Abstract
The synthesis and extensive experimental (X‐ray crystallography, UV/Vis spectroscopy, cyclic voltammetry) and theoretical (DFT calculations) characterization of two isomeric dithieno[b,f]phosphepines (DTPs) are presented herein. The relative orientation of the phosphepine and the thiophene moieties has a decisive impact on the electronic and structural properties of these compounds. Moreover, the thiophene units allow for a facile subsequent functionalization through direct Pd‐catalyzed C−H coupling, which renders DTPs highly promising building blocks for organophosphorus functional materials.
Keywords: cross-coupling, phosphepines, phosphorus heterocycles, ring fusion, thiophene
Similar yet different: A series of isomeric thiophene‐fused phosphepines has been synthesized and the impact of different ring fusions on structural, photophysical, and redox properties in parent as well as π‐expanded derivatives has been studied.

Heteroatoms are commonly implemented into π‐conjugated scaffolds to achieve diverse organic materials with desirable optical and electronic properties.1 In this context, phosphorus holds a privileged position due to its versatile electronic nature and reactivity, which allows for efficient molecular property tuning.2
For this reason, phospholes, the phosphorus‐containing analogues of cyclopentadiene, have emerged as promising building blocks for organic materials with unique electronic and redox properties.2a, 2b, 3 In contrast, the larger seven‐membered phosphorus heterocycles, so‐called phosphepines, have not been studied yet in detail. This is mainly due to synthetic challenges arising from the high thermodynamic instability of these compounds.4 As a matter of fact, early reports on the synthesis of unsubstituted phosphepines describe their rapid degradation already at room temperature.5 Stabilization can be achieved through steric shielding of the phosphorus center or annulation of the [b,d,f]‐C=C double bonds (compounds A–D, Scheme 1, left).6 Especially the fusion of (hetero)aromatic rings to the [b,f]‐bonds proved to be a desirable motif as the precursors for this strategy are easily made from simple starting materials.6b, 6d Only recently, the annulation strategy was exploited to furnish dibenzo[b,f]phosphepines (DBPs, B)7 and diazaphosphepines (D),8 both showing remarkable stability. The bottleneck of this chemistry is the variety of derivatization that can be applied to these systems, which up to now required mostly the preinstallation of suitable functionalities and carrying them through the entire synthetic sequence.8 Thus, the realization of phosphepine‐based small molecules that allow for π‐expansion by post‐functionalization methods is a desirable goal.
Scheme 1.

Schematic depiction of reported phosphepines (left) as well as isomeric dithieno[b,f]phosphepines (DTPs, right). The notations α‐ and β‐DTP refer to which positions of the thiophenes are bridged by the phosphorus.
In this connection, thiophene‐fused phosphepines are interesting synthetic targets due to the well‐established rich chemistry of thiophenes allowing for regioselective functionalization.9 Several dithieno[b,f]phosphepines (DTPs) are in principle feasible, depending on the relative orientation of the fused heterocycles. Among these, we were particularly interested in the two symmetric isomers α‐DTP (1α) and β‐DTP (1β, Scheme 1, right). Both scaffolds have been reported previously, however, no photophysical and electrochemical investigations of 1α have been described.6b Moreover, β‐DTP was only isolated as pentavalent phosphine oxide.10 With this work, we therefore provide a reliable and scalable synthesis of trivalent and pentavalent DTPs and discuss the influence of the isomerically different ring fusions on the electronic and structural properties of these compounds. Additionally, we explored the possibilities of post‐functionalization at the thiophene moieties to achieve novel π‐expanded DTPs.
