From Linear Molecular Chains to Extended Polycyclic Networks: Polymerization of Dicyanoacetylene

Article abstract of DOI:10.1021/acs.chemmater.7b01446

Dicyanoacetylene (C₄N₂) is an unusual energetic molecule with alternating triple and single bonds (think miniature, nitrogen-capped carbyne), which represents an interesting starting point for the transformation into extended carbon-nitrogen solids. While pressure-induced polymerization has been documented for a wide variety of related molecular solids, precise mechanistic details of reaction pathways are often poorly understood and the characterization of recovered products is typically incomplete. Here, we study the high-pressure behavior of C₄N₂ and demonstrate polymerization into a disordered carbon-nitrogen network that is recoverable to ambient conditions. The reaction proceeds via activation of linear molecules into buckled molecular chains, which spontaneously assemble into a polycyclic network that lacks long-range order. The recovered product was characterized using a variety of optical spectroscopies, X-ray methods, and theoretical simulations and is described as a predominately sp² network comprising "pyrrolic" and "pyridinic" rings with an overall tendency toward a two-dimensional structure. This understanding offers valuable mechanistic insights into design guidelines for next-generation carbon nitride materials with unique structures and compositions.

Full text of DOI:10.1021/acs.chemmater.7b01446

Subscriber access provided by NEW YORK UNIV
Article
From Linear Molecular Chains to Extended Polycyclic
Networks: Polymerization of Dicyanoacetylene
Huiyang Gou, LI Zhu, Haw-Tyng Huang, Arani Biswas, Derek W. Keefer, Brian L Chaloux,
Clemens Prescher, Liuxiang Yang, Duck Young Kim, Matthew D Ward, Jordan Lerach,
Shengnan Wang, Artem R Oganov, Albert Epshteyn, John V Badding, and Timothy A. Strobel
Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/acs.chemmater.7b01446 • Publication Date (Web): 26 Jun 2017
Downloaded from http://pubs.acs.org on June 26, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Chemistry of Materials is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
From Linear Molecular Chains to Extended Polycyclic Networks:
Polymerization of Dicyanoacetylene
1
,2,*,†
1,*
3
3
3
Huiyang Gou,
Li Zhu , HawꢀTyng Huang, Arani Biswas, Derek W. Keefer, Brian L.
4
5
1,2
1,2
1
Chaloux, Clemens Prescher, Liuxiang Yang, Duck Young Kim, Matthew D. Ward, Jordan
Lerach, Shengnan Wang, Artem R. Oganov, Albert Epshteyn, John V. Badding, Timothy
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
6
7,6
8
3
1
,‡
A. Strobel
1
Geophysical Laboratory, Carnegie Institution of Washington, Washington, DC 20015, USA
Center for High Pressure Science and Technology Advanced Research, Beijing 100094, China.
Department of Chemistry, Department of Physics, Department of Materials Science and Engineering,
and Materials Research Institute, The Pennsylvania State University, University Park, PA 16802, USA.
NRC Postdoctoral Associate, Naval Research Laboratory, 4555 Overlook Ave., SW, Washington, DC
2
3
4
20375, USA
5
Center for Advanced Radiation Sources, University of Chicago, Argonne, IL 60437, USA
Department of Geosciences, Center for Materials by Design, and Institute for Advanced Computational
Science, State University of New York, Stony Brook, NY 11794ꢀ2100
Skolkovo Institute of Science and Technology, Skolkovo Innovation Center, 3 Nobel St., Moscow
43026, Russia.
Naval Research Laboratory, Washington, DC20375, USA
6
7
1
8
*
†
‡
Contributed equally to this work
huiyang.gou@hpstar.ac.cn
tstrobel@ciw.edu
ABSTRACT
Dicyanoacetylene (C N ) is an unusual energetic molecule with alternating triple and single
4
2
bonds (think miniature, nitrogenꢀcapped carbyne), which represents an interesting starting point
for the transformation into extended carbonꢀnitrogen solids. While pressureꢀinduced
polymerization has been documented for a wide variety of related molecular solids, precise
mechanistic details of reaction pathways are often poorly understood and the characterization of
recovered products is typically incomplete. Here, we study the highꢀpressure behavior of C N
4
2
and demonstrate polymerization into a disordered carbonꢀnitrogen network that is recoverable to
ambient conditions. The reaction proceeds via activation of linear molecules into buckled
molecular chains, which spontaneously assemble into a polycyclic network that lacks longꢀrange
order. The recovered product was characterized using a variety of optical spectroscopies, Xꢀray
2
methods and theoretical simulations, and is described as a predominately sp network comprising
“
pyrrolic” and “pyridinic” rings with an overall tendency towards a twoꢀdimensional structure.
This understanding offers valuable mechanistic insights into design guidelines for nextꢀ
generation carbon nitride materials with unique structures and compositions.
1
ACS Paragon Plus Environment

Chemistry of Materials
Page 2 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
Carbon nitride materials have attracted continuous attention in research over past decades as
1
ꢀ2
3ꢀ19
indicated by extensive efforts from both theory and experiments . Sustained interest in
nitrogenꢀbearing, carbonꢀrich materials stems from their multiple functionalities and diverse
1
ꢀ19
applications.
For example, cubic carbon nitride (cꢀC N ) is predicted to have comparable
3 4
1
ꢀ2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
hardness to diamond, whereas graphitic carbon nitride (gꢀC N ) has been demonstrated to act
as a photocatalyst for various redox reactions.
3
4
4
ꢀ19
Furthermore, nitrogenꢀdoped carbon materials
often exhibit excellent properties as compared with their pure carbon counterparts, e.g.
conductivity, basicity, oxidation stability, and catalytic activity, although these properties are
4
ꢀ19
highly dependent on the amount of nitrogen incorporated within the structure.
When longꢀ
2
range order within carbon nitride materials is lacking, disordered structures containing sp and
3
20ꢀ26
sp bonding may still exhibit outstanding mechanical properties,
which is fundamentally
important for the realization of threeꢀdimensional, crystalline carbon nitride solids.
Previous attempts to produce novel CN solids have largely revolved around varying deposition
1
0ꢀ17
methods to manipulate particle size and texture, together with control of nitrogen content.
Nevertheless, adventitious hydrogen incorporation within these materials limits the development
2
7ꢀ
of structureꢀproperty relationships for precise stoichiometries and diverse CN building blocks.
3
2
Highꢀenergy molecules can provide an alternative path towards the formation of unique
structures, and highꢀpressure chemistry may be a viable approach to obtain novel extended CN
3
3
networks with predetermined architectures. Studies of the polymerization reactions of
3
4
35
36
37
38
cyanogen, cyanoacetylene, tetracyanoethylene, phosphorous tricyanide, and acetonitrile
under pressure have helped to clarify the influences of starting materials and reaction conditions
for the synthesis of new CN materials.
Dicyanoacetylene (C N ) is a linear molecule with alternating triple and single bonds (N≡Cꢀ
4
2
3
9
40
C≡CꢀC≡N). The standard enthalpy of formation of C N is 500 kJ/mol, leading to a flame
temperature of over 5000 K. Thus, C N represents an interesting, highꢀenergy starting point to
4
2
4
1
4
2
access novel extended CN networks. From the lowꢀpressure side, people have examined C N2
4
and found reactions with benzene to produce aromatic hydrocarbons via the DielsꢀAlder reaction,
42
in addition to explosive decomposition into dinitrogen and graphite. Chien and Carlini
synthesized polydicyanoacetylene via anionic polymerization of C N with nꢀbutyllithium in
4
2
4
2
THF, producing a linear polymer that cyclizes into a ladder polymer upon heating at 400 °C.
Interestingly, the physiochemical behavior of C N is also of astrophysical importance due to its
4
4ꢀ47
2
4
presence in Titan’s north polar stratosphere,
as detected by infrared spectroscopy. However,
the interpretation of Raman and IR spectra is challenging, especially with regards to conflicting
48ꢀ54
assignments for C≡N and C≡C stretching frequencies.
Inconsistencies between spectroscopic
assignments for C N may arise from difficulties associated with sample preparation and
4
2
handling due to its inherent instability and high reactivity; although it is of great importance for
the understanding of intrinsic properties and conclusions regarding its detection.
To our knowledge, C N has not yet been examined under highꢀpressure conditions, which is of
4
2
great interest from the perspective of novel carbon nitride materials and fundamental chemical
2
ACS Paragon Plus Environment

Page 3 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
transformations. The large fraction of unsaturated bonds suggests potential for pressureꢀinduced
ringꢀforming reactions, as has been documented for a large number of other nitrileꢀ and alkyneꢀ
3
4ꢀ38
based systems.
In addition, C N represents an interesting C:N stoichiometry to access
4 2
extended networks with nitrogen content intermediate between typical depositionꢀbased
syntheses (nitrogen doping within a carbon framework) and precursorꢀbased systems like
graphitic C N . Moreover, highꢀpressure behavior can help to understand the nature of chemical
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
4
bonding, reaction mechanisms, atomic structure and local environments of reaction products,
which will advance the general understanding of carbon nitride materials.
Here, we study the highꢀpressure behavior of linear dicyanoacetylene (C N ) up to 10 GPa in
4
2
diamond anvil cells. We first resolve the standing controversy regarding assignments of Raman
and IR spectra, aided by firstꢀprinciples phonon spectrum calculations. We next uncover a
pressureꢀinduced reaction process whereby discrete linear molecules polymerize into a
disordered extended network without significant change to the bulk composition. This novel,
amorphous material is fully recoverable to ambient conditions and its local structure,
composition and chemical bonding was established using a variety of optical and Xꢀray
scattering methods. The reaction mechanism was rationalized using molecular dynamics
simulations. The present results provide a framework to further the understanding of the local
mechanism of polymerization and offer valuable mechanistic insights into design guideline for
the nextꢀgeneration carbon nitride materials.
METHODS
Synthesis
5
5
C N was prepared according to literature procedures. All reactions were performed under an
4
2
inert atmosphere of either nitrogen or argon. Diethyl ether was dried and deoxygenated over
sodiumꢀbenzophenone ketyl prior to use. Sulfolane was purified by reaction with KMnO to
4
remove alkene impurities, dried over KOH, and vacuum distilled prior to use.
Acetylene dicarboxamide (1). 100 mL of NH (4.8 mol) were condensed into a Schlenk flask
3
and chilled to ꢀ78 °C. Separately, 13.41 g (94.36 mmol) of dimethyl acetylenedicarboxylate
(
SigmaꢀAldrich, 99%) were dissolved in 100 mL anhydrous diethyl ether (Et O) and transferred
2
dropwise to the chilled flask containing NH ; the reaction mixture quickly developed a red
3
coloration. The reaction mixture was warmed to ꢀ45 °C and stirred for 16 h at this temperature.
Residual NH was boiled off at ambient temperature and the remainder was evaporated to
3
dryness under vacuum, yielding a crude, orange product that was subsequently washed with cold
ethanol. The resulting beige powder was recrystallized from 150 mL boiling ethanol at ꢀ20 °C,
yielding 8.27 g (78% yield) of white, crystalline solid after filtering and drying the powder under
vacuum.
Dicyanoacetylene (2). 125 mL anhydrous sulfolane was added to a Schlenk flask equipped with
a solids addition funnel and a secondary (detachable) vacuum trap. Using a mortar and pestle,
3
ACS Paragon Plus Environment

