Toward unsaturated stannylenes Y₂Z=Sn: and related compounds with triplet electronic ground states

Abstract

A new series of unsaturated stannylenes is studied computationally. The singlet and triplet states of acyclic and cyclic stannylenes are fully optimized using the B3LYP, BHLYP, OPBE, and M06 functionals. The basis sets used are of double-ξ plus polarization quality with additional s- and p-type diffuse functions denoted DZP++. All cyclic and most acyclic stannylenes have been found to have triplet ground states. The most favored triplet state is that for the NHC (NMeCHCHNMe)Sn=Sn: system, where the triplet state lies ∼20 kcal mol⁻¹ below the singlet. The saturated cyclic systems are expected to be easier to synthesize, but the unsaturated cyclic counterparts have larger singlet–triplet splittings. A preparative outline based on retro-synthetic routes is briefly described.

---

1. Introduction

Carbenes, neutral molecules R₂C: containing divalent carbon atoms, have fascinated chemists for more than half a century because they exist in two accessible spin states, each with its own distinctive chemistry.^1–3 Their tin analogs, stannylenes, have been generated in solution only with singlet ground states, and, until 2009, when the preparation of 1-stannacyclopent-3-enes provided convenient precursors for a variety of stannylenes,⁴ most of these were produced with bulky substituents which discourage bimolecular reactions by steric shielding.^5,6

Reactions of sterically hindered stannylenes are comparatively slow and oligomerization is very rapid. Now stannylenes containing smaller substituents such as Me and Ph can be generated both thermally and photochemically, and the chemistry of ground state singlet stannylenes can now be more thoroughly explored, employing stannylenes whose intermolecular reactions are competitive with oligomerization.⁴

The motivation to study stannylenes R₂Sn:, is that the lowest singlet electronic states of the analogous carbenes, silylenes, and germylenes undergo synthetically useful stereospecific addition and insertion reactions with concerted formation of two bonds.^3,5 The chemistry of singlet stannylenes found by Neumann⁶ and Lappert⁷ also promises synthetic utilization. Calculations on the addition of Me₂Sn: to butadiene predict a concerted asynchronous addition, and predicted activation parameters for the retro-addition are in decent agreement with the experimental results.⁸ For the triplet stannylenes studied here, it is most desirable to have a sufficiently large singlet–triplet gap so that the reactions observed will be exclusively those of the triplet ground state.

Unsaturated heavier group 14 carbene analogs have already received some attention. Singlet states of unsaturated silylenes and germylenes have been studied. Worthington and Cramer⁹ performed DFT computations on the effects of substituents on the singlet–triplet energy gaps of unsaturated carbenes, the XYC=C:. Irikura, Goddard, and Beauchamp¹⁰ suggested that hyperconjugative electron-donation from the singly occupied carbene p-orbital to the C–F σ* orbital plays a major role in lowering the energy of the triplet states. A lesser stabilizing substituent effect was proposed to be inductive electron-withdrawal from the n-orbital at the carbene center.

In this work, double-bonded Y₂Z=Sn: (Z = C, Si, Ge, and Sn; Y = H, F, NMeCH₂– and NMeCH–) species are investigated, with the prime objective to reliably predict their singlet–triplet gaps. Many of these species were expected to contain stannylene centers sufficiently sterically unshielded to permit bimolecular reactions and to possess triplet ground states. It is hoped that density functional theory computations employing several functionals will provide useful information about their electronic properties that will facilitate their experimental characterization. Hopefully, exploration of the chemistry of triplet stannylenes should be possible for the first time. Interpretation of the computations presented should lead to answers for questions regarding computations on these novel species of vinylidenes. Specifically, the character of the multiple bond between the tin and Z atoms (Z = C, Si, Ge, and Sn) and the effects of substituents Y¹¹ will be explored.

