Abstract
1,3-Azaphospholo[1,5-a]pyridines named as 2-phosphaindolizines were obtained from [4+1] cyclocondensation and 1,5-electrocyclization methods. These compounds incorporate several reactive functionalities and undergo electrophilic substitution, Diels-Alder reaction and 1,2-addition with hydrogen sulfide and sulfur across the >C=P– functionality and form η1-P coordination compounds with metal carbonyls. Thus, 2-phosphaindolizines could be used as precursors for accessing a variety of organophosphorus compounds.
Introduction
An almost explosive growth in the field of low-coordination phosphorus compounds during the last three decades of the 20th century led the chemists realize that a remarkable analogy existed between these compounds and their carbon-analogues [1], [2], [3]. The analogy was so overwhelming that a new phrase “Phosphorus: The Carbon Copy” [4] was coined. It gave further impetus to the synthesis of more phospha-analogues of classical acyclic, cyclic and heterocyclic compounds and study their chemical properties. Thus, initially synthesized phosphaalkenes [5] and phosphaalkynes [6] were subsequently used as building blocks for obtaining a large variety of phosphaorganic compounds having carbon-phosphorus л–bond, such as phosphinines [7] and heterophospholes [8]. Schmidpeter [9] termed heterophospholes as a postscript chapter of the heterocyclic chemistry due to their close resemblance with the related compounds having no phosphorus atom. In an earlier review, we compared systematically the synthesis, structures and reactivity of the heterophospholes with those of the corresponding heteroles and it was emphasized that the synthetic methods used for heteroles including the annelated heteroles could be extended successfully to the synthesis of their phosphorus containing analogues [10]. As regards the reactivity, the presence of a >C=P– functionality in the five-membered ring may alter the reactivity of these compounds and additional reactions are observed involving this moiety leading to the access of a wide variety of organophosphorus compounds.
Thus, in analogy to the [4+1] cyclocondensation of a 2-substituted cycloiminium salt with a carboxylic acid derivative to afford annelated pyrrole and other azoles [11], we could develop a simple method for the preparation of annelated 1,3-azaphospholes involving [4+1] cyclocondensation of 2-substituted cycloiminium salts with PCl3 or P(NMe2)3 [12]. Another analogy we followed was the construction of the indolizine nucleus through 1,5-electrocyclization of N-allylpyridinium ylides [13]; we succeeded in introducing a phosphorus atom in the 2-position of indolizine by generating N-pyridinium dichlorophosphinomethylide in situ which underwent disproportionation followed by successive 1,5-electrocyclization and 1,2-elimination to afford 2-phosphaindolizines [14].
As mentioned earlier, introduction of a two-coordinated P-atom in the five-membered ring not only enhances the reactivity of these compounds, but they show many additional reactions, the centre of reactivity being the >C=P– functionality. In view of this, it could be possible to use the synthesized 2-phosphaindolizines as precursors to obtain a variety of organophosphorus compounds.
In this perspective, we present mainly the work carried out in our own laboratories with the emphasis on how interesting chemistry could be developed around 2-phosphaindolizines. The literature has been cited right from the beginning till the latest results.
Synthetic methods of 2-phosphaindolizines
[4+1] Cyclocondensation method
The reaction of 1,2-dialkylpyridinium bromides (1) with PCl3 in the presence of Et3N at room temperature (r.t.) afforded 1,3-azaphospholo[5,1-a]pyridines (2) in good to very good yields (Scheme 1) [15], [16], [17]. As mentioned earlier, the general skeleton was christened as 2-phosphaindolizine perceiving it as a “formal” CH/P exchange at the 2-position of indolizine [15].
![Scheme 1:
Synthesis of 2-phosphaindolizines through [4+1] cyclocondensation.](/document/doi/10.1515/pac-2018-0915/asset/graphic/j_pac-2018-0915_fig_002.jpg)
Synthesis of 2-phosphaindolizines through [4+1] cyclocondensation.