For the realization of α‐DTP and β‐DTP we combined the strategy reported by Lammertsma and co‐workers for DBP7 with the synthetic route towards thiophene‐fused borepines described by Tovar and co‐workers.11 In detail, DTPs were synthesized starting from isomeric (Z)‐dithienylethenes12 2 and 3 allowing for the incorporation of phosphorus at either the α‐ or the β‐positions of the thiophene moieties (Scheme 2, left). Given that oxidation of trivalent phosphepines during chromatography is a known issue, borane adducts 4α and 4β were synthesized first in 46 % and 69 % yield, respectively, through a sequence of lithiation using tBuLi, ring closure through treatment with PPhCl2, and protection of the phosphorus center with borane dimethylsulfide. Deprotection with 1,4‐diazabicyclo[2.2.2]octane (DABCO) gave the desired DTPs 1α and 1β in yields of 96 % and 91 %, respectively. Additionally, the phosphorus center was further modified through treatment with H2O2 to afford the corresponding DTP oxides 5α and 5β in 82 % and 93 % yield, respectively. All DTP derivatives were prepared on a gram scale and handled under ambient conditions, however, slow oxidation of the parent DTPs 1α and 1β was observed during storage of the isolated solids at room temperature.
Scheme 2.

Synthesis of α‐ and β‐DTPs starting from isomeric (Z)‐dithienylethenes together with ORTEP plots of DTPs drawn at the 50 % probability level (left). Hydrogen atoms are omitted for clarity. Conformational ring inversion of phosphepines (right, E=lone pair, BH3, O).
Single crystals of all DTP derivatives suitable for X‐ray crystallography were obtained by slow liquid diffusion of n‐pentane into CH2Cl2 solutions of the corresponding compounds at room temperature (Scheme 2, left). Owing to the nonplanar seven‐membered rings, all derivatives adopt butterfly shaped structures with an average dihedral angle of 145° between the thiophene moieties, whereby 1β has the most bent structure (134°) and 5β the most flattened one (156°). Regarding the environment of the phosphorus center, the sum of C‐P‐C bond angles (ΣC‐P‐C) ranges from 303.9° (1β) to 319.1° (5α), which is similar to the pyramidalizations of PPh3 (ΣC‐P‐C=308.4°)13 and PPh3O (ΣC‐P‐C=318.6°).14 Whereas the P−Cphenyl bond lengths of the DTP derivatives (P−Cphenyl 1.80–1.84 Å) lie in the same range as the P−C bond lengths of PPh3O (1.80 Å)14 and PPh3 (1.83 Å),13 the P−Cring bonds are found to be slightly shorter with an average length of 1.789 Å throughout the series. In this connection, it is worth mentioning that the P−Cring bonds in the α‐isomers are on average slightly shorter by 0.016 Å than in the β‐derivatives (Table S1, Supporting Information).
DTP oxide 5β adopts a different conformation than the other DTPs with its oxygen atom being oriented almost perpendicular to the DTP backbone (anti conformation). These features agree with those of the crystal structure reported for the same compound by Delouche et al.10 In contrast, the lone pairs, borane units, and oxo groups, respectively, are pointing in the same spatial direction as the thiophenes (syn conformation) in case of all other derivatives. It has been shown computationally that syn and anti conformations of phosphepines can interconvert through ring flip of the seven‐membered ring with rather small activation barriers ≤5 kcal mol−1 (Scheme 2, right).4b, 15 Based on calculations performed at the B3LYP‐D3/def2‐TZVPD level of theory (Figure S39, Supporting Information), we found the syn conformation of 1α and 1β to be lower in energy by 3.6 and 5.7 kcal mol−1, respectively, than the corresponding anti geometries. In case of the oxides 5α and 5β, this energy difference amounts to less than 1 kcal mol−1. With such small energetic preferences in mind, it can be assumed that the observed conformations in the crystal structures are a result of packing effects.
A common packing motif or distinct intermolecular interactions were not observable in the solid state packing of DTPs 1, 4, and 5, except for 1β showing S⋅⋅⋅S interactions with a distance of 3.6 Å (Figure S32, Supporting Information).
The electronic properties of DTPs 1 and 5 were investigated through UV/Vis absorption and fluorescence spectroscopy (in CH2Cl2) as well as cyclic voltammetry (in THF) and the results were corroborated by DFT calculations performed at the B3LYP‐D3/def2‐TZVPD level of theory.