Chemistry of Materials
Page 4 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
.960 g (17.49 mmol) of acetylene dicarboxamide (1) was ground with 15.456 g (54.06 mmol)
P O until homogeneous. Solids were added to the addition funnel, the apparatus was evacuated,
2
5
the trap was chilled to ꢀ196 °C, and the sulfolane was heated to 110 °C with vigorous stirring.
Solids were added slowly to the reaction mixture, which changed from colorless to yellow to red
over time. After bubbling had ceased (~3 h), the trap containing the crude product was removed,
transferred to an argonꢀfilled glovebox, and liquid was pipetted into a small Hꢀtube after melting.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The liquid was frozen with LN , a static vacuum of 50 mTorr was applied on the assembly. The
2
dewar was transferred to the empty arm of the Hꢀtube and the product was sublimed by warming
the arm. The volatile, pale beige (almost colorless) liquid product was weighed within the
glovebox at 0.794 g (60% yield). The product was stored at ꢀ35 °C under argon to prevent
decomposition.
Poly-C N . Liquid dicyanoacetylene (2) (~1 µl) was loaded into a ~150ꢀ210 ꢁm diameter hole in
4
2
a Re gasket using clean microsyringe. Before sample loading, Re gaskets were preꢀindented to a
thickness of 60ꢀ80 ꢁm and mounted on 300ꢀ500 µm anvil culets of symmetric diamond anvil
cells (DAC). Pressure was determined by measurement of fluorescence from a ruby standard
5
6
placed inside the gasket hole. To avoid any possible contamination of the highly reactive C N ,
4
2
a pressure medium was not used in the runs, but the molecular crystals are very soft and the
influence of deviatoric stresses is expected to be small. All sample loadings were conducted in
an inert Ar gas atmosphere glove box with oxygen and moisture concentrations of less than 1
ppm. All samples were sealed to a starting pressure of ~0.1 GPa within the inert Ar atmosphere
before removing from the glove box.
Raman spectroscopy
A system based around a Princeton Instruments spectrograph SP2750 (Trenton, NJ) with a 750
mm focal length was used for the Raman spectra collections. A 532 nm diode laser was used as
an excitation source and was focused through a 20× long working distance objective lens. The
laser power was optimized to be ~1 mW and a short exposure time of ≤60 s was used to avoid
local heating of the sample. Raman light was collected in the backscatter geometry through a 50
ꢁ
m confocal pinhole and two narrowꢀband notch filters (Ondax) were used to allow collection to
−1
within ~10 cm from the laser line. Raman light was collected though a 50 ꢁm slit and dispersed
off of an 1800 or 300 gr/mm grating onto a liquid nitrogen cooled charge coupled device (CCD)
detector providing a spectral resolution of about 2 cm . The spectrometer was calibrated using
−
1
−
1
the emission lines of Ne with an accuracy <1 cm .
Infrared spectroscopy
Infrared transmission spectra were collected using a Varian 670ꢀIR spectrometer system (Santa
−
1
Clara, CA) utilizing a Globar source and KBr mid/nearꢀIR beam splitter ca. 500ꢀ4000 cm .
Infrared light was focused on and recollected through the sample using reflecting objective
lenses. Transmitted light was collimated through a 50 ꢁm pinhole for increased spatial resolution
4
ACS Paragon Plus Environment

Page 5 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
and passed to an HgCdTe detector. IR spectra were recorded at varying pressure. The same
diamonds were used to obtain reference spectra after the cell was decompressed and cleaned.
X-ray diffraction
Powder Xꢀray diffraction (XRD) measurements up to 13.0 GPa were collected in ~0.2 GPa
pressure increments at the High Pressure Collaborative Access Team (HPCAT), beamline 16ꢀ
IDB, of the Advanced Photon Source (APS), Argonne National Laboratory. A monochromatic
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
beam (~4 × 5 ꢁm ) with λ = 0.406626 Å was focused on the sample, and data were recorded
using a MAR image plate. Diffraction images were processed using the FIT2D data analysis
5
7
program. Pressure was calibrated by the fluorescence from a Ruby sphere placed inside of the
56
sample chamber. The unitꢀcell volume and lattice parameters of C N were obtained through
4
2
5
8
59ꢀ60
full profile fitting using the Le Bail method, as implemented in GSAS with EXPGUI.
Singleꢀcrystal Xꢀray diffraction studies for C N were carried out at the Advanced Photon Source,
4
2
2
sector 16ꢀBMD. Monochromatic Xꢀrays (λ = 0.30998 Å) with a beam size of (~15 × 30 ꢁm )
were focused on the sample and roomꢀtemperature diffraction was collected using a MAR image
plate. The resulting data were processed and indexed using GSEꢀADA/RSV. Structure
solutions and refinements were completed with the use of the SHELXꢀ14 suite of programs.
Atom positions were standardized using the program STRUCTURE TIDY.
61
62ꢀ63
6
4ꢀ65
Crystallographic
images were made using the program CRYSTALMAKER. Further crystallographic details can
be found in the Supporting Information (SI) Table S1.
The structure factor of recovered amorphous C N was measured at the GSECARS 13ꢀIDD
4
2
beamline, Advanced Photon Source, ANL. The sample was measured with two different incident
monochromatic Xꢀray beam energies: 25 keV (λ = 0.4959 Å) for the lowꢀQ region and 60 keV (λ
2
2
= 0.2066 Å) for the highꢀQ region. The beam sizes were 2×3 ꢀm and 4×5 ꢀm , respectively.
Diffraction data were collected with a MAR345 image plate, whereby its position was calibrated
using a LaB standard. Detector calibration, image integration and intensity correction for
6
6
6
oblique Xꢀrayꢀtoꢀdetector angle was performed using the Dioptas software package. An empty
Reꢀgasket with same thickness as the recovered sample was measured as background. The
resulting data were stitched together from both measurements. The structure factor, S(Q), and
pair distribution function g(r) were calculated in the FaberꢀZiman formalism. Data were
ꢀ1
collected up to a Q
of 17 Å . The resulting data were optimized using a Kaplowꢀtype
max
6
7
correction with three iterations. A Lorch modification function was used to remove unphysical
oscillation in the resulting g(r), which are a result of cutoff effects in the Fourier transformation.
68
Density of the recovered sample was estimated by a method described by Eggert et al.
Scanning Electron Microscopy
The recovered samples within the Re gasket after Raman or Infrared spectral measurements were
mounted on an Al rod holder for chemical composition mapping and microstructure observations
using a field emission scanning electron microscope (FEꢀSEM; JEOL JSM 6500F) operating at
5
ACS Paragon Plus Environment

Chemistry of Materials
Page 6 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
5 kV and 1.5 nA. Compositional analyses were determined by energyꢀdispersive Xꢀray
spectroscopy (EDS) using graphite and BN standards.
Transmission Electron Microscopy and electron energy loss spectroscopy
TEM data collection and analysis were performed using both FEI Tecnai G2 LaB and a FEI
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Talos F200X transmission electron microscopes (TEM). The samples were dispersed on a SiO
grid with methanol and imaging was done under an accelerating voltage of 80 kV for both of the
microscopes. Representative EDS data from the TEM were collected by a Bruker QUANTAX
EDS system attached to the FEI TALOS TEM. Electron energy loss (EELS) / energyꢀfiltered
TEM measurements were carried out in diffraction mode under 80 kV using the Tecnai G2 LaB .
Carbon nano onions (CNOs) were used as a standard for pure 100% sp carbon. Nascent REGAL
6
2
2
50 carbon black was purchased from Cabot Corporation and graphitized at 3573 K for 5 hours
to form highly crystalline CNOs.
X-ray Photoelectron spectroscopy
XPS experiments were performed using a Physical Electronics VersaProbe II instrument
equipped with a monochromatic Al k xꢀray source (hν = 1,486.7 eV) and a concentric
hemispherical analyzer. Charge neutralization was performed using both lowꢀenergy electrons
(<5 eV) and Argon ions. The binding energy axis was calibrated using sputterꢀcleaned Cu foil
α
(Cu 2p3/2 = 932.7 eV, Cu 3p3/2 = 75.1 eV). We assigned the pyridinic-N band in the N 1s spectra
to the energy of 398.5 eV. Measurements were made at a takeoff angle of 45° with respect to the
sample surface plane. This resulted in a typical sampling depth of 3ꢀ6 nm. Quantification was
done using locallyꢀderived sensitivity factors from a pure polyvinylpyrrolidone (PVP) reference
+
sample. A soft surface cleaning was performed using 0.5 kV Ar .
TOF-SIMS
Timeꢀofꢀflight secondary ion mass spectrometry (TOFꢀSIMS) analysis was performed using a
Physical Electronics nanoTOF II instrument. A Bi primary ion beam with an ion dose of 5.0×
10 ions/cm was applied to an area of 100×100 µm and negative secondary ions were collected.
Charge compensation was performed using 10 eV Ar and 12 eV electrons. The surface sputter
cleaning was performed using 20 keV Ar
dose of 5.0×10 ions/cm . The melamine and highly oriented pyrolytic graphite (HOPG)
analytical standards were purchased from SigmaꢀAldrich and used without further purification.
2
3
+
1
1
2
2
+
+
2
. The sputtered area was 800×800 µm with an ion
2500
1
4
2
First-principles calculations
Ab initio molecular dynamics (AIMD) simulations for C N were performed within the NPT (N
is the number of particles, P the pressure, and T the temperature) ensemble implemented in
Vienna ab Initio Simulation Package (VASP) code with Langevin dynamics. The allꢀelectron
projectorꢀaugmented wave (PAW) potential was adopted with the PAW potentials taken from
4
2
6
9
7
0
71
6
ACS Paragon Plus Environment

