Last modified on 9 February 2012, at 19:04

Hydrogen-bonding-mediated directed osmium dihydroxylation

Hydrogen-bonding-mediated directed osmium dihydroxylation refers to the diastereoselective conversion of an alkene containing a proximal hydrogen bond donor to a syn 1,2-diol through the action of osmium tetroxide and a Lewis basic amine activator. Hydrogen bonding between the directing group and oxidant is invoked to explain the diastereoselectivity of this process.[1]

Introduction

The treatment of alkenes with osmium tetroxide is an effective method for the installation of syn 1,2-diols.[2] OsO4 is inert to many other functional groups, and over-oxidation of the products does not occur. Amine bases such as pyridine and quinuclidine are able to coordinate to osmium and increase the speed of the dihydroxylation process. A disadvantage of this method is that a stoichiometric amount of metallic oxidant is needed; however, employing an N-oxide (typically quinuclidine N-oxide, or QNO) as a co-oxidant permits the use of substoichiometric amounts of osmium.[3] Reduction of the N-oxide creates an amine that can coordinate to osmium to accelerate dihydroxylation. Kishi carried out studies of the stereoselectivity of dihydroxylations of chiral alkenes with these systems, and observed that dihydroxylations generally occur anti to a heteroatom attached to an allylic stereogenic center.[4] Donohoe discovered that the dihydroxylation of allylic alcohols and protected amines with osmium tetroxide and TMEDA gave the opposite, syn, selectivity (Eq. 1).

(1)
HBondOsDiOH-Gen.png

The origin of syn stereoselectivity in these reactions is a hydrogen bond between the directing group and the activated OsO4-amine adduct. Since its discovery, this reaction has served as an alternative to Kishi's anti-selective dihydroxylations and catalyst-controlled dihydroxylation with the Sharpless system.[5]

Mechanism and Stereochemistry

Stoichiometric Dihydroxylation

Coordination of bidentate tetramethylethylenediamine (TMEDA) to OsO4 creates a highly reactive adduct that rapidly dihydroxylates alkenes. In addition, this coordination also increases the electron density and hydrogen-bond-accepting ability of the oxo ligands on osmium, suggesting that appropriately positioned hydrogen bond donors in the substrate may be able to effectively direct hydroxylation by OsO4-amine adducts. This appears to be the case, as high diastereoselectivity is observed in reactions of constrained, cyclic allylic alcohols and amides (see below). Initially it was unclear whether the TMEDA ligand was promoting hydrogen bonding via transition structures such as A or B (Eq. 2). However, detailed mechanistic studies on the nature of the amines that allow hydrogen bonding control, together with infrared analysis of the osmium-TMEDA complex showed that reaction occurs via a bidentate complex A.[6]

(2)
HBondOsDiOH-Mech-1.png

Catalytic Dihydroxylation

TMEDA is an ideal ligand if high diastereoselectivity is desired because of its ability to form a bidentate complex with the transition metal. However, the use of TMEDA precludes the use of a catalytic amount of osmium tetroxide. Dihydroxylation can be rendered catalytic if a monoamine oxide is used as a stoichiometric terminal oxidant (Eq. 3).[7] The amine formed after reduction of the N-oxide is able to coordinate to osmium to accelerate dihydroxylation. Although the use of monodentate amines lowers diastereoselectivity with respect to the OsO4-TMEDA system, diastereoselectivity is often still high.

(3)
HBondOsDiOH-Mech-2.png

Scope and Limitations

This section addresses the influence of the amine ligand, directing group, steric effects in the substrate, and alkene configuration of the substrate on the diastereoselectivity of the directed dihydroxylation.

As mentioned previously, chelating amines such as TMEDA result in the highest diastereoselectivities. Removing the ability of the amine to form a bidentate complex lowers the diastereoselectivity, in general (Eq. 4).[6][7][8] Because the catalytic method requires the use of a monoamine oxide, diastereoselectivities are lower in general for this method.

(4)
HBondOsDiOH-Scope-1.png

Ideal directing groups are located close to the alkene. For open-chain alkenes, the directing group must be in an allylic or homoallylic position, and adjacent hydroxyl groups can lower selectivity.[9] In many cases, the observed diastereoselectivity is related to the pKa of the directing group. More acidic X–H bonds are better hydrogen bond donors and result in higher diastereoselectivities (Eq. 5).[6]

(5)
HBondOsDiOH-Scope-2.png

Steric effects are well understood for cyclic alkene substrates. When the directing group is homoallylic, it must be oriented in a pseudoaxial position to direct syn hydoxylation—directing groups locked in pseudoequatorial positions are ineffective. When the directing group is allylic, the opposite is true (Eq. 6). Pseudoaxial allylic directing groups cannot effectively direct dihydroxylation because of 1,3-diaxial interactions as the osmium approaches the alkene.[8]

