Synthesis and characterisation of new silicon–perfluoropropenyl compounds

Abstract

Novel, stable silicon–pentafluoropropane compounds have been synthesised from the direct reaction of hydrofluorocarbons Z-CFH=CFCF₃ (Z-HFC-1225ye) with ⁿBuLi, followed by appropriate silicon-halide. The compounds have been characterized by multinuclear NMR studies (¹⁹F, ¹H, ²⁹Si and ¹³C), DFT studies and structural confirmation was obtained by X-ray diffraction. Based on the outcome of treating synthetic silicon–pentafluoropropene compounds with different nucleophilic sources (ⁿBuLi, ^tBuLi, MeLi, and PhLi) and computed for this reaction DFT energetics, it is clear that the C–F_trans bond is more active than C–F_gem (F_gem and F_trans are labelled with respect to Si). This provides a route for efficient modification of pentafluoropropene group, that can be a crucial step in developing pharmaceuticals that include propenyl or vinyl groups, addressing the demand for medicines based on long carbonic chains.

---

Introduction

Fluorine plays an important role in the medical field, particularly in pharmacological developments ranging from perfluorinated fluids used as artificial blood and fluoropolymers used in grafts, through applications in drug delivery and in improving the metabolic stability of new medications.¹ It has been found that at least one fluorine moiety is present in 37% of all active small molecule pharmaceutical ingredients that have been approved by the FDA in 2020. Furthermore, between 2011 and 2020, a 26% increase in fluorine-containing pharmaceuticals in all pharmaceuticals approved by the FDA has been noted.²

Due to the importance of fluorocarbon fragments in pharmaceuticals, a number of studies covered the methods of attaching the fluorocarbon fragment into organic compounds³ or transition-metal complexes,⁴ as well as C–F bond activation have been reported.⁵ Unfortunately, there is a lack of studies of pentafluoropropene group (CF=CFCF₃) comparing to analogues perfluorocarbon groups, such as trifluoromethyl CF₃ and trifluoroethene (CF=CF₂).⁶ Therefore, this paper focuses on new silicon-based perfluoropropenyl compounds, which would be suitable for transferring that fluorocarbon fragment via a Hiyama cross coupling reaction into suitable organic substrates. Although these transfers have already been done for tin-containing compounds,⁷ the silicon analogues would be preferred because the majority of silicon compounds are non-toxic and commercially available.⁸ We synthesized a series of silicon–pentafluoropropene compounds in E configuration with the general formula presented in Fig. 1 and Table 1. The obtained compounds have been fully characterized by multinuclear NMR studies (¹⁹F, ¹H, ²⁹Si and ¹³C). The second part of this paper focuses on the investigation of the C–F bond activation through treating synthetic silicon–pentafluoropropene compounds with different nucleophilic sources (ⁿBuLi, ^tBuLi, MeLi, and PhLi). The DFT energetics have been also computed for this reaction.

Fig. 1 Skeleton of general structure of R₃Si(E-CF=CFCF₃).

Table 1 Summary of successfully synthesised compounds with ¹⁹F{¹H} NMR data (376 MHz, CDCl₃, 291 K), (F_gem and F_trans are labelled with respect to Si)

Compound	δ CF3	δ Fgem	δ Ftrans
(Et)3Si(E-CF=CFCF3) (a.1)	−68.13 ppm (d.d)	−137.16 ppm (q.d)	−141.20 ppm (q.d)
	^3J CF3 Ftrans = 13.8 Hz	^4J Fgem CF3 = 6.5 Hz	^3J Ftrans CF3 = 13.8 Hz
	^4J CF3 Fgem = 6.5 Hz	^3J Fgem Ftrans = 13.0 Hz	^3J Ftrans Fgem = 13.5 Hz
(Bu)3Si(E-CF=CFCF3) (a.2)	−67.91 ppm (d.d)	−136.57 ppm (q.d)	−141.33 ppm (q.d)
	^3J CF3 Ftrans = 13.7 Hz	^4J Fgem CF3 = 6.2 Hz	^3J Ftrans CF3 = 13.8 Hz
	^4J CF3 Fgem = 6.4 Hz	^3J Fgem Ftrans = 11.9 Hz	^3J Ftrans Fgem = 11.8 Hz
ClCH2(Me)2Si(E-CF=CFCF3) (a.3)	−67.83 ppm (d.d)	−139.02 ppm (m)	−139.14 ppm (m)
	^3J CF3 Ftrans = 13.3 Hz
	^4J CF3 Fgem = 6.2 Hz
^nBu(Me)2Si(E-CF=CFCF3) (a.4)	−67.62 ppm (d.d)	−137.26 ppm (q.d)	−141.89 ppm (q.d)
	^3J CF3 Ftrans = 13.7 Hz	^4J Fgem CF3 = 6.8 Hz	^3J Ftrans CF3 = 13.2 Hz
	^4J CF3 Fgem = 6.5 Hz	^3J Fgem Ftrans = 12.2 Hz	^3J Ftrans Fgem = 12.8 Hz
Ph(Me)2Si(E-CF=CFCF3) (a.5)	−67.36 ppm (d.d)	−136.30 ppm (q.d)	−140.89 ppm (q.d)
	^3J CF3 Ftrans = 13.2 Hz	^4J Fgem CF3 = 6.2 Hz	^3J Ftrans CF3 = 13.2 Hz
	^4J CF3 Fgem = 6.3 Hz	^3J Fgem Ftrans = 12.9 Hz	^3J Ftrans Fgem = 12.9 Hz
Me(Ph)2Si(E-CF=CFCF3) (a.6)	−67.06 ppm (d.d)	−134.12 ppm (q.d)	−138.22 ppm (q.d)
	^3J CF3 Ftrans = 13.7 Hz	^4J Fgem CF3 = 6.1 Hz	^3J Ftrans CF3 = 13.0 Hz
	^4J CF3 Fgem = 6.2 Hz	^3J Fgem Ftrans = 12.2 Hz	^3J Ftrans Fgem = 12.5 Hz
(Me)2Si(E-CF=CFCF3)2 (a.7)	−68.83 ppm (d.d)	−137.75 ppm (q.d)	−140.36 ppm (m)
	^3J CF3 Ftrans = 13.2 Hz	^4J Fgem CF3 = 5.8 Hz
	^4J CF3 Fgem = 5.4 Hz	^3J Fgem Ftrans = 12.7 Hz
(^iPr)2Si(E-CF=CFCF3)2 (a.8)	−68.97 ppm (d.d)	−136.95 ppm (q.d)	−138.54 ppm (m)
	^3J CF3 Ftrans = 14.1 Hz	^4J Fgem CF3 = 4.8 Hz
	^4J CF3 Fgem = 5.8 Hz	^3J Fgem Ftrans = 12.7 Hz
(Ph)2Si(E-CF=CFCF3)2 (a.9)	−67.95 ppm (d.d.d)	−133.95 ppm (q.d)	−136.94 ppm (m)
	^3J CF3 Ftrans = 13.0 Hz	^4J Fgem CF3 = 5.5 Hz
	^4J CF3 Fgem = 4.5 Hz	^3J Fgem Ftrans = 13.1 Hz
	J CF3 Fexternal = 3.9 Hz
PhSi(E-CF=CFCF3)3 (a.10)	−69.31 ppm (broad d)	−131.04 ppm (q.d)	−141.08 ppm (m)
	^3J CF3 Ftrans = 13.4 Hz	^4J Fgem CF3 = 5.9 Hz
		^3J Fgem Ftrans = 13.5 Hz
Si(E-CF=CFCF3)4 (a.11)	−70.05 ppm (d.m)	−127.32 ppm (q.d)	−145.45 ppm (m)
	^3J CF3 Ftrans = 16.1 Hz	^4J Fgem CF3 = 6.5 Hz
		^3J Ftrans Fgem = 13.0 Hz