Page 7 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
2
2
2
3
the VASP library where 2s 2p and 2s 2p are treated as valence electrons for C and N atoms,
respectively. We generated 2×3×3 supercell of a conventional unit cell, containing 36 formula
units (f. u.) in the simulation cell, which is big enough not to have a force field overlap with the
nearest cells. The planeꢀwave was expanded to an energy cutoff of 400 eV and Brillouin zone
sampling with a 1×1×1 kꢀmesh were employed. Exchange and correlation effects were treated in
7
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
the PerdewꢀBurkeꢀErnzerhof parameterization of generalized gradient approximation. We also
conducted Raman Intensity calculations using Quantum espresso software package based on
density functional perturbation theory . Calculations for vibrational modes need highly accurate
charge density and thus we carefully relaxed the atomic positions and the cell volume at a given
external pressure. For the structural relaxation, we used a 6×8×16 kꢀmesh with 700 eV energy
cutoff and for Raman Intensity calculation we used a 4×6×12 kꢀmesh with 700 eV energy cutoff.
7
3
7
4
Results and Discussion
ν7 + ν
8
ν4
×
×
1 ← IR Experiment → ×1
40 ← IR Calculation → ×1
ν5
ν6 + ν9
DAC absorption
region
νL
ν6 + ν8
ν6
ν1
×2 ← R Experiment → ×1
×40 ← R Calculation → ×1
2
ν5
ν7
ν3
ν2
*
*
0
250 500 750 1000 125 2100
2200
2300
2400
ꢀ1
Frequency (cm )
Fig. 1. Experimental and calculated Raman (bottom) and IR absorbance (top) spectra of solid
C N at room temperature (RT) and 0.2 GPa. Calculated spectra are represented by Gaussian
peak profiles with an arbitrary width. Spectral intensities are scaled by the factors indicated.
Asterisks show contributions from C isotopologues. The thick dashed line in the IR spectrum
indicates the diamond anvil absorption region.
4
2
1
3
Molecular C N
4
2
3
9
C N is a centrosymmetric linear molecule. Consequently, it has 3Nꢀ5=13 fundamental
vibrational modes, of which the Ramanꢀ and IRꢀactive modes can be classified as ΓRaman= 3Σg
4
2
+
+
(
v , v , v ) + 2Π (v , v ) and Γ = 2Σ (v , v ) + 2Π (v , v ), given the molecule’s D symmetry.
1 2 3 g 6 7 IR u 4 5 u 8 9 ∞h
Raman and IR spectra of solid C N recorded at 0.2 GPa (the solidification pressure at RT) are
shown in Fig.1. In order to provide accurate assignments of vibrational modes, we performed
4
2
7
ACS Paragon Plus Environment

Chemistry of Materials
Page 8 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
density functional perturbation theory calculations, based on the previouslyꢀreported lowꢀ
3
9
temperature crystal structure (confirmed under pressure, below), to obtain vibrational mode
frequencies and intensities. Experimentally, we observed all fundamental modes except the lowꢀ
frequency IRꢀactive bending modes (v , v ), although we are able to infer their experimental
8
9
frequencies based on observed combination bands. All experimental peak positions exhibit
agreement with calculations within a frequency difference of 10%, allowing for accurate
assignments.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Significant differences occur between previous mode assignments in the literature, notably in the
4
8ꢀ54
highꢀfrequency region and the distinction between C≡N and C≡C stretching.
Our
ꢀ1
observations indicate that the strongest Raman mode at 2270 (v ) cm can be assigned to C≡N
1
symmetric stretching, although it is not a pure vibration and shows mixing with C≡C stretching
ꢀ
1
character. The weaker intensity mode at 2122 (v ) cm is assigned to C≡C stretching. Previously,
2
4
8ꢀ49
inconsistent assignments of the C≡N and C≡C stretching modes were made by Miller et al.
at 2290 and 2119 cm , Khanna et al. at 2270 and 2331 cm , Smith et al. at 2204 and 2224
cm , Winther et al. at 2268 and 2121 cm , together with Bartel et al. at 2296 and 2287 cm
Two extremely weak peaks at 2310 and 2138 cm , also previously observed at 2309 and 2137
and 2307 and 2137 cm by Winther et al., correspond to the naturally
abundant C isotopologue. The assignments of 618 (v ), 505 (v ) and 262 (v ) cm are
ꢀ1
50
ꢀ1
51
ꢀ1
52
ꢀ1
54
ꢀ
1
ꢀ1
.
ꢀ1
48ꢀ49
ꢀ1
52
cm by Miller et al.,
1
3
ꢀ1
3
6
7
consistent with previously reports as shown in Table I. Interestingly, two new lattice modes (83
ꢀ1
and 51 cm ) are were observed for the first time in our Raman spectrum. The infrared spectrum
ꢀ
1
of C N exhibits a strong peak at 2241(v ) cm , corresponding to asymmetric C≡N and C≡C
4
2
4
ꢀ1
stretching modes. The observed mode at 1161 (v ) cm is appropriately assigned to be CꢀC
asymmetric stretching. The strong mode at 2338 cm in the Raman spectrum is almost twice the
5
ꢀ1
frequency of this mode, which can be assigned to 2v5.
Table I. Observed and calculated Raman and IR modes.
Experiment Theory
Assignments
ꢀ1
ꢀ1
(
cm )
(cm )
*
Raman
Raman
Raman
IR
v1 F. r.
2
2
6
270
122
18
2290
Symmetric C≡N+C≡C stretching
Symmetric C≡C stretching
Symmetric CꢀC stretching
*
v2
2117
*
v3
621
*
Asymmetric C≡N+C≡C
stretching
Asymmetric CꢀC stretching
C–C≡N bending
C≡C–C bending
Bending
v4
2
241
2234
*
IR
v5
1
5
2
4
161
05
1195
*
Raman
Raman
IR
v6
565
*
v7
62
285ꢀ296
#
*
v8
v9
77
509
#
*
IR
129
120
Bending
*
Most prominent frequency (theory predicts a number of modes of this character due to crystal field
#
splitting); Inferred from combinations. F.r. = Fermi resonance with 2ν .
5
8
ACS Paragon Plus Environment

Page 9 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
The in situ behavior of C N was also studied under pressure to 13.0 GPa using singleꢀ
4
2
crystal/powder Xꢀray diffraction. Singleꢀcrystal and powder XRD confirm that C N crystallizes
4
2
3
5
in the same P2 /c structure reported previously at low temperature. The lattice parameters
1
obtained at 0.2 GPa of a = 3.878(1) Å, b = 6.058(1) Å, c = 9.0262(2) Å and β=98.94(3)˚ are in
good agreement with the lowꢀpressure report (See SI, Tables S1ꢀS2 and Fig.S1). With increasing
pressure, the spectroscopic and diffraction data show that this structure is maintained to 5.5 GPa
with a corresponding volume decrease of ~28% (Fig.S2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(a)
(b)
(d)
(c)
Fig 2. Representative (a) Xꢀray diffraction patterns, (b) IR spectra and (cꢀd) photos showing the
onset of polymerization above ~6GPa.
Polymerization into an extended network
Above 5 GPa, changes in the Raman / IR spectra, as well as distinctive color changes of the
sample, indicate the onset of a chemical transformation. The fundamental vibrational modes of
C N decrease in intensity, while new, broad Raman/IR features (Fig.2b and Fig.S3 and S4) are
4
2
ꢀ1
observed near 1200ꢀ1600 cm . Above 6 GPa the sample transforms from a transparent molecular
crystal into a black and visibly opaque solid. Simultaneously, all crystalline diffraction is lost and
only a diffuse halo remains visible in Xꢀray diffraction patterns (Fig 2). These features are
2
indicative of the sp carbon and nitrogen reacting to form a disordered network with largely sp
character.
Upon releasing the pressure to ambient conditions, a solid phase with a black luster was
recovered, in stark constrast to the transparent starting liquid phase, indicating the irrevesibility
of the chemical transformation. The morphology and composition of recovered phase was
examined using SEM with compositional EDS mapping (Fig. 3a and Fig.S5). The recovered
sample exhibits a smooth, uniform texture with homogenous chemical composition near 30 at%
9
ACS Paragon Plus Environment

Chemistry of Materials
Page 10 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
nitrogen (C N1.8±0.1). This composition indicates that the majority of nitrogen remains chemically
bound within the sample with only a small degree of nitrogen loss, although EDS tends to
overestimate carbon content due to various sources of carbon contamination.
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Fig 3. (a) Backscattered electron SEM image showing smooth morphology with overlaid EDS
elemental mapping for recovered C N . The average composition was found to be C N1.8±0.1. (b)
4
2
4
Multiwavelength Raman spectra of recovered samples.
Raman spectra from the recovered sample obtained at various laser excitation wavelengths are
shown in Fig 3b. The spectra show three notable features, in contrast to the orginial spectra
ꢀ1
collected from the molecular crystal. A weak peak observed at 2222 (2227) cm with 633 (514)
nm indicates that a fraction of unconverted sp C≡N and/or C≡C bonding is preserved after the
pressureꢀinduced reaction. This sp vibration detected by UV Raman (257 nm) is much stronger
than under visible excitation due to resonance enhancement, as also observed in previously
87ꢀ90
amorphous carbon nitride thin film.
Broad Raman peaks are observed between 900ꢀ1800
ꢀ1
89ꢀ98
cm , as previously observed for amorphous cabons or carbon nitrides.
like” features are associated with C=N and/or C=C sp bonds and indicate the presence of
disorder or a range of graphitic domains in the recovered material.
These “Dꢀ“ and “Gꢀ
2
8
7ꢀ90
With increasing excitation
wavelengh, these broad peaks shift to lower wavenumber, similar to the G and D bands in
97ꢀ98
ꢀ1
amorphous carbon nitrides and carbon materials.
The peak at 696 cm with 633 nm
ꢀ1
excitation (700 cm with 514 nm) is attributed to ring bending modes; this peak is barely visible
in the UV Raman. We note that, due to overlap with the diamond T phonon, combined with
2
g
very weak Raman intensity above 6 GPa, we were not able to study the D and G features using
Raman under in situ, highꢀpressure conditions.
10
ACS Paragon Plus Environment

