The many phases of CaC2

Article abstract of DOI:10.1016/j.jssc.2016.04.030

Polymorphic CaC₂ was prepared by reacting mixtures of CaH₂ and graphite with molar ratios between 1:1.8 and 1:2.2 at temperatures between 700 and 1400 °C under dynamic vacuum. These conditions provided a well controlled, homogeneous, chemical environment and afforded products with high purity. The products, which were characterized by powder X-ray diffraction, solid state NMR and Raman spectroscopy, represented mixtures of the three known polymorphs, tetragonal CaC₂-I and monoclinic CaC₂-II and -III. Their proportion is dependent on the nominal C/CaH₂ ratio of the reaction mixture and temperature. Reactions with excess carbon produced a mixture virtually free from CaC₂-I, whereas high temperatures (above 1100 °C) and C-deficiency favored the formation of CaC₂-I. From first principles calculations it is shown that CaC₂-I is dynamically unstable within the harmonic approximation. This indicates that existing CaC₂-I is structurally/dynamically disordered and may possibly even occur as slightly carbon-deficient phase CaC_2-δ. It is proposed that monoclinic II is the ground state of CaC₂ and polymorph III is stable at temperatures above 200 °C. Tetragonal I represents a metastable, heterogeneous, phase of CaC₂. It is argued that a complete understanding of the occurrence of three room temperature modifications of CaC₂ will require a detailed characterization of compositional and structural heterogeneities within the high temperature form CaC₂-IV, which is stable above 450 °C. The effect of high pressure on the stability of the monoclinic forms of CaC₂ was studied in a diamond anvil cell using Raman spectroscopy. CaC₂-II and -III transform into tetragonal CaC₂-I at about 4 and 1GPa, respectively.

Full text of DOI:10.1016/j.jssc.2016.04.030

Author’s Accepted Manuscript
The Many Phases of CaC
2
Sumit Konar, Johanna Nylén, Gunnar Svensson,
Diana Bernin, Mattias Edén, Uwe Ruschewitz,
Ulrich. Häussermann
www.elsevier.com/locate/yjssc
PII:
S0022-4596(16)30162-1
http://dx.doi.org/10.1016/j.jssc.2016.04.030
YJSSC19376
DOI:
Reference:
To appear in:
Journal of Solid State Chemistry
Received date: 28 February 2016
Revised date: 19 April 2016
Accepted date: 19 April 2016
Cite this article as: Sumit Konar, Johanna Nylén, Gunnar Svensson, Diana
Bernin, Mattias Edén, Uwe Ruschewitz and Ulrich. Häussermann, The Many
Phases
of
CaC ,
Journal
of
Solid
State
Chemistry
2
http://dx.doi.org/10.1016/j.jssc.2016.04.030
This is a PDF file of an unedited manuscript that has been accepted fo
publication. As a service to our customers we are providing this early version o
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain

The Many Phases of CaC2
Sumit Konar,1 Johanna Nylén,1 Gunnar Svensson,1 Diana Bernin,1 Mattias Edén,1 Uwe
Ruschewitz,2 Ulrich Häussermann1,*
1Department of Materials and Environmental Chemistry, Stockholm University, SE-10691
Stockholm, Sweden
2Department of Chemistry, University of Cologne, Greinstrasse 6, 50939 Cologne, Germany
Abstract
Polymorphic CaC2 was prepared by reacting mixtures of CaH2 and graphite with molar ratios
between 1:1.8 and 1:2.2 at temperatures between 700 and 1400 °C under dynamic vacuum.
These conditions provided a well controlled, homogeneous, chemical environment and
afforded products with high purity. The products, which were characterized by powder X-ray
diffraction, solid state NMR and Raman spectroscopy, represented mixtures of the three
known polymorphs, tetragonal CaC2-I and monoclinic CaC2-II and -III. Their proportion is
dependent on the nominal C/CaH2 ratio of the reaction mixture and temperature. Reactions
with excess carbon produced a mixture virtually free from CaC2-I, whereas high temperatures
(above 1100 °C) and C-deficiency favored the formation of CaC2-I. From first principles
calculations it is shown that CaC2-I is dynamically unstable within the harmonic
approximation. This indicates that existing CaC2-I is structurally/dynamically disordered and
may possibly even occur as slightly carbon-deficient phase CaC2-. It is proposed that
monoclinic II is the ground state of CaC2 and polymorph III is stable at temperatures above
200 °C. Tetragonal I represents a metastable, heterogeneous, phase of CaC2. It is argued that a
complete understanding of the occurrence of three room temperature modifications of CaC2
will require a detailed characterization of compositional and structural heterogeneities within
the high temperature form CaC2-IV, which is stable above 450 °C. The effect of high pressure
on the stability of the monoclinic forms of CaC2 was studied in a diamond anvil cell using
Raman spectroscopy. CaC2-II and -III transform into tetragonal CaC2-I at about 0.5 and 2.5
GPa, respectively.
Keywords: acetylide carbides; polymorphism; structural stability
*email: Ulrich.Haussermann@mmk.su.se
1

I. Introduction
Calcium carbide, CaC2, is one of the most recognized inorganic compounds. In textbooks
it is often introduced as the prototype for salt-like carbides. CaC2 is a so-called acetylide,
2-
featuring C2 dumbbell anions isoelectronic to dinitrogen [1–3]. The simple tetragonal CaC2
type structure, in which dumbbells and Ca ions are arranged as in the NaCl structure, is very
prominent and probably known to any chemistry student. A century ago CaC2 represented a
major commodity chemical, used in the production of acetylene and the fertilizer calcium
cyanamide. Acetylene from coal derived CaC2 remained a main source for industrial
production of organic chemicals until the 1950s [4]. Today, CaC2 is still an important
commodity chemical, especially for China which has an annual production exceeding 20
million tons [5]. Its potential as reagent for current organic synthesis and nanotechnology has
recently been reviewed by Rodygin et al. [6].
During the first half of the last century the structural and physicochemical properties of
CaC2 were intensively investigated, which revealed a stunningly complex polymorphism.
Apart from a tetragonal form (CaC2-I, with the famous CaC2 type structure) [7] two
monoclinic phases were identified (II and III) [8]. All three phases exist at room temperature.
In addition, a high temperature form (IV) stable above 450 °C was established [9]. Significant
efforts were undertaken to understand the thermodynamic relationships of the four crystal
phases [7–9]. This included studies of the effects of oxygen, sulfur and nitrogen impurities in
stabilizing the three low temperature forms [9]. The structural characterization of the
monoclinic phases proved challenging [10] and remained inconclusive until around the year
2000 [11–13]. Today, the thermodynamic relations between the CaC2 polymorphs are still
uncertain. It is puzzling that the preparation of CaC2 typically affords a mixture of room
temperature phases [11,13]. The coexistence of several phases may be due to very small
enthalpy differences between the polymorphs. Alternatively, one could suspect that, because
of (small) compositional differences between the polymorphs, CaC2 does not represent a
single-component system.
From around the year 2010 CaC2 has also become the focus for theoretical investigations,
which uncovered further peculiarities. First, Kulkarni et al. explored the enthalpy landscape of
CaC2 by employing a simulated annealing algorithm and identified two more acteylide
structures that energetically closely compete with the established polymorphs I, II, III [14].
Further, computational studies showed that with pressure, acetylide structures (with C2
dumbbell units) become enthalpically unfavorable against polymorphs with novel polymeric
carbon structures [15,16]. The experimental verification of the existence of such polycarbides
is yet outstanding [17]. Most intriguing, however, is a recent report that questions the
enthalpic stability of the composition CaC2 in the ambient pressure Ca–C phase diagram [18].
From variable-composition evolutionary searches it was found that any CaC2 polymorph
should undergo phase segregation into a new carbide Ca2C3 and C. In light of this finding all
room temperature forms of CaC2 would be only metastable!
In this work we aim to shed some new light into the polymorphism of CaC2. We show that
tetragonal CaC2-I (with the famous type structure) is actually dynamically unstable. We argue
that its existence is only understandable when assuming CaC2-I disordered and/or even
slightly off-stoichiometric. We then report on the synthesis of CaC2 from reactions employing
CaH2 and graphite mixtures, showing that the distribution of room temperature polymorphs
depends on employed C/CaH2 ratio, dwelling temperature and dwelling time. We propose that
for a one-component system CaC2 the tetragonal form I is metastable at all temperatures apart
from a narrow interval prior transformation to the high temperature form CaC2-IV.
2