---

Results and discussion

Synthesis of silicon–perfluoropropenyl compounds

Based on the previously published method by Brisdon et al.^9,10 1,1,3,3,3-pentafluoropropene (CF₃CF=CFH) – known commercially as Z-HFC-1225ye – was used as a starting material to generate the intermediate Z-perfluoropropenyl lithium (CF₃CF=CFLi), followed by reaction with to prepare , as outlined in Scheme 1.

Scheme 1 The general synthesis of silicon–perfluoropropenyl compounds , (n = 1, 2, 3, 4; m = 1, 2, 3) and R = Me, Et, Bu, ⁱPr, Ph; R′ = ClCH₂, ⁿBu, Me, Ph.

The ¹⁹F{¹H} spectra of all of the silicon–perfluoropropenyl compounds produced the anticipated results: 3 signals with a relative intensity ratio of 3 : 1 : 1, and correlated with the expectations of the perfluoropropenyl fragment. Similarly to the published main-group⁹ and transition-metal perfluoropropenyl complexes,¹⁰ the CF₃ signal appeared between −65 ppm to −70 ppm with a higher intensity than F_gem and F_trans making its assignment straightforward.

In addition, the signal produced, on average, coupling constants of around 13 Hz between CF₃ and F_trans and around 6 Hz between CF₃ and F_gem. The signals for F_gem and F_trans were observed around (−127 to −137) ppm and (−136 to −145) ppm, respectively, and displayed mutual coupling with the CF₃ nuclei and each other. Interestingly, in the di-, tri- and tetra-perfluoropropenyl substituted compounds (Table 1), the CF₃ signal was observed sometime as a doublet of doublets of doublets. The presence of an additional coupling to the CF₃ group is thought to have occurred from a fluorine atom through space, or through the bonds, in addition to the coupling from F_gem and F_trans. This was similar to the coupling patterns that had been observed for [(COD)Pt(E-CF=CFCF₃)₂].¹⁰ The ¹⁹F{¹H} shows an instance of additional coupling in Si(E-CF=CFCF₃)₄ (a.11), wherein the CF₃ signal was observed as a doublet of multiplets instead of a doublet of doublets as expected if coupling only occurred to F_gem and F_trans.

The ¹³C{¹H} NMR spectra of the perfluoropropenyl part of the compounds were as expected, (see ESI†) wherein the C₃ signal was observed as a quartet of doublets of doublets due to the coupling to three equivalent F nuclei of the CF₃ group through one bond, coupling to F_trans through two bonds and to F_gem through three bonds and the coupling constants were found to be ca. 270, 37, and 10 Hz respectively. The C₂ signal appeared as a doublet of quartet of doublets due to the coupling to F_trans, the three equivalent CF₃ fluorines and F_gem, and the coupling constants were found to be ca. 270, 39, and 20 Hz respectively. The C₁ signal also appeared as a doublet of quartet of doublets, the coupling constants were found to be ca. 287, 4.7, and 1.7 Hz respectively.

The ²⁹Si{¹H} NMR spectra of the mono-perfluoropropenyl substituted compounds exhibited a significant coupling with F_gem nucleus, resulting in doublet splitting patterns, with coupling constant of between 20 to 30 Hz, which agrees with the coupling constant of the Si satellites in the ¹⁹F{H} spectrum of F_gem. For comparison, di-, tri- and tetra-substituted-perfluoropropenyl compounds exhibited multiplet splitting patterns as expected.

The majority of the silicon–perfluoropropenyl compounds were liquids, which limited the ability for structural characterisation by single crystal X-ray diffraction. However, (Ph)₂Si(E-CF=CFCF₃)₂ (a.9) was a solid and attempts to grow single crystals were successful. The crystallographic data for the obtained crystal presented in Fig. 2 (see ESI, Tables S11 and S12†) confirmed that (Ph)₂Si(E-CF=CFCF₃)₂ (a.9) was di-substituted, which correlated with the findings from the multinuclear NMR studies (Table 1). The bond lengths and angles for the perfluoropropenyl part of the compound showed similar data to the reported crystallographic values for the transition-metals perfluoropropenyl compounds.^9,10

Fig. 2 (A) ORTEP representation of the structure of (Ph)₂Si(E-CF=CFCF₃)₂ (a.9), (hydrogen atoms omitted for clarity), thermal ellipsoids are shown at 50% probability level. (B) Electrostatic potential map colour coded on the charge density (isovalue 0.01) showing electron rich/deficient (red/blue) regions.

