The elusive ethenediselone, Se=C=C=Se

Article abstract of DOI:10.1071/CH14098

The neutral ethenediselone, Se=C=C=Se, has been characterised by neutralisation-reionisation mass spectrometry, which implies a minimum lifetime of the order of microseconds. Tetraselenafulvalene 1 and tetramethyltetraselenafulvalene 2 were used as precursor molecules. Flash vacuum thermolysis (FVT) of these compounds with isolation of the products in Ar matrices permitted the identification of ethyne, 2-butyne, CSe₂, and selenoketene, H₂C=C=Se, but at best traces of Se=C=C=Se survived the FVT/matrix isolation experiment. Multiconfigurational calculations indicate that Se=C=C=Se is a ground state triplet molecule with a very small singlet-triplet gap.

Full text of DOI:10.1071/CH14098

RESEARCH FRONT
CSIRO PUBLISHING
Aust. J. Chem. 2014, 67, 1195–1200
Full Paper
http://dx.doi.org/10.1071/CH14098
The Elusive Ethenediselone, Se5C5C5Se
A F
,
B
C
Carl Th. Pedersen, Ming Wah Wong, Kazuo Takimiya,
D
E
Pascal Gerbaux, and Robert Flammang
A
Department of Physics, Chemistry, and Pharmacy, University of Southern Denmark,
Odense, DK-5230 Odense M, Denmark.
B
Department of Chemistry, National University of Singapore, 3 Science Drive 3,
Singapore 117543.
C
Emergent Molecular Function Research Group, RIKEN Center for Emergent Matter
Science (CEMS), 2-1 Hirosawa, Wako, Saitama, 351-0198, Japan.
D
Organic Synthesis and Mass Spectrometry Laboratory, University of Mons,
UMONS, B-7000 Mons, Belgium.
E
Deceased. Formerly of the Organic Synthesis and Mass Spectrometry Laboratory,
University of Mons, UMONS, B-7000 Mons, Belgium.
Corresponding author. Email: cthp@sdu.dk
F
The neutral ethenediselone, Se¼C¼C¼Se, has been characterised by neutralisation–reionisation mass spectrometry,
which implies a minimum lifetime of the order of microseconds. Tetraselenafulvalene 1 and tetramethyltetraselenaful-
valene 2 were used as precursor molecules. Flash vacuum thermolysis (FVT) of these compounds with isolation of the
products in Ar matrices permitted the identification of ethyne, 2-butyne, CSe2, and selenoketene, H2C¼C¼Se, but at best
traces of Se¼C¼C¼Se survived the FVT/matrix isolation experiment. Multiconfigurational calculations indicate that
Se¼C¼C¼Se is a ground state triplet molecule with a very small singlet-triplet gap.
Manuscript received: 25 February 2014.
Manuscript accepted: 7 March 2014.
Published online: 31 March 2014.
Introduction
Extended sulfur-containing heterocumulenes of the type
S¼C=(C)n¼C¼S have been studied intensively.[12] It is a
general observation that cumulenes of this type with an odd
number of atoms are relatively stable (e.g. CO2, C3O2, CS2,
C3S2, etc), whereas those with an even number are highly
unstable. In the selenium series only CSe2 is known. Com-
pounds of the type Se¼Cn¼Se are unknown, although calcula-
tions of their structures and properties have been reported.[13]
However, several selenoketenes R2C¼C¼Se have been
prepared.[14]
In this paper, we report an investigation into the existence
of the diselenium analogue, ethenediselone, Se¼C¼C¼Se
(ionised or neutral), by means of NRMS, FVT, and matrix-
isolation experiments. In addition, the electronic structures of
C2Se2 were examined by quantum chemical calculations.
Ethenedione, O¼C¼C¼O, has been described as ‘the holy grail
of cumulene chemistry’.[1] However, neither the singlet nor the
triplet state of this holy grail may ever be attained,[2] even
though there has been no shortage of attempts.[3] Calculations
indicate that the ground state is a triplet state.[4] Both the
molecular cation radical of ethenedione, O¼C¼C¼Oþꢀ, and the
corresponding anion can be produced in the mass spectrometer
where they are quite stable,[2] but a recovery signal was not
observable in neutralisation–reionisation mass spectrometry
(NRMS) experiments,[2] which means that the lifetime of the
putative neutral C2O2 is less than the microsecond timescale of
the NRMS experiment.