II. Experimental Section
Synthesis. All steps of sample preparations were carried out inside a glove box (Ar
atmosphere, O2 and H2O level <0.1 ppm). Starting materials were elemental Ca (Aldrich,
distilled, dendritic pieces, 99.99%, stored under Ar) and graphite powder (99.9995%, ABCR).
First CaH2 was produced by reacting Ca with pressurized H2 gas (20 bar) in a stainless steel
autoclave at 300 °C for 12 h. Graphite powder was degassed at 900 °C under dynamic
vacuum for 6 h prior to use. For CaH2–graphite reactions tantalum ampoules (L ~ 50 mm, OD
= 10 mm) were employed as containers. Calcium hydride and graphite powder were
thoroughly mixed with specific molar ratios (between 1:1.8 and 1:2.2) in an agate mortar and
subsequently pressed into pellets (~200 mg of weight). Reactant pellets were sealed in Ta
ampoules by welding. Ampoules were subsequently placed in fused silica jackets (L = 400
mm, OD = 26 mm).
For reactions at temperatures up to 1100 °C silica jackets were positioned in a vertical tube
furnace and connected to a vacuum line. The empty space of the furnace was filled with silica
insulation wool and a K-type thermocouple was introduced parallel to the silica jacket to
monitor temperature close to the location of the metal ampoule. Ampoules were heated to a
set temperature (between 700 and 1100 °C) at a rate of 10 °C/min and dwelled for 6 h and 24
h. During heating and dwelling ampoules were kept in a dynamic vacuum (p < 10-3 mbar),
which was applied to promote the release and removal of H2. Ta will become permeable to
hydrogen at temperatures above 600 °C. Reactions at temperatures higher than 1100 °C were
carried out in a RF furnace. Here, temperatures were monitored by a pyrometer.
At the end of the desired dwell time, ampoules were quenched to room temperature and
opened in the glove box. Typically, the initial shape of the reactant pellet remained intact after
reaction (see supporting information, Figure S1). The surface of the metal ampoule in contact
with the pellet during the reaction was carefully inspected, but never found to be attacked.
Recovered pellets were weighed (mg accuracy) and lightly ground for subsequent
characterization and further experiments.
Scanning electron microscopy (SEM): Powder samples were deposited on a holey carbon film
on a Cu grid and transferred to the sample holder which was then placed in a closed container
inside glove box. Sample was transferred to the microscope with air exposure less than 5 sec.
SEM images were obtained using a JEOL JSM-7000F electron microscope at an acceleration
voltage of 5 kV.
Powder X-ray diffraction. Powder X-ray diffraction (PXRD) patterns were collected on a
Panalytical X’Pert alpha1 diffractometer (Cu K radiation, λ=1.540598 Å) at room
temperature. Moisture sensitive powder samples were mounted on a Si wafer zero background
holder inside the glove box and covered with a Kapton foil. Diffraction patterns were
measured in a 2θ range 20 – 90° with a step size of 0.013°. Rietveld analysis of PXRD data
was performed using the program package FULLPROF [19]. Rietveld refinement proceeded
according to the following strategy: (1) Fit of the background using linear interpolation
between a set of background points; (2) Fit of the peak shape by Thompson-Cox-Hastings
pseudo-Voigt convoluted with axial divergence asymmetry function for individual phases; (3)
Spherical harmonics expansion of the crystallites shape function was employed for the major
phase in a multi-phase refinement (size model = 15 for monoclinic phases and size model =
21 for the tetragonal phase) to reduce preferred orientation in a flat plate geometry; (4) To
obtain a reliable phase fraction analysis, displacement parameters (Biso) of like atoms in
different phases were constrained to be the same.
3

NMR spectroscopy. Solid state 13C NMR spectra were acquired at 9.4 T for two different
CaC2 specimens (“C1.8” and C”2.2”) using a Bruker Avance-III spectrometer (–100.6 MHz
13C Larmor) equipped with a 7 mm double-resonance magic angle spinning (MAS)
probehead. Finely ground powders were packed in 7 mm zirconia rotors inside a glove box
13
(Ar atmosphere) and were spun at 7.0 kHz during the NMR experimentation. The C NMR
acquisitions involved single-pulse excitation with the radiofrequency (rf) pulses (60° flip
angle) operating at a nutation frequency of 47 kHz. The data collection involved a 5 h
equilibration delay, 45 minutes relaxation delay and a total of 56 (for C1.8) and 84 (for C2.2)
accumulated signal transients. A Gaussian window function was applied prior to Fourier
transformation of the recorded time-domain NMR signal. The absence of protons in
1
polymorphic CaC2 obtained from CaH2–graphite reactions was verified by H NMR
experiments performed at 14.1 T and 24.0 kHz MAS on four samples obtained from different
synthesis conditions, using a Bruker Avance-III spectrometer equipped with a 3.2 mm
probehead. The 13C and 1H rf pulses were calibrated on a powder of adamantane. 13C chemical
shifts (in ppm) are quoted relative to neat tetramethylsilane (TMS).
Raman spectroscopy. Raman spectra were measured using a Labram HR 800 spectrometer.
The instrument is equipped with an 800 mm focal length spectrograph and an air cooled (−70
oC), back thinned CCD detector. Samples were excited using an air cooled double frequency
Nd:YAG laser (532 nm) and an input laser power of 0.56 mW. Raman spectra were collected
with an exposure time of 60 s, accumulation number of 10, and using a 1800 grooves/mm
grating. In situ high pressure Raman spectroscopy experiments were performed at room
temperature with a screw-driven Diamond Anvil Cell (DAC) DiaCell DXR-6 using a culet
size of 300 microns. Powdered samples were packed into a 100–150 micron-sized hole drilled
in a stainless steel gasket. No pressure transmitting medium was employed because of the
interference from its luminescence and potential reactivity with the sample. CaC2 was
considered sufficiently soft to achieve satisfactory hydrostatic conditions, and care was taken
to measure spectra from sample areas in proximity to the ruby chips, which were used as
pressure indicators. Pressures were calculated from the fluorescence line shift according to
Mao et al. [20]. The pressure was increased at steps of about 0.5 GPa and Raman spectra
measured after a 15 min equilibration time.
Computations. Electronic structure calculations of polymorphs of CaC2 were performed using
the first principles all-electron projector augmented waves (PAW) method [21] as
implemented in the Vienna Ab Initio Simulation Package (VASP) [22]. Exchange-correlation
effects were treated within the generalized gradient approximation (GGA) using the Perdew-
Burke-Ernzerhof (PBE) parameterization [23]. The structures were relaxed with respect to
pressure, lattice parameters, and atomic positions. Forces were converged to better than 1×10-
3
eV/Å. The energy cutoff was set to 675 eV. The integration over the Brillouin zone was
done on a grid of special k-points of size 11×11×11 determined according to the Monkhorst-
Pack scheme [24] and the tetrahedron method with Blöchl correction [25]. Zone-centered
phonon calculations were performed using VASP’s density functional perturbation theory
approach. Extended phonon calculations were carried out for CaC2-I, -II, and -III using the
Quantum Espresso package [26] using a 6×6×6 k-point and 3×3×3 q-point grid according to
Monkhorst-Pack [24] and the PBE exchange-correlation [23]. The plane-wave cut off was set
at 65 Ry. Ultrasoft pseudopotentials used were generated by the Vanderbilt method with
3s23p64s2 and 2s22p2 as valence electrons for Ca and C, respectively [27].
4