Following similar strategy, 1,3,4-diazaphospholo[1,2-a]pyridines (1-aza-2-phosphaindolizines) [18], 1,2,3-diazaphospholo[1,5-a]pyridines (3-aza-2-phosphaindolizines) [19] and 1,2,4,3-diazaphospholo[1,5-a]pyridines (1,3-diaza-2-phosphaindolizines) [20] could also be obtained. It was possible to extend this method to the synthesis of 1,3 azaphospholes annelated to other heterocycles, namely pyrazine [15], oxazole [21], thiazole [22] and benzothiazoles [22] also.
Synthetic method involving 1,5-electrocyclization
1-(Alkoxycarbonylmethyl)pyridinium bromides (3) on reacting with PCl3 in the presence of Et3N at r.t. afforded 1,3-bis(alkoxycarbonyl)-2-phosphaindolizines (4) (Scheme 2) [14], [23].
Synthesis of 1,3-bis(alkoxycarbonyl)-2-phosphaindolizines.
On the basis of a crossed experiment, when a mixture of four products was obtained, the following mechanism was proposed (Scheme 3) [14].
Mechanism of the formation of 1,3-bis(alkoxycarbonyl)-2-phosphaindolizines.
Subsequently, we could prepare isoquinoline (5) [24] and phenanthridine (6) [25] analogues in similar manner.
Reactions
2-Phosphaindolizine nucleus incorporates several reactive functionalities (Fig. 1).
Reactive functionalities present in the 2-phosphaindolizine nucleus.
Electrophilic substitution reactions
Bromination
1-Unsubstituted 2-phosphaindolizines on reacting with N-bromosuccinimide (NBS) afforded 1-bromo derivatives (7) in good yields (Scheme 4) [26].
Bromination of 2-phosphaindolizines.
On hydrolysis, five membered ring of 1-bromo-2-phosphaindolizines opened and they finally changed into (2-pyridinio)methylphosphonates [26].
Chlorophosphinylation
1-Unsubstituted 2-phosphaindolizines reacted with PCl3 or PhPCl2 (but not with Ph2PCl) in the presence of Et3N to give 1-phosphinylated products (8). The dichlorophosphino derivative formed initially on reaction with PCl3 undergoes dismutation to form -PCl- bridged bis(2-phosphaindolzine) (9) (Scheme 5) [26].
Chlorophosphinylation of 2-phosphaindolizines.
1-Dichlorophosphino-2-phosphaindolizines (8, R=Cl) on reacting with MeOH in the presence of Et3N changed to phosphonite ester (10), which rearranged to phosphinate ester (11) or could be oxidized to thiophosphonate ester (12) (Scheme 6) [26].
Reaction of 1-dichlorophosphino-2-phosphaindolizine with methanol.
Reactions across the >C=P– functionality
Diels-Alder reaction
1,3-Bis(alkoxycarbonyl)-2-phosphaindolizines (4) reacted with 2,3-dimethyl-1,3-butadiene (DMB) and with isoprene in the presence of sulfur to give [2+4] cycloadducts (13) (Scheme 7) [27], [28]. In the absence of sulfur, the reaction is slow and takes about 10 days for completion. The role of sulfur was established to oxidize the phosphorus atom of the initially formed cycloadduct to push the reaction in the forward direction. The reaction with isoprene under these conditions occurred with total regioselectivity to afford only one regioisomer in each case.
DA reaction of 1,3-bis(alkoxycarbonyl)-2-phosphaindolizines.
Likewise, isoquinoline-[28] and phenanthridine [25] -analogues of 1,3-bis(alkoxycarbonyl)-2-phosphaindolizines underwent DA reactions. The reaction of 1,3-bis(ethoxycarbonyl)-1,3-azaphospholo[5,1-a]isoquinoline (14) with isoprene in the presence of methyl iodide, however, gave two regioisomers, the major (62 %) being 15 (Scheme 8) [28].