Page 11 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
Structure of the recovered material
In order to gain deeper insights into the local structure of the recovered C N , we performed
4
2
highꢀenergy Xꢀray scattering experiments to obtain S(Q) to a momentum transfer of Qmax = 17
ꢀ1
Å , and transformed this information into the radial distribution function, g(r). In order to help
interpret g(r), we performed AIMD simulations at 12 GPa and 300 K, and relaxed the reaction
product at ambient conditions. We note that a simulation pressure of 12 GPa was required to
observe the rare event chemical reaction on time scales suitable for AIMD simulations. The
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
obtained g(r) for the final disordered C N from both experiment and theory are shown in Fig. 4a.
4
2
The g(r) derrived from experiment yields an average firstꢀneighbor distance of 1.36 Å, smaller
2
87ꢀ88
than that of sp amorphous carbon (1.43ꢀ1.49 Å) and graphite (1.42 Å).
The calculated firstꢀ
neighbor distance is 1.35Å, which is very close to the experimental result. As suggested from
2
Raman analysis, the recovered sample is dominated by sp C=C and C=N bonds. Compared with
known C=C (1.42 Å in graphite and 1.399 Å in benzene) and C=N distances (1.30 Å in furazan,
8
7
1.33 Å in pyrazole, 1.34 Å in pyridine, 1.37 Å in pyrrole), the average firstꢀneighbor distance
observed here also confirms that the dominant bonding contributions are from C=C and C=N
bonds.
All general features of the experimental g(r) are reproduced by the calculations, suggesting that
the calculated structure is a strong candidate to understand the polymeric structure of the
2
disordered product. The structure consists of quasiꢀtwoꢀdimensional layers of sp rings that are
3
sporadically connected by a fraction of sp carbon atoms (< 5%) (Fig. 4b). This extended
polycyclic structure is comprised of mostly 5ꢀ and 6ꢀmembered rings, although a small fraction
of rings with four and ≥ 7 members were also observed. The density estimate from the S(Q) and
6
7
3
g(r) of the recovered material using the method of Eggert et al. is 2.12 ± 0.1 g/cm , compared
3
3
with 2.17 g/cm from the MD simulation. These values are greater than 2.0 g/cm reported for
CN(30% N), but smaller than 2.26 g/cm for crystalline graphite. The calculated average CꢀC
8
7
3
and CꢀN coordination numbers are 2.4 and 2.9, respectively. While 5% of the carbon atoms are
2
4
ꢀfold coordinated, a small fraction of unconverted sp and linear sp carbon keeps the average
coordination number below three.
The secondꢀnearestꢀneighbor peak in g(r) relates to the second coordination shell, and thus
provides information on bond angles. Experimentally, this peak is located at 2.43 Å, which is
slightly longer than 2.37 Å from calculations. These distances provide average bond angles of
124.8° and 122.3° for experiment and calculation, respectively, and indicate a disordered
graphitic ring structure. The absence of any distinct features between 2.8ꢀ3.0 Å indicates that
there is no significant fraction of ordered sixfold rings (the ideal crossꢀring, 6ꢀfold peak in
8
9
graphite occurs at 2.84 Å).
11
ACS Paragon Plus Environment

Chemistry of Materials
Page 12 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Fig 4. (a) Experimental and calculated g(r) including calculated partial correlations. (b) Relaxed
MD structure showing CN network with tendency towards 2D layered structure. (c) Common
local motifs with pyridinicꢀ and pyrrolicꢀlike characteristics and a fraction of partially reacted
material.
X-ray photoelectron spectroscopy
To provide additional information regarding the local structure and bonding features, we
performed XPS measurements. The coreꢀlevel XPS spectrum from the recovered sample is
shown in Fig.5. The pristine recovered sample displays both carbon (C 1s) and nitrogen (N 1s)
coreꢀlevel peaks, in addition to a small oxygen contribution, resulting in the approximate
chemical composition C N O . We confirmed that the bulk interior of the sample did not
4
1.4 0.24
+
contain any oxygen by performing a soft Ar surface milling to remove ~0.5 Å of material (~20
seconds at 0.5 kV), after which there was no detectable signal from the O 1s peak above the level
of noise (Fig 5a). Thus, the oxygen signal originates from surface oxidation and does not exist
within the bulk. After the surface cleaning, the chemical composition of the material indicated
partial nitrogen loss (C N ), and we therefore conclude that the chemical composition of the
4
1.3
recovered sample falls within the range of C N1.4+0.1 based on XPS analysis, which is in general
4
12
ACS Paragon Plus Environment

Page 13 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
agreement with the SEMꢀEDS composition. We note that both XPS (~26% N) and EDS (~31%
N) analyses indicate partial nitrogen loss as compared with the starting molecular C N , but both
4
2
techniques typically bias carbon due to environmental (adventitious) surface sources, and thus
we consider C N as a lower bound on the actual nitrogen concentration. Due to the
4
1.4+0.1
+
observed nitrogen loss after Ar milling, which suggests possible alteration of the local chemical
structure, we chose to analyze the raw XPS data, before cleaning. We note, however, that there
were no major differences between the spectra obtained before and after milling, aside from
minor changes in relative peak intensities (and the obvious absence of oxygen).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The carbon nitride literature contains numerous and varying reports concerning the interpretation
90ꢀ109
of XPS spectra.
XPS spectra provide information regarding the binding energies of coreꢀ
level electrons, which are a function of local environment including coordination number and
nearestꢀneighbor constituents. To a first approximation, partial charge, e.g., how much electron
density is withdrawn by neighbors, can be used to generalize trends in binding energies. All
2
other factors being equal, a sp carbon bound to another carbon atom will possess a lower
2
binding energy than a sp carbon bound to an oxygen atom. In wellꢀordered molecular or
crystalline systems, it is often straightforward to assign observed binding energies to specific
sites, given sufficient spectral resolution. Yet, in disordered materials, such as the recovered
C N presented here, numerous local environments exist, and spectra should be considered in a
4
2
statistical manor. Because we do not have sufficient spectral resolution to observe all possible
local chemical environments, we employed the following approach to understand the measured
XPS spectra. All XPS spectra were processed by first subtracting a linear background function.
A series of trial peak functions was tested for each energy region and it was determined that
simple Gaussian peak profiles could sufficiently describe the spectra as well as more complex
functions using the fewest number of adjustable parameters. We used the minimum number of
Gaussian peaks needed to adequately fit each spectrum, and each peak within the same binding
energy region was constrained to have the same FWHM. Using this approach we can
qualitatively understand the prominent local environments present within the recovered samples.
The N 1s coreꢀlevel region (Fig. 5b) was deconvoluted into two different Gaussian components
located at 398.5 and 400.3 eV, which are consistent with the general N 1s features reported for
9
0ꢀ109
amorphous carbon nitride and Nꢀdoped graphene.
The most intense feature at 398.5 eV is
attributed to “pyridinicꢀtype” local environments, whereas the lessꢀintense feature at 400.3 eV is
attributed to “pyrrolicꢀtype” environments. We note that, while this nomenclature is common in
the XPS literature, these terms do not explicitly mean pyridine or pyrrole, but refer to whether
9
0ꢀ112
the nitrogen lone pair participates in the π system.
We were unable to resolve the small
3
contribution from unreacted nitrile groups, or any fraction of sp nitrogen atoms. Analysis of
structures produced by MD simulations indicates that approximately 46% of the N atoms can be
classified as pyridinic (2ꢀfold coordinated), 29% as pyrrolic (3ꢀfold coordinated with short C=N
3
bonds), 19% as sp (3ꢀfold coordinated with long CꢀN bonds) and 6% as sp. The remaining
nitrile nitrogen likely contributes towards the “pryridinic” XPS contribution observed at low
13
ACS Paragon Plus Environment

Chemistry of Materials
Page 14 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
binding energy, whereas, the 3ꢀfold coordinate N atoms with long bonds likely contribute more
towards the “pyrrolic” contribution at higher binding energy. Adding these together from
calculation, the pyridininc+sp and pyrrolic+sp contributions are roughly 50% each, compared
3
with 69(5)% and 31(5)% averaged over four independent experimental observations. These
features are in general agreement, but we note again that it was not possible to precisely describe
9
0,100
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
all expected N 1s chemical environments in the peak fitting.
Fig 5. (a) XPS survey for recovered sample before and after ion milling. (b) Representative N 1s
and (c) C 1s XPS spectra of the raw surface before ion milling.
The C 1s coreꢀlevel spectrum (Fig. 5c) is characterized by a broad peak, tailing towards higher
2
90ꢀ109
binding energy, which is evidence for high sp carbon content in the material.
The spectrum
could be described by three Gaussian components at 284.7, 286.2 and 287.9 eV. The highest BE
peak is attributed to C=O bonds on the surface and is expected on the basis of a weak O 1s peak
2
in the survey scan. For sp dominated carbon materials, The XPS literature generally assigns the
93,100
low BE peak near 284.7 eV to C=C bonds and the higher 286.2 eV peak to C=N bonds.
But
this assignment is incomplete and the real situation is more complex. Nearly every carbon atom
in the recovered material is 3ꢀfold coordinated, thus the identity of these three neighbors is
needed for a baseꢀlevel interpretation, e.g., CCC, CCN, CNN, NNN. Our MD simulations reveal
no NNN environments, while approximately 39 % of 3ꢀcoordinated carbon atoms are CNN, 52%
are CCN and 9% are CCC. Experimentally, the peaks at 287.7 eV and 286.2 eV comprise 52(7)%
14
ACS Paragon Plus Environment

Page 15 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
and 48(7)% of the C 1s signal, respectively, when averaged over four independent observations
(discounting the contribution of oxygen). Thus, the 284.7 eV peak could represent predominantly
CCC and CCN environments, while the 286.2 eV peak could represent CNN environments.
While the interpretations across experiment and theory are fully consistent (MD:
CCC+CCN=61%; CNN=39%, experiment: CCC+CCN=52(7)%; CNN=48(7)%), we must
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
emphasize that we cannot fully resolve all C 1s chemical environments. A small fraction of sp
carbon and unreacted nitrile carbon will also contribute to the overall shape of the XPS spectrum,
although the sp features will dominate.
2
Transmission electron microscopy
Highꢀresolution TEM, together with selectedꢀarea electron diffraction (Fig 6), confirms that the
material lacks longꢀrange order, in agreement with the Xꢀray diffraction results. The TEM and
EELS measurements were carried out using low beam energy with short acquisition times in
order to minimize potential sample damage. HRTEM (Fig 6b) indicates that the material has a
smooth texture at the atomic scale and no local domains of order are present. In addition, STEMꢀ
EDS mapping (Fig 6c, d) of the recoved sample confirms that carbon and nitrogen are distributed
homogenously with no signs of chemical disproportionation. The fullꢀrange EELS spectrum
exhibits the characteristic 1s to σ* and 1s to π* transitions in both the carbon and nitrogen Kꢀ
102
edge regions, similar to previous measurements of CN films.
(
Zooming in on these regions
x
2
Fig 6 e, f ), the sample shows an abundance of sp bonding in the carbon and nitrogen atoms, as
indicated by the presence of 1s peaks at both C and N Kꢀedge. The sp ꢀcontent was calculated to
2
be ~90 +5% using graphitized carbon nanoꢀonions as a standard using the method proposed by
1
10
Brydson et al.
15
ACS Paragon Plus Environment