2. Theoretical methods

All geometries were fully optimized with the Gaussian 09 Program,¹² using four functionals, BHLYP, B3LYP, OPBE and M06.¹³ BHLYP is an HF/DFT hybrid method employing the Becke (B)¹⁴ half and half exchange functional (H)¹⁵ and the Lee, Yang, and Parr (LYP)¹⁶ non-local correlation functional. The ubiquitous B3LYP method combines Becke's three-parameter exchange functional (B3) with the LYP correlation functional. In all cases, an extended integration grid (199974) was used, with very tight convergence criteria applied to all computations. Harmonic vibrational frequencies were evaluated analytically to characterize all stationary points as minima.

Double-ζ basis sets with polarization and diffuse functions, denoted as DZP++, were used for all atoms, except that the valence basis set used for tin incorporates the fully-relativistic Stuttgart–Dresden small-core effective core potential.¹⁷ The double-ζ basis sets were constructed by augmenting the Huzinaga–Dunning–Hay^18–20 sets of contracted Gaussian functions with one set of p polarization functions for each H atom and one set of d polarization functions for each heavy atom, respectively [α_p(H) = 0.75, α_d(C) = 0.75, α_d(N) = 0.80, α_d(F) = 1.0, α_d(Si) = 0.5]. The above basis sets were further augmented with diffuse functions, where each atom received one additional s-type and one additional set of p-type functions. Each H atom basis set is augmented with one diffuse s-function. The diffuse functions were determined in an even-tempered fashion following the prescription of Lee,²¹

where α₁, α₂ and α₃ are the three smallest Gaussian orbital exponents of the s- or p-type primitive functions of a given atom (α₁ < α₂ < α₃). Thus α_s(H) = 0.04415, α_s(C) = 0.04302, α_p(C) = 0.03629, α_s(N) = 0.06029, α_p(N) = 0.05148, α_s(F) = 0.1049, α_p(F) = 0.0826, α_s(Si) = 0.02729, α_p(Si) = 0.025.

The DZP++ basis set for germanium was comprised of the Schafer–Horn–Ahlrichs double-ζ spd set plus a set of five pure d-type polarization functions with α_d(Ge) = 0.246, and augmented by a set of sp diffuse functions with α_s(Ge) = 0.024434 and α_p(Ge) = 0.023059.²² The overall contraction scheme for the all-electron basis sets is: H(5s1p/3s1p), C(10s6p1d/5s3p1d), F(10s6p1d/5s3p1d), Si(13s9p1d/7s5p1d), N(10s6p1d/5s3p1d), and Ge(15s12p6d/9s7p3d). Each (adiabatic) singlet–triplet splitting is predicted as the energy difference between the lowest singlet state and its lowest triplet state:

ΔE_S–T = E(optimized triplet) − E(optimized singlet).

3. Results and discussion

3.1. Geometries

The equilibrium geometries for the ¹A₁ and ³A₂ states of H₂Z=Sn: and F₂Z=Sn: (Z = C, Si, Ge, and Sn), are reported in Fig. 1–4. In Fig. 5–8 are the predicted equilibrium geometries for the ¹A and ³A states of (NMeCH₂CH₂NMe)Z=Sn: and (NMeCHCHNMe)Z=Sn:. The internal coordinates, theoretical harmonic vibrational frequencies (cm⁻¹), and energies (hartrees) at the BHLYP functional, are available as ESI.† There are large differences in the Z=Sn bond lengths along the series of ¹A₁ singlet states Y₂Z=Sn: (Y = H, F; Z = C, Si, Ge, and Sn) and their corresponding ³A₂ triplet states are presented in Table 1.

Fig. 1 Equilibrium geometries for the lowest lying singlet and triplet states of H₂C=Sn: and F₂C=Sn:.

Fig. 2 Equilibrium geometries for the lowest lying singlet and triplet electronic states of H₂Si=Sn: and F₂Si=Sn:.

Fig. 3 Equilibrium geometries for the lowest lying singlet and triplet states of H₂Ge=Sn: and F₂Ge=Sn:.

Fig. 4 Equilibrium geometries for the lowest lying singlet and triplet states of H₂Sn=Sn: and F₂Sn=Sn:.