III. Results and Discussion
The enigma of CaC2 phases. The polymorphism of CaC2 has been known for a long time
[7–9]. The existence at room temperature of three different crystal modifications of CaC2 was
established in 1937. In 1942 the discovery of a fourth modification, stable above 450 °C, was
reported. Whereas the “famous” tetragonal structure of CaC2-I was already determined in
1930 [28], the elucidation of the monoclinic forms II and III, either from twinned single
crystal X-ray diffraction data or multi-phase PXRD patterns, proved difficult [10]. Around the
year 2000 Meyer and co-workers [11,12] and Knapp et al. [13] finally reported on the
conclusive structure characterization of CaC2-II and –III.
Figure 1 displays the structures of CaC2-I, CaC2-II, and CaC2-III. Table 1 compares
structure parameters and interatomic distances. For this comparison we use the structures
obtained from computational relaxation (“DFT optimized” structures) which provides
structural parameters on the same ground. Experimental structures, which are discussed later
on, may be biased by the presence of disorder and/or compositional variations.
The structure of CaC2-I (space group I4/mmm, Z = 2) corresponds to a NaCl-type
arrangement of Ca ions and C2 acetylide dumbbells, which is tetragonally distorted due to the
alignment of dumbbells along one axis (i.e., the tetragonal c axis). In the monoclinic
structures the overall NaCl type, six-fold, coordination of dumbbells and Ca ions is
maintained. However, tetragonally elongated Ca6 octahedra (hosting dumbbells) are distorted
and, additionally, with respect to their long axis, oriented in two directions. In CaC2-II (space
group C2/c, Z = 4) dumbbells within the Ca6 octahedra are exclusively aligned approximately
along the long axis (Figure 1b). In CaC2-III (space group C2/m, Z = 4) half of the dumbbells
(those formed by the C2 atoms) are aligned toward opposite edges. When comparing
interatomic distances, the monoclinic forms seem to provide a more efficient Ca–C
coordination (that is, some Ca–C distances are shorter compared to tetragonal CaC2-I). At the
same time the mutual packing of Ca and dumbbells is slightly less effective. This is indicated
in the molar volume, which is 1.3% and 1.5% larger for CaC2-II and -III, respectively,
compared to CaC2-I. Lastly, the high temperature polymorph CaC2-IV adopts a cubic NaCl
type arrangement in which dumbbell units appear orientationally disordered [9a,13]. At 490
°C the molar volume of CaC2-IV is by about 2 % larger compared to CaC2-I [13].
Typically the synthesis of CaC2 yields a mixture of polymorphs. For example, technical
CaC2, which is prepared by reacting CaO or CaCO3 with carbon at high temperatures (2200
°C and higher) in an electric arc furnace, represents a blend of tetragonal I and monoclinic II
[13]. The high temperature reaction CaO + 3C = CaC2 + CO was extensively investigated by
Franck et al. [7,8] and Bredig [9]. It was established that the proportion of polymorphs can be
influenced by reaction conditions. “Excess of calcium” (or carbon deficient) reactions
afforded pure samples of CaC2-I. The formation of CaC2-I is also promoted by the presence of
impurities (being able to dissolve in CaC2 as solid solutions, like CaS or CaCN2 [7,8]). In the
absence of S and N impurities (and employing stoichiometric reactions) CaC2-I could be
completely suppressed and pure samples of CaC2-III were obtained [9]. With respect to
impurities, it is important to mention that CaO does not influence the stability of any CaC2
polymorph. It is practically insoluble in CaC2 [29]. The influence of the C/Ca ratio on
polymorph distribution was also observed for the formation reaction Ca + 2 C = CaC2, using
either sugar charcoal or graphite as carbon source [7,8]. Formation reactions can be performed
at lower temperatures. The temperature interval 925 – 1250 °C typically affords a mixture of I
and II. Franck et al. [7,8] and later Reckeweg et al. [11] showed that carbon deficiency and
excess favor the formation of CaC2-I and -II, respectively, whereas CaC2-III was not obtained
in this temperature interval. In contrast, Knapp et al. reported significant yields of CaC2-III
when carrying out formation reactions at 800 °C and employing conditions with excess C
5

[13]. As a matter of fact, CaC2-I, -II, and -III were obtained in almost equal proportions. With
this we conclude our review of earlier CaC2 synthesis experiments. An important point to
consider – and not so done earlier – is that the synthesis of CaC2 (at high temperatures)
initially leads to the formation of the high temperature phase CaC2-IV, from which the room
temperature stable polymorphs I, II, III (or rather their mixtures) somehow evolve.
In view of the puzzling outcome of the CaC2 synthesis experiments it appears difficult to
establish thermodynamic relations between the polymorphs. However, Bredig already in 1942
provided a splendid analysis, assuming that CaC2 is a one-component system [9a]. Later
Glaser et al. [12] and Knapp et al. [13] confirmed parts of Bredig’s analysis, and also added
new features. We show the relations in Figure 2. Interestingly, not only temperature but also
mechanical deformation (grinding) can induce transitions between polymorphs. Bredig
reported that pure CaC2-III (obtained from high temperature arc furnace reactions CaO + 3C =
CaC2 + CO) transformed quantitatively into CaC2-II when grinding at room temperature [9a].
Grinding CaC2-III at elevated temperatures, however, yielded CaC2-I. Later Glaser et al.
confirmed the mechanically induced III to II transition [12].
Turning to temperature induced transitions, Bredig showed that CaC2-II, obtained from
grinding CaC2-III, re-transforms to III upon heating [30]. Similarly, Glaser et al. obtained
pure samples of CaC2-III from temperature cycling mixtures of CaC2-I/II to 520 – 540 °C
[12]. Accordingly, there is no II-to-IV transition. Only CaC2-I and -III transform to CaC2-IV.
From DSC measurements, Glaser et al. obtained distinct temperatures, 480 and 460 °C for the
I-to-IV and the III-to-IV transition, respectively [12]. Transitions of CaC2-I into either CaC2-II
or -III are uncertain although Knapp et al. observed a partial transformation of I into II in a
low temperature [83 K (−190 °C)] PXRD experiment [13]. Today, Bredig’s analysis [9] is still
assumed valid and CaC2-II is considered the low temperature ground state of CaC2. CaC2-I is
regarded an ordered high temperature (HT) form, CaC2-IV is a disordered HT form, and
CaC2-III is an ordered metastable form.
Energies of formation of CaC2 phases. The evaluation of the structure stability of the
ordered forms provides instructive insight into the polymorphism of CaC2. Recently, Kulkarni
et al. performed an extensive search for structure candidates for CaC2 and identified in
addition to known CaC2-I, II, III, (IV) an orthorhombic (CaC2-V) and a further monoclinic
structure (CaC2-VI) as viable ambient pressure alternatives [14]. CaC2-V and -VI can be
derived from tetragonal CaC2-I by a simple rotational distortion of the C2 dumbbells. Figure 3
depicts the energies of formation E(CaC2) = E(CaC2) − E(fcc-Ca) − E(graphite) (referring to
the formation reaction Ca + 2 C(graphite) = CaC2) for the various polymorphs based on our
DFT calculations. We also include the recently proposed polycarbide Cmcm-CaC2 which is
the enthalpically most stable CaC2 polymorph at pressures 1 – 12 GPa [15].
All polymorphs have negative energies of formation, that is, they represent
thermodynamically stable compounds with respect to the constituting elements (note that this
is not the case for MgC2 [31]). It is remarkable that the enthalpies of the six phases differ by
no more than 25 meV/formula unit (about 2.4 kJ/mol). And one can easily envision that more
phases with (hypothetical) structures varying in the orientation of C2 dumbbells may fall in
this energy window. According to these calculations, the ground state corresponds to CaC2-III
which is essentially degenerate with CaC2-II. Although enthalpy differences are small, it is
surprising that tetragonal CaC2-I comes out as the least stable in this selection.
We therefore investigated the dynamic stability of CaC2-I by calculating phonon
dispersion relations (within the harmonic approximation). The criterion for a dynamically
stable structure is real eigenvalues (phonon energies/frequencies) for all k(q) points in the
Brillouin zone. This is the case for the monoclinic forms CaC2-II and -III. However, for
tetragonal CaC2-I it is surprisingly not. Figure 4 shows the phonon dispersion relations for
6