Chemistry of Materials
Page 16 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Fig. 6. (a) Obtained TEM with diffraction pattern (inset), (b) HRTEM with fast Fourier
transformation (inset), elemental analysis of (c) C and (d) N, and EELS of (e) carbon and (f)
nitrogen K edges for recovered sample of C N .
4
2
Mass Spectrometry
In order to further study the structure of the disordered carbon nitride, we carried out timeꢀofꢀ
flight secondary ion mass spectrometry measurements (TOFꢀSIMS). The sample surface was
+
cleaned using 20 keV Ar2500 ion sputtering to remove signals from surface contaminants and to
reveal the true characteristics of the bulk material. The mass spectra of graphitic materials, such
as highly oriented pyrolytic graphite (HOPG), do not contain monomeric molecular ions owing
to their extended structure. However, graphitic surfaces do exhibit characteristic fragmentation
patterns in the negative ion spectrum that consist of a series of prominent C H peaks with low
x
y
112
hydrogen content.
The negative ion spectrum of HOPG (as shown in Fig. S6) exhibits
features associated with C , CH , C , C H , C , C , C , and C . These prominent features
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
2
2
3
4
5
6
indicate a low hydrogen content and a fragmentation pattern consistent with an extended carbon
structure (2D graphite in this case). The negative ion spectrum for recovered C N is presented in
4
2
Fig. 7, together with the table of major fragments. In this case, the most prominent fragments are
ꢀ
ꢀ
ꢀ
ꢀ
CꢀN ions (such as CN , C N ), CꢀC ions (such as C , C ) and hydrogenated ions with low H
3
ꢀ
2
ꢀ
content (such as CH , C H ). Furthermore, larger C N fragments with high nitrogen content are
2
x
y
also observed. These results confirm the formation of an extended carbonꢀnitrogen network
16
ACS Paragon Plus Environment

Page 17 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
consisting predominantly of carbonꢀnitrogen and carbonꢀcarbon bonds, and are similar to
1
12
previously reported results for carbon nitride thin films. In addition, the negative spectrum
shows a similar fragmentation pattern to the TOFꢀSIMS spectrum of melamine, but with lower
hydrogen content (Fig. S7). Compared with the spectrum of melamine, recovered C N shows
4
2
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
similar lowꢀmass ion fragmentation (e.g., C , CH , C , C H , CN , CN , C N ), however, higherꢀ
2
2
2
2
3
ꢀ
ꢀ
ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
mass protonated fragments such as CN H , C N H , C N H are far less prominent.
2
2
3
2
3
2
+
Fig. 7. The TOFꢀSIMS negative ion spectrum of the sample after Ar2500 ion sputtering. C Ny
x
fragments with low content of hydrogen were observed. Although the hydrocarbon
contamination on the surface is largely removed by ion sputtering, the contamination is still
detected. However, the C N fragments peaks are prominent and therefore provides a direct
evidence of an extended carbonꢀnitrogen network presented. The negative fragment ions are
listed.
x
y
Polymerization mechanism
Previously reported static ab initio calculations suggest that below 14 GPa, there are no
1
13
114
thermodynamically stable crystalline carbon nitride phases
including graphitic C N
and its
3
4
1
15
modified allotropes . Here we work in the realm of metastability and examined potential
reaction pathways of crystalline, molecular C N using a combination of molecular dynamics
and metadynamics simulations, as well as evolutionary metadynamics, as implemented in the
4
2
1
16
1
17ꢀ118
USPEX code.
The former is intended to provide direct insights into the reaction mechanism,
while the two versions of metadynamics are used to examine a larger portion of the potential
energy landscape by overcoming rareꢀevent barriers and to draw conclusions as the system
evolves towards the ground state.
17
ACS Paragon Plus Environment

Chemistry of Materials
Page 18 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Molecular dynamics (NPT) trajectories were established from a supercell (36 molecules) of the
starting P2 /c structure at 300 K and 12 GPa. Initial polymerization was generally observed to
1
occur within the first 2 ps of the run (the same polymerized structures were equilibrated at 0 GPa
to obtain the g(r) data in Fig. 4). The molecular structure can be viewed as twoꢀdimensional
sheets of C N molecules that lie in a single plane when viewed approximately normal to the
4
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
201) direction (Fig. 8). Within these sheets, sideꢀaligned columns of C N molecules propagate
4 2
down the bꢀaxis. Two types of columns exhibit ABAB stacking within the ~(201) plane where
the A column is canted by ~67° with respect to the B column. The shortest interatomic distance
in the starting structure is between the nitrile nitrogen of a molecule in column A and the nitrile
carbon atom of an adjacent molecule in column B (~3.3 Å at 1 atm).
Fig. 8. (a) Molecular structure of C N . (b) View approximately normal to the (201) direction. (c)
4
2
Molecules form columns along the bꢀaxis with ABAB stacking. (d) The shortest distance
between columns is 3.3 Å at 1 atm.
When compressed to 12 GPa, the linear C N molecules began to kink. Some of the CꢀC≡C and
4
2
N≡CꢀC angles drop to as low as 140° within the first 2 ps, indicating enhanced interactions
between neighboring molecules. As expected from the starting crystal structure, the initial
polymerization initiates between the shortest N···C distance within a single 2D “sheet”. The
polymerization proceeds within individual sheets through cycloaddition reactions that produce
predominantly 5ꢀ and 6ꢀmembered rings, giving the overall structure a tendency to propagate in
two dimensions. Over time, unreacted bonds begin to form connections between the 2D layers,
producing a more interconnected structure that still remains largely twoꢀdimensional. Ultimately,
the final disordered C N structure is unchanged after ~20 picoseconds. This process is
4
2
summarized in Figure 9.
18
ACS Paragon Plus Environment

Page 19 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig.9. Polymerization process of C N at 12 GPa and 300K shown at (a) 1.8 ps, (b) 2.0 ps, (c)
4
2
2.1 ps and (d) 2.4 ps. Orange arrows indicate reaction points.
At 300K, MD trajectories become essentially trapped within a single configuration after
polymerization occurs – there is not enough energy relative to kT for the probable observation of
rareꢀevent, covalentꢀbond breaking at our simulation timescales. In order to gain insights into the
tendency of the system to evolve over longer time scales, we performed metadynamics
simulations from the starting molecular structure (Fig.S8). The metadynamics algorithm uses
scaled components of the edge vectors of the simulation cell as collective variables. The driving
force that guides the evolution of the simulation cell is the derivative of the Gibbs free energy
with respect to the six collective variables, which were updated in every metaꢀstep toward a lowꢀ
energy pathway to neighboring minima. In our metadynamics simulation, C N ultimately
4
2
transformed to a twoꢀdimensional layered structure. The final layered structure has a very low
symmetry with the space group of P1, which can be understood as an approximate amorphous
structure. The structural evolutions and calculated g(r) for the P1 structure and the experimental
data are shown in the supporting information (Fig. S9). The metadynamics reaction sequence and
calculated g(r) are very similar when compared with the MD simulations, confirming the
mechanism of the reaction process and the general tendency during the polymerization of C N is
4
2
towards a twoꢀdimensional structure comprised of heterocyclic rings. Interestingly, the
evolutionary metadynamics simulations also revealed lowꢀenergy 2D and 3D crystalline
structures, as reported in the supplementary material. These structures are reported in the Fig.S11
and S12 and represent interesting future targets for metastable synthesis.
CONCLUSIONS
In summary, by starting with a hydrogenꢀfree C N molecular precursor, an amorphous extended
4
2
carbon nitride with similar composition can be synthesized via highꢀpressure, solidꢀstate
chemistry. The fundamental vibrational modes of the molecular precursor were also definitively
assigned with the help of DFT calculations. The recovered product is a largely twoꢀdimensional
2
polycyclic network comprised of predominantly threeꢀfold coordinated sp carbon with
3
“
pyrrolicꢀ“ and “pyridinicꢀlike” nitrogen. A small fraction of sp carbon connects disordered 2D
19
ACS Paragon Plus Environment

Chemistry of Materials
Page 20 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
layers in three dimensions. The reaction proceeds by the activation of linear C N molecules into
4
2
buckled chains that spontaneous assemble through cycloaddition reactions. This understanding
of the highꢀpressure behavior and polymerization mechanism of C N will provide fundamental
4
2
contributions to carbon nitride chemistry.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at DOI.
Singleꢀcrystal Xꢀray diffraction analysis, Raman and infrared spectra with pressure, SEMꢀEDS
data, TOFꢀSIMS data, metadynamics structures.
ACKNOWLEDGEMENTS
This work was supported by DARPA under ARO Contract No. W31P4Qꢀ13ꢀIꢀ0005. Portions of
this work were performed at HPCAT (Sector 16), Advanced Photon Source (APS), Argonne
National Laboratory. HPCAT operations are supported by DOEꢀNNSA under Award No. DEꢀ
NA0001974 and DOEꢀBES under Award No. DEꢀFG02ꢀ99ER45775, with partial
instrumentation funding by NSF. Portions of this work were also carried out at
GeoSoilEnviroCARS (The University of Chicago, Sector 13), Advanced Photon Source (APS),
Argonne National Laboratory. GeoSoilEnviroCARS is supported by the National Science
Foundation ꢀ Earth Sciences (EARꢀ1128799) and Department of EnergyꢀGeoSciences (DEꢀ
FG02ꢀ94ER14466). This research used resources of the Advanced Photon Source, a U.S.
Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of
Science by Argonne National Laboratory under Contract No. DEꢀAC02ꢀ06CH11357. H.G.
acknowledges financial support from the National Natural Science Foundation of China (NSFC)
under Grants No. 51201148 and U1530402, as well as the Thousand Youth Talents Plan. D.Y.K.
acknowledges the Texas Advanced Computing Center (TACC) at the University of Texas at
Austin and Argonne Leadership Computing Facility (ALCF) which is a DOE Office of Science
User Facility supported under Contract No. DEꢀAC02ꢀ06CH11357 for providing high
performance computing resources.
REFERENCES
(1) Liu, A.Y.; Cohen, M. L. Prediction of new low compressibility solids. Science 1989,
245, 841ꢀ842.
(2) Teter, D. M.; Hemley, R. J. Lowꢀcompressibility carbon nitrides. Science 1996, 271, 53ꢀ55.
(3) Niu, C.; Lu, Y. Z.; Lieber, C. M. Experimental realization of the covalent solid carbon nitride.
Science 1993, 261, 334ꢀ337.
20
ACS Paragon Plus Environment