Fig. 5 Equilibrium geometries for the lowest lying singlet and triplet states of (NMeCH₂CH₂NMe)C=Sn: and (NMeCHCHNMe)C=Sn:.

Fig. 6 Equilibrium geometries for the lowest lying singlet and triplet states of (NMeCH₂CH₂NMe)Si=Sn: and (NMeCHCHNMe)Si=Sn:.

Fig. 7 Equilibrium geometries for the lowest lying singlet and triplet states of (NMeCH₂CH₂NMe)Ge=Sn: and (NMeCHCHNMe)Ge=Sn:.

Fig. 8 Equilibrium geometries for the lowest lying singlet and triplet states of (NMeCH₂CH₂NMe)Sn=Sn: and (NMeCHCHNMe)Sn=Sn:.

Table 1 Z=Sn predicted bond lengths (Å) and vibrational frequencies (cm⁻¹)

Predicted Z=Sn bond lengths, Å, employing BHLYP
	H2C=Sn:	H2Si=Sn:	H2Ge=Sn:	H2Sn=Sn:
^1A1	1.992	2.461	2.492	2.703
^3A2	2.176	2.591	2.629	2.840

---

	F2C=Sn:	F2Si=Sn:	F2Ge=Sn:	F2Sn=Sn:
^1A1	2.073	2.536	2.588	2.838
^3A2	2.195	2.635	2.715	3.022

---

	(NMeCH2CH2NMe)C=Sn:	(NMeCH2CH2NMe)Si=Sn:	(NMeCH2CH2NMe)Ge=Sn:	(NMeCH2CH2NMe)Sn=Sn:
^1A	2.173	2.584	2.608	2.833
^3A	2.356	2.659	2.715	2.961

---

	(NMeCHCHNMe)C=Sn:	(NMeCHCHNMe)Si=Sn:	(NMeCHCHNMe)Ge=Sn:	(NMeCHCHNMe)Sn=Sn:
^1A	2.215	2.627	2.669	2.925
^3A	2.355	2.670	2.731	2.970

---

Predicted Z=Sn stretching vibrational modes, cm^−1, employing BHLYP
	H2C=Sn:	H2Si=Sn:	H2Ge=Sn:	H2Sn=Sn:
^1A1	693	357	257	199
^3A2	530	295	212	166

---

	F2C=Sn:	F2Si=Sn:	F2Ge=Sn:	F2Sn=Sn:
^1A1	301	192	166	128
^3A2	246	163	129	86

---

3.2. Natural Lewis structures

Natural bond orbitals (NBO)²³ are computed to search for Lewis structures, representing non-bonding and bonding electron pairs. These may be termed as natural Lewis structures, explaining hybridizations and polarization coefficients for each NBO.²⁴

The Y₂Z: subsystem of the Y₂Z=Sn: (Z = C, Si, Ge, and Sn) molecules can exist in two low-lying electronic states: ¹A₁ and ³B₁, while the Sn atom itself exhibits a triplet ground state as depicted in Scheme 1. Hence, the final electronic state of the stannylene molecule depends on the distribution of the nonbonding electrons in the σ and π orbitals of the Y₂Z: moiety. In the singlet electronic state of the Y₂Z=Sn: molecule, one of the two nonbonding electrons of the Y₂C: moiety occupies the hybrid σ orbital, which is partly localized on the carbene carbon. The other nonbonding electron occupies the carbene 2pπ atomic orbital, and these electrons have parallel spins, hence ³A₂ symmetry. In the corresponding triplet electronic state there are two nonbonding electrons occupying the hybridized σ orbital, which is incompletely localized on the carbene carbon atom, thus a ¹A₁ state can be designated for the Y₂C: moiety. Since the σ orbital is a hybrid with a contribution from the carbon 2s atomic orbital, this σ orbital is lower in energy than the 2pπ orbital. The π orbital involves a carbon 2p atomic orbital, forming a partially occupied π orbital between the singlet ¹A₁ Y₂C: carbene moiety and the triplet Sn atom. It is expected that the ³A₂ electronic state of the Y₂Z=Sn: molecules, with a σ²π⁰ configuration in the H₂Z: moiety in the ¹A₁ singlet σ²π² electronic state of H₂Z=Sn:, can be favored relative to the ³B₁–σ¹π¹ configuration of the H₂Z: moiety by substituents lowering the energy of the pπ atomic orbital and raising the energy of the σ molecular orbital of the H₂Z: moiety. If the Y₂Z: subsystem presents a ³B₁–σ¹p¹ configuration in the ¹A₁ singlet state of Y₂Z=Sn:, a lower coulombic repulsion exists between the nonbonding electrons of the H₂Z: moiety. This allows a triplet ³A₂ ground state of Y₂Z=Sn: with ¹A₁–σ²p⁰ configuration at its Y₂Z: moiety to be stabilized selectively by substituents such as fluorine or a five-membered ring system containing nitrogen lone pair electrons which partially fill the p-orbital, encouraging the shift of an electron from π to p.

