Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Aug 15.
Published in final edited form as: J Org Chem. 2008 Jul 19;73(16):6137–6141. doi: 10.1021/jo800776z

Co-complexes Derived from Alkene Insertion to Alkyne-dicobaltpentacarbonyl complexes: Insight into the Regioselectivity of Pauson-Khand Reactions of Cyclopropenes

Mahesh K Pallerla 1, Glenn P A Yap 1, Joseph M Fox 1,*
PMCID: PMC2667904  NIHMSID: NIHMS105957  PMID: 18637694

Abstract

graphic file with name nihms-105957-f0001.jpg

Described are the X-ray crystallographic and spectral properties of Co-complexes that were isolated from two Pauson-Khand reactions of chiral cyclopropenes. These are the first examples of isolated Co-complexes derived from the putative alkene-insertion intermediates of Pauson-Khand reactions. The binuclear Co-complexes are coordinated to μ-bonded, five-carbon “flyover” carbene ligands. It is proposed that the complexes result from cyclopropane fragmentation subsequent to alkene insertion. The observation of these metal complexes provides a rationale for the origin of regioselectivity in Pauson-Khand reactions of cyclopropenes.

Introduction

The Pauson-Khand reaction is a formal [2+2+1] cyclocarbonylation reaction that results in the formation of cyclopentenone products.1,2 This transformation has emerged as a powerful tool for stereoselective syntheses because of the significant molecular complexity that is generated from simple starting materials.1 Cyclopropenes are excellent substrates for Pauson-Khand reactions.3,4 Recently, our group described general conditions for the utilization of chiral cyclopropenes in intermolecular Pauson-Khand reactions,4b and the enantioselective synthesis of pentalenene via an intramolecular cyclopropene Pauson-Khand reaction.4a In intermolecular Pauson-Khand reactions, 1,2-disubstituted cycloprop-1-enes are particularly efficient substrates. Such reactions proceed with excellent regio- and diastereoselectivity to produce cyclopentenones in good yield (57-82%). In many cases, red cobalt-containing side products were observed in the Pauson-Khand reactions of cyclopropenes. Described herein are the spectroscopic and crystallographic properties of two cobalt complexes that were isolated. Mechanistic implications of the formation of these complexes are also discussed.

According to the accepted mechanism,5 the first intermediate in the Pauson-Khand reaction is an alkyne-Co2(CO)5 complex that results from the loss of CO ligand from an alkyne-Co2(CO)6 complex. Structural evidence for this proposal first came from Krafft and coworkers, who isolated 2 as an intermediate to cyclopentenone 3 during studies on the directed intramolecular Pauson-Khand reaction (Scheme 1).6 A similar complex (4) was subsequently isolated by Perciás and coworkers while exploring Pauson-Khand reactions that employed tethered chiral auxiliaries.7 Later, alkyne•Co2(CO)5 complexes were characterized by IR in studies by Gordon and coworkers.8

Scheme 1.

Scheme 1

Evidence for alkyne•Co2(CO)5 complexes in Pauson-Khand reactions.

Several studies have also supported the proposal that alkene complexes of alkyne•Co2(CO)5 are intermediates in Pauson-Khand reactions. In an intermolecular Pauson-Khand reaction, Gimbert and coworkers used ESI-MS to monitor an intermolecular Pauson-Khand reaction, and detected a peak (781.1 Da) cooresponding to the alkyne-Co complex 5 + norbornene.9 More recently, Evans and McGlinchey reported the X-ray crystal structure of 6, an alkyne-pentacarbonyl complex with an alkene coordinated in intramolecular fashion.10 Although 6 is an alkene complex, structural constraints do not permit alkene insertion into the Co-alkyne complex.

Alkene insertion has been reported to be the rate determining step in Pauson-Khand reactions that have been studied mechanistically.1,5,11 Thus, it has been difficult to obtain information about the intermediates that form subsequent to alkene insertion. Herein, we describe the first examples of Co-complexes derived from the putative alkene-insertion intermediate of a Pauson-Khand reaction. The complexes are binuclear, and they are coordinated by μ-bonded, five-carbon “flyover” carbene ligands.12 It is postulated that these complexes result from fragmentation of the cyclopropane subsequent to the alkene insertion step. These observations provide insight into the origin of regioselectivity and the limitations of catalysis for the intermolecular Pauson-Khand reactions of cyclopropenes. Furthermore, these complexes provide indirect evidence for the alkene insertion intermediates that have been invoked in the Magnus-Schore mechanism5 for the Pauson-Khand reaction.