In contrast, thioxoethenone (O¼C¼C¼S)[5] and ethenedi-
thione (S¼C¼C¼S)[6,7] have been produced and characterised
by NRMS in the mass spectrometer as well as by matrix-
isolation infrared (IR) and UV spectroscopies. S¼C¼C¼S is
thermodynamically quite stable, surviving flash vacuum ther-
molysis (FVT) at high temperatures;[7b] yet it is kinetically
highly unstable, polymerising in the neat state at temperatures
above 40 K. This may be due to its triplet ground state as
predicted by theory.[4,8] Analysis of the IR spectrum of
13C-labelled O¼C¼C¼S indicated that this molecule has a
triplet ground state.[5] The monoimine, SCCNH,[9] has been
detected by NRMS, and the monoxime, OCCN–OH, has been
observed in an Ar matrix,[10] but the neutral monoamine,
OCCNH, does not survive the NRMS experiment.[11]
Results and Discussion
Mass Spectrometry
Tetraselenafulvalene 1 was the first precursor to be examined, as
it can be expected to fragment to ethyne (HCꢁCH), seleno-
ketene (H2C¼C¼Se), and ethenediselone (Se¼C¼C¼Se)
(Scheme 1). The 70 eV electron ionisation mass spectrum
(EIMS) of 1 is illustrated in Fig. 1a. This spectrum features a
base peak at m/z 394 corresponding to the molecular ion.
Owing to the natural isotope distribution of selenium (74Se :
76Se : 77Se : 78Se : 80Se : 82Se ¼ 0.9 : 9.4 : 7.6 : 23.8 : 49.6 : 8.7),
Journal compilation Ó CSIRO 2014
www.publish.csiro.au/journals/ajc

1196
C. Th. Pedersen et al.
R
R
92
R
R
Se
Se
Se
Se
(a)
FVT
SeꢀCꢀSe
ꢁ
ꢁ
R
R
SeꢀCꢀCꢀSe
R2CꢀCꢀSe
R ꢀ H
1: R ꢀ H
2: R ꢀ CH3
104
80
Scheme 1. Products of FVT of tetraselenafulvalenes.
394
160
172
(a)
80
(b)
184
40
236
288
368
130
94
92
104
(b)
184
160
26
m/z
Fig. 2. (a) CAMS (oxygen as collision gas) of the m/z 184 ions generated
by dissociative ionisation of tetramethyltetraselenafulvalene 2. Assign-
ments: 172 CSe2þꢀ, 160 Se2þꢀ, 104 C2Seþꢀ, 92 CSeþꢀ, and 80 Seþꢀ. (b)
NRMS of the same ions (ammonia and oxygen as collision gases for
172
160
394
neutralisation and reionisation, respectively). Assignments: 184 C2Se2þꢀ
,
80
160 Seþ2 ꢀ, 104 C2Seþꢀ, 92 CSeþꢀ, and 80 Seþꢀ
.
236
184
130
92
ꢁ
Se
Se
C
C
m/z
80
104
Fig. 1. (a) EIMS of tetraselenafulvalene 1 with an ion source temperature
of 2008C. Assignments: 394 1þꢀ, 368 loss of C2H2, 288 loss of C2H2Se, 236
loss of Se2, and 184 C2Seþ2 ꢀ. (b) The EIMS after FVT at 800–9008C.
Assignments: 394 Mþꢀ, 368 loss of C2H2, 288 loss of C2H2Se, 236 loss of
Chart 1. Fragmentation pattern of ethenediselone.
Se2, 184 C2Seþ2 ꢀ, 106 C2H2Seþꢀ, and 80 Seþꢀ
.
The tetramethyl derivative 2 is very stable thermally, under-
going only a small degree of pyrolysis at the highest tempera-
tures attainable in our apparatus, even when the contact time is
increased by placing a plug of quartz wool inside the pyrolysis
tube. However, the behaviour was similar to that of 1, affording
the molecular ions of 2-butyne and CSe2, whereas H2C¼C¼Se
was, of course, absent. The migration of a methyl group is less
likely than that of H, and there was no strong evidence for the
presence of Me2C¼C¼Se.
compositions are readily conferred to the various fragment
ions: m/z 368 (loss of ethyne), m/z 288 (loss of C2H2Se), m/z 236
(loss of Se2), m/z 130 [C4H2Se] þꢀ, and m/z 94 [CH2Se]þꢀ
The abundant m/z 184 ion has the composition C2Se2 and
.
therefore corresponds to the sought-after ethenediselone,
.
Se¼C¼C¼Seþꢀ
FVT of 1 at temperatures up to 800–9008C caused significant
changes in the EIMS (Fig. 1b), which can be explained by
thermal formation of selenium (m/z 80), diselenium (m/z 160),
and carbon diselenide CSe2 (m/z 172) ions. Note also that the
ratio of m/z 184 (C2Se2) to unpyrolysed material (m/z 394) has
increased significantly, thus suggesting the neutral C2Se2 mole-
cules are formed in the thermolysis. The intensity of m/z 368
([M–ethyne]þꢀ) is also stronger relative to m/z 394, thereby
indicating thermal loss of ethyne from 1. Also note that an
increased m/z 106 signal is present, corresponding to thermal
formation of selenoketene, H2C¼C¼Se.