CaC2-I, where imaginary frequencies are represented as negative numbers. In particular, the
transverse acoustic phonon branch is instable for the complete symmetry line D ( – X)
[which expresses the [110] polarization] and large parts of the line W (X – P). The imaginary
zone boundary X implies that CaC2-I is also elastically instable and one may deduce a
monoclinic distortion.
This result is unexpected: the famous tetragonal CaC2 structure type is not stable for
CaC2! Importantly, the dynamic instability becomes more pronounced with pressure (cf.
Figure 4, red curve, 10 GPa). There may be several reasons why the harmonic approximation
gives instable phonons. One possibility is that imaginary phonon modes will be
anharmonically stabilized at high temperatures. The dynamic stabilization of structures with
temperature has been demonstrated for transition metals, e.g. the fcc structures of Mo and W
[32]. This would imply that CaC2-I is only stable at high temperatures and with pronounced
anharmonic dynamics. Another possibility is that the actual structure of CaC2-I deviates from
the ideal model by e.g. the presence of structural disorder, defects, or even non-stoichiometry.
Synthesis and characterization of CaC2 from CaH2 and graphite. In light of the
disparate results of previous CaC2 synthesis experiments we decided to investigate reactions
between calcium hydride and graphite, according to CaH2 + 2C = CaC2 + H2. Our intention
was to perform carbide synthesis in a well controlled chemical environment, which at the
same time allows for low synthesis temperatures, and to compare against previous results
from CaO + C and Ca + C reactions. Employing CaH2 as Ca source has the advantage that
both reactants represent powders that can be intimately mixed and pressed to homogenous
pellets, thus allowing for stoichiometric control of C/Ca ratios. At temperatures around 700
°C CaH2 will decompose cleanly to Ca with a high surface area which greatly benefits
kinetics. Alkali and alkaline earth metal hydrides have been used previously as precursors for
the preparation of (Zintl phase) silicides and germanides [33], and also Li2C2 [34], which
afforded products at lower temperatures and with higher purity.
We considered the graphite/CaH2 molar ratios 1.8:1, 2:1, and 2.2:1 (slight excess of Ca,
stoichiometric, and slight excess of C). The temperature interval investigated was 700 – 1100
°C. Employed dwelling times were 6 and 24 hours. Additionally, reactions with short
dwelling periods (15 min) were carried out at 1400 °C. After dwelling, reactions were
quenched to room temperature and the products analyzed. The excess of Ca and C in the
mixtures 1.8:1 and 2.2:1, respectively, was not recognizable in PXRD patterns (see supporting
information, Figure S2). A dynamic vacuum was applied to promote the release and removal
of H2. No mass losses exceeding those expected from H2 release occurred.
So obtained CaC2 was virtually free from impurities. Notably, 1H MAS NMR revealed no
resonances beyond those associated with background signals from an empty rotor. If present
in trace amounts, the proton content of CaC2 obtained from CaH2–graphite reactions is below
100 wt. ppm. The samples represented a mixture of polymorphs. The proportions are clearly
dependent on the C/CaH2 ratio, but also dwelling temperature and time. From the color of the
product, the major phase could be assessed: CaC2-I black, CaC2-II grey, CaC2-III white.
Figure 5 shows representative Rietveld plots and the colors of products obtained from 1400
°C reactions (a complete compilation is given as supporting information, Figure S3). Figure 6
summarizes the phase fraction of CaC2 polymorphs in the products of the various reactions, as
determined by Rietveld refinement (for details, see supporting information Table S1). Three
scenarios can be distinguished: (i) Products from stoichiometric and C excess reactions were
almost free from CaC2-I when employing high temperatures (1100 °C and 1400 °C) and also
with the condition 900 °C/24 h. The ratios of II/III were very similar, perhaps there is a
tendency that excess C slightly favors the formation of CaC2-III. (ii) The formation of CaC2-I
was clearly favored in C deficient reactions at high temperatures (1100 °C and 1400 °C).
7

However, CaC2-I was essentially suppressed in the C deficient reaction 900 °C/24 h.
Generally, 900 °C/24 h reactions did not yield significant amounts of CaC2-I, instead, there is
an increase of the fraction of CaC2-III when going from C deficient to C excess. (iii) At lower
temperatures (700 °C and 900 °C) the polymorph distribution depends on the dwelling time.
This indicates that at these temperatures initially formed CaC2-IV (which is the stable
polymorph at temperatures above 450 °C) is “heterogeneous” and “homogenizes” (or
equilibrates) over time (cf. dwellings for 6 h and 24 h). It also indicates that heterogeneous
CaC2-IV favors the formation of CaC2-I.
Our synthesis reactions employing CaH2 confirm unambiguously the influence of the
C/Ca ratio on the proportion of polymorphs in the product, as described earlier by Bredig [9a]
and Reckeweg et al. [11]. Further, our reactions allow the synthesis of CaC2 at exceptionally
low temperatures (700 °C) which in turn revealed a dwelling time dependency of the
polymorph distribution.
We now turn to the analysis of the crystal structures of CaC2-I, -II, -III. The results from
our refinements are compiled in Tables 2 and 3, and compared with the structures previously
reported by Meyer and co-workers [11,12] and Knapp et al. [13]. (For details, see supporting
information). The literature structures differ somewhat for CaC2-II and CaC2-III, and our
results seem to be more similar to Meyer and co-workers. We note that refinement results are
slightly different across the products from the various reactions (see supporting information,
Table S2) indicating problems in obtaining precise structural parameters from Rietveld
refinements of multi-phase samples. When comparing Tables 2 and 3 with Table 1 it is
apparent that the experimental structures of CaC2-II and -III deviate significantly from the
DFT optimized ones (whereas the simple CaC2-I structure agrees well). We believe that these
discrepancies are due to shortcomings in the Rietveld refinements, possibly in conjunction
with structural disorder. It is worth mentioning that in the experimental structures the C–C
distance of dumbbell units ranges from 1.18 – 1.33 Å, whereas DFT calculations yield strictly
1.26 Å, independent of the polymorph.
Figure 7 shows representative SEM micrographs of particles from a CaC2-II/III phase
mixture and from a mixture dominated by CaC2-I. Crystallites are built from lamellae with
about 0.1 micron thickness, which are finely interwoven. For CaC2-I crystallites the lamellar
microstructure is peculiarly curved. The overall texture is most likely influenced by the
microstructure of CaC2-IV from which the room temperature phases evolve. We remind on
the multitude of structures with very similar energies that arise from various dumbbell
orientations (cf. Figure 3) and one can envision broad possibilities for structural disorder in
the room temperature phases. This includes (i) disorder from random alignment of dumbbells
within Ca6 octahedra and (ii) vibrational (dynamic) disorder. But also (iii) C2 defects and/or
2-
replacements 2 C2 ↔ C4- may be considered, which would result in compositional
variations. Traditional Rietveld refinement of CaC2 polymorphs (as also applied in this work)
will not capture the decisive structural features which relate to such disorder. There are
several clear indications for disorder described in the literature, especially for CaC2-I. These
include anomalous peak broadening of h0l and hkl reflections in PXRD patterns [13, 35] and
a pronounced broadening of signals in 13C MAS-NMR spectra [11,12,36].
Figure 8 displays the 13C MAS NMR spectra recorded from specimens with C/CaH2 ratio
of 1.8 and 2.2 that were quenched from 1400 °C; they are in the following referred to “C1.8”
and “C2.2”, respectively. Besides a broad signal centered around 110 ppm that stems from a
2-
minor impurity of graphite, the NMR spectra display resonances from the C2 environments
in three distinct CaC2 polymorphs (I, II and III). Owing to the sizable 13C chemical shift
13
2-
anisotropy [12], the resonance associated with each crystallographically unique C2 pair is
distributed over a set of spinning sidebands that results at the modest MAS rate (7.0 kHz)
utilized herein. The isotropic chemical shifts at each centerband position ~200 ppm in the
8