X-ray crystallography and DFT studies

For all compounds, the geometry optimization and electronic structure calculations were performed with DFT. For Ph₂Si(E-CF=CFCF₃)₂ (a.9), the reproduction of the observed solid-state distances by calculations of gas phase geometry using parameters that excluded diffuse functions from the basis set was imperfect, it was pleasing to see that many trends of the observed geometry were reproduced. As illustrated in Table 2, the differences between the observed bond lengths and calculated values were ca. 0.02 Å. The HOMO and LUMO (highest occupied molecular orbital and lowest unoccupied molecular orbital) orbitals are localized mainly on the carbons of the perfluoropropenyl group of the molecule in all cases except compounds with Ph group present, in which these orbitals are more spatially separated (see Fig. 3 and ESI-Table S10†). Thus, in the case of (Me)₂PhSi(E-CF=CFCF₃) (a.5), (Ph)₂MeSi(E-CF=CFCF₃) (a.6) and (Ph)₂Si(E-CF=CFCF₃)₂ (a.9) HOMO is localized on the Ph group, whereas the LUMO mainly on the CF=CFCF₃ part. The localization of the LUMO on the perfluoropropenyl group in all studied compounds indicates the electrophilic character of this group. Further, the calculations of the relative charges on the carbons of the perfluoropropenyl group were also performed. The positive charges on C₂ were found to be higher than those on C₁, (see Table 3). This suggested the likelihood for preferential nucleophilic attack at the C₂ of the perfluoropropenyl group. This charge distribution appears to be consistent irrespective of the other groups coordinated to the silicon centre. For all of the compounds for which calculations were performed, as illustrated in Table 3, Me₂Si(E-CF=CFCF₃)₂ (a.7), Ph₂Si(E-CF=CFCF₃)₂ (a.9) and PhSi(E-CF=CFCF₃)₃ (a.10) the C₂ centre is more positive than C₁. The electrostatic potential maps computed for (a.1), (a.2), (a.3), (a.5), (a.6), (a.7), (a.8), and (a.9) confirm this trend (see ESI Table S13†).

Table 2 Selected bond lengths (Å) for (Ph)₂Si(E-CF=CFCF₃)₂ (a.9) from the crystallographic data (solid phase) with estimated standard deviations in parentheses; and the calculated with DFT bond lengths (Å) for (Ph)₂Si(E-CF=CFCF₃)₂ (a.9) in the gas phase

Solid phase		Gas phase
Atoms	Bond length Å	Atoms	Bond length Å
Si1–C1	1.895(2)	Si3–C26	1.921
C1–C2	1.323(3)	C26–C33	1.317
C2–C3	1.491(4)	C33–C35	1.499
F1–C1	1.362(2)	F34–C26	1.341
F2–C2	1.342(3)	F36–C33	1.324
F3a—C3	1.330(3)	F37–C35	1.320
F3b—C3	1.326(3)	F38–C35	1.316
F3c—C3	1.319(3)	F39–C35	1.319
Si1–C4	1.906(2)	Si3–C23	1.925
C4–C5	1.319(3)	C23–C24	1.317
C5–C6	1.484(4)	C24–C28	1.497
F4–C4	1.365(3)	F25–C23	1.340
F5–C5	1.349(3)	F27–C24	1.323
F6a—C6	1.334(3)	F29–C28	1.320
F6b—C6	1.314(3)	F31–C28	1.316
F6c—C6	1.329(3)	F30–C28	1.319

---

Fig. 3 Visualization of HOMO and LUMO orbitals computed for Et₃Si(E-CF=CFCF₃) (a.1) and (Ph)₂Si(E-CF=CFCF₃)₂ (a.9). Ball and stick representation of the structure: Si – orange, C – grey, F – green, H – white.

Table 3 Calculated Mulliken charges for selected atoms in Me₂Si(E-CF=CFCF₃)₂ (a.7), Ph₂Si(E-CF=CFCF₃)₂ (a.9), and PhSi(E-CF=CFCF₃)₃ (a.10)

Atoms	(a.10)	(a.9)	(a.7)
Si	0.707970	0.768787	0.452303
C1	0.157803	0.156063	0.157768
C2	0.225322	0.253306	0.223625
C1a	0.154999	0.157344	0.156894
C2a	0.275160	0.259685	0.207578
C1b	0.152841
C2b	0.287702

---

Next, calculations of thermodynamic reaction energies between compounds and nucleophilic reagent R′′Li, according to reaction summarized in Scheme 2, were conducted. As an example, DFT reaction energetics for Et₃Si(E-CF=CFCF₃) (a.1) are presented in Table 4. For the energetics of the rest of the studied compounds in reaction with R′′Li see ESI-Table S10.† These calculations revealed that the nucleophilic attack at C₂ position giving Z-isomer as a product is energetically more favourable in all studied cases. Generally, the preference of R′′ to attack the C₂ position over C₁ position increases in the order of: Ph < Me < ⁿBu < ^tBu. Furthermore, the reactivity of nucleophiles in the reaction producing Z-isomer increases in the following order: Ph < ^tBu ≤ Me < ⁿBu, except for (Me)₂Si(E-CF=CFCF₃)₂ (a.7), where using ^tBuLi is energetically the most favourable.

Scheme 2 The general reaction of silicon–perfluoropropenyl compounds with R′′Li (R = Me, Et, ⁿBu, ⁱPr, Ph; R′′ = ⁿBu, ^tBu, Me, Ph).

Table 4 DFT energetics [kcal mol⁻¹] computed for the nucleophilic attack of R′Li at C₂/C₁ position of Et₃Si(E-CF=CFCF₃) (a.1) giving Z-/E-isomer, respectively

Nucleophile R′Li	ΔE [kcal mol^−1] for product
	R3Si(Z-CF=CR′CF3)n	R3Si(E-CR′=CFCF3)n
^nBuLi	−76.35	−70.27
MeLi	−73.55	−67.31
^tBuLi	−72.86	−59.34
PhLi	−69.99	−66.70

---

C–F activation via nucleophilic attack

The outcome from DFT calculation suggests that C–F_trans bond has a higher tendency to be replaced than C–F_gem bond. This theory has been tested by treating the silicon–perfluoropropenyl compounds with different nucleophilic sources (ⁿBuLi, ^tBuLi, MeLi, and PhLi), as shown in Scheme 2. In most of the reactions studied (see Table 6), ¹⁹F{¹H} spectra showed four peaks, and the J values confirmed the existence of a mixture of two compounds for the reaction between R₃Si(E-CF=CFCF₃) and the organolithium reagents. For example, the reaction of Et₃Si(E-CF=CFCF₃) (a.1) and ⁿBuLi, the ¹⁹F{¹H} NMR spectrum showed a total of four peaks, two in the CF₃ region and two others. Based on their integration values they could be divided into two sets of peaks with relative intensities of 3 : 1. For the first pair of peaks (Z-isomer in Scheme 2), a mutual J coupling of approximately 8 Hz was measured, while for the E-isomer, the mutual J coupling was slightly larger (around 10 Hz). In both cases the ¹⁹F and ¹⁹F{¹H} NMR spectra exhibited the same splitting patterns, which excluded the probability of couplings to H. The ²⁹Si{¹H} NMR spectrum showed two peaks with similar intensity ratios as those found in the ¹⁹F{¹H} spectra. Each signal was a doublet with J values of approximately 30 Hz and 9 Hz.