Page 21 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
(
4) Badding, J. V. Solidꢀstate carbon nitrides. Adv. Mater. 1997, 9, 877ꢀ886.
(5) Thomas, A.; Fischer, A.; Goettmann, F.; Markus Antonietti, M.; Müller, J. F.; Robert
Schlögl, Carlsson, J. M. Graphitic carbon nitride materials: variation of structure and
morphology and their use as metalꢀfree catalysts. J. Mater. Chem. 2008, 18, 4893ꢀ4908.
(6) Wang, X.; Maeda, K.; Thomas, A.; Takanabe, K.; Xin, G.; Carlsson, J. M.; Domen, K.;
Antonietti, M. A metalꢀfree polymeric photocatalyst for hydrogen production from water under
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
visible light. Nature Mater. 2009, 8, 76ꢀ80.
(7) Liu, G.; Niu, P.; Sun, C.; Smith, S. C.; Chen, Z.; Lu, G.; Cheng, H. M. Unique electronic
structure induced high photoreactivity of sulfurꢀdoped graphitic C N . J Am. Chem. Soc. 2010,
3
4
132, 11642ꢀ11648.
8) Wang, Y.; Wang, X.; Antonietti, M. Polymeric graphitic carbon nitride as a heterogeneous
(
organocatalyst: from photochemistry to multipurpose catalysis to sustainable chemistry. Angew.
Chem. Int. Ed. 2012, 51, 68ꢀ89.
(
9) Zheng, Y.; Liu, J.; Liang, J.; Jaroniecc, M.; Qiao, S. Graphitic carbon nitride materials:
controllable synthesis and applications in fuel cells and photocatalysis. Energy Environ. Sci.,
012, 5, 6717ꢀ6731.
10) Li, X. H.; Antonietti, M. Metal nanoparticles at mesoporous Nꢀdoped carbons and carbon
nitrides: functional MottꢀSchottky heterojunctions for catalysis. Chem. Soc. Rev. 2013, 42, 6593ꢀ
604.
2
(
6
(11) Zhu, J.; Xiao, P.; Li, H.; Carabineiro, S. A. C. Graphitic carbon nitride: synthesis, properties,
and applications in catalysis. ACS Appl. Mater. Interfaces 2014, 6, 16449ꢀ16465.
(
12) Zhao, Z.; Sun, Y.; Dong, F.; Graphitic carbon nitride based nanocomposites: a review.
Nanoscale 2015, 7, 15ꢀ37.
13) Cao, S.; Low, J.; Yu, J.; Jaroniec, M. Polymeric photocatalysts based on graphitic carbon
(
nitride. Adv. Mater.2015,27, 2150ꢀ2176.
(14) Zheng, Y.; Lin, L.; Wang, B.; Wang, X. Graphitic carbon nitride polymers toward
sustainable photoredox catalysis. Angew. Chem. Int. Ed. 2015, 54, 12868ꢀ12884.
(
15) Liu, J.; Wang, H.; Antonietti, M. Graphitic carbon nitride “reloaded”: emerging
applications beyond (photo) catalysis. Chem. Soc. Rev. 2016, 45, 2308ꢀ2326.
16) Ong, W. J.; Tan, L.; Ng, Y. H.; Yong, S. T.; Chai, S. P. Graphitic carbon nitride (gꢀC N )ꢀ
(
3
4
based photocatalysts for artificial photosynthesis and environmental remediation: are we a step
closer to achieving sustainability? Chem. Rev. 2016, 116, 7159ꢀ7329.
(17) Patnaik, S.; Satyabadi, M.; Saumyaprava, A.; Parida, K. M. An overview of the
modification of gꢀC N with high carbon containing materials for photocatalytic applications.
3
4
Inorg. Chem. Front. 2016, 3, 336ꢀ347.
18) Kang, Y.; Yang, Y.; Yin, L.; Kang, X.; Liu, G.; Cheng, H. Efficient C N /graphene oxide
(
3
4
macroscopic aerogel visibleꢀlight photocatalyst. Adv. Mater. 2015, 27, 4572ꢀ4577.
19) Fan, J.; Chen, J.; Zhang, Q.; Chen, B.; Zang, J.; Zheng, M.; Dong, Q. An Amorphous
(
Carbon Nitride Composite Derived from ZIFꢀ8 as Anode Material for SodiumꢀIon Batteries.
ChemSusChem 2015, 8, 1856ꢀ1861.
(20) Sjostrom, H.; Hultman, L.; Sundgren, J. E.; Hainsworth, S. V.; Page, T. F.; Theunissen, G.
Structural and mechanical properties of carbon nitride CN (0.2⩽ x⩽ 0.35) films. J. Vac. Sci.
Technol. A 1996,14, 56ꢀ62.
x
(21) Hellgren, N.; Johansson, M. P.; Broitman, E.; Hultman, L.; Sundgren, J. E. Role of nitrogen
in the formation of hard and elastic CN thin films by reactive magnetron sputtering. Phys. Rev.
x
B 1999,59, 5162ꢀ5169.
21
ACS Paragon Plus Environment

Chemistry of Materials
Page 22 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
22) Holloway, B. C.; Kraft, O.; Shuh, D. K.; Nix, W. D.; Kelly, M. A.; Pianetta, P.; Hagstrom, S.
Interpretation of xꢀray photoelectron spectra of elastic amorphous carbon nitride thin films. Appl.
Phys. Lett. 1999, 74, 3290ꢀ3292.
(23) Wan, L.; Egerton, R. F. Preparation and characterization of carbon nitride thin films. Thin
Solid Films 1996,279, 34ꢀ42.
(
24) Spaeth, C.; Kuhn, M.; Richter, F.; Falke, U.; Hietschold, M.; Kilper, R.; Kreissig, U. A
comparative study of elastic recoil detection analysis (ERDA), electron energy loss spectroscopy
EELS) and Xꢀray photoelectron spectroscopy (XPS) for structural analysis of amorphous carbon
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
nitride films. Diamond Relat. Mater. 1998, 7, 1727ꢀ1733.
(25) Zheng, W. T.; Yu, W. X.; Li, H. B.; Wang, Y. M.; Cao, P. J.; Jin, Z. S.; Broitman, E.;
Sundgren, J. E. Chemical bonding, structure, and hardness of carbon nitride thin films. Diamond
Relat. Mater. 2000, 9, 1790ꢀ1794.
(26) Gammon, W. D.; Malyarenko, I.; Kraft, O.; Hoatson, G. L.; Reilly, A. C.; Holloway, B.C.
1
3
Hard and elastic amorphous carbon nitride thin films studied by C nuclear magnetic resonance
spectroscopy. Phys. Rev. B 2002,66, 153402:1ꢀ4.
(27) Zhang, Z.; Leinenweber , K.; Bauer, M.; Laurence A. J. Garvie, L. A. J.; McMillan, P. F.;
Wolf, G. H. HighꢀPressure Bulk Synthesis of Crystalline C N H •HCl:ꢂ A Novel C N Graphitic
6
9
3
3
4
Derivative. J. Am. Chem. Soc., 2001, 123, 7788ꢀ7796.
28) Jurgens, B.; Irran, E.; Serker, J.; Kroll, P.; Muller, H.; Schnick, W. Melem (2, 5, 8ꢀtriaminoꢀ
(
triꢀsꢀtriazine), an important intermediate during condensation of melamine rings to graphitic
carbon nitride: Synthesis, structure determination by Xꢀray powder diffractometry, solidꢀstate
NMR, and theoretical studies. J. Am. Chem. Soc., 2003, 125, 10288ꢀ10300.
(29) Lotsch B.V.; Schnick, W. Thermal conversion of guanylurea dicyanamide into graphitic
carbon nitride via prototype CN precursors. Chem. Mater., 2005, 17, 3976ꢀ3982.
(30) Lotsch B. V.; Schnick, W. From Triazines to Heptazines: Novel Nonmetal
Tricyanomelaminates as Precursors for Graphitic Carbon Nitride. Chem. Mater. 2006, 18, 1891ꢀ
1
(31) HorvathꢀBordon, E., Riedel, R., McMillan, P. F., Kroll, P., Miehe, G., van Aken, P. A.;
Zerr, A.; Hoppe, P.; Shebanova, O.; Mclaren. I.; Lauterback, S.; Kroke, E.; Boehler, R. Highꢀ
Pressure Synthesis of Crystalline Carbon Nitride Imide, C N (NH). Angew. Chem. Int.
x
900.
2
2
Edit. 2007,46,1476ꢀ1480.
32) Salamat, A.; Deifallah, M.; Carbrera, R.Q.; Cora, R.; McMillan, P. F. Identification of new
pillaredꢀlayered carbon nitride materials at high pressure. Sci. Rep. 2013, 3, 2122:1ꢀ6.
33) Schettino, V.; Bini, R. Molecules under extreme conditions: Chemical reactions at high
pressure. Phys. Chem. Chem. Phys.2003,5, 1951ꢀ1965.
34) Yoo, C. S.; Nicol, M. F. Chemical and phase transformations of cyanogen at high
pressures. J. Phys. Chem. 1986, 90, 6726ꢀ6731.
35) Aoki, K.; Kakudate, Y.; Yoshida, M.; Usuba, S.; Fujiwara, S. Solid state polymerization of
cyanoacetylene into conjugated linear chains under pressure. J. Chem. Phys.1989, 91, 778ꢀ782.
36) Tomasino, D.; Chen, J. Y.; Minesob Kim, M.; Yoo, C. S. Pressureꢀinduced phase transition
and polymerization of tetracyanoethylene (TCNE). J. Chem. Phys.2013, 138, 094506:1ꢀ10.
37) Gou, H.; Yonke, B. L.; Epshteyn, A.; Kim, D. K.; Smith, J. S.; Strobel T. A. Pressureꢀ
(
(
(
(
(
(
induced polymerization of P(CN) . J. Chem. Phys.2015, 142, 194503:1ꢀ9.
3
22
ACS Paragon Plus Environment