The lowest‐energy absorption wavelengths (λ max) of the β‐isomers are considerably red‐shifted by 31–39 nm as compared with the α‐derivatives (Figure 1, Table 1). The optical band gaps obtained from UV/Vis spectroscopy are in qualitative agreement with the calculated HOMO–LUMO gaps, which are found to be around 0.5 eV smaller for the β‐DTPs (Figure 2). These findings point to the presence of a longer effective conjugation pathway in the β‐DTPs, as a result of the relative orientation of the thiophenes allowing for electronic communication between the outer α‐positions through the [d]‐bond of the seven‐membered ring (wing‐to‐wing communication, Figure 1). It is worth mentioning that Tovar and co‐workers have reported similar observations in isomeric dithieno[b,f]borepines (DTBs), however, the effect in this case was much less pronounced with a redshift of only 12 nm of the λ max value of the β‐isomer.11
Figure 1.

a) UV/Vis absorption (solid lines) and fluorescence spectra (dashed lines) of DTPs 1 and 5 recorded in CH2Cl2 at room temperature. The inset shows the different conjugation pathways within α‐ and β‐isomers (E=lone pair, O). b) Normalized (with respect to the lowest energy absorption bands) UV/Vis absorption spectra of π‐expanded DTPs 6 recorded in CH2Cl2 at rt.
Table 1.
UV/Vis spectroscopic, electrochemical, and theoretical data of DTPs 1, 5, and 6.
|
Compound |
λ max [nm][a] |
ϵ [m −1 cm−1] |
E gap [eV][b] |
HOMO–LUMO gap [eV][c] |
λ em [nm][a] |
Stokes shift [nm] ([cm−1])[d] |
Φ F [e] [%] |
E red,1 [V][f] |
|---|---|---|---|---|---|---|---|---|
|
1α |
306 |
4900 |
4.05 |
4.38 |
390 |
84 (7050) |
<1 |
−2.74 |
|
5α |
314 |
7800 |
3.95 |
4.33 |
392 |
78 (6340) |
2 |
−2.40 |
|
1β |
345 |
10 200 |
3.59 |
3.87 |
425 |
80 (5460) |
<1 |
−2.52 |
|
5β |
345 |
14 500 |
3.59 |
3.86 |
424 |
79 (5400) |
5 |
−2.22 |
|
6α |
314 |
– |
3.95 |
– |
424 |
110 (8260) |
1 |
−2.16 |
|
6β |
407 |
– |
3.05 |
– |
479 |
72 (3690) |
13 |
−1.93 |
[a] Measured in CH2Cl2 at room temperature. [b] Optical band gap calculated using E gap=h⋅c⋅λ max −1. [c] Obtained from DFT calculations performed at the B3LYP‐D3/def2‐TZVPD level of theory. [d] Calculated as λ em−λ max. [e] Fluorescence quantum yields. [f] Half‐wave potentials were obtained from CV measurements in 0.1 m solutions of nBu4NPF6 in THF (scan rate 0.1 V s−1, vs. Fc/Fc+).
Figure 2.

Kohn–Sham FMOs of isomeric DTPs 1 and 5 as obtained from DFT calculations at the B3LYP‐D3/def2‐TZVPD level of theory.
All DTPs feature weak (Φ F<5 %) blue light fluorescence in solution (CH2Cl2). The fluorescence spectra are mirror‐imaged to the corresponding lowest‐energy absorption bands with large Stokes shifts of approximately 80 nm (5500–7000 cm−1). Recent studies have shown that dibenzannulated arsepines as well as other 8 π‐electron heteropines undergo planarization upon photoexcitation due to a gain of excited‐state aromaticity (Baird's rule)16 resulting in unusually large Stokes shifts.17 Although computations suggest that phosphepines might behave similarly,17a our experimental results provide no indication for such effect as the fluorescence of trivalent and pentavalent DTP derivatives is essentially the same.