NMR spectra are shown in Figures 8a and b, whereas Figure 8c assigns each resonance to its
corresponding CaC2 polymorph (I, II, and III).
According to the PXRD analysis, the C1.8 and C2.2 specimens comprises polymorph
ratios I:II:III of 60:20:20 and 3:47:50, respectively (cf. Figure 6). The NMR spectrum from
C2.2 reveals all expected resonances, with that ~200 ppm is assigned to CaC2-II, while the
peaks at isotropic chemical shifts of ~203 ppm and ~197 ppm are attributed to the two
crystallographically distinct C2 dumbbells in CaC2-III. Compared to Glaser et al. [12] we
observe a slightly higher chemical shift (by 2 ppm) for CaC2-II whereas the ones for CaC2-III
accord very well (203 ppm and 197 ppm). The NMR signal with an isotropic chemical shift
~206 ppm is attributed to CaC2-I. This shift value is in excellent agreement with that reported
by Wrackmeyer et al. for “technical grade” CaC2 [36] but is significantly higher than those
(196–199 ppm) reported by Meyer and co-workers [11,12].
Besides the different signal intensities stemming from variable contributions of each CaC2
13
polymorph to the specimens, the primary distinction between the two C NMR spectra is the
overall broader resonances observed from the C1.8 sample relative to those in the C2.2
counterpart; this is evident by comparing the two NMR responses from the CaC2-II
component in Figure 8c. We attribute the resonance broadening to distributions of isotropic
chemical shifts stemming from different extents of structural disorder in the two samples. In
this respect especially the CaC2-I phase appears to be heterogeneous. The reported chemical
shift dispersion from 196 to 206 ppm points to significant structural and/or even
compositional variations. In addition, as already described by Reckeweg et al. [11], there is a
pronounced dependence of 13C NMR peak-width of CaC2-I on applied synthesis conditions.
Next we discuss the results of the Raman spectroscopy analysis, probing sample areas of
1–2 m. The Raman spectra of I, II, III are shown in Figure 9. Additionally, the wavenumbers
of Raman active modes are compiled in Table 4 and compared with the ones obtained from
DFT calculations. The spectra can be divided into two regions: C–C stretches occur at
wavenumbers between 1850 and 1900 cm-1, whereas libration modes of C2 dumbbells and
translation modes, which constitute largely of displacements of the Ca ions, occur at
wavenumbers below 400 cm-1. Just two bands constitute the Raman spectrum of tetragonal
CaC2-I: the stretching mode at 1858 cm-1 and the double degenerate libration at 305 cm-1.
These wavenumbers agree well with the DFT calculated ones (at 1859 and 297 cm-1). And
there is also good agreement between experimental bands and calculated wavenumbers for
CaC2-II and -III. Note that for CaC2-II and -III the primitive unit cell contains two formula
units. The interaction (vibrational coupling) between two dumbbell units in the crystal
(correlation splitting) is clearly resolved in our spectra.
The C–C stretches of CaC2-II and -III are at significantly higher wavenumbers compared
to CaC2-I. The larger unit cell and lower symmetry of the monoclinic forms implies a larger
number of libration and translation modes. The associated bands in the spectrum are broad
and overlapping. The excellent agreement of DFT calculated C–C stretching modes with
experimentally observed bands strongly supports the validity of the DFT optimized structures
and, thus, the presence of uniform C2 dumbbell units (with a C–C distance near 1.26 Å) in all
three polymorphs. The disorder established from NMR on CaC2 bulk samples is, thus,
difficult to reconcile on the m scale. The observation of narrow stretching bands and
correlation splitting implies crystalline order within the probed domains. However, the broad
libration/translation bands (in contrast with sharp stretching ones) indicate significant
anharmonic behavior.
Pressure stability of CaC2 phases. Before returning to the influence of T to phase
relations we also address the pressure stability of CaC2 polymorphs. In recent years this topic
has attracted considerable attention because of a potential pressure induced transformation of
9

the acetylide carbide into a polycarbide [15,16]. Previously it has been reported that tetragonal
CaC2-I can be pressurized up to about 18 GPa with only minor structural changes. At higher
pressures an irreversible amorphization takes place [17,37]. This extended pressure stability
contrasts the results from the phonon calculations which revealed an increasing dynamic
instability of CaC2-I with pressure (cf. Figure 4). The existence of CaC2-I at high pressures is
especially difficult to reconcile without assuming structural/compositional disorder. The high
pressure behavior of the monoclinic forms CaC2-II and -III is not clear. To study their
pressure stability Raman spectra of samples free from CaC2-I were recorded in a DAC.
Figure 10a shows Raman spectra of a II/III mixture with increasing pressure (restricted to
the spectral region of C–C stretches). It is clearly noticeable that with pressure both
monoclinic modifications transform independently into CaC2-I. The transformation of CaC2-
III is seen already at very low pressures (below 0.9 GPa). At 3.6 GPa CaC2-III disappeared
and only II and I coexist. At 5.5 GPa also CaC2-II has completely transformed into I. The
onset of the II-to-I transition is presumably at lower pressures. Note that with pressure bands
of C–C stretches broaden considerably. A rather interesting observation was made upon
decompression (Figure 10b): At around 2 GPa C–C stretching bands of monoclinic CaC2 II
became visible again which indicates the reversibility of the pressure induced II-I transition.
The pressure induced transformation of III and II into I is expected when considering molar
volumes (cf. Table 1). Importantly, the II-to-I transformation and its partial reversal upon
pressure release shows unambiguously that there is a direct phase relation between I and II,
with the latter being the more stable form at ambient conditions.
Relation of CaC2 phases. So, what are the thermodynamic relations between CaC2
phases? Bredig in 1942 developed an apparently reasonable view – based on the assumption
that CaC2 is a one-component system – where II represents the ground state, I a medium
temperature phase, IV the high temperature phase, and III is metastable at all temperatures [9].
This view was questioned by Meyer and co-workers [11,12] who seem to prefer III as stable
medium temperature phase in a sequence II – III – IV and posed the question how I should be
related to this sequence. First of all, it is not clear whether CaC2 actually represents a single-
component system because of possible (small) compositional differences between the
polymorphs. Especially, the preferred formation of CaC2-I from Ca excess reaction mixtures
may suggest that this phase possesses a homogeneity range allowing for slight C deficiency. If
nonetheless assuming a one-component system we suggest an altered view to the one by
Bredig.
Figure 11 depicts a schematic representation of our suggested phase relations. Compared
to Bredig’s view, I and III are interchanged. That is, the sequence of stable phases with
increasing T is II –(~200 °C)– III –(~460 °C) – IV as suggested by Meyer. I is assumed to be
generally metastable, but is presented as a band (with variable free energy (F) width) in the
plot. The band representation accounts for various dynamical and structural scenarios as well
as compositional heterogeneities that may be realized within phase I. This includes also
effects induced by stress and microstrain. The boundary of the band toward low F is
uncertain. For pure samples of CaC2 the F variability of I is assumed narrow, in line with the
width of the band drawn in Figure 11, allowing a small window of thermodynamic stability at
high temperatures slightly below the transformation to IV. The presence of impurities, such as
2-
S2- and CN2 , can stabilize I tremendously (as indicated by a dotted line), making it more
stable than III [7–9].
For reasons unknown, metastable I has a distinct kinetic stability. Upon heating from
room temperature the only transition of I is into IV. Depending on its form (i.e. location
within the F band), the transition temperature varies (between 450 and 480 °C). Most
important is the role of CaC2-IV for understanding the phase relations of CaC2. IV exists over
10

