Skip to main content
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2007 Feb 22;104(9):3026–3030. doi: 10.1073/pnas.0611309104

Effect of carbonates/phosphates as nucleophilic catalysts in dimethylformamide for efficient cyanosilylation of aldehydes and ketones

G K Surya Prakash 1, Habiba Vaghoo 1, Chiradeep Panja 1, Vijayalakshmi Surampudi 1, Roman Kultyshev 1, Thomas Mathew 1, George A Olah 1,
PMCID: PMC1805593  PMID: 17360603

Abstract

Cyanosilylation of aldehydes and aliphatic ketones can be carried out in dimethylformamide even without the use of any catalyst. In the presence of nucleophilic catalysts such as carbonate and phosphate salts, the reaction rate is significantly enhanced.

Keywords: nature of solvent


Cyanohydrins (1) serve as key intermediates in the synthesis of biologically important compounds such as β-amino alcohols, α-hydroxy acids, α-hydroxy ketones, and α-amino acids. The possibilities to obtain a wide range of 1,2-bifunctional compounds by using both the hydroxyl and nitrile groups enhance their use as versatile building blocks in synthetic organic chemistry. Hydrogen cyanide (HCN) is the most commonly industrially used reagent for cyano transfer to carbonyl compounds (2). However, due to its toxicity and difficulty in handling, new methods have been developed to substitute HCN with other potentially less harmful and yet easily manageable reagents.

Trimethylsilyl cyanide (TMSCN) is widely used as a cyanide source with various catalysts. The use of TMSCN dates back to the early 1970s. In 1973, Evans and Truesdale (3, 4) were among the first to report the use of anionic catalysts with TMSCN. Since then, a multitude of different catalysts has been reported in the literature for both the racemic and asymmetric addition of the cyanide to carbonyls. Majority of these catalysts are based on metallic Lewis acidic systems (522) containing a variety of ligands that enable enantioselective transfer of CN to carbonyls. Izumi et al. (23) have reported cyanosilylation of carbonyl compounds with TMSCN using inorganic solid acids (varioius ion exchanged montmorillonites) and bases (basic solids such as CaF2, CaO, MgO, etc.) as catalysts. Kagan and coworkers (24, 25) have reported the use of mono- and dilithium salts of binol. Recently, Ishiahara and coworkers (26) have modified this system by introducing a water/alcohol mixture as a coactivator. There has been also an increased interest in nonmetallic catalytic systems. Deng (27, 28), Plummet (29, 30), Corey (31), Jacobsen (32), and Feng (33) have reported systems based on chincona alkaloids, phosphonium salts, oxazaborolidinium ions, modified thiourea, and chiral amino acids, respectively, as good catalysts for cyano transfer from TMSCN. With the exception of the phosphonium salts, the rest are chiral catalysts and afford the products in high enantioselectivities. Recently, Denmark and Chung (34) conducted a brief survey of effective solvents, catalysts, and kinetics of Lewis base catalyzed addition of TMSCN to aldehydes. We now report our study of these reactions in dimethylformamide (DMF) using nucleophilic catalysts such as carbonates and phosphates. The convenient and inexpensive reaction conditions do not require any air and moisture free environment. We found that the CN to carbonyl transfer in N,N-dimethylformamide can be carried out even in the absence of a catalyst. With the addition of K2CO3 or organic phosphate as catalyst, the rate of the reaction has been significantly enhanced. Herein, we describe our results on cyanosilylation using K2CO3 or organic phosphate in DMF.

Results and Discussion

Use of DMF as a solvent for the cyanosilylation reaction using a metallic catalyst was reported in 1978 by Rasmussen and Heilmann (35) while in situ preparing TMSCN using KCN under reflux conditions. Our present studies have shown that DMF is not only a more suitable solvent, but itself can catalyze the addition of TMSCN to aldehydes, whereas for ketones, the use of a nucleophilic catalyst is necessary for faster reaction and complete conversion.

Because silicon is highly oxophilic, any solvent that can coordinate to silicon in TMSCN was expected to facilitate CN transfer. Therefore, we explored the use of suitable oxygenated solvent and found DMF as the preferred solvent. Furthermore, as DMF increases the dielectric constant of the medium and thus stabilizes the charged transition state, it also facilitates the reactions. For comparison a series of reactions has been performed by using different solvents and solvent systems. From the result of these studies (Table 1), DMF has been identified as the most suitable solvent for the cyanosilylation of carbonyl compounds (Scheme 1).