The collisional activation mass spectrum (CAMS) of the
mass-selected m/z 184 ions originating from the molecular ion
of 2 (dissociative ionisation) featured intense peaks at m/z 104,
92, and 80, in agreement with the expected connectivity (Fig. 2a
and Chart 1).
Peaks at m/z 172 (loss of C) and 160 (loss of C2) in Fig. 2a are
less readily explained, but it is noted that loss of atoms from
‘internal’ positions of ionised cumulenes is not unprecedented.
For example, the CAMS of S¼C¼C¼S featured peaks at m/z
76 (CS2) and 64 (S2) as well as 32 (S) and 44 (CS).[7b]

The Elusive Ethenediselone, Se¼C¼C¼Se
1197
80
70
90
85
80
75
70
65
60
55
50
45
40
35
30
D
B
B
60
B
B
C
B
C
50
40
30
20
10
0
B
B
B
A
A
4000
3000
2000
1500
1000
700
4000
3000
2000
[cmꢂ1
1500
1000
700
[cmꢂ1
]
]
Fig. 4. IR spectrum (10 K, Ar matrix) of the products of pyrolysis of
tetramethyltetrasenafulvalene 2 (10008C, 10ꢂ5 hPa). A: CSe2 1298 cmꢂ1; B:
2-butyne 2976, 2936, 2872, 1447, and 1036 cmꢂ1 (see text for the possible
presence of CSe). C: CO at 2138 cmꢂ1. D: A weak peak at 906 cmꢂ1 may be
ascribed to Se¼C¼C¼Se (calculated 930 cmꢂ1).
Fig. 3. IR spectrum (10 K, Ar matrix) of the products of pyrolysis of
compound 1 (9008C, 10ꢂ5 hpa). A: CSe2 1298 cmꢂ1; B: ethyne 3302, 3288,
and 735 cmꢂ1; C: CH2¼C¼Se 1699 cmꢂ1
.
The neutralisation–reionisation mass spectrum of the m/z
184 ions is shown in Fig. 2b. The clear observation of a recovery
signal at m/z 184 with qualitatively the same fragmentation
pattern as in the CAMS provides strong evidence that neutral
ethenediselone is stable in the gas phase on the microsecond
timescale of the NRMS experiment. One should not expect an
exact agreement between the CAMS and NRMS ion intensities;
the strong m/z 80 peak in the NRMS is undoubtedly due to
reionisation of selenium atoms formed in the neutralisation cell.
Note however that the m/z 92 peak is not particularly strong in
the NRMS, thereby indicating that Se¼C¼C¼Se is reasonably
stable towards fragmentation to CSe molecules. This is con-
firmed by calculations reported in the section Calculations.
Table 1. Calculated relative energies (kJ mol21) of the three lowest
states of C2Se2 using various single-determinant methods using aug-cc-
pVTZ basis set
LevelA
3Sꢂg
1DBg
1Sþg B
HF
0.0
0.0
0.0
0.0
0.0
0.0
0.0
27.4
13.2
14.0
15.6
10.1
15.3
11.2
98.5
56.7
25.1
34.9
31.0
38.8
29.3
B3LYP
MP2
MP3C
MP4C
CCSD
CCSD(T)C
ABased on fully optimised geometries unless otherwise noted.
BUsing an unrestricted Hartree–Fock (UHF) starting point.
Matrix-Isolation IR Spectroscopy
The IR spectrum of the products of FVT of the parent compound
1 isolated in an argon matrix (Fig. 3) is dominated by a strong
absorption of CSe2 at 1298 cmꢂ1 as well as bands due to ethyne.
The gas phase fundamental n3, the only strong absorption of
CBased on CCSD/aug-cc-pVTZ optimised geometries [r(C¼C) ¼ 1.269,
1.270, and 1.269 for 3Sgꢂ, 1Dg, and 1Sþg , respectively; r(C¼Se) ¼ 1.720,
3
ꢂ
1
1
þ
˚
1.718, and 1.719 A, for Sg , Dg, and Sg , respectively].
CSe2, occurs at 1303 cmꢂ1 [15]
.
Selenoketene H2C¼C¼Se is
of ketene (ꢂ52.4 kJ molꢂ1), thioketene (193.6 kJ molꢂ1
)
positively identified by a medium-strength band at 1699 cmꢂ1
and in excellent agreement with the previously determined value
and selenoketene (240.3 kJ molꢂ1) at the G2(MP2) level
of theory.[20]
of 1699.7 cm
is obscured by the ethyne band at 735 cmꢂ1
.