Scheme 1 Distribution of electrons between the Y₂Z: moiety and the triplet Sn atom in Y₂Z=Sn: and representations of singlet and triplet Y₂Z=Sn: and Y₂Z: electronic configurations.

As seen in Table 2, the singlet ground states of the unsubstituted H₂Z=Sn: (Z = C, Si, Ge, and Sn) are predicted to be energetically preferred to their corresponding triplet states except for H₂Sn=Sn: at BHLYP. The trends in the predicted decreasing energy gaps follow the standard Pauling electronegativities²⁵ (χ_F = 3.98; χ_C = 2.55, χ_H = 2.20, χ_Ge = 2.01, χ_Sn = 1.96, and χ_Si = 1.90) fairly regularly with the carbon substituent standing apart in predicting the largest singlet–triplet gaps, compared to its analogs in the series studied. The predicted ΔE_S–T values (kcal mol⁻¹) for H₂C=Sn: are 20.9 (B3LYP), 15.3 (BHLYP), 21.3 (OPBE), and 24.0 (M06), while the Si, Ge, and Sn analogues result in decreases in the ΔE_S–T (kcal mol⁻¹) to 5.4 (H₂Si=Sn), 5.2 (H₂Ge=Sn), and 1.1 (H₂Sn=Sn) with the B3LYP functional. Similar decreases are predicted with the OPBE functional, while computations with the BHLYP functional predict a triplet ground state for H₂Sn=Sn:.

Table 2 Singlet–triplet gaps for stannylenes in eV (kcal mol⁻¹ in parentheses)

	B3LYP		BHLYP		OPBE		M06
H2C=Sn:	0.91	(20.9)	0.67	(15.3)	0.92	(21.3)	1.04	(24.0)
H2Si=Sn:	0.23	(5.4)	0.13	(2.9)	0.17	(3.9)	0.43	(9.8)
H2Ge=Sn:	0.23	(5.2)	0.13	(3.0)	0.12	(2.7)	0.51	(11.8)
H2Sn=Sn:	0.05	(1.1)	−0.04	(−0.9)	0.03	(0.6)	0.33	(7.6)
F2C=Sn:	0.12	(2.9)	−0.06	(−1.4)	0.06	(1.3)	0.35	(8.1)
F2Si=Sn:	−1.34	(−30.8)	−0.36	(−8.2)	−0.40	(−9.3)	−0.01	(−0.2)
F2Ge=Sn:	−0.31	(−7.2)	−0.36	(−8.4)	−0.43	(−10.0)	−0.14	(−3.2)
F2Sn=Sn:	−0.44	(−10.0)	−0.51	(−11.9)	−0.50	(−11.6)	−0.34	(−7.9)
(NMeCH2CH2NMe)C=Sn:	−0.40	(−9.3)	−0.52	(−11.9)	−0.54	(−12.4)	−0.04	(−1.0)
(NMeCH2CH2NMe)Si=Sn:	−0.60	(−13.8)	−0.65	(−14.9)	−0.72	(−16.6)	−0.28	(−6.4)
(NMeCH2CH2NMe)Ge=Sn:	−0.55	(−12.6)	−0.59	(−13.6)	−0.67	(−15.4)	−0.21	(−4.8)
(NMeCH2CH2NMe)Sn=Sn:	−0.60	(−13.8)	−0.66	(−15.2)	−0.65	(−15.1)	−0.26	(−5.9)
(NMeCHCHNMe)C=Sn:	−0.56	(−12.9)	−0.67	(−15.5)	−0.70	(−16.2)	−0.22	(−5.0)
(NMeCHCHNMe)Si=Sn:	−0.68	(−15.6)	−0.81	(−18.8)	−0.94	(−21.7)	−0.46	(−10.7)
(NMeCHCHNMe)Ge=Sn:	−0.80	(−18.6)	−0.82	(−18.9)	−0.94	(−21.7)	−0.46	(−10.7)
(NMeCHCHNMe)Sn=Sn:	−0.89	(−20.4)	−0.92	(−21.1)	−1.00	(−23.1)	−0.54	(−12.5)