![Scheme 8:
Regioselective DA reaction of 1,3-bis(ethoxycarbonyl)-1,3-azaphopholo[5,1-a] isoquinoline with isoprene in the presence of MeI.](/document/doi/10.1515/pac-2018-0915/asset/graphic/j_pac-2018-0915_fig_011.jpg)
Regioselective DA reaction of 1,3-bis(ethoxycarbonyl)-1,3-azaphopholo[5,1-a] isoquinoline with isoprene in the presence of MeI.
On the other hand, 2-phosphaindolizine having electron-withdrawing group (EWG) only at 3-position, namely 3-(ethoxycarbonyl)-1-methyl-2-phosphaindolizine (17) did not undergo DA reaction with DMB even in boiling toluene in the presence of sulfur [25]. Theoretical investigation at the DFT level indicated that the presence of EWG at the 3-position makes the nitrogen lone pair transferred more effectively to the azaphosphole ring, as a result of which >C=P– moiety becomes electron-rich and hence, it does not undergo DA reaction with an electron-rich diene, such as DMB. But an additional EWG at 1-position as in 4 acts as a trap for the negative charge due to which character of the >C=P– functionality remains electron-deficient and it reacts with DMB [29].
Organoaluminium reagent catalyzed Diels-Alder reaction
We first investigated the effect of AlCl3 as catalyst on the dienophilic reactivity of indolizine and 2-phosphaindolizine towards DA reaction with 1,3-butadiene theoretically at the DFT (B3LYP/6-31+G**) level [30]. Unexpectedly it was found that co-ordination of AlCl3 with the oxygen atom of the carbonyl group of indolizine or 2-phosphaindolizine, instead of lowering the activation barrier, raised it further as compared to the activation barrier for the DA reaction of the uncomplexed indolizine or 2-phosphaindolizine with 1,3-butadiene (Scheme 9) [30].
Computed model DA reactions of indolizines (Z=CH) and 2-phosphaindolizines (Z=P) with 1,3-butadiene at B3LYP/6-31+G**.
The NBO analysis revealed that co-ordination of AlCl3 to the carbonyl group accentuates the transfer of the bridgehead nitrogen atom lone-pair to the azaphosphole ring still more efficiently than in the uncomplexed indolizine or 2-phosphaindolizine, thus making the dienophilic moiety, >C=Z– (Z=CH or P) more electron-rich thereby raising the activation barrier of the DA reaction further.
2-Phosphaindolizine, however, has another site for co-ordination: σ2,λ3-P atom. On computing the DA reaction of 2-phosphaindolizine complexed with AlCl3 through the σ2,λ3-P atom (20), with 1,3-butadiene, the activation barrier is lowered remarkably; it comes down to 22.57 kcal mol−1 only (Scheme 10) [30].
Computed model DA reaction of (σ2,λ3-P-AlCl3)-2-phosphaindolizine with 1,3-butadiene at B3LYP/6-31+G**.
Guided by the above theoretical findings, we attempted DA reaction of 3-(alkoxycarbonyl/acyl)-1-methyl-2-phosphaindolizines (22) with DMB in the presence of ethylaluminium dichloride as a promoter (i.e. using its equimolar quantity) when a clean reaction occurred to afford [2+4] cycloadducts (23) (Scheme 11) [31]. The 31P NMR indicated formation of a single product in each case; the chemical shift of δ=−16.6 to −5.3 ppm accords well with the 31P NMR chemical shifts reported for the [2+4] cycloadducts obtained from the DA reaction of P-W(CO)5 complexes of λ3-phosphinines [32].
DA reaction of 3-(alkoxycarbonyl/acyl)-1-methyl-2-phosphoindolizines with DMB in the presence of ethylaluminium dichloride as promoter.