ꢂ1 [16] The other expected absorption at 736 cmꢂ1
.
Calculations
For Se¼C¼C¼Se the only allowed absorption in the IR
spectrum is predicted at 950 cmꢂ1, and it is relatively weak
(53 kJ molꢂ1) at the B3LYP/6–31G* level.[13] A value of
930–932 cmꢂ1 is predicted for all three electronic states at
the B3LYP/6–311þþG(3df,2p) level (see Table S1, Supple-
First we examine the ground-state electronic configuration of
C2Se2. For the sulfur analogue (C2S2, ethenedithione), Ma and
Wong,[8] through high-level multiconfiguration calculations,
have clearly established that the triplet 3Sgꢂ state is the ground
state, in contrast to an early prediction of a singlet ground sta-
te.[7a] There are three possible electronic states of linear C2Se2,
namely 3Sgꢂ, 1Dg, and 1Sgþ, arising from two electrons occupying
two degenerate p orbitals. Initially, we examined the relative
energies of these three states of C2Se2 using various single
determinant methods together with the triple-zeta aug-cc-pVTZ
basis set (Table 1). For the singlet C2Se2, the open-shell 1Dg state
mentary Material). This predicted frequency value (,930 cmꢂ1
)
is readily confirmed by MP2, CCSD, and CASSCF calculations.
Most interestingly, a weak peak at 906 cmꢂ1 matches predic-
tions for Se¼C¼C¼Se in the analogous matrix-IR spectrum
from the FVT of 2 (Fig. 4). In agreement with the mass
spectrometry results, CSe2 and 2-butyne are also identified
in Fig. 4. Both 2-butyne[17] and CSe[18] are reported to have
weak absorptions at 1036 cmꢂ1. A weak absorption at 1036
and 1022 cmꢂ1 in Fig. 4 could be due to both of these com-
pounds, but at any rate, the amount of CSe potentially formed is
very small.
is significantly more stable than the Sþg state at all levels of
1
theory by more than 18 kJ molꢂ1 (Table 1). Hence, we examined
1
only the Dg state in the following multiconfiguration calcula-
tions. All the single determinant methods, namely HF, B3LYP,
MP2, MP3, MP4, CCSD, and CCSD(T), predicted a triplet 3Sgꢂ
ground state for C2Se2 (Table 1). Similar relative energies
between the singlet 1Dg and triplet 3Sꢂg states were predicted by
all the correlated methods. Our best estimate of the singlet-
triplet gap for C2Se2 using the single determinant method
It is relevant to mention that the analogous FVT reactions of
tetrathiafulvalene afforded acetylene, CS, CS2, C3S2, and thio-
ketene H2C¼C¼S.[19] Sulfur has a lower tendency than oxygen
to form double bonds to carbon, and selenium even less so, as
suggested by the steeply increasing calculated heatsof formation

1198
C. Th. Pedersen et al.
Table 2. Calculated S-T gapA (kJ mol21) of C2Se2 using various multiconfiguration methods together with aug-cc-pVDZ basis set
CASB
Active orbitals
CASSCFC
CASPT2C
CASPT3C
MRCIC
(2,2)
5p2u
46.5
20.9
39.9
14.9
12.9
13.7
16.0
15.0
ꢂ1.6
14.5
ꢂ1.8
0.3
ꢂ1.0
4.4
ꢂ0.4
5.8
(6,4)
4p4g 5p2u
(6,6)
4p4g 5p2u 5p0g
4p4u 4p4g 5p2u 5p0g
ꢂ0.3
0.5
ꢂ0.3
0.5
(10,8)
(12,10)
(12,10)D
(14,12)
(20,16)E
11sg2 4pu4 4p4g 5pu2 5pg0 11su0
0.5
0.8
0.4
11sg2 4pu4 4p4g 5pu2 5pg0 11su0
0.6
0.7
0.6
10su2 11s2g 4pu4 4p4g 5pu2 5p0g 11su0 12sg0
9s2g 9s2u 10sg2 10s2u 11sg2 4p4u 4p4g 5p2u 5p0g 11s0u 12s0g 12su0
0.6
0.6
0.7
1.7
ARelative energy between the singlet 1Dg state and triplet 3Sgꢂ state.
BNumber of electrons and number of orbitals in the active space.
CBased on CASSCF optimised geometries.
DLarger aug-cc-pVTZ basis set.
ECAS(20,16) corresponds to the full-valence CAS of C2Se2.