The electrochemical data of DTPs 1 and 5 is summarized in Table 1. Anodic events were not observable within the electrochemical window (<1.0 V) except for 1β, which shows an irreversible oxidation at +0.67 V. All derivatives feature a reversible or quasi‐reversible reduction between −2.2 and −2.8 V (vs. Fc/Fc+). Going from trivalent DTPs 1 to oxides 5 lowers the reduction potential by approximately 0.3 V. Regarding the influence of the relative orientation of thiophene and phosphepine it is discernible that the reduction potentials of the α‐DTPs are cathodically shifted by 0.2 V as compared with their isomeric counterparts. In analogy to this observation, the calculated LUMO levels of 1α and 5α are higher in energy by 0.25–0.3 eV than those of their isomeric relatives (Figure 2). These trends can be explained by a higher electron density around the phosphorus center in the α‐isomers. In detail, the ring fusion in the latter facilitates the delocalization of electron density from the electron‐rich thiophenes towards phosphorus (wing‐to‐phosphorus communication, Figure 1), which consequently renders their reductions more difficult.
Additional support for the observed conjugation pathways is provided by the frontier molecular orbitals (FMOs, Figure 2). In case of α‐DTPs 1α and 5α, the orbitals are mostly concentrated on the central ring and on the sulfur atoms, which reflects the strong wing‐to‐phosphorus conjugation. In contrast, the orbital coefficients at the outer α‐positions at thiophene are noticeably larger in β‐DTPs 1β and 5β, which is in agreement with wing‐to‐wing communication.
With DTPs 1, 4, and 5 in hand, we targeted the functionalization of the thiophene moieties. In this connection, the introduction of functional groups at the α‐positions of the thiophenes was probed first. Given the fact that DTPs possess a phosphine‐like reactivity rendering them susceptible towards oxidation, our efforts were focused on the borane adducts 4 and phosphepine oxides 5. In principle, two reaction pathways are feasible, that is, first, electrophilic aromatic substitution and, second, metalation followed by quenching with suitable electrophiles. Unfortunately, both routes showed to be unsuitable for our DTPs with the exception of borane adduct 4β, which allowed for the introduction of bromines and formyl groups at the α‐position of thiophene through treatment with tBuLi and subsequent quenching with elemental bromine and DMF, respectively (compounds S1 and S2 in the Supporting Information). Although both functionalities should in principle allow for subsequent derivatization, we did not pursue this approach any further and turned our attention towards more straightforward methods such as direct oxidative cross‐couplings. Very recently, Delouche et al. reported on the direct Pd‐catalyzed arylation of 5β,10 however, we found that this approach is not applicable to the corresponding α‐isomer. Instead, we were able to introduce phenylethynyl units through direct C−H functionalization of thiophene. Suitable conditions (Pd2(dba)3, PivOH, Cs2CO3, Et3N, Ag2O in DME at 100 °C)9a were adapted from Jie et al. and afforded the desired ethynylated DTP oxides 6α and 6β in over 40 % yield (Scheme 3).
Scheme 3.

Functionalization of DTP oxides 5 through direct Pd‐catalyzed C−H coupling.
The influence of the π‐expansion through ethynylation was investigated through UV/Vis spectroscopy (Figure 1, Table 1). Strikingly, α‐derivative 6α has the same λ max value (314 nm) as unfunctionalized 5α. This observation is comprehensible remembering that the outer α‐positions in the α‐DTPs are cross‐conjugated, which prevents efficient communication of the phenylethynyl substituents in 6α. Turning to β‐isomer 6β, a strong bathochromic shift of the lowest‐energy absorption band of 62 nm is observed as compared to 5β. In this case, the acetylenes are conjugated through the olefinic backbone, which leads to an effectively expanded π‐system.