a very wide T range; CaC2 melting is estimated at around 2200 °C [7,8]. We propose that
CaC2-IV has to be considered as “heterogeneous”. That is IV actually comprises several
phases because of various states of disorder in addition to a possible small homogeneity
range. At high temperatures (close to the melting point) dumbbells will be dynamically
disordered, possibly freely rotating in the average NaCl structure. Upon lowering temperature
different states of static or mixed static/dynamic disorder could freeze out, opening for
complex non-equilibrium scenarios. Depending on synthesis conditions (C/Ca ratio, dwelling
temperature and time, annealing rate) the transition into the low temperature form of CaC2
may occur from a different state of IV, preferring either the metastable I or stable III pathway
(red circle in Figure 11). III is known to be kinetically stable with respect to transformation
into II upon cooling. Consequently, II has to form from I upon cooling (i.e. the location of the
red cross in Figure 11 at about ~250 °C). If true then temperature cycling I between 200 and
300 °C should produce increasing amounts of II. To the best of our knowledge, this
experiment has not yet been performed. Also, valuable insight could be obtained from
calorimetric studies.
IV. Conclusion
The polymorphism of the commodity chemical CaC2 is a puzzling and long standing
problem. The existence – and especially coexistence – of three different crystal modifications
at ambient conditions may point to compositional differences between polymorphs or
structural peculiarities. In this work tetragonal CaC2-I was identified as dynamically unstable
and its existence attributed to “disorder”. Disorder may be structural (crystal structure and
microstructure), compositional or dynamical (large anharmonicities). Accordingly,
monoclinic CaC2-II represents the ground state of stoichiometric CaC2 and monoclinic CaC2-
III a high temperature phase. With pressure both monoclinic forms transform into CaC2-I.
Above 500 °C CaC2 occurs as cubic CaC2-IV for which C2 dumbbell moieties are
orientationally disordered. It is assumed that CaC2-IV actually constitutes multiple phases.
The characterization of possible various degrees/states of disorder, along with
compositional variations of phase IV, e.g. from temperature dependent in situ experiments at
synchrotron and neutron facilities, is considered to be key for a conclusive understanding of
the formation of I, II, III and their thermodynamic relation. Such investigations are under way.
In addition, the presence of structural disorder in I, II, III as well as possibly carbon deficiency
for CaC2-I needs more detailed investigations. Here PDF analysis of total scattering data (both
synchrotron X-ray and neutron) would allow probing local structure and help extracting
structural features related to disorder. Lastly we note that the peculiar polymorphism of CaC2
may originate from its enthalpic instability toward decomposition into Ca2C3 and Ca, as
recently suggested from theory [18]. In this respect, complex transformation processes of
CaC2-IV upon cooling may be interpreted as an onset to phase segregation.
Acknowledgement. This work was supported by the Swedish Research Council (grant no.
2012-2956). In addition we acknowledge NMR equipment grants from the Swedish Research
Council and the Knut and Alice Wallenberg Foundation.
Appendix A. Supporting information
Supplementary data associated with this article can be found in the online version at XXXXX
11

References.
[1] M. Weller, T. Overton, J. Rourke, F. Armstrong, Inorganic Chemistry, 6th Edition, Oxford
University Press, 2014.
[2] C.E. Housecroft, A.G. Sharpe, Inorganic Chemistry, 4th edition, Prentice Hall, 2012.
[3] G.L. Miessler, P.J. Fischer, D.A. Tarr, Inorganic Chemistry, 5th edition, Prentice Hall,
2013.
[4] Commercialization of Calcium Carbide and Acetylene - Landmark - American Chemical
Society
http://www.acs.org/content/acs/en/education/whatischemistry/landmarks/calciumcarbideacety
lene.html (accessed Aug 12, 2015).
[5] W. Sun, Econ. Anal. Chin. Petrol. Chem. Ind. (Chin.) 4 (2013) 29−31.
[6] K.S. Rodygin, G. Werner, F.A. Kucherov, V.P. Ananikov, Chem. Asian. J. 11 (2016) 965–
976.
[7] H.H. Franck, M.A. Bredig, G. Hoffmann, Z. Anorg. Allg. Chem. 232 (1937) 61–74.
[8] H.H. Franck, M.A. Bredig, K.-H. Kou, Z. Anorg. Allg. Chem. 232 (1937) 75–111.
[9] (a) M.A. Bredig, J. Phys. Chem. 46 (1942) 801–819; (b) M.A. Bredig, Z. Anorg. Allg.
Chem. 298 (1959) 255–257; (c) M.A. Bredig, Z. Anorg. Allg. Chem. 310 (1961) 338–340.
[10] (a) N.-G. Vannerberg, Acta Chem. Scand 15 (1961) 769–774.; (b) N.-G. Vannerberg,
Acta Chem. Scand. 16 (1962) 1212–1220.
[11] O. Reckeweg, A. Baumann, H.A. Mayer, J. Glaser, H.-J. Meyer, Z. Anorg. Allg. Chem.
625 (1999) 1686–1692.
[12] J. Glaser, S. Dill, M. Marzini, H.A. Mayer, H.-J. Meyer, Z. Anorg. Allg. Chem. 627
(2001) 1090–1094.
[13] M. Knapp, U. Ruschewitz, Chem.-Eur. J. 7 (2001) 874–880.
[14] A. Kulkarni, K. Doll, .C. Sch n, . ansen, . Phys. Chem. B 114 (2010) 155 3–15581.
[15] D. Benson, Y. Li, W. Luo, R. Ahuja, G. Svensson, U. Häussermann, Inorg. Chem. 52
(2013) 6402–6406.
[16] Y.-L. Li, W. Luo, Z. Zeng, H.-Q. Lin, H. Mao, R. Ahuja, Proc. Natl. Acad. Sci. U.S.A.
110 (2013) 9289–9294.
[17] J. Nylén, S. Konar, P. Lazor, D. Benson, U. Häussermann, J. Chem. Phys. 137 (2012)
224507.
12