However, aluminium chloride reagent being a hard acid is expected to co-ordinate preferentially with a harder nucleophilic site, namely the carbonyl oxygen atom rather than with a softer P atom. We calculated theoretically the total energies of the C=O– and P- co-ordinated AlCl3 complexes of 3-methoxycarbonyl-1-methyl-2-phosphaindolizine (17, Z=P, X=AlCl3), when the total energy of the former was found smaller than that of the latter by 3.43 kcal mol−1 only. But the activation energy barrier for the DA reaction of the σ2,λ3-P-AlCl3 complexed 2-phosphaindolizine derivative (20) is lower than for the DA reaction of the C=O co-ordinated AlCl3 complex by 24.20 kcal mol−1. In our opinion, in solution, a small percentage of the C=O-AlEtCl2 complexed species changes into the σ2,λ3-P-AlCl3 complexed species which undergoes DA reaction due to a small activation barrier to afford the [2+4] cycloadduct thereby pushing the equilibrium in the right direction (Scheme 12).
Equilibrium between C=O-AlEtCl2 and =P-AlEtCl2 species accompanied by DA reaction across the C=P– bond.
Asymmetric Diels-Alder reaction of 2-phosphaindolizines
3-[Alkoxycarbonyl/acyl)-1-methyl-2-phosphaindolizines (22) reacted with DMB in the presence of a chiral catalyst, O-menthoxyaluminium dichloride, generated in situ, to afford DA adducts (25) with complete diastereoselectivity (Scheme 13) [33].
DA reaction of 3-(alkoxycarbonyl/acyl)-1-methyl-2-phosphoindolizines with DMB in the presence of O-menthoxyaluminium dichloride as promoter.
Addition of hydrogen sulfide
On reacting with hydrogen sulfide and elemental sulfur, 2-phosphaindolizines afforded zwitterionic pyridinio dithiophosphinate derivatives (26) (Scheme 14) [34]. The latter could be used as chelating ligands for the preparation of Pd(II) complexes [35].
Reaction of 2-phosphaindolizines with hydrogen sulfide and sulfur.
Co-ordination compounds
It has been possible to obtain (η1-2-phosphaindolizine)M(CO)₅ (M=Cr, Mo, W) (27) complexes by reacting 2-phosphaindolizines with M(CO)5. THF [36] (Scheme 15). In one case, (cyclooctene)Cr(CO)₅ was used as the transfer agent [15].
Preparation of η1-P co-ordination compounds of 2-phosphaindolizines.
Conclusion
A good number of 1,3-azaphospholo[1,5-a]pyridines, i.e. 2-phosphaindolizines belonging basically to two classes: those having alkoxycarbonyl groups on both sides of the σ2,λ3-P atom and other having EWG only at the 3-position, can be obtained from two simple synthetic methods. Both types of 2-phosphaindolizines incorporate several reactive functionalities that can be worked upon to access a wide variety of organophosphorus compounds. 2-Phosphaindolizines having EWG only at 3-position can be made undergo DA reaction in the presence of a chiral catalyst leading to the formation of a single product with complete diastereoselectivity.
Future perspectives
Several possibilities can be perceived for future. 1-Bromo-2-phosphaindolizines can be used for preparing Grignard’s reagent, which can be subsequently employed for introducing 1,3-azaphospholo[5,1-a]pyridine nucleus in different substrates. The products obtained from the DA reactions of 2-phosphaindolizines accomplished without the use of sulfur can furnish chiral phosphines. Thus, method will have to be developed to speed up DA reaction so that sulfur is not required. Likewise, if the catalytic moiety can be removed from the DA adducts, it can furnish chiral phosphines enantioselectively. The use of the co-ordination compounds as catalysts appear to be another possibility.
Article note
A collection of invited papers based on presentations at 22nd International Conference on Phosphorus Chemistry (ICPC-22) held in Budapest, Hungary, 8–13 July 2018.
