Pyrolysis of these pyrazolines in the gas phase at 292° allows a complete study of the stereochemistry of 1-pyrazoline decomposition. 34T affords cis-1-ethyl-2-methylcyclopropane (48C) in nearly racemic form and trans-1-ethyl-2-methylcyclopropane (48T) 22.5% optically active with predominant inversion of the alkyl groups. Further thermal reactions of azabutayaenes to form olefins and nitriles and dihydroisoquinolines were also studied.
The net single inversion involved in the formation of the cyclopropane major product is not limited to the thermal decomposition of 1-pyrazolines, but appears to be present in several other 4-electron fragmentation processes as well. Interesting work by Freeman and co-workers involving the thermal decomposition of a series of diazenes shows a stereospecificity very similar to that of 1-pyrazolines (6. Trost and co-workers cite the similarity of the decomposition of 1-pyrazolines and (probably) tetraalkyl sulfur compounds llC and 11 T at -78 °C (~} (7).
The normal central C-C-C bond angle for the pyrazoline in the sheath conformation is 109°, implying that no significant change in geometry is required for conversion to the 0,0 diradical. Crawford has observed that at least 6% of the cyclopropane products from the pyrolysis of optically active 3, S-dimethyl-1-.
Our study of the thermal decomposition of these pyrazolines was designed to answer the following questions. Resolution of both series was accomplished by conversion of 37 to the phthalate half-ester followed by formation of the brucine salt and tedious fractional recrystallization from ethyl acetate. After reduction of the brucine salt and saponification of the half-ester, the optically active l -hexen-4-ol was obtained with 38 .
Since accurate knowledge of the optical purity of ~ is imperative, its purity was determined by two independent methods. We checked the correlation results using Mosher's (39) method of converting the alcohol to its ester (52) with Ct-methoxy-Ct-trifluoromethylphenylacetyl chloride (MTPA-Cl), followed by integration of diastereomeric CF. Optical purity for both series 38 was confirmed by agreement of polarimetric determination with nmr integration at.
The purity of the cis- or trans-pyrazoline is assumed to be identical to the purity of the dibromide precursor. The nmr spectra of cis- and trans-pyrazolines provide limited confirmation of our assumption.
6T (Sh) and double retention in 34C -+ 48C conversion; and 5) possible simultaneous cleavage of the C-N bond in the rate-determining step of the reaction. 075 mol) of 38 was added over a period of ten minutes, during which time the yellow suspension became a colorless solution •. The solution was stirred for another twenty-five minutes. of an aqueous 3N NaOH solution was added dropwise over ten. minutes and the solution turned yellow again. The reaction mixture was saturated with NaCl and extracted four times with a total of 200 mL of tetrahydrofuran; the extracts were dried over K. The solvent was distilled off through a Vigreaux column and the residue was distilled to give 6. the spectra are what one would expect by comparison with 40H prepared by Clarke {la).
The ether layer was washed with aqueous HCl, followed by water, and was dried over Na2SO4. • The ether was removed by rotary evaporation and 100 mg of. Polymerization of some of the pyrolysis products in the collection trap was a troublesome problem in this work. However, these results were not reproducible; recovery of acetonitrile in a 10% yield relative to styrene was a more typical result.. 123 . the stability of diethyl ether to the reaction conditions.
The conversion of the azabutadiene (~) into the dihydroisoquinoline. 2,!) is in itself a new transformation. This pyrolysis product was isolated by preparative vpc of a concentrated ether solution of the pyrolysate (column f, 100°C, 100 ml/min). Conversion of 4 to the silanol (10) and subsequent treatment with strong acid provides another route to the desired cation. The existence of species that utilize silicon-carbon (prr-prr) bonding has been postulated (11).
Barton has also argued for the equilibrium shown in Scheme II (16). West's research group attempted catalytic oxidation of ~ and §_ in an effort to "generate aromatic compounds containing the silicon-olefinic bond {15)".
Electrophilic displacement of the zirconium moiety with dilute acid or halogens creates alkanes or haloalkanes. An initial addition of Zr at the t -butyl carbon of l followed by elimination of the most accessible D1 would generate 2 (Scheme III). Elimination of (Zr)H from & followed by hydrozirconation would place two Ds on one carbon of the alkyl fragment; this was not observed.
Consequently, at least for olefins (1) and (~), the zirconium moiety appears to initially add the least hindered carbon. The regioselectivity of hydrozirconation reactions; the deformed tetrahedral geometry of the zirconium hydride and the lack of dissociative behavior of Zr(IV) complexes (16 electrons. The number of available electrophilic cleavage processes of the alkylzirconium complex makes development of such a reagent even more attractive.
Six of the hybrid orbitals are directed toward the cyclopentadienyl ligands, with the other three orbitals lying in the x,y plane. Excluding as unlikely the initial cleavage of the 16-electron complex ligand, zirconium hydride, Zr-C bond formation is best thought to proceed through a type of four-. The 1f orbital of the approaching olefin may lie in the H-Zr-Cl (2) plane or be perpendicular to this plane (~) (Scheme IV).
Elimination of the olefin by zirconium hydride regeneration from !!_ should be microscopically the reverse of the formation of 11. Consequently, the zirconium hydride must retain its original chirality even if elimination toward the more sterically hindered carbon is an important process. Recently, electrophilic cleavage of the carbon-zirconium complex (3) was found to proceed with forward attack on the carbon atom for H+ and Br+ and for CO insertion (3). the stereochemistry of this attack has no effect on the chiral carbon).
It may be possible to reduce cycloalkenes to chiral cycloalkanes, even if the zirconium moiety must remain attached to a ring carbon. After synthesis of 14, the zirconium hydride can be synthesized by established methods shown in Scheme VII (11, 12). The chiral ligand in ~ (after fractional recrystallization) should have no apparent effect on the asymmetric induction process of the zirconium complex.
