Cu2O (a), Cu3N (b), and
CuO (c). Total DOS is depicted in black. Partial DOS (pDOS) are
given in copper (Cu), red (O), and blue (N).
2c). Scherrer particle sizes obtained from XRD are 16 and 12
nm for CuO nanocrystals derived from NH3- and urea-made
Cu3N, respectively. TEM shows that the CuO produced is
made of highly aggregated clusters (100−140 nm) of small
nanocrystals (13 3 nm) (Figure 4c,e and Table 1).
Reported experimental band gap values for CuO lie in the
1.2−1.5 eV range (see below).43,44 The black-colored CuO
produced here has a slightly redder absorption edge (ca. 850
nm) than Cu3N (Figure 5). DOS calculations show that the
electronic structure of CuO is much more complicated than
Cu2O or Cu3N (Figure 5). The most notable feature of the
DOS curve for CuO is the peak located at the Fermi level. In
traditional band theory, this would indicate CuO is metallic.
Along with CoO and NiO, CuO belongs to a class of late metal
oxides known as Mott insulators. In addition to the strong
Coulombic interactions mentioned above, a recent inves-
tigation, focused on elucidating the electronic structure of CuO,
found that spin−orbit coupling (SOC) also plays an important
role.45 Without SOC, the valence band ends 1.17 eV above the
Fermi level, after which there is a small gap of 0.49 eV,
consistent with prior theoretical estimates.41 However,
calculations accounting for SOC found the band gap to start
at the Fermi level and be 1.2 eV in magnitude.45 This value
agrees better with experimental values and supports the claim
that SOC is an important factor for this class of compounds.
To elucidate what caused the decomposition of Cu3N
nanocrystals, we attempted to react them with water under
Figure 4. Representative TEM micrographs of nanocrystalline Cu2O
(a), Cu3N made from NH3 (b) and its decomposition product (c), and
Cu3N made from urea (d) and its decomposition product (e).
Spontaneous Decomposition of Cu3N to CuO. During
the humid summer time, we observed spontaneous decom-
position of the Cu3N nanocrystals over a period of several days.
To track their fate, we characterized the product using powder
XRD, and determined that it was made of nanocrystalline
monoclinic (tenorite) CuO (Figure S1, Scheme 2). The
tenorite CuO unit cell contains tetracoordinate cupric (Cu2+)
ions in an approximately square planar configuration (Figure
Table 1. Synthesis and Instability Studies of Cuprous Nitride Nanocrystals
dimensions (nm)
c
no.
1
precursors
conditions XRD
TEM
XRD phase(s) (% composition)
cubic Cu2O (100)
b
CuCl2(aq) + NaOH +
R.T.,
20 min
25 43 10
a
C6H8O6
2
3
4
5
Cu2O + NH3(g)
250 °C,
15 22
7
cubic (anti-ReO3) Cu3N (94) + cubic Cu2O (6)
21 h
Cu2O + H2NCONH2
(urea)
190 °C,
22 370 160 (rods) 310 160 (sheets)
220 140 (aggregates)
cubic (anti-ReO3) Cu3N (98) + cubic Cu2O (1.6) +
monoclinic (tenorite) CuO (0.4)
6 h
b
Cu3N (from NH3) + H2O R.T., 15
+ O2 (air)
16 13 3 (dots) 140 60 (aggregates)
monoclinic (tenorite) CuO (99.1) + cubic Cu2O (0.7)
+ cubic (anti-ReO3) Cu3N (0.2)
days
b
Cu3N (from urea) + H2O R.T., 15
+ O2 (air)
12 100 70 (aggregates)
CuO monoclinic (100)
days
a
b
c
Ascorbic acid. 21−24 °C. Sample percent composition was determined using PowderCell 2.4 (PCW) refined against standard XRD patterns for
Cu2O, Cu3N, and CuO.