Table 1.

Screening of solvents for the cyanosilylation reaction of 9-anthraldehyde using K2CO3 and TMSCN

graphic file with name zpq00507511300t1.jpg
Solvents Time Conversion, % Yield, %
DMF 5 min 100 95
THF 48 h 60
THF + DMF (4:1) 6 h 100 90
CH2Cl2 48 h 47
CH2Cl2 + DMF (4:1) 48 h 10

Catalyst, 1 mol%; TMSCN, 1.2 equiv.

Scheme 1.

Scheme 1.

Cyanosilylation of aldehydes and ketones in DMF.

Although Izumi et al. (23) reported the use of some carbonates and phosphates (but not potassium carbonate or organic phosphates) in the solid base catalyzed cyanosilylation, our present study indicates that cyanosilylation of carbonyl compounds in DMF is efficiently catalyzed by potassium carbonate or easily synthesized organic phosphates such as (Bu)4N+(MeO)2P(O)O. Ketones are generally less reactive in cyanosilylation. Indeed we found that these reactions are rather slow without a suitable catalyst. However, cyanosilylation of aldehydes occurs efficiently without any catalyst when DMF is the solvent (because DMF itself acts as a nucleophilic activator). However, ketones also undergo reaction in DMF without the catalyst but require longer reaction times, which can be reduced considerably by the presence of catalyst. Use of a readily prepared metal-free organic phosphate catalyst allows to avoid metal based initiators in reactions where metal coordination with the substrates hinder the progress of the cyanosilylation. All aldehydes we studied gave high yields in 5–20 min (except 1d) with or without catalyst (Table 2).

Table 2.

Cyanosilylation of aldehydes in DMF

Entry Aldehydes (1a-g) Time, min Product (2a-g) Yields, %
No catalyst K2CO3
graphic file with name zpq00507511300t2.jpg

*At low temperature (−50 to 0°C).

Cyanosilylation under solvent-free conditions using K2CO3 catalyst has been reported. However, this was restricted to reactions with liquid substrates which give liquid product (11, 36, 37). In Kurono's work, THF has been used as a solvent with LiCl catalyst (11). In a practical sense these “solvent free” methods however require solvent at least in one of their work up stages after the reaction. These reactions are also slow and required large amount of salt to catalyze the reactions of ketones. Reaction of acetophenone using K2CO3 under solvent free conditions required 30 mol% of K2CO3 and 24-h reaction time to give 91% product. When we have carried out the reaction using K2CO3 in DMF the reaction is complete in 30 min with an isolated yield of 98% (Table 3, entry g). In the case of all ketones studied, the reactions are fast (reaction time, 5 min to 2 h) except in the case of 3e (Table 3).

Table 3.

Cyanosilylation of ketones with TMSCN in DMF using K2CO3 as catalyst

Entry Ketone (3a-k) Time, min Product (4a-k) Yield, %
graphic file with name zpq00507511300t3.jpg

Catalyst, 5 mol%.

*After 6.0 h, 0.5 eqv excess TMSCN was added to complete the reaction.

Song et al. (38) used N-heterocyclic carbenes (NHCs) as efficient initiators for cyanosilylation. Because NHCs such as 1,3-bis(tert-butyl)imidazol-2-ylidene or 1,3-bis(cyclohexyl)imidazol-2-ylidene are costly, easily available K2CO3 and DMF are used preferentially in our cyanosilylations. Recently, Feng et al. (39) reported an efficient solvent-free cyanosilylation using tetramethylaminoguanidine, but it requires prolonged reaction times (4–15 h).

Our quest to find other salts that can efficiently catalyze cyanosilylation led us to investigate a number of nitrates, sulfates, and phosphates. Results showed that phosphate salts were able to catalyze the cyanosilylation very efficiently. There have been some reports in the literature in which phosphate moiety is a part of metal-based Lewis base system (40) or a Brønsted acid system (41). To our knowledge, use of organic phosphates alone as catalysts in cyanosilylation has not been reported in the literature. An initial screening of the catalytic effect of different phosphates revealed 100% conversion (3a) in most cases in a reasonable amount of time with 5 mol% catalyst loading. In addition to the common inorganic phosphate salts, we also studied organic phosphate salt (Bu)4N+(MeO)2P(O)O (Table 4, entry 3) because of its high solubility in organic solvents, thus reaping the benefits of homogeneous catalysis. Reactions of various ketones catalyzed by this organic phosphate in DMF afforded the cyanosilylated products in excellent yields (Table 5).