[18] Y.-L. Li, S.-N. Wang, A.R. Oganov, H. Gou, J.S. Smith, T.A. Strobel, Nat. Commun. 6
(2015) 6974.
[19] J. Rodriguez-Carvajal, Abstracts of the Satellite Meeting on Powder Diffraction of the
XV Congress of the IUCr, (1990) p. 127.
[20] H.K. Mao, J. Xu, P.M. Bell, J. Geophys. Res. 91 (1986) 4673–4676.
[21] (a) P.E. Blöchl, Phys. Rev. B 50 (1994) 17953–17979. (b) G. Kresse, D. Joubert, Phys.
Rev. B 59 (1999) 1758.
[22] (a) G. Kresse, J. Hafner, Phys. Rev. B 48 (1993) 13115–13118. (b) G. Kresse, J.
Furthmüller, Comp. Mater. Sci. 6 (1996) 15–50.
[23] J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77 (1996) 3865–3868.
[24] H.J. Monkhorst, J.D. Pack, Phys. Rev. B 13 (1976) 5188–5192.
[25] P.E. Blöchl, O. Jepsen, O.K. Andersen, Phys. Rev. B 49 (1994) 16223–16233.
[26] P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, G.L.
Chiarotti, M. Cococcioni, I. Dabo, A.D. Corso, S. de Gironcoli, S. Fabris, G. Fratesi, R.
Gebauer, U. Gerstmann, C. Gougoussis, A. Kokalj, M. Lazzeri, L. Martin-Samos, N. Marzari,
F. Mauri, R. Mazzarello, S. Paolini, A. Pasquarello, L. Paulatto, C. Sbraccia, S. Scandolo, G.
Sclauzero, A.P. Seitsonen, A. Smogunov, P. Umari, R.M. Wentzcovitch, J. Phys. Condens.
Matter 21 (2009) 395502.
[27] D. Vanderbilt, Phys. Rev. B 41 (1990) 7892–7895.
[28] (a) M. V. Stackelberg, Naturwissenschaften. 18 (1930), 305–306; (b) M. V. Stackelberg,
Z. Phys. Chem. Abt. B 9 (1930) 437.
[29] R. Juza, H.-U. Schuster, Z. Anorg. Allg. Chem. 311 (1961) 62–75.
[30] The transformation of II to III at around 200 °C occurs without noticeable thermal event
in DSC measurement [12].
[31] O.O. Kurakevych, T.A. Strobel, D.Y. Kim, G.D. Cody, Angew. Chem. Int. Ed. 52 (2013)
8930–8933.
[32] (a) G. Grimvall, B. Magyari-K pe, V. Ozoliņš, K.A. Persson, Rev. od. Phys. 84 (2012)
945–986.; (b) C. Asker, A.B. Belonoshko, A.S. Mikhaylushkin, I.A. Abrikosov, Phys. Rev. B
77 (2008) 220102; (c) V. Ozolins, Phys. Rev. Lett. 102 (2009) 065702.
[33] (a) X. Ma, F. Xu, T.M. Atkins, A.M. Goforth, D. Neiner, A. Navrotsky, S.M. Kauzlarich,
Dalton Trans. (2009) 10250–10255; (b) X. Ma, F. Wu, S.M. Kauzlarich, J. Solid State Chem.
181 (2008) 1628–1633; (c) T. Yi, S. Chen, S. Li, H. Yang, S. Bux, Z. Bian, N.A. Katcho, A.
Shakouri, N. Mingo, J.-P. Fleurial, N.D. Browning, S.M. Kauzlarich, J. Mater. Chem. 22
(2012) 24805–24813.
13

[34] S. Konar, U. Häusserman, G. Svensson, Chem. Mater. 27 (2015) 2566–2575.
[35] (a) V. Vohn, W. Kockelmann, U. Ruschewitz, J. Alloys Compd. 284 (1999) 132–137; (b)
V. Vohn, M. Knapp, U. Ruschewitz, J. Solid State Chem. 151 (2000) 111–116.
[36] B. Wrackmeyer, K. Horchler, A. Sebald, L. H. Merwin, C. Ross II, Angew. Chem. Int.
Ed. 29 (1990) 807–809.
[37] I. Efthimiopoulos, G. V. Vajenine, E. Stavrou, K. Syassen, St. Liebig, U. Ruschewitz, M.
Hanfland, Acta Cryst. A66 (2010) s197.
.
14

Figure captions.
Figure 1. Crystal structures of the ambient temperature polymorphs of CaC2: a) tetragonal
CaC2-I (space group I4/mmm, Z = 2), b) monoclinic CaC2-II (space group C2/c, Z = 4), c)
monoclinic CaC2-III (space group C2/m, Z = 4). The red and gray spheres correspond to C
and Ca atoms, respectively. The octahedral environment of C2 dumbbells by Ca atoms is
emphasized.
Figure 2. Scheme of phase relations among CaC2 polymorphs, according to Refs. [9, 12, 13].
Note, the high temperature form IV transforms to either III or I upon cooling. CaC2-III upon
grinding turns to II or I depending on temperature of grinding.
Figure 3. Compilation of the formation energies E(CaC2) = E(CaC2) − E(fcc-Ca) −
E(graphite) for various CaC2 polymorphs based on DFT calculations.
Figure 4. Calculated phonon dispersion relations of CaC2-I at ambient pressure (black line)
and at 10 GPa (red line).
Figure 5. Multiphase Rietveld fits to powder X-ray diffraction patterns (Cu Kα1 radiation) of
products obtained from reactions at 1400 °C using three starting material compositions
(C/CaH2 = 1.8, 2.0 and 2.2) and a dwelling time of 0.25 h. The location of Bragg peaks are
marked by vertical bars. CaC2-I red, CaC2-II blue, CaC2-III green; this color code is
maintained in the following figures.
Figure 6. Summary of synthesis results. The relative phase fractions of CaC2 polymorphs are
presented as pie charts, with different colors corresponding to different phases (CaC2-I red,
CaC2-II blue, CaC2-III green): (a) 1400, (b) 1100, (c) 900, and (d) 700 °C. For absolute phase
fractions (wt %) and details, see the Supporting Information.
Figure 7. Scanning electron micrographs of CaC2 specimens obtained from reactions at 1100
ºC and using a dwelling time of 24 h: a) predominantly tetragonal form CaC2-I (mixture
C/CaH2 = 1.8), b) monoclinic forms CaC2-II and –III (mixture C/CaH2 = 2.2).
Figure 8. 13C MAS NMR spectra recorded from C/CaH2 = 2.2 (sample C2.2, (a)) and C/CaH2
= 1.8 (sample C1.8, (b)) reactions quenched from 1400 °C. Each number marks the
corresponding peak maximum. (c) Zoom around the first order sideband region indicated by
the rectangle in (a, b), showing the assignment of each 13C resonance to its corresponding
CaC2 polymorph in C2.2 (black trace) and C1.8 (red trace). The data collection involved a 5 h
equilibration delay, 45 min relaxation delay and a total of 56 (sample C1.8) and 84 (sample
C2.2) accumulated signal transients.
Figure 9. Raman spectra of CaC2 polymorphs from multi-phase products obtained from
reactions at 1100 °C and for a dwelling time of 24 h (cf. Figure 6). The upper spectrum
corresponds to a mixture of II and III. The middle spectrum is mainly form II (with a very
minor III contribution). The lower one is mainly form I (with a minor II contribution).
Calculated Raman frequencies are marked by vertical bars (I = red, II = blue, III = green; Ag,
A1g = single bar, E2g = double bar, Bg = dashed bar). Asterisks in the right hand figure indicate
the contribution from 13C and 12C to the C–C stretch vibration.
15