The elemental analysis data for the resulting mixture from treatment of Et₃Si(E-CF=CFCF₃) (a.1) with ⁿBuLi was C: 54.55% and H: 8.23%. These values are close to the values calculated for (a.1) in which one of the fluorines has been replaced by a ⁿBu group, which are C: 54.92% and H: 8.51% (see Table 5). This suggests that the two compounds observed are isomers, rather than two different products.

Table 5 Elemental analysis of (a.1) and the outcome of reaction with RLi

Compound	Molecular formula	% weight (theory)	% weight (found)
(a.1)	C9H15F5Si	C, 43.89; H, 6.14	C, 43.20; H, 5.95
(a.1) + ^nBuLi	C13H24F4Si	C, 54.90; H, 8.51	C, 54.55; H, 8.23
(a.1) + ^tBuLi	C13H24F4Si	C, 54.90; H, 8.51	C, 54.89; H, 8.03
(a.1) + MeLi	C10H18F4Si	C, 49.56; H, 7.49	C, 49.73; H, 7.12
(a.1) + PhLi	C15H20F4Si	C, 59.19; H, 6.63	C, 59.63; H, 6.91

---

For some reactions the ¹⁹F{¹H} and ²⁹Si{¹H} spectra confirmed that only one compound had been generated, as was the case for the reaction between Et₃Si(E-CF=CFCF₃) (a.1) and ^tBuLi. In this case, the mutual F–F coupling in the ¹⁹F{¹H} spectrum was around 7 Hz, while the Si–F coupling observed in the ²⁹Si{¹H} spectrum was 32.9 Hz. These values are similar to those observed for the more intense of the two sets of signals in the mixture that resulted from the reaction between Et₃Si(E-CF=CFCF₃) (a.1) and ⁿBuLi.

When compared to similar systems (see Fig. 4), a range of coupling constants between CF₃ and F are observed, but generally the CF₃–F_trans coupling constants are bigger (>10 Hz) than the coupling constants between CF₃ and F_gem (<10 Hz). The ⁴J(CF₃–F) coupling between substituents on the same side of the double bond are between 15–20 Hz. The ¹⁹F{¹H} NMR data for the major product formed in the reaction of ⁿBuLi with Et₃Si(E-CF=CFCF₃) (a.1) and the only product formed when ^tBuLi was used, had F–F coupling constants of ca. 8.0 and 7.0 Hz respectively. This suggests that the first compound in the mixture occurred as a result of the F_trans substitution by the ⁿBu group leaving F_gem to couple with the CF₃ signal. This substitution produced Et₃Si(Z-CF=CⁿBuCF₃) (12Z), see Scheme 2 and Table 6, which correlated with the ²⁹Si{¹H} NMR data, where the coupling between Si and F was approximately 30 Hz. This is similar to the coupling between Si and F_gem in Et₃Si(E-CF=CFCF₃) (a.1). On the other hand, the low abundance product was Et₃Si(E-CⁿBu=CFCF₃) (12E) which likely results from substitution of F_gem with the ⁿBu group. This is consistent with both the coupling between CF₃ and F, which is ca. 11.0 Hz, in agreement with b.4,⁵ b.5 ⁶ and b.6 ⁸ (see Fig. 4 and Table 6) and the smaller Si–F coupling since the fluorine is now further from the silicon centre. The fluorine NMR data also indicated that it was unlikely for the second compound to be formed with the cis geometry. The results of the ¹³C{¹H} NMR spectra also correlate well with the suggested interpretation of the other NMR data. For the major product of the reaction involving ⁿBuLi, expansions of the signals for the perfluoropropenyl carbons are shown in Fig. 5, while the corresponding signals for the minor product are shown in Fig. 6. For the major species, the carbon of the CF₃ couples to the other fluorine with J = 24.3 Hz, whereas in the minor product the coupling between the carbon nucleus of the CF₃ and the other F is 50.7 Hz, which indicates that the CF₃ is separated from the F by fewer bonds in the minor species (E-isomer) compared with the major product (Z-isomer). Similarly, the coupling between the carbon directly bonded to the unique fluorine atom exhibits a larger quartet coupling in E-isomer (39.5 Hz) than in Z-isomer (6.6 Hz).

Fig. 4 Some examples of published compounds formally derived from substitution of a fluorine atom from a sp² hybridised carbon atom of a perfluoropropenyl group: (b.1),¹ (b.2),¹¹ (b.3),¹¹ (b.4),⁵ (b.5),⁶ (b.6),⁸ (b.7),¹¹ (b.8),¹¹ (b.9),⁵ (TsO = CH₃C₆H₄SO₂).

Table 6 Summary of the results of reactions between and nucleophilic sources with product ratio (given as %), and ¹⁹F{¹H}, ²⁹Si{¹H} NMR data (CDCl₃)