Page 23 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
(
38) Zheng, H.; Li, K.; Cody, G.; Tulk, C. A.; Dong, X.; Gao, G.; Molaison, J. J.; Liu, Z.;
Feygenson, M.; Yang, W.; Ivanov, I. N.; Basile, L.; Idrobo, J. C.; Guthrie, M.; Mao, H. K. Phase
Transitions and Polymerization of C H ꢀC F Coꢀcrystal under Extreme Conditions. Angew.
6
6
6 6
Chem. Int. Ed. 2016, 55, 2040ꢀ2044.
39) Hanan, R. B.; Collin, R. L. The crystal structure of dicyanoacetylene. Acta Cryst. 1953, 6,
50ꢀ352.
40) Armstrong, G.T.; Marantz, S. The Heat of Combustion of Dicyanoacetylene. J. Phys. Chem.
963, 67, 2888
(
3
(
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(41) Kirshenbaum, A. D.; Grosse, A. V. The combustion of carbon subnitride, C4N2, and a
chemical method for the production of continuous tempertures in the range of 5000ꢀ6000° K. J.
Am. Chem. Soc., 1956, 78, 2020.
(
42) Ciganek, E. DielsꢀAlder additions of dicyanoacetylene to aromatic hydrocarbons.
Tetrahedron Lett. 1967,8, 3321ꢀ3325.
43) Chien, J. C. W.; Carlini, C. Semiconducting poly(dicyanoacetylenes). J. Polym. Sci.: Polym.
Chem. Edt. 1985, 23, 1383ꢀ1393.
44) Coll, P.; Guilleminc, J. C.; Gazeaua, M. C.; Raulin, F. Report and implications of the first
observation of C N in laboratory simulations of Titan's atmosphere. Planet. Space Sci. 1999,47,
(
(
4
2
1433ꢀ1440.
45) Nishio, M.; Paillous, P.; Khlifi, M.; Bruston, P.; Raulin, F. Infrared spectra of gaseous
(
−
1
ethanenitrile in the 3500ꢀ250 cm region: absolute band intensity and implications for the
atmosphere of Titan. Spectrochimica Acta 1995, 51A, 617ꢀ622.
(46) Samuelson, R. E.; Mayo, L. A.; Knuckles, M. A.; Khanna, R. J. C N ice in Titan's north
4 2
polar stratosphere. Planet. Space Sci. 1997, 45, 941ꢀ948.
(47) de Kok, R. J.; Teanby, N. A.; Maltagliati, L.; Irwin, P. G. J.; Vinatier, S. HCN ice in Titan's
highꢀaltitude southern polar cloud. Nature 2014, 514, 65ꢀ67.
(
48) Miller, F. A.; Hannan, Jr. R. B. The Infrared and Raman Spectra of Dicyanoacetylene. J.
Chem. Phys. 1953, 21, 110ꢀ114.
49) Miller, F. A.; Hannan, Jr. R. B.; Cousins, L. R. Infrared and Raman Spectra of
(
Dicyanoacetylene. II. J. Chem. Phys. 1955,23, 2127ꢀ2129.
(50) Khanna, R. K.; PereraꢀJarmer, M. A.; Ospina, M. J. Vibrational infrared and Raman spectra
of dicyanoacetylene. Spectrochimica Acta. 1987, 43A, 421ꢀ425.
(
51) Smith, A. M.; Schallmoser, G., Thoma A.; Bondybey, V. E. Infrared spectral evidence of
N≡CꢀC≡CꢀN≡C:Photoisomerization of N≡CꢀC≡CꢀN≡C in an argon matrix. J. Chem. Phys. 1993,
8, 1776ꢀ1785.
52) Winther, F.; Schonhoff, M. The Fundamental Vibrations of N≡CꢀC≡CꢀN≡C
(Dicyanoacetylene). J. Mol. Spectrosc.1997,186, 54ꢀ65.
53) Khlifi, M.; Paillous, P.; Bruston, P.; Guillemin, J. C.; Benilan, Y.; Daoudi, A.; Raulin, F.
9
(
(
−1
Gas infrared spectra, assignments, and absolute IR band intensities of C N in the 250ꢀ3500 cm
4
2
region: implications for Titan's stratosphere. Spectrochimica Acta 1997, 53A, 707ꢀ712.
54) Bartel, C.; Botschwina, P.; Bürger, H.; Guarnieri, A.; Heyl, A.; Huckauf, A.; Lentz, D.;
(
Merzliak, T.; Mkadmi, E. B. Cyanoisocyanoacetylene, N≡CꢀC≡CꢀN≡C. Angew. Chem. Int. Ed.
1998, 37, 2879ꢀ2882.
(55)Saggiomo, A. J. The Dinitriles of Acetylenedicarboxylic and Polyacetylenedicarboxylic
Acids. 1 I. 2 Dicyanoacetylene and Dicyanodiacetylene. J. Org. Chem., 1957, 22, 1171ꢀ1175.
23
ACS Paragon Plus Environment

Chemistry of Materials
Page 24 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
56) Mao, H. K.; Xu, J.; Bell, P. M. Calibration of the ruby pressure gauge to 800 kbar under
quasiꢀhydrostatic conditions. J. Geophys. Res. 1986, 91, 4673ꢀ4676.
57) Hammersley, A. P.; Svensson, S. O.; Hanfland, M.; Fitch, A. N.; Häsermann, D. Twoꢀ
(
dimensional detector software: from real detector to idealized image or twoꢀtheta scan. High
Pressure Res. 1996, 14, 235ꢀ248.
(
58) Le Bail, A. ESPOIR: a program for solving structures by Monte Carlo from powder
diffraction data. Mater. Sci. Forum 2001,378-381, 65ꢀ70.
59) Larson, A. C.; Von Dreele, R. B. Los Alamos National Lab. Report 2000, LAUR 86-748.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
(60) Desgreniers, S.; Lagarec, K. XRDA3.1ꢀa program for Xꢀray diffraction analysis on a PC. J.
Appl. Crystallogr. 1998,31, 109ꢀ110.
(61) Dera, P.; Zhuravlev, K.; Prakapenka, V.; Rivers, M. L.; Finkelstein, G. J.; GruborꢀUrosevic,
O.; Tschauner, O.; Clark, S. M.; Downs, R. T. High pressure singleꢀcrystal micro Xꢀray
diffraction analysis with GSE_ADA/RSV software. High Pressure Research 2013, 33 (3), 466ꢀ
4
84.
(62) Sheldrick, G. A short history of SHELX. Acta Crystallogr. Section A 2008, 64 (1), 112ꢀ122.
63) Sheldrick, G. Crystal structure refinement with SHELXL. Acta Crystallogr. Section C 2015,
1 (1), 3ꢀ8.
64) Parthe, E.; Gelato, L. M. The standardization of inorganic crystalꢀstructure data. Acta
(
7
(
Crystallogr. Section A 1984, 40 (3), 169ꢀ183.
(65) Gelato, L. M.; Parthe, E. STRUCTURE TIDY–a computer program to standardize crystal
structure data. J. Applied Crystallogr. 1987, 20 (2), 139ꢀ143.
(
66) Prescher, C.; Prakapenka, V. B. DIOPTAS: a program for reduction of twoꢀdimensional Xꢀ
ray diffraction data and data exploration. High Press. Res. 2015, 35, 223ꢀ230.
67) Eggert, J.; Weck, G.; Loubeyre, P.; Mezouar, M. Quantitative structure factor and density
(
measurements of highꢀpressure fluids in diamond anvil cells by xꢀray diffraction: Argon and
water. Phys. Rev. B 2002, 65, 174105ꢀ174116.
(
68) Kaplow, R.; Strong, S. L.; Averbach, B. L. Radial density functions for liquid mercury and
lead. Phy. Rev.1965, 138, A1336ꢀ1345.
69) Hernández, E. Metricꢀtensor flexibleꢀcell algorithm for isothermal–isobaric molecular
(
dynamics simulations. J. Chem. Phys. 2001, 115, 10282ꢀ10290.
(70) Kresse, G.; Furthmüller, J. Efficient iterative schemes for ab initio totalꢀenergy calculations
using a planeꢀwave basis set. Phys. Rev. B 1996, 54, 11169ꢀ11186.
(
71) Kresse, G.; Joubert, D. From ultrasoft pseudopotentials to the projector augmentedꢀwave
method. Phys. Rev. B 1999, 59, 1758ꢀ1775.
72) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized gradient approximation made simple.
(
Phys. Rev. Lett. 1996, 77, 3865ꢀ3868.
(73) Giannozzi, P.; Baroni, S.; Bonini, N. et al. QUANTUM ESPRESSO: a modular and openꢀ
source software project for quantum simulations of materials. J. Phys.: Condens. Matter 2009,
2
1, 395502:1ꢀ19.
74) Baroni, S.; de Gironcol, S.; Dal Corso, A.; Giannozzi, P. Phonons and related crystal
properties from densityꢀfunctional perturbation theory. Rev. Mod. Phys. 2001,73, 515ꢀ562.
(
−
1
(75) Ferrari, A. C.; Robertson, J. Origin of the 1150 cm Raman mode in nanocrystalline
diamond. Phys. Rev. B 2001,63, 121405:1ꢀ4.
(76) Ferrari, A. C.; Rodil, S. E.; Robertson, J. Interpretation of infrared and Raman spectra of
amorphous carbon nitrides. Phys. Rev. B. 2003, 67, 155306:1ꢀ20.
24
ACS Paragon Plus Environment