References
[1] F. Mathey. Acc. Chem. Res.25, 90 (1992).10.1021/ar00014a006Search in Google Scholar
[2] In Multiple Bonds and Low Coordination in Phosphorus Chemistry, M. Regitz, O. J. Scherer (Eds.), Thieme, Stuttgart (1990).Search in Google Scholar
[3] R. K. Bansal, R. Gupta, P. Maheshwari, M. Kour. Curr. Org. Chem.20, 2099 (2016).10.2174/1385272820666160407154204Search in Google Scholar
[4] K. B. Dillon, F. Mathey, J. F. Nixon. Phosphorus: The Cabon Copy, Wiley, Chichester (1998).Search in Google Scholar
[5] R. Appel, F. Knoll, I. Ruppert. Angew. Chem. Int. Ed.20, 731 (1981).10.1002/anie.198107311Search in Google Scholar
[6] M. Regitz. Chem. Rev.90, 191 (1990).10.1021/cr00099a007Search in Google Scholar
[7] G. Märkl. In Multiple Bonds and Low Coordination in Phosphorus Chemistry, M. Regitz, O. J. Scherer (Eds.), pp. 220–257, Thieme, Stuttgart (1990).Search in Google Scholar
[8] A. Schmidpeter. “Heterophospholes”, in Phosphorus-Carbon Heterocyclic Chemistry: The Rise of a New Domain, F. Mathey (Ed.), pp. 363–461, Pergamon, Oxford (2001).10.1016/B978-008043952-5/50011-5Search in Google Scholar
[9] A. Schmidpeter, Heteroat. Chem.10, 529 (1999).10.1002/(SICI)1098-1071(1999)10:7<529::AID-HC2>3.0.CO;2-FSearch in Google Scholar
[10] R. K. Bansal, J. Heinicke. Chem. Rev.101, 3549 (2001).10.1021/cr000434cSearch in Google Scholar
[11] (a) W. Flitsch. In Comprehensive Heterocyclic Chemistry-I, A. R. Katritzky, C. W. Rees (Eds.), vol. 4, p. 443, Pergamon, Oxford (1984). (b) W. Flitsch. In Comprehensive Heterocyclic Chemistry-II, A. R. Katritzky, C. W. Rees, E. F. V. Scriven (Eds.), Vol. 8, p. 237, Pergamon, Oxford (1996).10.1016/B978-008096519-2.00058-8Search in Google Scholar
[12] R. K. Bansal, K. Karaghiosoff, N. Gandhi, A. Schmidpeter. Synthesis 361 (1995).10.1055/s-1995-3935Search in Google Scholar
[13] E. C. Taylor, I. J. Turchi. Chem. Rev.79, 2 (1979).Search in Google Scholar
[14] R. K. Bansal, A. Surana, N. Gupta. Tetrahedron Lett.40, 1565 (1999).10.1016/S0040-4039(98)02645-8Search in Google Scholar
[15] R. K. Bansal, N. Gupta, K. Karaghiosoff, A. Schmidpeter, C. Spindler. Chem Ber.124, 475 (1991).10.1002/cber.19911240309Search in Google Scholar
[16] R. K. Bansal, V. Kabra, N. Gupta, K. Karaghiosoff. Indian J. Chem.31, 254 (1992).Search in Google Scholar
[17] N. Gupta, C. B. Jain, J. Heinicke, N. Bharatiya, R. K. Bansal, P. G. Jones. Heteroat. Chem.9, 333 (1998).10.1002/(SICI)1098-1071(1998)9:3<333::AID-HC10>3.0.CO;2-SSearch in Google Scholar
[18] K. Karaghiosoff, R. K. Bansal, N. Gupta. Z. Naturforsh B476, 373 (1992).10.1515/znb-1992-0312Search in Google Scholar
[19] R. K. Bansal, G. Pandey, K. Karaghiosoff, A. Schmidpeter. Synthesis 173 (1995).10.1055/s-1995-3865Search in Google Scholar