Reactant 1	Reactant 2	Ratio	Result	δ CF3	δ F	δ Si
Et3Si(E-CF=CFCF3) (a.1)	^nBuLi	88	Et3Si(Z-CF=C^nBuCF3) (12Z)	−60.28 ppm, d, ^4J = 8.6 Hz	−99.31 ppm, q, ^4J = 8.2 Hz	6.20 ppm, d, ^2J = 31.2 Hz
		12	Et3Si(E-C^nBu=CFCF3) (12E)	−67.34 ppm, d, ^3J = 10.6 Hz	−106.75 ppm, ^3J = 10.9 Hz	4.96 ppm, d, ^3J = 9.4 Hz
	^tBuLi	100	Et3Si(Z-CF=C^tBuCF3) (13Z)	−55.85 ppm, d, ^4J = 7.4 Hz	−87.06 ppm, q, ^4J = 7.1 Hz	8.70 ppm, d, ^2J = 32.8 Hz
	MeLi	79	Et3Si(Z-CF=CMeCF3) (14Z)	−62.50 ppm, d, ^4J = 8.6 Hz	−98.47 ppm, q, ^4J = 8.7 Hz	6.15 ppm, d, ^2J = 30.7 Hz
		21	Et3Si(E-CMe=CFCF3) (14E)	−67.07 ppm, d, ^3J = 10.3 Hz	−105.47 ppm, q, ^3J = 10.3 Hz	8.95 ppm, d, ^3J = 6.8 Hz
	PhLi	77	Et3Si(Z-CF=CPhCF3) (15Z)	−58.61 ppm, d, ^4J = 9.0 Hz	−93.33 ppm, q, ^4J = 9.2 Hz	7.90 ppm, d, ^2J = 30.3 Hz
		23	Et3Si(E-CPh=CFCF3) (15E)	−67.77 ppm, d, ^3J = 10.6 Hz	−100.27 ppm, q, ^3J = 10.5 Hz	5.73 ppm, d, ^3J = 5.7 Hz
^nBu3Si(E-CF=CFCF3)(a.2)	^nBuLi	88	^nBu3Si(Z-CF=C^nBuCF3) (16Z)	−60.18 ppm, d, ^4J = 8.4 Hz	−98.95 ppm, q, ^4J = 8.5 Hz	2.1 ppm, d, ^2J = 31.7 Hz
		12	^nBu3Si(E-C^nBu=CFCF3) (16E)	−67.22 ppm, d, ^3J = 10.6 Hz	−106.90 ppm, q, ^3J = 10.9 Hz	0.93 ppm, d, ^3J = 9.3 Hz
	^tBuLi	100	^nBu3Si(Z-CF=C^tBuCF3) (17Z)	−55.55 ppm, d, ^4J = 7.1 Hz	−86.66 ppm, q, ^3J = 7.2 Hz	2.15 ppm, d, ^2J = 31.1 Hz
	MeLi	83	^nBu3Si(Z-CF=CMeCF3) (18Z)	−62.42 ppm, d, ^4J = 8.4 Hz	−98.11 ppm, q, ^4J = 8.5 Hz	2.0 ppm, d, ^2J = 31.0 Hz
		17	^nBu3Si(E-CMe=CFCF3) (18E)	−67.06 ppm, d, ^3J = 10.6 Hz	−105.64 ppm, q, ^3J = 10.5 Hz	0.90 ppm, d, ^3J = 8.7 Hz
	PhLi	69	^nBu3Si(Z-CF=CPhCF3) (19Z)	−58.51 ppm, d, ^4J = 9.1 Hz	−93.02 ppm, q, ^4J = 9.0 Hz	3.71 ppm, d, ^2J = 30.6 Hz
		31	^nBu3Si(E-CPh=CFCF3) (19E)	−67.67 ppm, d, ^3J = 10.8 Hz	−100.34 ppm, q, ^3J = 10.8 Hz	−0.01 ppm, d, ^3J = 5.4 Hz
^nBuMe2Si(E-CF=CFCF3) (a.4)	^nBuLi	57	^nBuMe2Si(Z-CF=C^nBuCF3) (20Z)	−59.74 ppm, d, ^4J = 8.5 Hz	−100.36 ppm, q, ^4J = 8.6 Hz	3.71 ppm, d, ^2J = 30.6 Hz
		43	^nBuMe2Si(E-C^nBu=CFCF3) (20E)	−66.88 ppm, d, ^3J = 10.5 Hz	−108.49 ppm, q, ^3J = 10.4 Hz	−0.01 ppm, d, ^3J = 5.4 Hz
	^tBuLi	100	^nBuMe2Si(Z-CF=C^tBuCF3) (21Z)	−55.20 ppm, d, ^4J = 7.3 Hz	−88.24 ppm, q, ^4J = 7.1 Hz	−1.85 ppm, d, ^2J = 37.2 Hz
	MeLi	40	^nBuMe2Si(Z-CF=CMeCF3) (22Z)	−61.59 ppm, d, ^4J = 8.3 Hz	−99.39 ppm, q, ^4J = 8.5 Hz	−21.95 ppm, d, ^2J = 37.0 Hz
		60	^nBuMe2Si(E-CMe=CFCF3) (22E)	−66.65 ppm, d, ^3J = 10.2 Hz	−107.12 ppm, q, ^3J = 10.5 Hz	7.25 ppm, d, ^3J = 5.6 Hz
Me2PhSi(E-CF=CFCF3) (a.5)	^nBuLi	55	Me2PhSi(Z-CF=C^nBuCF3) (23Z)	−59.38 ppm, d, ^4J = 8.2 Hz	−99.54 ppm, q, ^4J = 8.6 Hz	−7.7 ppm, d, ^2J = 37.8 Hz
		45	Me2PhSi(E-C^nBu=CFCF3) (23E)	−66.48 ppm, d, ^3J = 10.0 Hz	−107.00 ppm, q, ^3J = 10.2 Hz	−6.22 ppm, d, ^3J = 11.5 Hz
	^tBuLi	62	Me2PhSi(Z-CF=C^tBuCF3) (24Z)	−59.17 ppm, d, ^4J = 8.2 Hz	−99.55 ppm, q, ^4J = 9.0 Hz	−2.45 ppm, d, ^2J = 38.0 Hz
		38	Me2PhSi(E-C^tBu=CFCF3) (24E)	−66.27 ppm, d, ^3J = 10.3 Hz	−106.72 ppm, q, ^3J = 10.9 Hz	−1.14 ppm, d, ^3J = 6.6 Hz
	MeLi	72	Me2PhSi(Z-CF=CMeCF3) (25Z)	−60.17 ppm, d, ^4J = 8.2 Hz	−98.95 ppm, q, ^4J = 8.7 Hz	6.10 ppm, d, ^2J = 37.4 Hz
		28	Me2PhSi(E-CMe=CFCF3) (25E)	−67.20 ppm, d, ^3J = 10.5 Hz	−106.91 ppm, q, ^3J = 10.9 Hz	1.88 ppm, d, ^3J = 5.6 Hz
Ph2MeSi(E-CF=CFCF3) (a.6)	^nBuLi	70	Ph2MeSi(Z-CF=C^nBuCF3) (26Z)	−59.31 ppm, d, ^4J = 8.3 Hz	−97.03 ppm, q, ^4J = 8.5 Hz	−12.60 ppm, d, ^2J = 38.0 Hz
		30	Ph2MeSi(E-C^nBu=CFCF3) (26E)	−66.65 ppm, d, ^3J = 10.4 Hz	−105.47 ppm, q, ^3J = 10.3 Hz	−9.34 ppm, d, ^3J = 12.9 Hz
	^tBuLi	100	Ph2MeSi(Z-CF=C^tBuCF3) (27Z)	−54.77 ppm, d, ^4J = 6.7 Hz	−84.52 ppm, q, ^4J = 6.7 Hz	−10.03 ppm, d, ^2J = 37.5 Hz
	MeLi	52	Ph2MeSi(Z-CF=CMeCF3) (28Z)	−61.46 ppm, d, ^4J = 7.9 Hz	−96.13 ppm, q, ^4J = 7.7 Hz	−21.91 ppm, d, ^2J = 37.3 Hz
		48	Ph2MeSi(E-CMe=CFCF3) (28E)	−66.37 ppm, d, ^3J = 10.4 Hz	−103.41 ppm, q, ^3J = 10.7 Hz	−10.99 ppm, d, ^3J = 5.3 Hz
	PhLi	100	Ph2MeSi(Z-CF=CPhCF3) (29Z)	−61.13 ppm, d, ^4J = 9.6 Hz	−79.91 ppm, q, ^4J = 9.5 Hz	−22.29 ppm, d, ^2J = 36.3 Hz
Me2Si(E-CF=CFCF3)2 (a.7)	^nBuLi	50	Me2Si(Z-CF=C^nBuCF3)2 (30Z)	−59.71 ppm, d, ^4J = 8.4 Hz	−100.32 ppm, q, ^3J = 8.5 Hz	−21.93 ppm, m
		50	Me2Si(E-C^nBu=CFCF3)2 (30E)	−66.79 ppm, d, ^3J = 10.5 Hz	−108.51 ppm, q, ^4J = 10.0 Hz	−0.57 ppm, m
	^tBuLi	100	Me2Si(Z-C^tBu=CFCF3)2 (31Z)	−55.18 ppm, d, ^4J = 7.2 Hz	−88.23 ppm, q, ^4J = 7.5 Hz	1.80 ppm, m
	MeLi	40	Me2Si(Z-CMe=CFCF3)2 (32Z)	−62.05 ppm, d, ^4J = 8.6 Hz	−99.44 ppm, q, ^4J = 8.5 Hz	−22.76 ppm, m
		60	Me2Si(E-CF=CMeCF3)2 (32E)	−66.75 ppm, d, ^3J = 10.3 Hz	−107.19 ppm, q, ^3J = 10.4 Hz	−21.94 ppm, m
^iPr2Si(E-CF=CFCF3)2 (a.8)	^nBuLi	100	^iPr2Si(Z-CF=C^nBuCF3)2 (33Z)	−61.98 ppm, d, ^4J = 8.6 Hz	−101.26 ppm, q, ^4J = 8.4 Hz	−3.27 ppm, m
	^tBuLi	100	^iPr2Si(Z-CF=C^tBuCF3)2 (34Z)	−56.61 ppm, d, ^4J = 6.5 Hz	−89.0 ppm, q, ^4J = 5.9 Hz	−12.80 ppm, m
	MeLi	100	^iPr2Si(Z-CF=CMeCF3)2 (35Z)	−67.53 ppm, m	−139.77 ppm, m	−13.16 ppm, m
Ph2Si(E-CF=CFCF3)2 (a.9)	^nBuLi	77	Ph2Si(Z-CF=C^nBuCF3)2 (36Z)	−59.93 ppm, d, ^4J = 7.7 Hz	−97.61 ppm, q, ^4J = 8.0 Hz	−11.95 ppm, m
		23	Ph2Si(E-C^nBu=CFCF3)2 (36E)	−67.02 ppm, d, ^3J = 10.5 Hz	−105.09 ppm, q, ^3J = 10.8 Hz	−9.25 ppm, m
	PhLi	100	Ph2Si(Z-CF=CPhCF3)2 (37Z)	−57.84 ppm, d, ^4J = 8.4 Hz	−90.35 ppm, q, ^4J = 8.8 Hz	−21.80 ppm, m