---

In order to verify the energies and structures of the ¹A₁ singlet ground state H₂C=Sn: and its corresponding ³A₂ triplet state, CCSD(T) computations were performed. The predicted singlet triplet gap is predicted to be 23.8 kcal mol⁻¹, which is very close to the M06 result. The BHLYP functional provides the best agreement of the predicted structures with experimental geometrical parameters. The same functional is found to be the most reliable in predicting structural parameters and electron affinities of germylenes.²⁶ These may be correlated with the fact that the BHLYP functional incorporates the largest fraction of the Hartree–Fock method.¹⁵

3.3. Design of unsaturated stannylenes

Despite the careful analysis by Worthington and Cramer,⁹ the nature of the stabilization of the triplet states of unsaturated carbenes remains incompletely understood. Therefore, the goals of our present computations include contributions to that understanding, as well as to the question of unsaturated stannylenes with triplet ground states. Such desirable structures should be sufficiently sterically unhindered at their two-coordinate tin atoms to allow the exploration of the chemistry of the neutral triplet stannylenes.

The theoretical description of Momeni and Shakib²⁷ of triplet silylenes related to H₂Si=Si: provided some of the motivation for the present exploration of novel unsaturated stannylenes. The substituent effect favoring the triplet state was attributed to the donation of lone pair electrons from Y substituents on Y₂Si=Si: to the half-filled π molecular orbital. Stabilization of the triplet ground states was also credited to the acceptor properties of the β-substituent-silicon antibonding orbitals. The fusion of an Arduengo carbene or silylene onto the unsaturated silylene was suggested by Momeni and Shakib²⁷ to reduce the rate of silylene loss by rearrangements and to increase the size of the energy gap favoring ground triplet states. Our computations describe the effects of changing the Y-substituent of the Y₂Z=Sn: (Z = C, Si, Ge, and Sn) triplet electronic structures. The relative changes in the geometrical parameters and the predicted singlet–triplet gaps obtained with the different functionals used are found to be reasonably consistent. The BHLYP values are used in discussions within this paper, since this functional appears better than the other functionals employed in the prediction of geometries, harmonic vibrational frequencies, and singlet–triplet splittings,²⁸ for related molecules.

3.4. Substitution in stannylenes

The triplet state unsaturated stannylenes arise in a simple picture from the excitation of an electron from the Z=Sn π orbital to the empty 5p orbital component at the Sn atom of the singlet Y₂Z=Sn:. Due to the presence of heavier group 14 elements as α-substituents, there is a lower tendency for the tin atom to form hybrids of the 5s and 5p orbitals on the stannylene centre. Because of this low tendency for hybridization, the expected singlet–triplet energy gaps between the Sn 5p vacant orbitals and the occupied orbitals are expected to be larger compared to the predicted results for silylenes and carbenes.²⁷ Synthetic realization of these predictions would allow to the experimental study of triplet stannylenes. Substitutions of the hydrogen atoms by atoms or groups of greater electronegativity are predicted to enhance the double-bond lengths of these stannylenes along with changes in the ΔE_S–T.