Table 4.

Cyanosilylation of benzophenone using phosphate salts as catalyst

graphic file with name zpq00507511300t4.jpg
Entry Phosphate salt TMSCN, equiv. Conversion, %
1 K3PO4 2 100
2 K2HPO4·3H2O 2 100
3 graphic file with name zpq0050751130t4b.jpg 1.5 100
4 KH2PO4 2 90

Reaction time, 1 h; catalyst, 5 mol%.

Table 5.

Cyanosilylation of ketones with TMSCN in DMF using (Bu)4N+ (MeO)2P(O)O as catalyst

Entry Ketone (3a-k) Time, min Product (4a-k) Yield, %
graphic file with name zpq00507511300t5.jpg

Catalyst, 5 mol%.

*After 6.0 h, 0.5 eqv excess TMSCN was added to complete the reaction.

Two possible pathways may be considered for the mechanism for cyanosilylation of carbonyl compounds using the above mentioned catalysts: an autocatalytic (initiated) and a catalytic pathway. The autocatalytic pathway can be initiated by either a neutral (AX) or an ionic (BY) species. In the case of initiation by neutral species, DMF or the substrate itself can activate TMSCN by coordinating to silicon. Thus, the Si-CN bond is weakened and polarized to give rise to a zwitter ionic intermediate 1, which can then attack the electrophilic carbonyl, affording another zwitter ionic alkoxide species 2. Species 2 can attack a new TMSCN molecule, giving the cyanosilylated product. The cycle is repeated until all of the carbonyl substrate is consumed. In the case of initiation by ionic species, the anion of an ion pair can attack the TMSCN forming a trigonal bipyramidal intermediate 3, from which the cyanide ion is transferred to the carbonyl compound to give rise to an alkoxide ion that can continue the autocatalytic cycle till all of the starting material is consumed (Scheme 2).

Scheme 2.

Scheme 2.

Autocatalytic pathway initiated by neutral (a) and ionic (b) species.

In the case of carbonate catalyzed reaction, two pathways can be considered (Scheme 3). In the first (pathway a), the bidentate carbonate catalyst can interact with the TMSCN to give rise to the pentacoordinate species i. Then the silicon atom can extend its coordination sphere to give rise to the hexacoordinate intermediate ii. This type of valency expansion is well known in silicon chemistry (42). The cyanide can then add to the carbonyl compound, and the resulting alkoxide ion can coordinate to the silicon atom of the coordination sphere. Again a penta- or hexacoordinate intermediate (iii or iv, respectively) is possible, from which the TMS protected cyanohydrin is released with the regeneration of the catalyst. Another possibility involves the simultaneous coordination of two anionic sites of the catalyst with two TMSCN moities to give intermediate vi in the (a+b) pathway. As mentioned, cyanide is then transferred to another molecule of the carbonyl compound, and the resulting alkoxide replaces the cyanide by coordinating to the silicon atom. Finally, the TMS-protected cyanohydrin is released with the regeneration of the catalyst doubling the catalytic effciency. In our recently reported trifluoromethylsilylation studies, calculations at the B3LYP/6–311+G* level showed that intermediates similar to ii (hexacooordinate intermediate) and vi (pentacoordinate intetmediate) are stable minima (43).

Scheme 3.

Scheme 3.

Catalytic pathway involving a bidentate catalyst.

In conclusion, we have described the role of DMF as the preferred solvent (acting also as catalyst) and carbonate/phosphate salts as efficient nucleophilic catalysts for cyanosilylation of aldehydes and ketones. The efficiency of an organic phosphate as a homogenous catalyst was also manifested. Although the reactions of many aldehydes and aliphatic ketones in DMF proceed smoothly, even in the absence of a catalyst, the reactions were found to be faster and more efficient in the presence of nucleophilic catalysts such as carbonate and phosphate salts.