---

Fig. 5 Expansions of C₁ to C₃ signals in the ¹³C{¹H} NMR spectrum for the major product from the reaction of Et₃Si(E-CF=CFCF₃) (a.1) with ⁿBuLi, (100 MHz, CDCl₃, 298 K).

Fig. 6 Expansions of C₁ to C₃ signals in the ¹³C{¹H} NMR spectrum for the minor product from the reaction of Et₃Si(E-CF=CFCF₃) (a.1) with ⁿBuLi, (100 MHz, CDCl₃, 298 K).

The ¹³C{¹H} NMR spectrum for the single product formed in the reaction with ^tBuLi, shown in Fig. 7, is very similar to that observed for Z-isomer, both in terms of the J coupling and splitting patterns. This suggests that the only product formed when using ^tBuLi is (Et)₃Si(Z-CF=C^tBuCF₃) (13Z). According to the DFT study of Ph₂Si(E-CF=CFCF₃)₂, of the carbons in the perfluoropropenyl group C₂ is energetically the most likely site for attack by the incoming nucleophile. Therefore, substitution of F_trans is the most likely result of nucleophilic attack, while attack at C₁ to give the F_gem substituted compound is less favoured. This is consistent with the observation of a small amount of (Et)₃Si(E-C(ⁿBu)=CFCF₃) (13E). However, in case of bigger group such as ^tBu,¹⁰ the steric hindrance prevents any attack on C₁.

Fig. 7 Expansions of C₁ to C₃ signals in the ¹³C{¹H} NMR spectrum for the product from the reaction of Et₃Si(E-CF=CFCF₃) (a.1) with ^tBuLi, (100 MHz, CDCl₃, 298 K).

In Table 6, a summary of all successful attempts to substitute one fluorine atom with an organic group, by reaction with organolithium compounds is listed. PhSi(E-CF=CFCF₃)₃ (a.10) and Si(E-CF=CFCF₃)₄ (a.11) were found not to result in substitution and therefore characterization was not possible. In these cases treatment with ⁿBuLi and ^tBuLi resulted in ¹⁹F{¹H} NMR spectra that showed a number of signals in the CF₃ area, which were no longer present after work up of the reaction. Reactions with MeLi and PhLi resulted in ¹⁹F{¹H} NMR spectra which showed that there was no reaction. The reason that these two substrates reacted differently could be based on steric or electronic factors, or both. For example the large size of the ^tBu group could affect the ability to add to the presumably already sterically crowded PhSi(E-CF=CFCF₃)₃ (a.10) and Si(E-CF=CFCF₃)₄ (a.11) molecules, while MeLi and PhLi are considered to be less reactive compared to ⁿBuLi and ^tBuLi, consistent with DFT calculations. Moreover, the presence of a large number of sterically demanding perfluoropropenyl groups attached to the silicon centre make these substitution reactions less likely, instead allowing for alternative reactions, that result in breaking the Si perfluoropropenyl bond and producing small volatile fluorocarbon molecules which could be easily removed on work up.