The replacement of the carbon substituent by silicon, germanium, and tin atoms, which are more electropositive than carbon, affects the bonding characteristics of the ground state triplet structures as shown in Table 1, the predicted Z=Sn bond lengths for the triplet state structures calculated with the BHLYP functional are comparable to the typical C–Sn, Si–Sn, Ge–Sn, and Sn=Sn single bonds of 2.19 Å, 2.60 Å, 2.64 Å, and 2.96 Å, respectively.

One of the two π NBOs for the H₂Si=Sn: system includes the overlapping of the Sn p_x orbital with a hybrid pd^0.5 orbital of the adjacent H₂Si: substituent.²⁷ The d orbitals on Sn do not participate significantly in this NBO and instead serve as polarization functions, consistent with results for the H₂Si–Si: molecule.²⁷ Therefore, this NBO is primarily an Sn 5p orbital which is partly polarized towards the adjacent substituent; hence an asymmetric π bond is formed. The replacement of the carbon atom with heavier group 14 atoms may decrease the singlet–triplet splitting values by a significant margin because their lower electronegativity eases promotion of an electron from a π-orbital.²⁷

3.5. Fluorine substitution

In the difluoro substituted stannylenes, F₂Y=Sn: (Y = C, Si, Ge, and Sn), the F–Y σ molecular orbital of the F₂Y: moiety contains contributions from the two fluorine atoms. With respect to the standard Pauling electronegativities,²⁵ the greater electronegativity of fluorine χ_F = 3.98 compared to hydrogen χ_H = 2.20 provides stabilization to the σ molecular orbital of the F₂Z: moiety relative to the σ molecular orbital of H₂Z:. The p–π atomic orbital on the Z centre of the Y₂Z: moiety in the lowest singlet state of Y₂Z=Sn: is destabilized by the delocalisation of a pair of fluorine lone-pair electrons. The pair of electrons that is delocalised into the p–π atomic orbital comes from the in-phase combination of the 2p–π lone pair atomic orbitals on the fluorine atoms. The decrease in the F–Z–F bond angle compared to the predicted H–Z–H bond angles at the Y₂Z centre is a result of an increase in the 2s character of the Z–F σ molecular orbital, which is thus stabilized compared to the σ molecular orbital of ZH₂. The geometries of the lowest lying triplet states of Y₂C=Sn: (Y = H and F) incorporate the lowest lying singlet states of the H₂C: or F₂C: moiety and should be similar. However, the donation of electrons from the fluorine lone pairs into the empty 2pπ orbital on the carbene centre raises the energy of the LUMO of the σ²π⁰ singlet state of the F₂C:. Hence, the triplet ground state of F₂C=Sn: displays a longer C=Sn bond. In replacing the carbon atom with Si, Ge, or Sn, the triplet states are found to be the ground states. The singlet–triplet gaps of F₂Y=Sn: (Y = C, Si, Ge, and Sn) decrease with a decrease in the Pauling electronegativities. The effect of the fluoro substituents reduces the energy of the triplet state structures compared to the unsubstituted H₂Z=Sn: (Z = C, Si, Ge, and Sn) molecules. The singlet–triplet splitting of F₂C=Sn: is significantly decreased compared to H₂C=Sn:, with the computed values (kcal mol⁻¹) being 2.9 (B3LYP), 1.3 (OPBE), and 8.1 (M06), and at the BHLYP functional the triplet state is found to be lower in energy than its singlet state by −1.4 kcal mol⁻¹. The computed singlet–triplet splitting values with the BHLYP functional for the F₂Z=Sn: (Z = C, Si, Ge, and Sn) are −1.4, −8.2, −8.4, and −11.9 kcal mol⁻¹, respectively.