Similar to their unfunctionalized counterparts, 6 show fluorescence in solution (CH2Cl2) with Stokes shifts >70 nm (3700–8300 cm−1, Figure S37, Supporting Information). Noteworthy, the fluorescence quantum yield of 6α (Φ F=1 %) was found be as low as for the unfunctionalized DTPs, whereas the fluorescence of 6β was significantly stronger with Φ F=13 %. Cyclic voltammetry measurements revealed that the first reduction potentials of 6 are lowered roughly equally by 0.2 V as compared to 5. Thus, the significantly redshifted absorption of the β‐isomer is most likely to be ascribed to a more pronounced increase of the energy of its HOMO. With the different electronic structures of α‐ and β‐DTPs in mind, the coupling of different acetylenic groups offers great possibilities for target‐oriented molecular property tuning.
In summary, we presented a reliable and scalable synthetic route towards a series of two isomeric dithieno[b,f]phosphepines, that is, α‐ and β‐DTPs. Comprehensive experimental and theoretical investigations revealed the different electronic nature of the isomeric DTPs regarding the relative orientation of phosphepine and thiophene moieties. In particular, wing‐to‐phosphorus communication turned out to be the preferred conjugation pathway in α‐DTPs, whereas wing‐to‐wing communication was found to be dominant in β‐DTPs. Finally, we showed that it is possible to introduce acetylenes through direct C−H functionalization at thiophene, which is a promising strategy for the synthesis of novel functional materials based on the DTP scaffold.
Conflict of interest
The authors declare no conflict of interest.
Supporting information
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Supplementary
Acknowledgements
Generous funding by the Deutsche Forschungsgemeinschaft (DFG)—Project number 182849149‐SFB 953 and project number 401247651‐KI 1662/3‐1 is gratefully acknowledged.
K. Padberg, J. D. R. Ascherl, F. Hampel, M. Kivala, Chem. Eur. J. 2020, 26, 3474.
References
- 1.
- 1a. Hirai M., Tanaka N., Sakai M., Yamaguchi S., Chem. Rev. 2019, 119, 8291–8331; [DOI] [PubMed] [Google Scholar]
- 1b. Stępień M., Gońka E., Żyła M., Sprutta N., Chem. Rev. 2017, 117, 3479–3716. [DOI] [PubMed] [Google Scholar]
- 2.
- 2a. Baumgartner T., Acc. Chem. Res. 2014, 47, 1613–1622; [DOI] [PubMed] [Google Scholar]
- 2b. Baumgartner T., Réau R., Chem. Rev. 2006, 106, 4681–4727. [DOI] [PubMed] [Google Scholar]
- 3.
- 3a. Duffy M. P., Delaunay W., Bouit P. A., Hissler M., Chem. Soc. Rev. 2016, 45, 5296–5310; [DOI] [PubMed] [Google Scholar]
- 3b. Matano Y., Chem. Rec. 2015, 15, 636–650; [DOI] [PubMed] [Google Scholar]
- 3c. Ren Y., Baumgartner T., Dalton Trans. 2012, 41, 7792–7800; [DOI] [PubMed] [Google Scholar]
- 3d. Matano Y., Imahori H., Org. Biomol. Chem. 2009, 7, 1258–1271. [DOI] [PubMed] [Google Scholar]
- 4.
- 4a. Jansen H., Slootweg J. C., Lammertsma K., Beilstein J. Org. Chem. 2011, 7, 1713–1721; [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4b. Jansen H., Slootweg J. C., Ehlers A. W., Lammertsma K., Organometallics 2010, 29, 6653–6659. [Google Scholar]
- 5.
- 5a. Kurita J., Shiratori S., Yasuike S., Tsuchiya T., J. Chem. Soc. Chem. Commun. 1991, 1227–1228; [Google Scholar]
- 5b. Märkl G., Burger W., Tetrahedron Lett. 1983, 24, 2545–2548; [Google Scholar]
- 5c. Märkl G., Schubert H., Tetrahedron Lett. 1970, 11, 1273–1276. [Google Scholar]
- 6.
- 6a. Yasuike S., Shiratori S.-i., Kurita J., Tsuchiya T., Chem. Pharm. Bull. 1999, 47, 1108–1114; [Google Scholar]
- 6b. Yasuike S., Nakashima F., Kurita J., Tsuchiya T., Heterocycles 1997, 45, 1899–1902; [Google Scholar]
- 6c. Yasuike S., Kiharada T., Kurita J., Tsuchiya T., Chem. Commun. 1996, 2183–2184; [Google Scholar]
- 6d. Yasuike S., Ohta H., Shiratori S., Kurita J., Tsuchiya T., J. Chem. Soc. Chem. Commun. 1993, 1817–1819; [Google Scholar]
- 6e. Märkl G., Burger W., Angew. Chem. Int. Ed. Engl. 1984, 23, 894–895; [Google Scholar]; Angew. Chem. 1984, 96, 896–897; [Google Scholar]
- 6f. Winter W., Chem. Ber. 1976, 109, 2405–2419. [Google Scholar]
- 7. Lyaskovskyy V., van Dijk-Moes R. J. A., Burck S., Dzik W. I., Lutz M., Ehlers A. W., Slootweg J. C., de Bruin B., Lammertsma K., Organometallics 2013, 32, 363–373. [Google Scholar]
- 8. Ren Y., Sezen M., Guo F., Jakle F., Loo Y.-L., Chem. Sci. 2016, 7, 4211–4219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.
- 9a. Jie X., Shang Y., Hu P., Su W., Angew. Chem. Int. Ed. 2013, 52, 3630–3633; [DOI] [PubMed] [Google Scholar]; Angew. Chem. 2013, 125, 3718–3721; [Google Scholar]
- 9b. Baghbanzadeh M., Pilger C., Kappe C. O., J. Org. Chem. 2011, 76, 8138–8142; [DOI] [PubMed] [Google Scholar]
- 9c. McCullough R. D., Adv. Mater. 1998, 10, 93–116. [Google Scholar]
- 10. Delouche T., Mocanu A., Roisnel T., Szűcs R., Jacques E., Benkő Z., Nyulászi L., Bouit P.-A., Hissler M., Org. Lett. 2019, 21, 802–806. [DOI] [PubMed] [Google Scholar]
- 11. Levine D. R., Siegler M. A., Tovar J. D., J. Am. Chem. Soc. 2014, 136, 7132–7139. [DOI] [PubMed] [Google Scholar]
- 12. Imrich H.-G., Conrad J., Beifuss U., Eur. J. Org. Chem. 2015, 7718–7734. [DOI] [PubMed] [Google Scholar]
- 13. Dunne B. J., Orpen A. G., Acta Crystallogr. Sect. C 1991, 47, 345–347. [Google Scholar]
- 14. Al-Farhan K. A., J. Crystallogr. Spectrosc. Res. 1992, 22, 687–689. [Google Scholar]
- 15. Kassaee M. Z., Cheshmehkani A., Musavi S. M., Majdi M., Motamedi E., J. Mol. Struct. THEOCHEM 2008, 865, 73–78. [Google Scholar]
- 16.
- 16a. Baird N. C., J. Am. Chem. Soc. 1972, 94, 4941–4948; [Google Scholar]
- 16b. Ottosson H., Nat. Chem. 2012, 4, 969. [DOI] [PubMed] [Google Scholar]
- 17.
- 17a. Kawashima I., Imoto H., Ishida M., Furuta H., Yamamoto S., Mitsuishi M., Tanaka S., Fujii T., Naka K., Angew. Chem. Int. Ed. 2019, 58, 11686–11690; [DOI] [PubMed] [Google Scholar]; Angew. Chem. 2019, 131, 11812–11816; [Google Scholar]
- 17b. Toldo J., El Bakouri O., Solà M., Norrby P.-O., Ottosson H., ChemPlusChem 2019, 84, 712–721. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Supplementary