Reaction with ⁿBuLi. The substitute of one fluorine atom of the perfluoropropenyl-containing silicon compounds with the ⁿBu group using ⁿBuLi resulted in a mixture of products and in varying proportions. Most substrates resulted in the major products being the Z-isomers, such as was the case for (Et)₃Si(E-CF=CFCF₃) (a.1), (ⁿBu)₃Si(E-CF=CFCF₃) (a.2), (Ph)₂MeSi(E-CF=CFCF₃) (a.6) and (Ph)₂Si(E-CF=CFCF₃)₂ (a.9). However, for ⁿBu(Me)₂Si(E-CF=CFCF₃) (a.4) and (Me)₂Si(E-CF=CFCF₃)₂ (a.7) similar proportions of the E- and Z-isomeric products were formed. By contrast, (ⁱPr)₂Si(E-CF=CFCF₃)₂ (a.8) gave exclusively the Z isomer (ⁱPr)₂Si(Z-CF=CⁿBuCF₃)₂ (33Z). This variance in the proportion of E and Z isomers formed could be due to steric hindrance due to the sizes of the R groups attached to Si. For example, in the case of the most sterically demanding substituent, ⁱPr, it was found that (ⁱPr)₂Si(E-CF=CFCF₃)₂ (a.8) gave only the isomer (ⁱPr)₂Si(Z-CF=CⁿBuCF₃)₂ (33Z). As the groups become smaller the proportion of Z isomer decreases and the E-isomeric product increases. For example: (ⁱPr)₂Si(E-CF=CFCF₃)₂ (a.8) gave 100% of the Z-isomeric product, (Ph)₂Si(E-CF=CFCF₃)₂ (a.9) gave 77% Z-isomer and (Me)₂Si(E-CF=CFCF₃)₂ (a.7) 50% Z-isomer.

Reaction with ^tBuLi. Given the above argument it would be anticipated that increasing the size of the incoming nucleophile is most likely to result in more production of the Z-isomer since the size of the ^tBu group limits the possibility of generating the E-isomer by replacing F_gem on C₁. In the cases of the reactions of ^tBuLi with (Et)₃Si(E-CF=CFCF₃) (a.1), (ⁿBu)₃Si(E-CF=CFCF₃) (a.2), (ⁿBu)Me₂Si(E-CF=CFCF₃) (a.4) (Ph)₂MeSi(E-CF=CFCF₃) (a.6), (Me)₂Si(E-CF=CFCF₃)₂ (a.7), and (ⁱPr)₂Si(E-CF=CFCF₃)₂ (a.8) all reactions resulted in exclusive formation of the Z-isomeric product. However, based on the NMR data reaction with (Me)₂PhSi(E-CF=CFCF₃) (a.5) gave a mixture of both (Me)₂PhSi(Z-CF=C^tBuCF₃) (24Z) and (Me)₂PhSi(E-C^tBu=CFCF₃) (24E) in the ratio 62 : 38. In the case of the reaction of ^tBuLi with (Ph)₂Si(E-CF=CFCF₃)₂ (a.9) the ¹⁹F{¹H} NMR data suggested fragmentation, due to the observation of many signals around the CF₃ region in the ¹⁹F{¹H} NMR spectrum of the crude reaction sample. However, these signals disappeared after the reaction had been worked up and it is suggested that they are therefore small volatile fluorocarbon species.

Reaction with MeLi. When the nucleophilic substitution reactions were performed with a much smaller nucleophile, such as MeLi, a mixture of E- and Z-isomeric products was always formed. Similar to reaction with ⁿBuLi, the reactions involving (Et)₃Si(E-CF=CFCF₃) (a.1) and (ⁿBu)₃Si(E-CF=CFCF₃) (a.2) gave a mixture of both products with a high proportion of Z-isomers. Smaller differences in the proportions of E- and Z-isomers were found in the mixtures that came from reacting MeLi with ⁿBu(Me)₂Si(E-CF=CFCF₃) (a.4), (Ph)₂MeSi(E-CF=CFCF₃) (a.6) and (Me)₂Si(E-CF=CFCF₃)₂ (a.7). However, by analysis of the ¹⁹F{¹H} NMR spectrum, only for (ⁱPr)₂Si(E-CF=CFCF₃)₂ (a.8) was 100% of the Z product obtained, (ⁱPr)₂Si(Z-CF=CMeCF₃)₂ (25Z). Finally, (Ph)₂Si(E-CF=CFCF₃)₂ (a.9) did not show any reaction, even after extending the reaction time to 5 days and the amount of MeLi added has been increased.

Reaction with PhLi. When using phenyllithium, like with ⁿBuLi and MeLi, (Et)₃Si(E-CF=CFCF₃) (a.1) and (ⁿBu)₃Si(E-CF=CFCF₃) (a.2) gave a mixture of the two isomeric products, with a high proportion of the Z-isomers. The reactions with (Ph)₂MeSi(E-CF=CFCF₃) (a.6) and (Ph)₂Si(E-CF=CFCF₃)₂ (a.9) gave the single Z-isomer exclusively. Unfortunately, the substitution of F with the Ph group was unsuccessful for ⁿBu(Me)₂Si(E-CF=CFCF₃) (a.4), (Me)₂PhSi(E-CF=CFCF₃) (a.5), (Me)₂Si(E-CF=CFCF₃)₂ (a.7) and (ⁱPr)₂Si(E-CF=CFCF₃)₂ (a.8) according to multinuclear NMR spectroscopy.

Experimental

Materials and methods

All reagents and solvents were purchased from Sigma-Aldrich (purity 97–98%) and used without purification. Non-chlorinated solvents were dried over sodium wire for at least 24 h prior to use. Z-HFC-1225ye was kindly donated by Mexichem Fluor. NMR spectra were recorded at 20 °C on a Bruker Avance III 400 MHz spectrometer operating at 400.00, 100.61, 376.46, and 79 MHz for ¹H, ¹³C, ¹⁹F, and ²⁹Si respectively using CDCl₃ as solvent. Chemical shift values are quoted relative to TMS and CFCl₃ in parts per million (ppm) on the δ scale and coupling constant (J) values are reported in Hz. Elemental analysis was conducted by the University of Manchester's School of Chemistry Micro-Analytical service. Single crystal was grown by slow evaporation of a chloroform solution and X-ray structures were obtained using SuperNova diffractometers using Mo Kα radiation (λ = 0.71073 Å). All the raw data frames were reduced and corrections were applied for Lorentz, polarisation and absorption using the multi-scan methods with CrysAlisPro.¹²

Computational methods

The X-ray structural data were solved by direct methods, with full-matrix least-squares refinement of F² using: Olex2,¹³ Shelx¹⁴ and Shelxtl¹⁵ programs. Mercury¹⁶ was used to generate the graphical representations. The geometry of Ph₂Si(E-CF=CFCF₃)₂ (a.9) was optimised using hybrid Density Functional Theory (DFT) at the B3LYP/6-31G(d,p) level;^17,18 using the GAMESS software¹⁹ to calculate the bond lengths of (Ph)₂Si(E-CF=CFCF₃)₂ (a.9) in gas phase and the Mulliken charges for Me₂Si(E-CF=CFCF₃)₂ (a.7), Ph₂Si(E-CF=CFCF₃)₂ (a.9), PhSi(E-CF=CFCF₃)₃ (a.10), and Si(E-CF=CFCF₃)₄ (a.11). The electronic structure of (a.1, a.2, a.4, a.5, a.6, a.7, a.8, a.9) compounds and reaction energetics for nucleophilic attack was obtained with B3-LYP/TZVPP²⁰ using TURBOMOLE V7.3 2018 suite of quantum chemical programs.²¹