Results and Discussion

In the Pauson-Khand reaction of 7 with 8, red cobalt complex 10 accompanied formation of cyclopentanone 9 (Scheme 3). The conditions in Scheme 3 led to maximum yields of the red cobalt complex, whereas conditions previously reported (BuSMe, 100 °C) maximized the yield of the cyclopentenone.4b The red cobalt complex could be purified on a column of silica gel and isolated along with the cyclic enone product. While crude TLC analysis indicated that the red complex was the major product, only 13% of 10 could be isolated because of partial decomposition upon silica gel chromatography.

Scheme 3.

Scheme 3

Isolation of a Co-complex in an intermolecular Pauson-Khand reaction.

It was possible to obtain X-ray quality crystals for 10. A molecular diagram of the X-ray structure of 10 and a table of selected bond distances is displayed in Figure 1. The binuclear complex is coordinated by a μ-bonded, five-carbon ligand. “Flyover” 2,4-pentadienylidene ligands of this type have been described previously for dicobalt complexes, but they were prepared by unrelated routes.12 Structure 10 contains two ligands that bridge both cobalt atoms: a bridging carbonyl (C8) and the bridging “flyover” carbene (C3) of the five carbon ligand (C3 to C7). Thus, C3 is both σ-bonded to Co1 and also a part of the allylic bond to Co2. Bonding of Co2 to C4 and C5 completes the allylic bond. Co1 is π-bonded to C6 and C7.

Figure 1.

Figure 1

X-ray crystal structure and selected properties of 10.

The IR spectrum displays four peaks that are attributed to the five carbonyls of 10. Peaks at 2067, 2038, and 2008 cm-1 are attributed to the terminal carbonyl ligands: the peak at 2008 cm -1 is roughly twice the intensity of the other carbonyls and is attributed to two overlapping carbonyl peaks. A peak at 1853 cm-1 was attributed to the bridging carbonyl. The frequency of the terminal and bridging carbonyl peaks correlates to those of related dinuclear cobalt complexes in the literature.12a,13 The 13C NMR spectrum of 10 at 25 °C contains three peaks at 197 ppm, 202 ppm, and 212 ppm that integrate with a ratio of 1:1:3, respectively (Figure 2). These peaks are associated with the carbonyls attached to the cobalt atoms. However, at -25 °C the peak at 212 ppm is not observed. Two new peaks at 230 ppm and 206 ppm appear when the temperature is lowered to -60 °C. The new peak at 230 ppm is assigned to the bridging carbonyl.13 For the spectrum acquired at -60 °C, the peaks at 230 ppm, 206 ppm, 203 ppm, and 198 ppm integrate with a ratio of 1:2:1:1, respectively. Our interpretation is that the exchange of the bridging and the terminal carbonyls is slow at -60 °C, but exchange between the bridging carbonyl and two of the terminal carbonyls is fast on the NMR time scale at 25 °C. Coalescence is observed at -25 °C in the 13C NMR of 10. Analogous observations have been made by Haenel and coworkers for a hexacarbonyldicobalt complex coordinated by 4,6-bis(diphenylphosphino)dibenzofuran.13 The Haenel complex contains two, non-equivalent bridging carbonyls, and 13C NMR studies on this complex demonstrated that carbonyl exchange is rapid at rt but slow at -80 °C, and that coalescence is observed at -30 °C.13

Figure 2.