Page 25 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
(
77) Ferrari, A. C.; Rodil, S. E.; Robertson, J. Resonant Raman spectra of amorphous carbon
nitrides: the G peak dispersion. Diamond and Related Mater. 2003, 12, 905ꢀ910.
78) Ferrari, A. C.; Robertson, J. Raman spectroscopy of amorphous, nanostructured, diamond–
like carbon, and nanodiamond. Phil. Trans.R. Soc.Lond. A: 2004, 362, 2477ꢀ2512.
79) Rodil, S. E.; Muhl, S. Bonding in amorphous carbon nitride. Diamond and Related Mater.
2004, 13, 1521ꢀ1531.
80) Tamor, M. A.; Wu, C. H. Graphitic network models of “diamondlike” carbon. J. Appl. Phys.
990, 67, 1007ꢀ1012.
81) Kaufman, J. H.; Metin, S.; Saperstein, D. D. Symmetry breaking in nitrogenꢀdoped
(
(
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
1
(
amorphous carbon: Infrared observation of the Ramanꢀactive G and D bands. Phys. Rev. B
1989,39, 13053ꢀ13060.
(
82) Beeman, D.; Silverman, J.; Lynds, R.; Andserson, M. R. Modeling studies of amorphous
carbon. Phys. Rev. B 1984, 30, 870ꢀ875.
83) Shi, J. R.; Shi, X.; Sun, Z.; Liu, E.; Tay, B. K.; Lau, S. P. Ultraviolet and visible Raman
(
studies of nitrogenated tetrahedral amorphous carbon films. Thin Solid Films 2000,366,169ꢀ174.
(84) Ferrari, A. C.; Robertson, J. Interpretation of Raman spectra of disordered and amorphous
carbon. Phys. Rev. B 2000, 61, 14095ꢀ14107.
(
85) Ferrari, A. C.; Robertson, J. Resonant Raman spectroscopy of disordered, amorphous, and
diamondlike carbon. Phys. Rev. B 2001,64, 075414:1ꢀ13.
86) Casari, C. S.; Li Bassi, A.; Baserga, A.; Ravagnan, L.; Piseri, P.; Lenardi, C.; Tommasini,
(
M.; Milani, A.; Fazzi, D.; Bottani, C. E.; Milani, P. Lowꢀfrequency modes in the Raman
spectrum of sp− sp nanostructured carbon. Phys. Rev. B 2008, 77, 195444ꢀ1:7.
2
(87)Walters, J. K.; Kühn, M.; Spaeth, C.; Fischer, H.; Richter, F.; R. J. Newport. Neutronꢀ
diffraction studies of amorphous CN materials. Phys. Rev. B 1997,56, 14315ꢀ14320
x
(
88) Clark, S. J.; Crain, J.; Ackland, G. J. Comparison of bonding in amorphous silicon and
carbon. Phys. Rev. B 1997, 55, 14059ꢀ14062.
89) Li, F.; Lannin, J. S. Radial distribution function of amorphous carbon. Phys. Rev. Lett. 1990,
5,1905ꢀ1908.
90) Titantah, J.T.; Lamoen, D. Carbon and nitrogen 1s energy levels in amorphous carbon
nitride systems: XPS interpretation using firstꢀprinciples. Diamond Related Mater. 2007,16, 581ꢀ
88.
91) Bartle, K. D.; Perry, D. L.; Wallace, S. The functionality of nitrogen in coal and derived
liquids: an XPS study. Fuel Process. Technol. 1987,15, 351ꢀ361.
92) Kelemen, S. R.; Gorbaty, M. L.; Kwiatek, P. J. Quantification of nitrogen forms in Argonne
premium coals. Energy & Fuels. 1994,8, 896ꢀ906.
93) Hellgrena, N.; Haaschb, R. T.; Schmidtc, S.; Hultmanc, L.; Petrov, I. Interpretation of Xꢀ
(
6
(
5
(
(
(
ray photoelectron spectra of carbonꢀnitride thin films: New insights from in situ XPS. Carbon
2016, 108, 242ꢀ252.
(
CꢀN binding states in carbon nitride films deposited by reactive XeCl laser ablation. Thin Solid
Films 1999,347, 72ꢀ77.
94) E D'Anna, E.; De Giorgi, M.L.; Luches, A.; Martino, M.; Perrone, A.; Zocco, A. Study of
(
95) Echꢀchamikh, E.; Essafti, A.; Ijdiyaou, Y.; Azizan, M. XPS study of amorphous carbon
nitride (aC:N) thin films deposited by reactive RF sputtering. Solar Energy Mater. Solar Cells
006, 90,1420ꢀ1423.
2
25
ACS Paragon Plus Environment

Chemistry of Materials
Page 26 of 27
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
96) Pels, J. R., Kapteijn, F., Moulijn, J. A., Zhu, Q., Thomas, K. M. Evolution of nitrogen
functionalities in carbonaceous materials during pyrolysis. Carbon 1995, 33, 1641ꢀ1653.
97) EstradeꢀSzwarckopf, H. XPS photoemission in carbonaceous materials: A “defect” peak
beside the graphitic asymmetric peak. Carbon 2004, 42, 1713ꢀ1721.
98) Zheng, W. T.; Xing, K. Z.; Hellgrena, N.; Lögdlunda, M.; Johanssona, Å.; Gelivsb, U.;
(
(
Salanecka, W.R.; Sundgrena, J.ꢀE. Nitrogen 1s electron binding energy assignment in carbon
nitride thin films with different structures. J. Electron Spectros. Rel. Phenom, 1997,87, 45ꢀ49.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(99) Hellgren, N.; Johansson, M. P.; Broitman, E.; Hultman, L.; Sundgren, J. E. Role of nitrogen
in the formation of hard and elastic CN thin films by reactive magnetron sputtering. Phys. Rev.
B 1999,59, 5162ꢀ5169.
x
(
100) Ronning, C.; Feldermann, H.; Merk, R.; Hofsass, H.; Reinke, P.; Thiele, J. U. Carbon
nitride deposited using energetic species: a review on XPS studies. Phys. Rev. B 1998,58, 2207ꢀ
215.
2
(101) Ripalda, J. M.; Román, E.; Díaz, N.; Galán, L.; Montero, I.; Comelli, G.; Baraldi, A.;
Lizzit, S.; Goldoni, A.; Paolucci, G. Correlation of Xꢀray absorption and Xꢀray photoemission
spectroscopies in amorphous carbon nitride. Phys. Rev. B 1999, 60, 3705ꢀ3708.
(102) Braun, A.; Huggins, F.E.; Shah, N.; Chen, Y.; Wirick, S.; Mun, S.B.; Jacobsen, C.;
Huffman, G.P. Advantages of soft Xꢀray absorption over TEMꢀEELS for solid carbon studies––a
comparative study on diesel soot with EELS and NEXAFS. Carbon 2005,43, 117ꢀ124.
(103) Qu, L.; Liu, Y.; Baek, J.; Dai, L. Nitrogenꢀdoped graphene as efficient metalꢀfree
electrocatalyst for oxygen reduction in fuel cells. ACS Nano, 2010,4, 1321ꢀ1321
(
104) Geng, D.; Chen, Y.; Chen, Y.; Li, Y.; Li, R.; Sun, X.; Ye, S.; Knights, S. High oxygenꢀ
reduction activity and durability of nitrogenꢀdoped grapheme. Energy Environ. Sci. 2011,4, 760ꢀ
64.
105) Wang, H.; Maiyalagan, T.; Wang, X. Review on recent progress in nitrogenꢀdoped
graphene: synthesis, characterization, and its potential applications. ACS Catal. 2012,2, 781ꢀ794
106) Warner, J. H.; Lin, H. C.; He, K.; Koshino, M.; Suenaga, K. Stability and spectroscopy of
single nitrogen dopants in graphene at elevated temperatures. ACS Nano, 2014, 8, 11806ꢀ11815.
107) Arenal, R.; March, K.; Ewels, C. P.; Rocquefelte, X.; Kociak, M.; Loiseau, A.; Stephan, O.
7
(
(
(
Atomic configuration of nitrogenꢀdoped singleꢀwalled carbon nanotubes. Nano lett. 2014,14,
5509ꢀ5516
(108) Lin, Y.; Teng, P.; Yeh, C.; Koshino, M.; Chiu, P. ; Suenaga, K. Structural and chemical
dynamics of pyridinicꢀnitrogen defects in grapheme. Nano lett. 2015,15, 7408ꢀ7413.
(109) Gupta, N. K.; Peng, B. ; Haller, G. L.; Ember, E. E.; Lercher, J. A. Nitrogen Modified
Carbon NanoꢀMaterials as Stable Catalysts for Phosgene Synthesis. ACS Catal. 2016,6, 5843ꢀ
5
855.
110) Brydson, R.; Zhili, Z.; Brown, A. Revisiting the determination of carbon sp /sp ratios via
analysis of the EELS carbon Kꢀedge. Vol. 1: Instrumentation and Methods 2008,12, 357ꢀ358.
111) Deslandes, A.; Jasieniak, M.; Ionescu, M.; Shapter, J. G.; Fairman, C.; Gooding, J. J.;
2
3
(
(
Quinton, J. S. ToF SIMS characterisation of methane and hydrogen plasma modified graphite
using principal component analysis. Surface and Interface Analysis 2009, 41, 216ꢀ224.
26
ACS Paragon Plus Environment

Page 27 of 27
Chemistry of Materials
1
2
3
4
5
6
7
8
9
(
112) Huang, L. J.; Hung, Y.; Chang, S.; Massoumi, G. R.; Lennard, W. N.; Mitchell, I. V.
Characterization of carbon and carbon nitride thin films using timeꢀofꢀflight secondaryꢀionꢀmass
spectrometry. J. Vacuum Science Technology a-Vacuum Surfaces and Films 1997, 15, 2196ꢀ
2201.
(113) Dong, H.; Oganov, A.R.; Zhu, W.; Qian, G. Prediction of a new ground state of superhard
compound B6O at ambient conditions. Sci. Rep.2016, 5, 9870ꢀ9874.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
2
114) Kroke, E. gtꢀC N ꢀThe First Stable Binary Carbon (IV) Nitride, Angew. Chem., Int. Ed.
014,53, 11134ꢀ11135.
3
4
(115) Pickard, C. J.; Salamat, A.; Bojdys, M. J.; Needs, R. J.; McMillan, P. F. Carbon nitride
frameworks and dense crystalline polymorphs. Phys. Rev. B 2016, 94, 094104:1ꢀ10.
(116) Zhu, Q.; Oganov, A.R.; Lyakhov, A.O. Evolutionary metadynamics: a novel method to
predict crystal structures. Cryst. Eng. Comm. 2012, 14, 3596ꢀ3601.
(117) Oganov, A.R., Glass, C.W. Crystal structure prediction using ab initio evolutionary
techniques: Principles and applications. J. Chem. Phys. 2006, 124, 244704.
(118) Lyakhov, A.O.; Oganov, A.R.; Stokes, H.T.; Zhu Q. New developments in evolutionary
structure prediction algorithm USPEX. Comp. Phys. Comm. 2013. 184, 1172ꢀ1182.
Table of Contents artwork
27
ACS Paragon Plus Environment