Experimental Methods

The general procedure for the addition of TMSCN to carbonyl compounds under nucleophilic catalysis is as follows. TMSCN (1.2–1.5 equivalent) was added to a solution of carbonyl compound (1 mmol) and the nucleophilic catalyst (1–5 mol%) in dry DMF (3 ml), and the reaction mixture was stirred vigorously at room temperature. After the completion of the reaction (monitored by TLC), the mixture was poured in brine solution (15 ml) and extracted with CH2Cl2 (30 ml, three times). The organic extract was washed with brine solution and dried over anhydrous Na2SO4. Removal of the solvent and purification by column chromatography (using hexane-ethyl acetate; 4:1) afforded the pure TMS protected alcohols. Products were characterized by spectral analysis ([1H]NMR, [13C]NMR, [19F]NMR, and HRMS) and the spectral data are included in supporting information (SI) Text.

Supplementary Material

Supporting Text

Acknowledgments

Dedicated to Professor Niel Bartlett with friendship and admiration on the occasion of his 75th birthday. This work was supported by the Loker Hydrocarbon Research Institute.

Abbreviations

DMF

dimethylformamide

TMSCN

trimethylsilyl cyanide.

Footnotes

The authors declare no conflict of interest.

This article contains supporting information online at www.pnas.org/cgi/content/full/0611309104/DC1.