Figure 2

Variable temperature 13C NMR spectra of 10 (100 MHz, CDCl3). Because overnight acquisition times were necessary and because 10 is only moderately stable over prolonged periods in solution, freshly prepared solutions of 10 were prepared for spectral acquisition at different temperatures. Thus, the signal-to-noise ratios differ for the three spectra displayed.

graphic file with name nihms-105957-f0002.jpg (1)

It was possible to utilize substoichiometric amounts of Co2(CO)8 in some of the Pauson-Khand reactions while using tetramethyl thiourea (TMTU)14d as the promotor of cyclopentenone (64%) formation (Eq. 1).15 However, in no case was it possible to use less than 0.5 equiv of Co2(CO)8. Red side products were generally observed in the Pauson-Khand reactions of cyclopropenes. It was speculated that the formation of these side products represented a catalytically non-productive pathway by which cobalt was consumed. To support this hypothesis, 10 was treated with NMO14a,b under conditions that promote Pauson-Khand reactions of cyclopropenes:4b 10 remained unaltered and cyclopentenone 11 did not form (Scheme 4). Similarly, complex 10 did not return the alkyne complex 8 when treated with trimethylsilylacetylene under an atmosphere of CO. These experiments suggest that the formation of 10 takes cobalt out of the catalytic cycle, and provide one reason for the high catalyst loadings that are needed.

Scheme 4.

Scheme 4

Formation of cobalt complex 10 is unproductive for catalysis.

Intermolecular Pauson-Khand reactions of cyclopropenes proceed with high regioselectivity to produce cyclopentenone products.4b Intriguingly, the complex 10 has the opposite regioselectivity relative to the cyclopentenone product 9 (Scheme 5). It seems that the high regioselectivity in cyclopropene Pauson-Khand reactions is only partly due to the selectivity in alkene insertion. The regioselectivity is further accentuated by a kinetic discrimination after alkene insertion. Thus, the putative insertion product 13 leads to 10, and does not produce cyclopentenone 11. In contrast, cyclopentenone 9 arises from 12 and/or diastereomeric 12'. It is intriguing that the putative intermediate 13 leads to reductive ring-opening, whereas the regioisomeric structures 12 lead to Pauson-Khand products. Ring-opened product 10 contains a carbon-metal bond that is α-to the trimethylsilyl group that was originally bound to the cyclopropane moiety of 13. Silicon is known to stabilize α-carbon-metal bonds.16 This stability may be reflected in a relatively low barrier for the transformation of 13 into 10, and provide a plausible explanation for the susceptibility of 13 to undergo reductive cleavage.

Scheme 5.

Scheme 5

The origin of regiocontrol for intermolecular Pauson-Khand reactions.

The NMO-promoted reaction of phenylacetylene complex 14 with 15 gave Co-complexes 17 (15%) and 18 (41%) along with cyclopentenone 16 (19%) (Scheme 6). The conditions in Scheme 6 led to maximum yields of the red cobalt complexes, whereas conditions previously reported (BuSMe, 100 °C) maximized the yield of the cyclopentenone.4b Compound 18 was characterized by X-ray crystallography and was found to be structurally similar to 10 (Figure 3). The spectral properties of compound 17 were roughly similar to 18, including the observation of bridging carbonyl (1866 cm-1) in the IR spectrum. However, it was not possible to unambiguously characterize 17 because it was limited in quantity. The spectral properties of cobalt complex 17 may be consistent with a structure that is regioisomeric with 18.

Scheme 6.

Scheme 6

Two cobalt complexes formed as side products along with cyclopentenone 16.

Figure 3.

Figure 3

X-ray crystal structure and selected properties of 18

In summary, red cobalt complexes were isolated as side products in the intermolecular Pauson-Khand reactions of cyclopropenes. The complexes were characterized by X-ray crystallography, 13C NMR and IR spectroscopy, and they each possess a bridging carbonyl and a five-carbon, μ-bonded “flyover” carbene ligand. These structures result from fragmentation of the cyclopropane ring after alkene insertion. A conclusion of the study is that the high regioselectivity of intermolecular cyclopropene Pauson-Khand reactions is not solely a consequence of selective alkene insertion. Rather, the regioselectivity is augmented by a kinetic discrimination after alkene insertion, in which only the minor diastereomer leads to the ring-opened side product. As the first characterized examples of Co-complexes derived from putative alkene insertion intermediates, these complexes provide indirect evidence for the Magnus-Schore mechanism for the Pauson-Khand reaction.