Synthesis of silicon–perfluoropropenyl compounds

Was prepared with same procedure described in¹⁰ but on different scales (see ESI, Scheme S1†). In a three-necked round-bottom flask under a positive pressure of nitrogen cooled to between −75 to −80 °C were placed dry diethyl ether (150 mL) and one equivalent of Z-HFC-1225ye. One equivalent of ⁿBuLi (2.5 M solution in hexanes) was added slowly so as to maintain the temperature below −78 °C. The solution was stirred for 1 h to ensure formation of perfluoropropenyl lithium. In the next step, a solution of the appropriate silicon-halide was added slowly. The mixture was left to stir and warm slowly to room temperature overnight. Hexane (25 mL) was added to the reaction mixture and the resulting solution was filtered through a pad of Celite, and solvent was removed using a rotary evaporator.

Reactions between silicon–perfluoropropenyl compounds and nucleophilic sources

A solution of dry THF (150 mL) and was placed in a three-necked round-bottom flask under a positive pressure of nitrogen. The solution was cooled to −30 °C, and then RLi (solution in hexanes) was added slowly. The mixture was slowly warmed to room temperature and left stirring for 24 hours. The reaction was worked up with hexanes (10 mL), followed by filtration through Celite, and solvent was removed using a rotary evaporator (see ESI, Scheme S2†).

Conclusions

Derived from (HFC-1225 ye), eleven new and stable silicon–perfluoropropenyl compounds have been successfully prepared, and fully characterised by multinuclear NMR spectroscopy. The compounds formed are generally liquids at room temperature, except Ph₂Si(E-CF=CFCF₃)₂ (a.9) which was solid and structural confirmation was obtained by X-ray diffraction studies.

The investigation of silicon–perfluoropropenyl compounds was extended to study the substitution reactions using a wide range of organolithium nucleophilic sources: ⁿBuLi, ^tBuLi, MeLi, and PhLi, leading to twenty-six new compounds. Two types of products were identified: one where carbolithiation had occurred at C₁ and one at C₂, leading to two isomers of the and formula, respectively. The outcomes of these reactions were rationalised based on steric arguments. Bulky groups around the silicon centre or in the incoming nucleophile (e.g. ⁿBuLi vs. ^tBuLi) led to a greater proportion of the Z-isomer. Due to uneven charges on the carbons of the pentafluoropropene group, where C₂ attached to F_trans has higher positive charge than C₁ attached to F_gem, the nucleophilic attack preferred C₂–F_trans to generate Z-isomer. The calculated reaction energetics between silicon–perfluoropropenyl compounds and organolithium reagents, confirmed that the Z-isomer is energetically more favoured product.

Author contributions

L. Alluhaibi: conceptualization, investigation, methodology, writing – original draft, and writing – review & editing; A. Brisdon: supervision; S. Klejna: investigation, visualization, writing – review & editing; A. Muneer: formal analysis.

Conflicts of interest

The authors declare there are no conflicts of interest.

Acknowledgements

This research was partly supported by program “Excellence Initiative-Research University” for the AGH University of Science and Technology. L. Alluhaibi would like to acknowledge the financial support of the Saudi Arabian Government (King Abdullah Scholarship Program). S. Klejna was supported by the National Science Centre Poland under grant agreement UMO-2019/35/D/ST5/02964. S. Klejna acknowledge PLGrid and Cyfronet for providing computer facilities and support.

References

1. R. B. Larichev, V. A. Petrov, G. J. Grier, M. J. Nappa, W. J. Marshall, A. A. Marchione and R. J. Dooley, Org. Process Res. Dev., 2014, 18, 1060–1066.
2. P. J. Morgan, G. C. Saunders, S. A. Macgregor, A. C. Marr and P. Licence, Organometallics, 2022, 41, 883–891.
3. H. J. Böhm, D. Banner, S. Bendels, M. Kansy, B. Kuhn, K. Müller, U. Obst-Sander and M. Stahl, ChemBioChem, 2004, 5, 637–643.
4. P. J. Morgan, M. W. D. Hanson-Heine, H. P. Thomas, G. C. Saunders, A. C. Marr and P. Licence, Organometallics, 2020, 39, 2116–2124.
5. T. A. Gray, W. R. Dolbier and K. Onnishi, Synthesis, 1987, 10, 956–958.
6. W. Dmowski, J. Fluorine Chem., 1984, 26, 223–241.
7. D. O'Hagan, J. Fluorine Chem., 2010, 131, 1071–1081.
8. K. Funabiki, T. Ohtsuki, T. Ishihara and H. Yamanaka, J. Chem. Soc., Perkin Trans. 1, 1998, 2413–2423.
9. A. K. Brisdon, R. G. Pritchard and A. Thomas, Organometallics, 2012, 31, 1341–1348.
10. L. M. Alluhaibi, A. K. Brisdon and R. G. Pritchard, J. Fluorine Chem., 2017, 203, 146–154.
11. R. N. Haszeldine, C. R. Pool and A. E. Tipping, J. Chem. Soc., Perkin Trans. 1, 1974, 2293–2296.
12. P. W. Betteridge, J. R. Carruthers, R. I. Cooper, K. Prout and D. J. Watkin, J. Appl. Crystallogr., 2003, 36, 1487.
13. O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. Howard and H. Puschmann, J. Appl. Crystallogr., 2009, 42, 339–341.
14. G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008, 64, 112–122.
15. T. R. Schneider and G. M. Sheldrick, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2002, 58, 1772–1779.
16. C. F. Macrae, P. R. Edgington, P. McCabe, E. Pidcock, G. P. Shields, R. Taylor, M. Towler and J. Streek, J. Appl. Crystallogr., 2006, 39, 453–457.
17. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785.
18. A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
19. M. D. Hanwell, D. E. Curtis, D. C. Lonie, T. Vandermeersch, E. Zurek and G. R. Hutchison, J. Cheminf., 2012, 4, 17.
20. A. Schäfer, H. Horn and R. Ahlrichs, J. Chem. Phys., 1992, 97, 2571–2577.
21. R. Ahlrichs, M. Bär, M. Häser, H. Horn and C. Kölmel, Chem. Phys. Lett., 1989, 162, 165–169.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 2240893. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d3ra01353g

---