[20] R. K. Bansal, N. Gandhi, K. Karaghiosoff, A. Schmidpeter. Z. Naturforsh B50b, 558 (1995).10.1515/znb-1995-0414Search in Google Scholar
[21] R. K. Bansal, C. B. Jain, N. Gupta, V. Kabra, K. Karaghiosoff, A. Schmidpeter. Phosphorus Sulfur Silicon86, 139 (1994).10.1080/10426509408018397Search in Google Scholar
[22] R. K. Bansal, R. Mahnot, D. C. Sharma, K. Karaghiosoff, A. Schmidpeter. Heteroat. Chem.3, 351 (1992).10.1002/hc.520030408Search in Google Scholar
[23] R. K. Bansal, N. Gupta, M. Baweja, L. Hemrajani, V. K. Jain. Heteroat. Chem.12, 602 (2001).10.1002/hc.1091Search in Google Scholar
[24] R. K. Bansal, L. Hemrajani, N. Gupta. Heteroat. Chem.10, 598 (1999).10.1002/(SICI)1098-1071(1999)10:7<598::AID-HC13>3.0.CO;2-HSearch in Google Scholar
[25] D. Singh, P. Sinha, N. Gupta, R. K. Bansal. Phosphorus Sulfur Silicon Relat. Elem.191, 488 (2016).10.1080/10426507.2015.1091834Search in Google Scholar
[26] R. K. Bansal, N. Gupta, V. Kabra, C. Spindler, K. Karaghiosoff, A. Schmidpeter. Heteroat. Chem.3, 359 (1992).10.1002/hc.520030409Search in Google Scholar
[27] R. K. Bansal, N. Gupta, S. K. Kumawat, G. Dixit. Tetrahedron64, 6395 (2008).10.1016/j.tet.2008.04.080Search in Google Scholar
[28] R. K. Bansal, V. K. Jain, N. Gupta, N. Gupta, L. Hemrajani, M. Baweja, P. G. Jones. Tetrahedran58, 1573 (2002).10.1016/S0040-4020(02)00004-2Search in Google Scholar
[29] R. K. Bansal, N. Gupta, G. Dixit, S. K. Kumawat. J. Phys. Org. Chem.22, 125 (2009).10.1002/poc.1436Search in Google Scholar
[30] N. Gupta, R. K. Jangid, R. K. Bansal, M. V. Hopffgarten. Current Cat.1, 93 (2012).10.2174/2211544711201020093Search in Google Scholar
[31] R. K. Jangid, N. Gupta, R. K. Bansal, M. V. Hopffgasten, G. Frenking. Tetrahedron Lett.52, 1721 (2011).10.1016/j.tetlet.2011.02.007Search in Google Scholar
[32] J. P. Alcaraz, F. Mathey. Tetrahedron Lett.25, 207 (1984).10.1016/S0040-4039(00)99841-1Search in Google Scholar
[33] R. K. Jangid, N. Sogani, N. Gupta, R. K. Bansal, M. V. Hopffgasten, G. Frenking. Beilstein J. Org. Chem.9, 392 (2013).10.3762/bjoc.9.40Search in Google Scholar
[34] R. K. Bansal, N. Gupta, N. Bharatiya, G. Gupta, A. Surana, G. Hackenbracht, K. Karaghiosoff. Heteroat. Chem.9, 445 (1998).10.1002/(SICI)1098-1071(1998)9:4<445::AID-HC14>3.0.CO;2-RSearch in Google Scholar
[35] V. K. Jain, L. Hemrajani, R. K. Bansal. Indian J. Chem.41A, 550 (2002).Search in Google Scholar
[36] N. Gupta, C. B. Jain, J. Heinicke, R. K. Bansal, P. G. Jones. Eur. J. Inorg. Chem. 1079 (1998).10.1002/(SICI)1099-0682(199808)1998:8<1079::AID-EJIC1079>3.0.CO;2-USearch in Google Scholar
©2018 IUPAC & De Gruyter. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. For more information, please visit: http://creativecommons.org/licenses/by-nc-nd/4.0/