Experimental Section

The experimental procedures for the preparation of alkyne-dicobalt hexacarbonyl complexes 8 and 14 are as described in literature.4b The Pauson-Khand reaction conditions that maximize the formation of red cobalt complexes are described below. The spectral data for 9 and 16 were previously reported.4b

General Procedure for the formation of red cobalt complexes in intermolecular Pauson-Khand reactions

A dry 50-mL round bottomed flask was outfitted with a stir bar, gas inlet adapter and a rubber septum. The assembly was flame dried under vacuum and cooled under nitrogen. The alkyne-cobalt complex (2.0 equiv) and the cyclopropene (1.0 equiv) were then added, and the apparatus was evacuated and filled with nitrogen. CH2Cl2 was added and the mixture was cooled by an ice-bath (0 °C). NMO (5.0 equiv) was added and the reaction mixture was allowed to stir at 0 °C for 5 min. The cold bath was then removed and the mixture was allowed to stir while warming to rt for 45 min. Volatile solvents were removed under reduced pressure and the residue was chromatographed on silica gel.

Pentacarbonyl-μ-[1,3-bis(trimethylsilyl)-5-ethoxycarbonyl-4-methyl-cis-2,4-pentadienylidene-1(1,2,3η3):2(4,5η2)-1κC1:2κC1]-dicobalt (Co-Co) (10)

The general procedure was followed using 7 (255 mg, 1.29 mmol), 8 (990 mg, 2.58 mmol), NMO (1.51 g, 12.9 mmol), and CH2Cl2 (25 mL). Chromatography on silica gel (2% CH2Cl2 in hexanes) provided 81 mg (0.25 mmol, 19%) of 94b and 95 mg (0.17 mmol, 13%) of 10 as a red solid. The red cobalt complex 10 crystallized from hexanes, mp 118 °C; 1H NMR (CDCl3, 400 MHz, 298 K)δ 4.63 (s, 1H), 4.19-4.14 (m, 1H), 4.01-3.96 (m, 1H), 2.52 (m, 4H), 1.23 (t, J=8.0 Hz, 3H), 0.34 (s, 9H), 0.29 (s, 9H); 13C NMR (CDCl3, 100 MHz) δ 212.5, 202.8, 197.2, 173.7, 159.2, 110.9, 106.4, 92.0, 59.9, 49.1, 24.5, 14.2, 0.2, -0.4. IR (CHCl3, cm-1): 2066, 2038, 2008, 1852, 1691, 1215; HRMS-EI m/z: [M+Na], calcd for C20H28Co2O7Si2Na, 576.9935, found 576.9940. Anal. Calcd. For C20H28Co2O7Si2: C, 43.32; H, 5.09. Found: C, 43.62; H, 5.03.

Cobalt complex 17 and Pentacarbonyl-μ-{4-methyl-1-phenyl-5-[4(S)-4-phenyloxazolidinatocarbonyl]-3-trimethylsilyl-cis-2,4-pentadienylidene-1(1,2,3η3):2(4,5η2)-1κC1:2κC1}-dicobalt (Co-Co) (18)

The general procedure was followed using 15 (150 mg, 0.39 mmol), 14 (300 mg, 0.78 mmol), NMO (228 mg, 1.95 mmol), and CH2Cl2 (10 mL). Chromatography on silica gel (10% ethyl acetate in hexanes) provided 45 mg (0.06 mmol, 15%) of 17 as a red solid, 120 mg (0.161 mmol, 41%) of 18 as a red solid and 39 mg (0.76 mmol, 19%) of 164b as a colorless semi-solid. Compound 18 crystallized from hexanes. The structure of 17 could not be definitively assigned; spectral properties of 17 are listed in the Supporting Information.

Spectral properties of 18: mp 70-73 °C; 1H NMR (CDCl3, 400 MHz)δ 7.50-7.26 (m, 10H), 5.43 (m, 1H), 4.78 (br s, 1H), 4.60 (m, 1H), 4.28 (br s, 1H), 4.17 (m, 1H), 2.67 (m, 1H), 2.43 (m, 1H), 1.51-0.81 (m, 11H), 0.33 (s, 9H); 13C NMR (CDCl3, 100 MHz) δ 212.3 (br), 201.6, 196.7, 172.0, 163.0, 153.5, 148.2, 139.7, 129.0, 128.4 (2 peaks), 128.3, 127.4, 125.6, 107.6, 101.5, 99.9, 69.4, 58.0, 49.8, 37.9, 32.1, 31.4, 29.4, 22.4, 14.1, 0.3; IR (CHCl3, cm-1): 2958, 2070, 2043, 2009, 1855, 1776, 1665, 1443,1383, 1213; HRMS-EI m/z: [M+Na], calcd for C35H37Co2NO8SiNa, 768.0850, found 768.0842. Anal. Calcd. For C35H37Co2NO8Si: C, 56.38; H, 5.00; N, 1.88. Found: C, 56.22; H, 5.09; N, 1.85.

Supplementary Material

1_si_001
2_si_002
3_si_003

Scheme 2.

Scheme 2

Evidence for alkene coordination to alkyne•Co2(CO)5 complexes.

Acknowledgment

This work was supported by NIH grant GM068640-01.

Footnotes

Supporting Information Available: Full experimental details are provided, as are characterization details for all compounds, copies of 1H NMR, 13C NMR and IR spectra for new compounds, and CIF files for compounds 10 and 18. This material is available free of charge via the Internet at http://pubs.acs.org.

References

  • (1).Selected reviews: Brummond KM, Kent JL. Tetrahedron. 2000;56:3263–3283.. Gibson SE, Mainolfi N. Angew. Chem., Int. Ed. 2005;44:3022–3037. doi: 10.1002/anie.200462235.. Gibson SE, Stevenazzi A. Angew. Chem. Int. Ed. 2003;42:1800–1810. doi: 10.1002/anie.200200547.. Alcaide B, Almendros P. Eur. J. Org. Chem. 2004:3377–3383.. Ingate ST, Marco-Contelles J. Org. Prep. Proced. Int. 1998;30:121–143.. Shibata T. Adv. Synth. Catal. 2006;348:2328–2336.. Laschat S, Becheanu A, Bell T, Baro A. Synlett. 2005:2547–2470.. Boñaga LVR, Krafft ME. Tetrahedron. 2004;60:9795–9833..
  • (2).(a) Khand IU, Knox GR, Pauson PL, Watts WE. J. Chem. Soc.,Chem. Commun. 1971:36. [Google Scholar]; (b) Khand IU, Knox GR, Pauson PL, Watts WE, Foreman MI. J. Chem. Soc., Perkin Trans. 1973;1:977–981. [Google Scholar]
  • (3).(a) Kireev SL, Smit VA, Ugrak BI, Nefedov OM. Bull. Acad. Sci. USSR (Engl. Transl.) 1991;40:2240–2246. [Google Scholar]; (b) Witulski B, Göβmann M. Synlett. 2000:1793–1797. [Google Scholar]; (c) Nüske H, Bräse S, de Meijere A. Synlett. 2000:1467–1469. [Google Scholar]; (d) Marchueta I, Verdaguer X, Moyano A, Pericás MA, Riera A. Org. Lett. 2001;3:3193–3196. doi: 10.1021/ol016505r. [DOI] [PubMed] [Google Scholar]
  • (4).(a) Pallerla MK, Fox JM. Org. Lett. 2007;9:5625–5628. doi: 10.1021/ol702597y. [DOI] [PubMed] [Google Scholar]; (b) Pallerla MK, Fox JM. Org. Lett. 2005;7:3593–3595. doi: 10.1021/ol051456u. [DOI] [PubMed] [Google Scholar]
  • (5).(a) Magnus P, Principe LM. Tetrahedron Lett. 1985;26:4851–4854. [Google Scholar]; (b) Magnus P, Exon C, Albaugh-Robertson P. Tetrahedron. 1985;41:5861–5869. [Google Scholar]; (c) Magnus P, Principe LM, Slater MJ. J. Org. Chem. 1987;52:1483–1486. [Google Scholar]; (d) La Belle BE, Knudsen MJ, Olmstead MM, Hope H, Yanuck MD, Schore NE. J. Org. Chem. 1985;50:5215–5222. [Google Scholar]
  • (6).Krafft ME, Scott IL, Romero RH, Feibelmann S, Van Pelt CE. J. Am. Chem. Soc. 1993;115:7199–7207. [Google Scholar]
  • (7).(a) Verdaguer X, Moyano A, Pericás MA, Riera A, Bernardes V, Greene AE, Alvarez-Larena A, Piniella JF. J. Am. Chem. Soc. 1994;116:2153–2154. [Google Scholar]; (b) Marchueta I, Montenegro E, Panov D, Poch M, Verdaguer X, Moyano A, Pericás MA, Riera A. J. Org. Chem. 2001;66:6400–6409. doi: 10.1021/jo015790p. [DOI] [PubMed] [Google Scholar]
  • (8).Gordon CM, Kiszka M, Dunkin IR, Kerr WJ, Scott JS, Gebicki J. J. Organomet. Chem. 1998;554:147–154. [Google Scholar]
  • (9).Gimbert Y, Lesage D, Milet A, Fournier F, Greene AE, Tabet J-C. Org. Lett. 2003;5:4073–4075. doi: 10.1021/ol035503z. [DOI] [PubMed] [Google Scholar]
  • (10).Banide EV, Müller-Bunz H, Manning AR, Evans P, McGlinchey MJ. Angew. Chem. Int. Ed. 2007;46:2907–2910. doi: 10.1002/anie.200605171. [DOI] [PubMed] [Google Scholar]
  • (11).Computational studies on the Pauson-Khand reaction: Yamanaka M, Nakamura E. J. Am. Chem. Soc. 2001;123:1703–1708. doi: 10.1021/ja005565+.; Robert F, Milet A, Gimbert Y, Konya D, Greene AE. J. Am. Chem. Soc. 2001;123:5396–5400. doi: 10.1021/ja004170n.; de Bruin TJM, Milet A, Robert F, Gimbert Y, Greene AE. J. Am. Chem. Soc. 2001;123:7184–7185. doi: 10.1021/ja015781y.; de Bruin TJM, Milet A, Greene AE, Gimbert Y. J. Org. Chem. 2004;69:1075–1080. doi: 10.1021/jo035294w.. de Bruin TJM, Michel C, Vekey K, Greene AE, Gimbert Y, Milet A. J. Organomet. Chem. 2006;691:4281–4288.. Kinetic studies on the Pauson-Khand reaction: Cabot R, Lledó A, Revés M, Riera A, Verdaguer X. Organometallics. 2007;26:1134–1142..
  • (12).Closest crystallographically characterized precedents of 10 and 18: Váradi G, Horváth IT, Pályi G, Markó L, Slovokhotov YL, Struchkov YT. J. Organomet. Chem. 1981;206:119–130.. King JA, Jr., Vollhardt KPC. J. Am. Chem. Soc. 1983;105:4846–4848.. King JA, Jr., Vollhardt KPC. J. Organomet. Chem. 1994;470:207–222.. Hong P, Aoki K, Yamazaki H. J. Organomet. Chem. 1978;150:279–293..
  • (13).Vogl EM, Bruckmann J, Krüger C, Haenel MW. J. Organomet. Chem. 1996;520:249–252. [Google Scholar]
  • (14).(a) Shambayati S, Crowe WE, Schreiber SL. Tetrahedron Lett. 1990;31:5289–5292. [Google Scholar]; b) Jeong N, Chung YK, Lee BY, Lee SH, Yoo S-E. Synlett. 1991:204–206. [Google Scholar]; c) Sugihara T, Yamada M, Yamaguchi M, Nishizawa M. Synlett. 1999:771–773. [Google Scholar]; d) Tang Y, Deng L, Zhang Y, Dong G, Chen J, Yang Z. Org. Lett. 2005;7:593–595. doi: 10.1021/ol047651a. [DOI] [PubMed] [Google Scholar]
  • (15).It was also possible to utilized substoichiometric Co2(CO)8 in the TMTU promoted intramolecular Pauson-Khand reaction of a cyclopropen-yne. See ref 4a.
  • (16).Paquette LA. Science. 1982;217:793. doi: 10.1126/science.217.4562.793. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1_si_001
2_si_002
3_si_003

RESOURCES