(at CCSD(T)/aug-cc-pVTZ level) is 11.2 kJ molꢂ1. The pre-
dicted C¼C and C¼Se bond lengths of the triplet ground state
Se
C
C
Se
S
C
C
S
˚
are 1.269 and 1.720 A, respectively (at CCSD/aug-cc-pVTZ
level). The central C¼C double bond is shorter than that in C2S2
˚
(1.278 A) while the C¼Se bonds are weaker than the corre-
˚
C2Se2
C2S2
sponding C¼S (1.572 A). This is not surprising, as the C¼Se p
bond overlap is weaker than that in C¼S.
Fig. 5. Highest occupied molecular orbitals (HOMOs) of C2Se2 and C2S2.
Next, the singlet-triplet (S-T) gap of C2Se2 was investigated
using multiconfiguration methods, which are expected to pro-
vide a proper description of the degeneracy problem involved.
We have carried out a series of complete active space self-
consistent field (CASSCF). CASSCF calculations with active
space systematically expanded from the two-electron, two-
orbital CAS (CASSCF(2,2)) to the 20-electron, 16-orbital
CAS (CASSCF(20,16)), which represents the full-valence
CASSCF, using the aug-cc-pVDZ basis set. The active orbitals
involved in these CAS calculations are given in Table 2. The
dominant configurations arise mainly from the excitation of the
valence p electrons. Hence, one would expect an active space
including all p orbitals (i.e. 4pu, 4pg, 5pu, and 5pg), for
example CASSCF(10,8), to yield reliable result. In fact, further
expansion of the active space to include the valence s orbitals
leads to small changes in the relative energies (Table 2). For our
analogue, C2S2, is 23.0 kJ molꢂ1 [8]
The large reduction in the
.
S-T gap may be attributed to the fact the highest occupied
molecular orbital (HOMO) of C2Se2 (namely 5pu) has a
relatively larger contribution of the p orbitals of the terminal
Se atoms (Fig. 5). In addition, the electron density of the HOMO
is more spread out in C2Se2 because selenium has a larger atomic
radius than sulfur. As a consequence, the electron-electron
1
repulsion in the open-shell singlet Dg state is less severe in
C2Se2. It is important to note that the HOMOs of both singlet and
triplet states of C2S2 and C2Se2 are similar.
The free energies of reaction and activation at 9008C for the
various possible fragmentation pathways of tetraselenafulva-
lene 1 were evaluated at the B3LYP/6–311þþG(3df,2p) level
(Scheme 2). The fragmentation of 1 to CSe2 þ C2H2 is the most
favourable pathway both kinetically and thermodynamically.
This dissociation process involves the loss of C2H2 to form a
diselone-thioketene intermediate 3 in the first step, via transition
state TS1. The observation of m/z 368 ions in EIMS (Fig. 1b) is
consistent with the proposed mechanism. A cleavage of the
central C¼C bond in 1 to form two singlet 1,3-diselenolylidenes
(4, ‘diselenole carbenes’), stabilised by the neighbouring
selenium lone pairs, requires a high activation barrier of
275 kJ molꢂ1 (via transition state 2). A minor peak at 784 cmꢂ1
in the IR spectrum (Fig. 3) may possibly be due to 4 (predicted
frequency of 785 cmꢂ1). Dissociation of 1 to C2Se2 þ selenirene
is the least favourable pathway both kinetically and thermody-
namically. This dissociation process involves initially the for-
mation of a cyclic thioketene intermediate 5 (Scheme 2), which
may correspond to the m/z 288 ions in EIMS (Fig. 1b). This step
also requires a high activation barrier of 281 kJ molꢂ1. The
two possible fragmentations of C2Se2 require high energies
(Scheme 2). Not surprisingly, therefore, there is little or no
evidence for the formation of CSe and C2Se in the experimental
spectra. The corresponding standard free energies of reaction
and activation at room temperature (298 K) and the optimised
1
largest CASSCF calculation, CASSCF(20,16), the singlet Dg
state lies 15.0 kJ molꢂ1 above the triplet 3Sꢂg ground state.
However, inclusion of dynamic electron correlation at MP2
(CASPT2), MP3 (CASPT3), and CI (MRCI) levels significantly
lower the S-T energy gap (Table 2). This clearly indicates that
the CASSCF methods are not sufficient to describe properly the
singlet-triplet gap. For the smaller active space which includes
only the p orbitals, both CASPT3 and MRCI levels yield almost
identical energy for the 3Sgꢂ and 1Dg states. Further expansion of
the p active space leads to a very small preference for the triplet
state (by ,1 kJ molꢂ1). At our best level of theory, CASPT2
(20,16), the triplet 3Sgꢂ state is 1.7 kJ molꢂ1 more stable than the
singlet 1Dg state. To examine the effect of basis set on the S-T
gap, we have also performed CAS(12,10) calculations using the
larger aug-cc-pVTZ basis set. The changes in relative energies
on going from the aug-cc-pVDZ to aug-cc-pVTZ are very small,
particularly at the correlated levels (Table 2). Hence, we have
confidence in the predicted S-T gap of 1.7 kJ molꢂ1. In summa-
ry, all the single- and multi-determinant methods predict a triplet
ground for C2Se2, and there is no evidence for Hund’s rule being
violated. It is clear, however, that the singlet-triplet splitting is
very small. For comparison, the calculated S-T gap for the sulfur

The Elusive Ethenediselone, Se¼C¼C¼Se
1199
Se
ꢀCH
Se
Se
SeꢀCꢀCꢀSe ꢁ 2
ꢁ
CꢀSe
HC
HCꢀCH
Se
(222.3)
5
(39.2)
(2 steps: TS3: 196.2,
intermediate: 156.8; TS4: 280.5)
ꢁ
2H2CꢀCꢀSe
SeꢀCꢀCꢀSe
(ꢂ144.4)
Se
Se
Se
Se
Se
Se
2
(S)
2 CSe ꢁ 2H2CꢀCꢀSe
(TS2: 274.9)
(12.5)
(80.0)
4
(0)
1
(TS1: 204.3)
Se
Se
Se
Se
2 CSe2 ꢁ 2HCϵCH
(ꢂ252.6)
ꢁ HCϵCH
C2Se4 ꢁ 2HCϵCH
(ꢂ64.1)
6
(17.8)
3
CSe ꢁ CSe
C2Se ꢁ Se
SeꢀCꢀCꢀSe
(156.9)
(354.6)
(0)
Scheme 2. Fragmentation pathways of tetraselenafulvalene 1 and C2Se2. Calculated relative free energies
(DG1173, kJ molꢂ1) at the B3LYP/6–311þþG(3df,2p) level are given in parentheses.
geometries of transition states (TS1–TS4) and intermediates
(3–6) are available in the Supplementary Material.
The IR spectra of the possible fragmentation products of 1
and 2 were also calculated at the B3LYP/6–311þþG(3df,2p)
level and showed good agreement with the experimental fre-
quencies (Table S1 in the Supplementary Material).
Multiconfiguration SCF calculations, including CASSCF,[28]
second-order and third-order multireference perturbations
CASPT2[29] and CASPT3,[29] and multireference configuration
interaction (MRCI)[30] were carried out for the Sgꢂ and Dg,
states of C2Se2. Restricted Hartree–Fock (RHF) was used for
closed-shell species and the unrestricted Hartree–Fock (UHF)
formalism was employed for the open-shell systems (namely
3
1
3Sꢂg and Dg states). Harmonic fundamental vibrations were
1
Conclusion
calculated for C2Se2 and its fragmented products at the B3LYP/
6–311þþG(3df,2p) level. The directly calculated vibrational
frequencies were scaled by a factor of 0.967.[31] For the frag-
mentation pathways of tetraselenafulvalenes, the energetics was
examined at B3LYP/6–311þþG(3df,2p) level.
The formation of selenoketene CH2¼C¼Se in the FVT of tet-
raselenafulvalene 1 is confirmed by mass spectrometry and
IR spectroscopy. The existence of neutral ethenediselone,
Se¼C¼C¼Se, is established on the microsecond timescale of
the NRMS experiment. This compound barely survives the
millisecond timescale[21] of FVT/matrix isolation experiments,
but a weak absorption at 906 cmꢂ1 in the IR spectrum of the
thermolyzate from 2 is in agreement with theoretical predictions
for C2Se2. C2Se2 is predicted to be a ground state triplet mole-
cule with a small singlet-triplet gap (1.7 kJ molꢂ1 at the CASPT2
(20,16) level).
Supplementary Material
Computational details are available from the Journal’s website.
Acknowledgements
R.F. and P.G. acknowledge the FRS-FNRS for financial support for the
acquisition of the AutoSpec 6F and for continuing support. The authors
thank Professor Curt Wentrup, The University of Queensland, for valuable
discussions and suggestions.
Experimental
The apparatus for argon matrix isolation and IR spectroscopy
employed a quartz thermolysis tube (150 mm length and 8 mm
internal diameter) as described earlier.[22] BaF2 optics were used
for IR spectroscopy. The six-sector tandem mass spectrometer
(Waters AutoSpec 6F) with E1 B1 c1 E2 c2 c3 E3 B2 c4 E4
geometry was fitted with a quartz thermolysis tube (50 mm
length, 3 mm inner diameter) directly connected to the outer ion
source as previously described.[23] The NRMS technique has
been reported.[24] Compounds 1 and 2 were prepared according
to literature procedures.[25]
References
[1] N. Goldberg, H. Schwarz, Acc. Chem. Res. 1994, 27, 347. doi:10.1021/
AR00047A005
[2] D. Schro¨der, C. Heinemann, H. Schwarz, J. N. Harvey,
S. Dua, S. J. Blanksby, J. H. Bowie, Chem. – Eur. J. 1998, 4,
2550.
doi:10.1002/(SICI)1521-3765(19981204)4:12,2550::AID-
CHEM2550.3.0.CO;2-E
[3] (a) J. A. Berson, D. M. Birney, W. P. Dailey III, J. F. Liebman, in
Modern Models of Bonding and Delocalization, Molecular Structures
and Energetics (Eds J. F. Liebman, A. Greenberg) 1988, Vol. 6,
pp. 391–441 (Wiley VCH: Weinheim).
Computational Methods
(b) D. M. Birney, J. A. Berson, Tetrahedron 1986, 42, 1561 and
references therein. doi:10.1016/S0040-4020(01)87571-2
(c) R. C. Haddon, D. Poppinger, L. Radom, J. Am. Chem. Soc. 1975,
97, 1645 and references therein. doi:10.1021/JA00840A003
[4] G. P. Raine, H. F. Schaefer, R. C. Haddon, J. Am. Chem. Soc. 1983,
105, 194. doi:10.1021/JA00340A009
Ab initio and multiconfiguration self-consistent field calcula-
tions were carried out with GAUSSIAN 09[26] and MOLPRO
2010[27] programs. The structures and energies of the three
electronic states of C2Se2 were investigated at the Hartree–Fock
(HF), B3LYP, MP2, MP3, MP4, CCSD, and CCSD(T) levels.

1200
C. Th. Pedersen et al.
[5] (a) D. Su¨lzle, J. K. Terlouw, H. Schwarz, J. Am. Chem. Soc. 1990, 112,
628. doi:10.1021/JA00158A020
(b) G. Maier, H. P. Reisenauer, R. Ruppel, Angew. Chem. Int. Ed. 1997,
36, 1862. doi:10.1002/ANIE.199718621
[6] D. Su¨lzle, H. Schwarz, Angew. Chem. Int. Ed. 1988, 27, 1337.
doi:10.1002/ANIE.198813371
[23] (a) R. H. Bateman, J. Brown, M. Lefevere, R. Flammang, Y. Van
Haverbeke, Int. J. Mass Spectrom. 1992, 115, 205. doi:10.1016/0168-
1176(92)85042-X
(b) J. Brown, R. Flammang, Y. Govaert, M. Plisnier, C. Wentrup,
Y. Van Haverbeke, Rapid Commun. Mass Spectrom. 1992, 6, 249.
doi:10.1002/RCM.1290060405
[7] (a) G. Maier, H. P. Reisenauer, J. Schrot, R. Janoschek, Angew. Chem.
Int. Ed. 1990, 29, 1464. doi:10.1002/ANIE.199014641
(b) C. Wentrup, P. Kambouris, R. A. Evans, D. Owen, G. Macfarlane,
J. Chuche, J. C. Pommelet, A. Ben Cheikh, M. Plisnier, R. Flammang,
J. Am. Chem. Soc. 1991, 113, 3130. doi:10.1021/JA00008A048
(c) R. B. Bohn, Y. Hannachi, L. Andrews, J. Am. Chem. Soc. 1992, 114,
6452. doi:10.1021/JA00042A024
(d) M. W. Wong, C. Wentrup, R. Flammang, J. Phys. Chem. 1995, 99,
16849. doi:10.1021/J100046A010
[8] N. L. Ma, M. W. Wong, Angew. Chem. Int. Ed. 1998, 37,
3402. doi:10.1002/(SICI)1521-3773(19981231)37:24,3402::AID-
ANIE3402.3.0.CO;2-S
[9] R. Flammang, D. Landu, S. Laurent, M. Barbieux-Flammang,
C. O. Kappe, M. W. Wong, C. Wentrup, J. Am. Chem. Soc. 1994, 116,
2005. doi:10.1021/JA00084A046
[10] G. Maier, H. P. Reisenauer, B. Ro¨ther, J. Eckwert, Liebigs Ann. Chem.
1996, 303.
[11] R. Flammang, Y. Van Haverbeke, S. Laurent, M. Barbieux-
Flammang, M. W. Wong, C. Wentrup, J. Phys. Chem. 1994, 98,
5801. doi:10.1021/J100074A002
[12] (a) W. Weltner, Jr, R. J. Van Zee, Chem. Rev. 1989, 89, 1713.
doi:10.1021/CR00098A005
(c) P. Gerbaux, L. Lamote, Y. Van Haverbeke, R. Flammang,
J. M. Brown, Eur. J. Mass Spectrom. 2012, 18, 93. doi:10.1255/
EJMS.1174
[24] (a) F. Turecˇek, Top. Curr. Chem. 2003, 225, 77. doi:10.1007/
3-540-36113-8_3
(b) P. Gerbaux, Y. Van Haverbeke, R. Flammang, J. Mass Spectrom.