3.6. Analogues of N-heterocyclic carbenes

Turning to the development of experimentally more viable structures, the cyclization of nitrogen-containing five-membered structures should reduce the probability for any rearrangement of unsaturated stannylenes Z₂Y=Sn:. For this reason we selected the (NMeCHCHNMe) subsystem to favour the triplet systems²⁷ and to study the effects of σ-donating substituents on the relative energies of the cyclic triplet ground state systems. In order to achieve a change in the relative energies of the N-heterocyclic carbenes the σ molecular orbital at the carbene centre can be raised. The destabilization of this molecular orbital can be achieved by using substituents at the carbene centre which contain lone pairs of electrons. The properties of the cyclic structures engender ring strain due to the presence of the nitrogen atoms. The nitrogen lone pairs have a large effect on reducing the energy differences between the ¹A₁–σ²π⁰ and ³B₁–σ¹π¹ configurations of the carbene centre.

The computed C–N, Si–N, Ge–N, and Sn–N bond lengths (BHLYP) for the saturated cyclic singlet state structures are 1.364 Å, 1.708 Å, 1.825 Å, and 2.028 Å. For the low-lying triplet states the predicted bond lengths are 1.341 Å, 1.702 Å, 1.815 Å, and 2.019 Å, respectively. The C–N bond length for the corresponding triplet unsaturated cyclic structure is predicted to be shorter than that for the singlet while the Si–N, Ge–N, and Sn–N bond lengths are predicted to be longer. However, there are only modest differences in the N–Z–N bond angles in the saturated and unsaturated structures, between the singlet and triplet states (0.1–1.7°). The differences in the predicted bond lengths result from the lower ability of the nitrogen atoms in the unsaturated rings to donate electron density to the Z centre. A larger increase in ΔE_S–T favoring the triplet state was found for the unsaturated five-membered rings (NMeCHCHNMe)Z=Sn: (Z = C, Si, Ge, and Sn).

4. Summary

The work reported here adds understanding of the nature of the energetics of triplet states of unsaturated stannylenes. Another goal of these computations was the design of unsaturated stannylenes with triplet ground states. The research reported here suggests that unsaturated triplet stannylenes Y₂Sn=Sn: are favored by donation of electron density from a :Y lone pair to a partially occupied 5p-orbital on Sn. This analysis is consistent with the explanation by Momeni and Shakib²⁷ of the effects of electronegative substituents on triplet unsaturated silylenes, Y₂Si=Si:. Table 2 reports the Z=Sn bond lengths and stretching vibrational modes of several unsaturated stannylenes predicted to possess triplet ground states, and their predicted structures are shown in Fig. 1–8. Table 2 reports singlet–triplet energy gaps for unsaturated stannylenes.

Synthetic routes to these unsaturated triplet stannylenes should be achievable. A retro-synthetic analysis containing potentially successful synthetic routes to a family of unsaturated ground-state triplet silylenes, germylenes, and stannylenes is shown in Scheme 2.

Scheme 2 Scheme for the retro-synthetic routes to the unsaturated ground state triplet silylenes, germylenes, and stannylenes. (a) For thermal dissociation of disilene to two silylenes, see ref. 29. (b) For addition of chlorine to silicon–silicon pi-bonds of disilenes, see ref. 30. (c) For a computational study of an insertion by an N-heterocyclic silylene into a Cl–Si bond, see ref. 31.

Acknowledgements

AB and PR acknowledge the facilities at the University of Mauritius. PPG would like to thank the United States National Science Foundation for financial support of this research under NSF grant CHE-1213696. HFS would like to acknowledge generous support from NSF grant CHE-1361178. HFS thanks the Alexander von Humboldt Foundation for allowing a scientific visit to LMU Munich, in the laboratory of Professor Christian Ochsenfeld. The authors are thankful to anonymous reviewers for their useful comments to improve the manuscript.

References

1. S. V. O'Neil, H. F. Schaefer and C. F. Bender, J. Chem. Phys., 1971, 55, 162–169.
2. P. P. Gaspar and G. S. Hammond, The Spin States of Carbenes, Carbene Chemistry, ed. W. Kirmse, Academic Press, Inc., N.Y., 1964, pp. 235–274.
3. P. P. Gaspar and G. S. Hammond, Spin States in Carbene Chemistry, in Carbenes Vol. II, ed. R. A. Moss and M. Jones Jr, John Wiley & Sons, New York, 1975, ch. 6, pp. 207–362.
4. D. Zhou, C. Reiche, M. Nag, J. A. Soderquist and P. P. Gaspar, Organometallics, 2009, 28, 2595–2608.
5. N. Tokitoh and W. Ando, Silylenes (and Germylenes, Stannylenes, Plumbylenes), in Reactive Intermediate Chemistry, ed. R. A. Moss, M. S. Platz and M. Jones Jr, Wiley-Interscience, Hoboken, NJ, 2004, ch. 14, pp. 651–715.
6. W. P. Neumann, Chem. Rev., 1991, 91, 311–334.
7. B. Gehrhus and M. F. Lappert, Polyhedron, 1998, 17, 999–1000.
8. M. Nag and P. P. Gaspar, Organometallics, 2009, 28, 5612–5622.
9. S. E. Worthington and C. J. Cramer, J. Phys. Org. Chem., 1997, 10, 755–767.
10. K. K. Irikura, W. A. Goddard and J. L. Beauchamp, J. Am. Chem. Soc., 1992, 114, 48–51.
11. B.-Y. Li and M.-D. Su, Organometallics, 2011, 30, 6189–6200.
12. M. J. Frisch, et al., Gaussian 09, Revision D.01, Gaussian, Inc., Wallingford CT, 2009.
13. Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215–241.
14. A. D. Becke, J. Chem. Phys., 1988, 38, 3098–3100.
15. A. D. Becke, J. Chem. Phys., 1993, 98, 1372–1377.
16. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785–789.
17. For pseudopotentials of the Stuttgart–Dresden, see http://www.theochem.uni-stuttgart.de/pseudopotentials/index.en.html.
18. S. Huzinaga, J. Chem. Phys., 1965, 42, 1293–1302.
19. T. H. Dunning and P. J. Hay, in Modern Theoretical Chemistry, ed. H. F. Schaefer, Plenum, New York 1977, vol. 3, pp. 1–28.
20. S. Huzinaga, Approximate Atomic Wavefunctions II, University of Alberta, Edmonton, Alberta, 1971.
21. T. J. Lee and H. F. Schaefer, J. Chem. Phys., 1985, 83, 1784–1794.
22. A. Schafer, H. Horn and R. Ahlrichs, J. Chem. Phys., 1992, 97, 2571–2577.
23. E. D. Glendening, A. E. Reed, J. E. Carpenter and F. Weinhold, NBO Version 3.1.
24. F. Weinhold and C. R. Landis, Valency and Bonding: A Natural Bond Orbital Donor–Acceptor Perspective, Cambridge University Press, 2005, p. 760.
25. A. L. Allred, J. Inorg. Nucl. Chem., 1961, 17, 215–221.
26. A. Bundhun, P. Ramasami and H. F. Schaefer, J. Phys. Chem. A, 2009, 113, 8080–8090.
27. M. R. Momeni and F. A. Shakib, Organometallics, 2011, 30, 5027–5032.
28. A. Bundhun, P. Ramasami, P. P. Gaspar and H. F. Schaefer, Inorg. Chem., 2012, 51, 851–863.
29. S. Hiroyuki, T. Norihiro and O. Renji, Bull. Chem. Soc. Jpn., 1995, 68, 2471–2481.
30. V. Y. Lee, K. McNeice, Y. Ito and A. Sekiguchi, Chem. Commun., 2011, 47, 3272–3274.
31. M. Driess, S. Yao, M. Brym, C. van Wuellen and D. Lentz, J. Am. Chem. Soc., 2006, 128, 9628–9629.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra00492j

---