6359
DOI: 10.1021/acs.inorgchem.5b00679
Inorg. Chem. 2015, 54, 6356−6362

Inorganic Chemistry
Article
argon (oxygen-free), with dry oxygen (water-free), or with
water and oxygen together as control experiments (eqs 3, 4,
and 5, respectively, in Scheme 2). As shown in Figure 6,
Scheme 3. Colorimetric Indophenol Quantification of NH3
Scheme 2
Table 2. Colorimetric Determination of NH3 Produced from
Cu3N Decomposition Using the Indophenol Method
a
Cu3N
NH3
a
b
c
moles
moles
% yield
63.7 0.9
56
Cu3N (NH3)
Cu3N (urea)
5.13 0.07 × 10−5
5.33 0.05 × 10−5
3.27 0.04 × 10−5
2.98 0.04 × 10−5
1
a
Decomposition reactions were carried out in 10 mL of deionized
b
water. Small, 100 μL aliquots were further diluted with 10 mL of
deionized water to produce the solutions in Figure S2. Based on the
c
total amount of nitrogen (from Cu3N).
Thermodynamic Rationale of CuO Formation. In
principle, formation of N2 or NH3 from Cu3N decomposition
(oxidation to CuO) could be rationalized by either of the
balanced eqs 7 or 8 (Scheme 4). The enthalpies of formation
Scheme 4
Figure 6. Experimental powder XRD patterns obtained before and
after treating Cu3N nanocrystals (NCs) with H2O under Ar (O2-free),
dry O2 (H2O-free), and H2O plus O2. Reported powder XRD patterns
for bulk cubic Cu2O, cubic (anti-ReO3) Cu3N, and monoclinic
(tenorite) CuO are shown for comparison (all reactions were
performed at R.T. for 15 days).
(ΔHf) of all reactants and products in these equations are
known, including those of Cu3N (ΔHf = 77 kJ/mol)48,49 and
CuO (ΔHf = −157.3 kJ/mol)50 (near standard conditions).
Using these values, the enthalpies of reactions (ΔHr) (7) and
(8) are calculated to be −549 kJ/mol and −197 kJ/mol,
respectively. Therefore, decomposition of Cu3N exclusively to
N2 vs NH3 (reaction 7 vs 8) is favored enthalpically by ca. 350
kJ/mol (Figure 7). In other words, there is an inverse
relationship between the enthalpy of the Cu3N decomposition
treating Cu3N nanocrystals with water under an oxygen-free Ar
atmosphere at room temperature (R.T.) for 15 days did not
result in any observable change. The same was true when dry
(moisture-free) O2 was used. In contrast, Cu3N completely
transformed to nanocrystalline CuO in the presence of both
water and O2 (Figure 6). This reproducible result is
independent of the original nitridation source (NH3 or urea)
that was employed in making Cu3N with very different sizes
and morphologies. This result is also consistent with our
observation that Cu3N decomposition occurred faster during
the humid summer months than during the dry winter months.
Tracing the Fate of Nitrogen after Cu3N Decom-
position. Two nonmutually exclusive, possible nitrogen-
containing products of Cu3N decomposition are ammonia
(NH3) and molecular nitrogen (N2). We first suspected that
NH3 was produced after observing an increase in the pH of the
original solution exposed to the Cu3N nanocrystals from
slightly acidic or neutral before reaction, to basic after the
decomposition of Cu3N to CuO was complete. To further
examine this issue, we sought to confirm whether NH3 was
produced and, if so, to measure exactly how much NH3 was
produced, using the colorimetric indophenol method (eq 6 in
Scheme 3 and Figure S2).46,47 As shown in Table 2, similar
amounts of ammonia were released upon Cu3N decomposition
regardless of the source of nitrogen originally used for
nitridation (NH3 or urea). In all cases, the experimentally
measured chemical yield of NH3 remained constant and equal
to 60% relative to the initial amount of Cu3N.
Figure 7. Change in enthalpy (ΔHr) and free energy (ΔGr) of Cu3N
decomposition to CuO as a function of NH3 chemical yield (percent =
100x). The circled cross marker in red represents the NH3 yield (x)
observed in this study.
6360
DOI: 10.1021/acs.inorgchem.5b00679
Inorg. Chem. 2015, 54, 6356−6362

Inorganic Chemistry
Article
and the amount of N2 produced by this reaction (relative to
NH3). Unfortunately, we are presently unable to exactly
calculate the free energies (ΔGr) of reactions (7) and (8),
because the free energy of formation (ΔGf) of Cu3N remains
unknown.
In spite of the large enthalpic preference for exclusive N2
production, this was not observed experimentally. On the basis
of our observations that H2O is required for Cu3N
decomposition to occur and that NH3 is produced in 60%
yield based on the initial amount of Cu3N (x = 0.6, Table 2),
we conclude that this reaction proceeds through a complex
stoichiometry that involves the concomitant release of both N2
and NH3, according to eq 9. We are able to calculate the true
thermodynamic cost of ammonia production by considering the
difference in Gibb’s free energies between reactions 9 and 7
(ΔΔG9−7), as follows:
as measured using the colorimetric indophenol method.
Because Cu3N decomposition requires H2O and produces
substoichiometric amounts of NH3, we conclude that this
reaction proceeds through a complex stoichiometry that
involves the concomitant release of both N2 and NH3. This
is a thermodynamically unfavorable outcome, strongly indicat-
ing that H2O (and thus NH3 production) facilitate the kinetics
of the reaction by lowering the energy barrier for Cu3N
decomposition. Additional studies will be required to assess the
potential reversibility and catalytic usefulness of this trans-
formation.
ASSOCIATED CONTENT
■
S
* Supporting Information
Additional XRD, selected area electron-diffraction (SAED)
analyses, and solution-phase optical calibration data. The
Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.5b00679.
f
f
+
−
ΔΔG9−7 = ΔG9 − ΔG7 = xΔG
+ xΔGHO (aq)
NH4 aq
(
)
5x
2
x
2
3x
4
−
ΔGHf O(l)
+
ΔGNf (g)
−
ΔGOf (g)
2
2
2
In this equation, 3ΔGfCuO(s) and ΔGfCu N(s) have been
AUTHOR INFORMATION
■
3
canceled out. In addition, by definition both ΔGfN (g) and
Corresponding Author
*E-mail: vela@iastate.edu.
2
ΔGfO (g) are zero under ideal or near-ideal conditions (25 °C, 1
2
Notes
atm). This yields
The authors declare no competing financial interest.
5x
2
f
f
f
+
−
ΔΔG9−7 = xΔG
+ xΔGHO (aq)
−
ΔG
NH4 aq
H2O(l)
(
)
ACKNOWLEDGMENTS
■
= − x(79 kJ/mol) − x(157 kJ/mol)
+ (5x/2)(237 kJ/mol)47 = x(358 kJ/mol)
J.V. gratefully acknowledges the National Science Foundation
for funding of this work through the Division of Materials
Research, Solid State and Materials Chemistry program (NSF-
DMR-1309510). M.A.W. thanks Yuemei Zhang for assistance
with calculations.
Figure 7 shows a plot of ΔΔG9−7 as a function of chemical
yield of NH3 based on the initial amount of Cu3N. The
experimentally observed value of 60% is 72 kJ/mol less
favorable by free energy compared to when (if) no ammonia is
produced. This strongly indicates that the role of H2O (and
thus NH3 production) is to facilitate the kinetics of the reaction
by lowering the energy barrier for Cu3N decomposition to
CuO.
REFERENCES
■
(1) Wang, D.; Li, Y. Chem. Commun. 2011, 47, 3604−3606.
(2) Xi, P.; Xu, Z.; Gao, D.; Chen, F.; Xue, D.; Tao, C.-L.; Chen, Z.-N.
RSC Adv. 2014, 4, 14206−14209.
(3) Nakamura, T.; Hayashi, H.; Hanaoka, T.-A.; Ebina, T. Inorg.
Chem. 2013, 53, 710−715.
CONCLUSION
■
(4) Wu, H.; Chen, W. J. Am. Chem. Soc. 2011, 133, 15236−15239.
(5) In, S.-I.; Vaughn, D. D., II; Schaak, R. E. Angew. Chem., Int. Ed.
2012, 51, 3915−3918.
In summary, nanocrystalline Cu3N was synthesized from
nanocrystalline Cu2O and one of two nitrogen sources,
ammonia (NH3) or urea (H2NCONH2). Using urea leads to
slightly more phase-pure Cu3N, as observed by XRD. In
addition, the use of ammonia as the nitridation (ammonolysis)
reagent requires a longer reaction time and higher temperature.
TEM reveals that the cubic morphology of the initial Cu2O
nanocrystals is lost upon nitridation to Cu3N. Nitridation with
ammonia produces clusters of spherical particles whereas
nitridation with urea produced a mixture of several different
morphologies (rods, sheets, and aggregates). Electronic
structure calculations of the DOS appear to underestimate,
but agree with, the presence of a band gap of under 1 eV for
both Cu2O and Cu3N phases.
Cu3N nanocrystals spontaneously decompose to nanocrystal-
line CuO at room temperature. Control experiments show that
both air (O2, an oxidant) and moisture (H2O, for hydrolysis to
NH3) are necessary for this transformation to take place. Cu3N
decomposition occurs within 15 days to produce substoichio-
metric amounts of ammonia. Specifically, ammonia was
produced in 60% chemical yield during Cu3N decomposition,
(6) Lee, B. S.; Yi, M.; Chu, S. Y.; Lee, J. Y.; Kwon, H. R.; Lee, K. R.;
Kang, D.; Kim, W. S.; Lim, H. B.; Lee, J.; Youn, H.-J.; Chi, D. Y.; Hur,
N. H. Chem. Commun. 2010, 46, 3935−3937.
(7) Desmoulins-Krawiec, S.; Aymonier, C.; Loppinet-Serani, A.;
Weill, F.; Gorsse, S.; Etourneau, J.; Cansell, F. J. Mater. Chem. 2004,
14, 228−232.
(8) Cansell, F.; Aymonier, C.; Loppinet-Serani, A. Curr. Opin. Solid
State Mater. Sci. 2003, 7, 331−340.
(9) Cansell, F.; Chevalier, B.; Demourgeues, A.; Etourneau, J.; Even,
C.; Garrabos, Y.; Pessey, V.; Petit, S.; Tressaud, A.; Weill, F. J. Mater.
Chem. 1999, 9, 67−75.
(10) Choi, J.; Gillan, E. G. Inorg. Chem. 2005, 44, 7385−7393.
(11) Trivedi, M.; Singh, G.; Kumar, A.; Rath, N. P. RSC Adv. 2014, 4,
34110−34116.
(12) Paniconi, G.; Stoeva, Z.; Doberstein, H.; Smith, R. I.; Gallagher,
B. L.; Gregory, D. H. Solid State Sci. 2007, 9, 907−913.
(13) Gordillo, N.; Gonzalez-Arrabal, R.; Martin-Gonzalez, M. S.;
Olivares, J.; Rivera, A.; Briones, F.; Agullo-
Cryst. Growth 2008, 310, 4362−4367.
(14) Hadian, F.; Rahmati, A.; Movla, H.; Khaksar, M. Vacuum 2012,
86, 1067−1072.
́ ́
Lopez, F.; Boerma, D. O. J.
6361
DOI: 10.1021/acs.inorgchem.5b00679
Inorg. Chem. 2015, 54, 6356−6362

Inorganic Chemistry
Article
(15) Gordillo, N.; Rivera, A.; Grotzschel, R.; Munnik, F.; Guttler, D.;
̈
̈
́ ́
Crespillo, M. L.; Agullo-Lopez, F.; Gonzalez-Arrabal, R. J. Phys. D:
Appl. Phys. 2010, 43, 345301−1−9.
(16) Gonzalez-Arrabal, R.; Gordillo, N.; Martin-Gonzalez, M. S.;
Ruiz-Bustos, R.; Agullo-
7.
́ ́
Lopez, F. J. Appl. Phys. 2010, 107, 103513−1−
(17) Deshmukh, R.; Schubert, U. Eur. J. Inorg. Chem. 2013, 14,
2498−2504.
(18) Deshmukh, R.; Schubert, U. J. Mater. Chem. 2011, 21, 18534−
18536.
(19) Baiker, A.; Maciejewski, M. J. Chem. Soc., Faraday Trans. 1 1984,
80, 2331−2341.
̌
(20) Filipic, G.; Cvelbar, U. Nanotechnology 2012, 23, 194001−1−16.
(21) Zhang, Q.; Zhang, K.; Xu, D.; Yang, G.; Huang, H.; Nie, F.; Liu,
C.; Yang, S. Prog. Mater. Sci. 2014, 60, 208−337.
(22) Kuo, C.-H.; Huang, M. H. Nano Today 2010, 5, 106−116.
(23) Lignier, P.; Bellabarba, R.; Tooze, R. P. Chem. Soc. Rev. 2012, 41,
1708−1720.
(24) Fan, X. Y.; Wu, Z. G.; Yan, P. X.; Geng, B. S.; Li, H. J.; Li, C.;
Zhang, P. J. Mater. Lett. 2008, 62, 1805−1808.
(25) Lesch, N.; Karduck, P.; Cremer, R.; Richthofen, A. V. Fresenius J.
Anal. Chem. 1998, 361, 604−607.
(26) Yue, G. H.; Yan, P. X.; Wang, J. J. Cryst. Growth 2005, 274,
464−468.
(27) Asano, M.; Umeda, K.; Tasaki, A. Jpn. J. Appl. Phys. 1990, 29,
1985−1986.
(28) Wang, Z.; Wang, H.; Wang, L.; Pan, L. Cryst. Res. Technol. 2009,
44, 624−628.
(29) Weatherburn, M. W. Anal. Chem. 1967, 39, 971−974.
(30) Searle, P. L. Analyst 1984, 109, 549−568.
(31) Kresse, G.; Furthmuller, J. Comput. Mater. Sci. 1996, 6, 15−50.
̈
(32) Kresse, G. Phys. Rev. B 1996, 54, 11169−11186.
(33) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758−1775.
(34) Monkhorst, H. J.; Pack, J. D. Phys. Rev. B 1976, 13, 5188−5192.
(35) Blochl, P. E. Phys. Rev. B 1994, 50, 17953−17979.
̈
(36) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77,
3865−3868.
(37) Scanlon, D.; Walsh, A.; Morgan, B.; Watson, G.; Payne, D.;
Egdell, R. Phys. Rev. B 2009, 79, 035101−1−7.
(38) Raebiger, H.; Lany, S.; Zunger, A. Phys. Rev. B 2007, 76,
045209−1−5.
(39) Xiao, J.; Li, Y.; Jiang, A. J. Mater. Sci. Technol. 2011, 27, 403−
407.
(40) Ching, W.; Xu, Y.-N.; Wong, K. Phys. Rev. B 1989, 40, 7684−
7695.
(41) Hahn, U.; Weber, W. Phys. Rev. B 1996, 53, 12684−12693.
(42) Elliott, R. Phys. Rev. 1957, 108, 1384−1389.
(43) Marabelli, F.; Parravicini, G. B.; Salghetti-Drioli, F. Phys. Rev. B
1995, 52, 1433−1436.
(44) Serin, N.; Serin, T.; Horzum, S
2005, 20, 398−401.
̧.; Çelik, Y. Semicond. Sci. Technol.
(45) Ekuma, C. E.; Anisimov, V. I.; Moreno, J.; Jarrell, M. Eur. Phys. J.
B 2014, 87, 23−1−6.
(46) Vela, J.; Stoian, S.; Flaschenriem, C. J.; Munck, E.; Holland, P. L.
̈
J. Am. Chem. Soc. 2004, 126, 4522−4523.
(47) Yandulov, D. V.; Schrock, R. R. Science 2003, 301, 76−78.
(48) Elder, S. H.; DiSalvo, F. J.; Topor, L.; Navrotsky, A. Chem.
Mater. 1993, 5, 1545−1553.
(49) Caskey, C. M.; Richards, R. M.; Ginleya, D. S.; Zakutayeva, A.
Mater. Horiz. 2014, 1, 424−430.
(50) CRC Handbook of Chemistry and Physics, 95th ed.; CRC: Boca
Raton, FL, 2013.
6362
DOI: 10.1021/acs.inorgchem.5b00679
Inorg. Chem. 2015, 54, 6356−6362