1997, 32, 1170. doi:10.1002/(SICI)1096-9888(199711)32:11,1170::
AID-JMS573.3.0.CO;2-4
[25] A. Morikami, T. Takimiya, Y. Aso, T. Otsubo, J. Mater. Chem. 2001,
11, 2431. doi:10.1039/B104402H
[26] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci,
G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada,
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima,
Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr,
J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers,
K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand,
K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi,
M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross,
V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,
R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth,
P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas,
J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09
(Revision A.2) 2009 (Gaussian, Inc.: Wallingford CT).
[27] H.-J. Werner, P. J. Knowles, G. Knizia, F. R. Manby, M. Schu¨tz,
P. Celani, T. Korona, R. Lindh, A. Mitrushenkov, G. Rauhut,
K. R. Shamasundar, T. B. Adler, R. D. Amos, A. Bernhardsson,
A. Berning, D. L. Cooper, M. J. O. Deegan, A. J. Dobbyn, F. Eckert,
E. Goll, C. Hampel, A. Hesselmann, G. Hetzer, T. Hrenar, G. Jansen,
C. Ko¨ppl, Y. Liu, A. W. Lloyd, R. A. Mata, A. J. May,
S. J. McNicholas, W. Meyer, M. E. Mura, A. Nicklass, D. P. O’Neill,
P. Palmieri, K. Pflu¨ger, R. Pitzer, M. Reiher, T. Shiozaki, H. Stoll,
A. J. Stone, R. Tarroni, T. Thorsteinsson, M. Wang, A. Wolf,
MOLPRO, Version 2010.1 2010 (University of Birmingham:
Birmingham).
(b) G. Maier, H. P. Reisenauer, R. Ruppel, Sulfur Rep. 1999, 21, 335.
doi:10.1080/01961779908047947
(c) R. Flammang, C. Wentrup, Sulfur Rep. 1999, 21, 357. doi:10.1080/
01961779908047948
(d) S. Saito, Sulfur Rep. 1999, 21, 401. doi:10.1080/
01961779908047950
(e) R. Flammang, C. Wentrup, Sulfur Rep. 1997, 20, 255. doi:10.1080/
01961779708047921
[13] H.-Y. Wang, X. Lu, R.-B. Huang, L.-S. Zheng, J. Mol.
Struct. THEOCHEM 2002, 593, 187. doi:10.1016/S0166-
1280(02)00317-2
[14] H. Ulrich, Cumulenes in Click Reactions 2009 (Wiley: Chichester).
[15] G. W. King, K. Srikameswaran, J. Mol. Spectrosc. 1969, 29, 491.
doi:10.1016/0022-2852(69)90127-1
[16] J. L. Laboy, D. J. Clouthier, Chem. Phys. Lett. 1995, 236, 211.
doi:10.1016/0009-2614(95)00220-X
[17] I. S. Butler, M. L. Newbury, Spectrochim. Acta A 1980, 36A, 458.
[18] R. F. Barrow, W. Jevons, Proc. Phys. Soc. 1948, 61, 99. doi:10.1088/
0959-5309/61/1/115
[28] B. O. Roos, Adv. Chem. Phys. 1987, 69, 399. doi:10.1002/
9780470142943.CH7
[29] H.-J. Werner, Mol. Phys. 1996, 89, 645. doi:10.1080/
00268979609482499
[19] C. Th. Pedersen, Asian Chem. Lett. 2003, 7, 41.
[20] N. L. Ma, M. W. Wong, Eur. J. Org. Chem. 2000, 1411. doi:10.1002/
(SICI)1099-0690(200004)2000:8,1411::AID-EJOC1411.3.0.
CO;2-N
[30] H.-J. Werner, P. J. Knowles, J. Chem. Phys. 1988, 89, 5803.
doi:10.1063/1.455556
[31] (a) M. W. Wong, Chem. Phys. Lett. 1996, 256, 391. doi:10.1016/0009-
2614(96)00483-6
[21] R. F. C. Brown, Pyrolytic Methods in Organic Chemistry 1980
(Academic Press: New York, NY).
(b) Precomputed vibrational scaling factors from the Computational
Chemistry Comparison and Benchmark Database (CCCBC) in
National Institute of Standards and Technology (NIST). Available at:
http://cccbdb.nist.gov/vibscalejust.asp (accessed 10 February 2014).
[22] R. Flammang, M. Barbieux-Flammang, P. Gerbaux, C. Th. Pedersen,
J. Chem. Soc., Perkin Trans. 2 1997, 1261. doi:10.1039/A700502D