(6)
HBondOsDiOH-Scope-3.png

The configuration of the alkene also plays an interesting role in the diastereoselectivity of directed hydroxylations of open-chain alkenes. The diastereoselectivity of reactions of cis alkenes is significantly higher than reactions of analogous trans alkenes (Eq. 7). This observation has been attributed to the development of prohibitive A[1,3] strain in the most likely transition state for anti dihydroxylation. The transition state for syn hydroxylation avoids this strain, and in the absence of a large group in the cis position (i.e., in the trans substrate), A[1,3] strain is less important.[10]

(7)
HBondOsDiOH-Scope-4.png

Comparison to Other Methods

Relatively few methods exist for the conversion of alkenes to syn 1,2-diols. A notable example is the Woodward modification of the Prévost reaction, which exhibits moderate syn diastereoselectivity when alkenes containing an allylic stereocenter are employed (Eq. 8).[11] Iodonium ion formation occurs on the face opposite the acetoxy group, and subsequent opening by acetate, neighboring group participation, and hydrolysis produce the syn diol.

(8)
HBondOsDiOH-Alt-1.png

In an example typical of high-valent transition metals other than osmium, ruthenium-mediated dihydroxylation proceeds with anti diastereoselectivity (Eq. 10).[12]

(9)
HBondOsDiOH-Alt-2.png

Experimental Conditions and Procedure

Typical Conditions

The osmium-mediated dihydroxylation reaction is carried out under an inert atmosphere such as argon or nitrogen and solvents (CH2Cl2, acetone, THF) must be anhydrous. Osmium tetroxide is toxic, volatile, and sublimes quite easily; it should therefore be handled in a well-ventilated fume-hood. The aqueous layers from the osmium-mediated reactions and any other waste materials should be disposed of with care.

Example Procedure[6]

(10)
HBondOsDiOH-Ex.png

To a solution of 2-cyclohexene-1-ol (50 mg, 0.51 mmol) (0.50 mmol) and TMEDA (0.55 mmol) in CH2Cl2 precooled to –78 °C was added a solution of OsO4 (0.53 mmol) in CH2Cl2 (~1 mL). The solution turned deep red and then brown-black. It was stirred until the reaction was complete (TLC analysis, ca. 1 h) before being allowed to warm to rt. After completion of the oxidation, the solvent was removed under vacuum and replaced with THF (10 mL) and sodium sulfite (aq saturated solution, 10 mL). This mixture was heated at reflux for 3 h. The crude reaction mixture was then concentrated under vacuum to afford a grey powder; EtOH (30 mL) was added and the suspension stirred at rt for 1 h. Filtration of the resulting suspension through Celite and concentration of the filtrate under vacuum gave a colorless solid (80 mg). Purification by column chromatography (SiO2, EtOAc/petroleum ether 7:1) afforded the compound shown as an inseparable mixture of isomers (66 mg, 98%, syn/anti 9:1): IR (film) 3192, 2927 cm–1; 1H NMR (300 MHz, D2O) δ 3.81 (t, J = 2.6 Hz, 1H), 3.52 (ddd, J = 10.0, 4.6, 2.6 Hz, 2H), 1.80–1.00 (m, 6H); 13C NMR (75 MHz, D2O) δ 72.6, 70.3, 26.3, 18.8; CIMS (m/z): [M + NH4]+ 150(100); CI (m/z): [M + NH4]+ calcd for C6H16NO3, 150.1130; found, 150.1128.

References

  1. Donohoe, T. J.; Bataille, C. J. R.; Innocenti, P. Org. React. 2012, 76, 1. doi: (10.1002/0471264180.or076.01)
  2. Criegee, R. Liebigs Ann. 1936, 522, 75.
  3. Kolb, H. C.; Vannieuwenhze, M. S.; Sharpless, K. B. Chem. Rev. 1994, 94', 2483.
  4. Cha, J. K.; Kim, N. S. Chem. Rev. 1995, 95, 1761.
  5. Jacobsen, E. N.; Marko, I.; Mungall, W. S.; Schroeder, G.; Sharpless, K. B. J. Am. Chem. Soc. 1988, 110, 1968.
  6. a b c d Donohoe, T. J.; Blades, K.; Moore, P. R.; Waring, M. J.; Winter, J. J. G.; Helliwell, M.; Newcombe, N. J.; Stemp, G. J. Org. Chem. 2002, 67, 7946.
  7. a b Blades, K.; Donohoe, T. J.; Winter, J. J. G.; Stemp, G. Tetrahedron Lett. 2000, 41, 4701.
  8. a b Donohoe, T. J.; Blades, K.; Helliwell, M.; Moore, P. R.; Winter, J. J. G.; Stemp, G. J. Org. Chem. 1999, 64, 2980.
  9. Donohoe, T. J.; Moore, P. R.; Waring, M. J. Tetrahedron Lett. 1997, 38, 5027.
  10. Hoveyda, A.; Evans, D. A.; Fu, G. C. Chem. Rev. 1993, 93, 1307.
  11. Kallatsa, O. A.; Koskinen, A. M. P. Tetrahedron Lett. 1997, 38, 8895.
  12. Plietker, B.; Niggemann, M. Org. Lett. 2003, 5, 3353.