Figure 10. Raman spectra of a mixture of CaC2-II and –III upon compression (a) and
decompression (b). Inset numbers denotes pressure in GPa.
Figure 11. Sketch of the CaC2 phase relations (modified from Bredig [9a]) assuming a one-
component system, but with a heterogeneous phase I. (I = red, II = blue, III = green, IV =
black). The existence of different forms of I is indicated by the red band. The lower boundary
(highest possible stability) is indicated by a red dotted line. The circle marks the position of
IV-to-I and IV-to-III transitions, the cross marks the position of the I-to-II transition. The blue
arrow indicates the pressure induced II-to-I transition. The green arrows indicate the pressure
induced III-to-I and mechanically induced III-to-II transitions.
16

Figure 1
17

Figure 2
Figure 3
18

Figure 4
Figure 5
19

Figure 6
Figure 7
20

Figure 8
21

Figure 9
Figure 10
22

Figure 11
23

Table 1. Compilation of structure parameters from DFT calculations.
Structure parameter
Distance
2.569 × 2
Phase
a (Å)
c (Å)
3.8846
6.3990
48.28
Ca–C
2.818 × 8
2.747 × 4
3.200 × 2
1.262
CaC2-I
I4/mmm
V/Z (Å3)
C 4e
CC–Ca[a]
C–C
0.4014
(0 0 z)
a (Å)
b (Å)
c (Å)
6.6158
4.1513
7.4669
107.5
2.574 × 2
2.638 × 2
2.790 × 2
2.825 × 2
3.191 × 2
Ca–C
 (°)
V/Z (Å3)
Ca 4e
(0 x ¼)
48.90
CaC2-II
C2/c
0.1757
2.736 × 2
2.859 × 2
3.170 × 2
CC–Ca[a]
0.2211
0.1421
-0.0618
C 8f
(x y z)
C–C
1.260
2.567
a (Å)
b (Å)
c (Å)
7.1548
3.8367
7.4569
106.9
Ca–C1
2.718 × 2
2.947 × 2
 (°)
2.607 × 2
2.762
2.777
V/Z (Å3)
Ca 4i
49.0
0.2036
0.2471
0.4532
0.0598
Ca–C2
(x 0 z)
C1 4i
(x 0 z)
CaC2-III
C2/m
2.697 × 4
3.068 × 2
2.764 × 4
3.187 × 2
1.261
CC1–Ca[a]
CC2–Ca[a]
C2 4i
(x 0 z)
0.0793
0.5648
C1–C1
C2–C2
1.259
[a] “CC” refers to the center of a C2 dumbbell.
24

Table 2. Compilation of experimental structure parameters.
CaC2-I,II,III =
I (1400C-0.25h-1.8)
II (900C-6h-2)
III (900C-24h-2.2)
This work
I (SC, RT)
(ref. [11])
III (powder, RT)
(ref. [12])
Structure
parameter
synchrotron XRD at 83 K
IV = NPD
Phase
(ref. [13])
(powder, RT)
a (Å)
c (Å)
3.8658(1)
6.4007(2)
3.8863(4)
6.3862(9)
3.8845(1)
6.3969(2)
48.26
V/Z (Å3)
47.83
48.2
CaC2-I
I4/mmm
Ca 2a
(0 0 0)
C 4e
0.407(1)
0.406(2)
0.408(1)
(0 0 z)
a (Å)
b (Å)
6.5995(2)
4.1899(2)
7.3109(3)
107.034(2)
48.32
6.632(1)
4.1889(7)
7.311(1)
107.24(1)
48.5
6.6479(1)
4.2007(3)
7.3335(5)
107.259(3)
48.89
c (Å)
 (°)
V/Z (Å3)
CaC2-II
C2/c
0
0
Ca 4e
(0 x ¼)
0.1821(5)
0.25
0.185(1)
0.25
0.282(1)
0.146(1)
0.0562(8)
7.2076(3)
3.8283(2)
7.3677(4)
107.19
48.55
0.276(3)
0.135(3)
0.051(3)
7.2317(4)
3.8540(1)
7.3780(3)
107.366(2)
49.1
C 8f
(x y z)
a (Å)
b (Å)
c (Å)
7.226(1)
3.8526(7)
7.376(1)
107.345(1)
49.0
 (°)
V/Z (Å3)
0.1969(7)
0
0.2544(7)
0.421(2)
0
0.021(3)
0.048(3)
0
0.2055(2)
0
0.2464(3)
0.4506(9)
0
0.061(1)
0.0799(8)
0
0.206(1)
0
0.248(2)
0.444(4)
0
0.052(4)
0.075(4)
0
Ca 4i
(x 0 z)
CaC2-III
C2/m
C1 4i
(x 0 z)
C2 4i
(x 0 z)
0.58(2)
0.562(1)
0.563(4)
25

Table 3. Compilation of experimental interatomic distances.
I (SC-XRD, RT)
Synchrotron
I (1400C-0.25h-1.8)
II (900C-6h-2)
III (900C-24h-2.2)
This work (PXRD, RT)
2.610(6)
(ref. [11])
Phase
Distance
PXRD at 83 K
(ref. [13])
III (PXRD, RT)
(ref. [12])
2.60(1)
2.605(6)
2.798(1)
2.728(0)
3.2004(1)
1.190(9)
Ca–C
2.812(2)
2.809(1)
2.7468(0)
3.1985(1)
1.177(9)
CaC2-I
2.7480(2)
3.1931(4)
1.20(1)
CC–Ca[a]
C–C
2.581(5)
2.653(7)
2.815(5)
2.816(5)
3.080(7)
2.58(2)
2.67(2)
2.76(2)
2.90(1)
3.10(2)
Ca–C
CaC2-II
2.752(1)
2.8125(2)
3.156(2)
2.767(3)
2.8295(4)
3.154(3)
CC–Ca[a]
C-C
1.187(7)
1.21(2)
2.68(2)
2.74(1)
2.92(1)
2.68(1)
2.42(2)
2.90(2)
2.609(6)
3.062(3)
2.760(3)
3.272(6)
1.27(3)
1.18(2)
2.540(8)
2.733(4)
2.932(5)
2.609(4)
2.739(8)
2.826(8)
2.713(2)
3.060(1)
2.759(1)
3.184(2)
1.30(1)
2.56(3)
2.79(2)
2.89(2)
2.62(2)
2.76(3)
2.79(3)
2.707(8)
3.055(5)
2.766(4)
3.192(8)
1.27(5)
1.20(3)
Ca–C1
Ca–C2
CaC2-III
CC1–Ca[a]
CC2–Ca[a]
C1–C1
C2–C2
1.241(8)
[a] “CC” refers to the center of a C2 dumbbell.
26

Table 4. Compilation of Raman active modes (in cm-1) for CaC2-I, CaC2-II, and CaC2-III. (lib
= libration mode, trans = translation mode).
CaC2-I (D4h)
CaC2-II (C2h)
CaC2-III(C2h)
exp.
calc.
exp.
calc.
exp.
calc.
1859
1859 A1g
1874
1871
346
1877 Bg
1876
1869
1879 Ag
1874 Ag
1869 Ag
305
297 Eg (lib)
339 Bg (lib)
304 Ag (lib)
244 Bg (lib)
233 Ag (lib)
193 Bg (trans)
165 Ag (trans)
73 Bg (trans)
337 Ag (lib)
298 Ag (lib)
255 Bg (lib)
228 Bg (lib)
192 Ag (trans)
163 Ag (trans)
90 Bg (trans)
306
244
260
232
169
79
165
90
27

TOC
The famous tetragonal form of CaC2, CaC2-I, is dynamically unstable. Why does it exist?
Highlights:
 Critical review of complex polymorphism of the commodity chemical CaC2.
 Tetragonal CaC2-I is identified as dynamical instable and its existence attributed to
disorder.
 With pressure monoclinic CaC2-II and –III transform into tetragonal CaC2-I.
 The high temperature form CaC2-IV is assigned a key role for understanding the
polymorphism of CaC2.
28