References

  • 1.Gregory RJH. Chem Rev. 1999;99:3649–3682. doi: 10.1021/cr9902906. [DOI] [PubMed] [Google Scholar]
  • 2.North M. Synlett. 1993:807–820. [Google Scholar]
  • 3.Evans DA, Truesdale LK. Tetrahedron Lett. 1973;49:4929–4932. [Google Scholar]
  • 4.Evans DA, Carroll GL, Truesdale LK. J Org Chem. 1974;39:914–917. [Google Scholar]
  • 5.North M. Tetrahedron Asymmetry. 2003;14:147–176. [Google Scholar]
  • 6.Shen ZL, Ji S-J, Loh T-P. Tetrahedron Lett. 2005;46:3137–3139. [Google Scholar]
  • 7.Li Z-B, Rajaram AR, Decharin N, Qin Y-Z, Pu L. Tetrahedron Lett. 2005;46:2223–2226. [Google Scholar]
  • 8.Lundgren S, Wingstrand E, Penhoat M, Moberg C. J Am Chem Soc. 2005;127:11592–11593. doi: 10.1021/ja052804q. [DOI] [PubMed] [Google Scholar]
  • 9.Trost BM, Martinez-Sanchez S. Synlett. 2005:627–630. [Google Scholar]
  • 10.Qin Y-Z, Liu L, Pu L. Org Lett. 2005;7:2381–2383. doi: 10.1021/ol050660e. [DOI] [PubMed] [Google Scholar]
  • 11.Kurono N, Yamaguchi M, Suzuki K, Ohkuma T. J Org Chem. 2005;70:6530–6532. doi: 10.1021/jo050791t. [DOI] [PubMed] [Google Scholar]
  • 12.Wilkinson HS, Grover PT, Vandenbossche CP, Bakale RP, Bhongle NN, Wald SA, Senanayake CH. Org Lett. 2001;3:553–556. doi: 10.1021/ol0069608. [DOI] [PubMed] [Google Scholar]
  • 13.Utimoto K, Wakabayashi Y, Horiie T, Inoue M, Shishiyama Y, Obayashi M, Nozaki H. Tetrahedron. 1983;39:967–973. [Google Scholar]
  • 14.Grunewald GL, Brouillette WJ, Finney JA. Tetrahedron Lett. 1980;21:1219–1220. [Google Scholar]
  • 15.Gassman PG, Talley JJ. Tetrahedron Lett. 1978;19:3773–3776. [Google Scholar]
  • 16.Bandini M, Cozzi PG, Melchiorre P, Umani-Romchi A. Tetrahedron Lett. 2001;42:3041–3043. [Google Scholar]
  • 17.Bandini M, Cozzi PG, Garelli A, Melchiorre P, Umani-Ronchi A. Eur J Org Chem. 2002:3243–3249. doi: 10.1021/jo0256235. [DOI] [PubMed] [Google Scholar]
  • 18.Whitesell JK, Apodaca R. Tetrahedron Lett. 1996;37:2525–2528. [Google Scholar]
  • 19.Jenner G. Tetrahedron Lett. 1999;40:491–494. [Google Scholar]
  • 20.Saravanan P, Anand RV, Singh VK. Tetrahedron Lett. 1998;39:3823–3824. [Google Scholar]
  • 21.Karimi B, Ma'Mani L. Org Lett. 2004;6:4813–4815. doi: 10.1021/ol0482083. [DOI] [PubMed] [Google Scholar]
  • 22.Chen F, Feng X, Qin B, Zhang G, Jiang Y. Org Lett. 2003;5:949–952. doi: 10.1021/ol034158a. [DOI] [PubMed] [Google Scholar]
  • 23.Higuchi K, Onaka M, Izumi Y. Bull Chem Soc Jpn. 1993;66:2016–2032. [Google Scholar]
  • 24.Holmes IP, Kagan HB. Tetrahedron Lett. 2000;41:7453–7456. [Google Scholar]
  • 25.Holmes IP, Kagan HB. Tetrahedron Lett. 2000;41:7457–7460. [Google Scholar]
  • 26.Hatano M, Ikeno T, Miyamoto T, Ishihara K. J Am Chem Soc. 2005;127:10776–10777. doi: 10.1021/ja051125c. [DOI] [PubMed] [Google Scholar]
  • 27.Tian S-K, Deng L. J Am Chem Soc. 2001;123:6195–6196. doi: 10.1021/ja010690m. [DOI] [PubMed] [Google Scholar]
  • 28.Tian S-K, Hong R, Deng L. J Am Chem Soc. 2003;125:9900–9901. doi: 10.1021/ja036222p. [DOI] [PubMed] [Google Scholar]
  • 29.Córdoba R, Plumet J. Tetrahedron Lett. 2003;44:6157–6159. [Google Scholar]
  • 30.Amurrio I, Córdoba R, Csákÿ AG, Plumet J. Tetrahedron. 2004;60:10521–10524. [Google Scholar]
  • 31.Ryu DH, Corey EJ. J Am Chem Soc. 2004;126:8106–8107. doi: 10.1021/ja0475959. [DOI] [PubMed] [Google Scholar]
  • 32.Fuerst DE, Jacobsen EN. J Am Chem Soc. 2005;127:8964–8965. doi: 10.1021/ja052511x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Liu X, Qin B, Zhou X, He B, Feng X. J Am Chem Soc. 2005;127:12224–12225. doi: 10.1021/ja052629d. [DOI] [PubMed] [Google Scholar]
  • 34.Denmark SE, Chung W-j. J Org Chem. 2006;71:4002–4005. doi: 10.1021/jo060153q. [DOI] [PubMed] [Google Scholar]
  • 35.Rasmussen JK, Heilmann SM. Synthesis. 1978:219–221. [Google Scholar]
  • 36.He , Li Y, Feng X, Zhang G. Synlett. 2004:1776–1778. [Google Scholar]
  • 37.Chen F-X, Feng X. Synlett. 2005:892–899. [Google Scholar]
  • 38.Song JJ, Gallou F, Reeves JT, Tan Z, Yee NK, Senanayake CH. J Org Chem. 2006;71:1273–1276. doi: 10.1021/jo052206u. [DOI] [PubMed] [Google Scholar]
  • 39.Wang L, Huang X, Jiang J, Liu X, Feng X. Tetrahedron Lett. 2006;47:1581–1584. [Google Scholar]
  • 40.Curini M, Epifano F, Marcotullio MC, Rosati O, Rossi M. Synlett. 1999:315–316. [Google Scholar]
  • 41.Firouzabadi H, Iranpoor N, Jafari AA. J Organomet Chem. 2005;690:1556–1559. [Google Scholar]
  • 42.Chuit C, Corriu RJP, Reye C, Young JC. Chem Rev. 1993;93:1371–1448. [Google Scholar]
  • 43.Prakash GKS, Panja C, Vaghoo H, Surampudi V, Kultyshev R, Mandal M, Rasul G, Mathew T, Olah GA. J Org Chem. 2006;71:6806–6813. doi: 10.1021/jo060835d. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supporting Text

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES