1.
Tetrahydrofuran was gold label from Aldrich Chemical Company, Inc.
2.
(+)-Pulegone was obtained from SCM Corporation, Jacksonville, FL or Givaudan Corporation, Clifton, NJ; the specific rotation ranged from +21.85 to 22.6°. The material is also available from Aldrich Chemical Company, Inc. Pure pulegone has2 [α]D23 + 22.5°. The discrepancy, if any, is probably due to chemical impurities since the pulegone used has been shown to be enantiomerically pure.3,4 The checkers used (+)-pulegone from Givaudan Corporation, [α]D20 +25.7°, which was 94.4% pure (GLC) and contained 4.8% of isopulegone or carvone.
3.
Benzyl mercaptan was used as received from Aldrich Chemical Company, Inc.
4.
Spectral data; IR (film), cm−1: 1708, 1620, 1500, 1458, 1382, 1363, 1120, 710, 695; 13C NMR (50 MHz, CDCl3) δ: 22.2, 23.8, 27.8, 29.6, 33.1, 34.5, 36.6, 48.0, 52.3, 57.8, 126.8, 128.4 (2 C), 128.9 (2 C), 138.7, and 210.2; 1H NMR (200 MHz, CDCl3) δ, partial: singlets at 1.38, 1.60, and 3.73, and an intense doublet at 0.97 (J = 5.9). The checkers noted that the product distillate was a pale-blue color, which turned pale yellow after a few minutes.
5.
The submitters noted that at least 15 kg of dry ice was required for a reaction of this scale. Using the described apparatus the checkers found that the condensation of the ammonia required 8–10 hr, which could be reduced to ca. 5 hr by using the following assembly. An over-dried, 5-L, three-necked, round-bottomed flask was equipped with a glass stirrer shaft fitted with a sleeve joint and a large Teflon blade and the shaft was connected to an overhead motor drive. The flask was also connected to two 450-mL Dewar condensers with a large soda-lime drying tube attached to the tube connector of one condenser while the tube connector of the other condenser was attached with Tygon tubing that led through a tower of solid potassium hydroxide pellets to a tank of anhydrous ammonia. While ammonia was slowly flushed through the entire assembly, the Dewar condensers were filled with dry ice–acetone, and a dry ice–acetone bath was raised to cool the vessel by immersion. The ammonia flow rate was increased to condense ca. 3000 mL of ammonia into the flask. The condenser with the inlet connection from the ammonia was removed and the flask sealed with a glass stopper.
6.
Sodium was stored under mineral oil and washed with pentane before use. For convenience the checkers used 1/6–¼-in. sodium spheres (Matheson Coleman and Bell) that were weighed in mineral oil, then wiped free of oil, rinsed in hexane, cut in half, rinsed in hexane again, and immediately added to the reaction over a 2-hr period, during which time the dark-black mixture became extremely viscous.
7.
At this point the checkers charged an oven-dried, 1000-mL pressure-equalizing addition funnel with the ketone in methanol and ether, and then quickly mounted the sealed funnel on the reaction flask by removing the flask's glass stopper.
8.
Methanol was used as received from Fisher Scientific, Inc. Anhydrous ether was used as received from freshly opened containers from Mallinckrodt, Inc. and Fisher Scientific, Inc.
9.
It is important to let the reaction mixture warm slowly; otherwise the ammonia will boil violently and carry some of the reaction material out of the flask.
10.
This reaction is highly exothermic, and caution should be exercised since some active sodium may occasionally be left on the sides of the flask.
11.
The intense 13C NMR (50 MHz, CDCl3) signals of the major isomer are at 21.9, 26.9, 29.0, 31.3, 34.5, 34.6, 45.4, 47.2, 54.6, and 72.9 ppm relative to TMS; 1H NMR (200 MHz, CDCl3) δ, partial: 0.91 (d, 3 H, J = 6.5), 1.40 (s, 3 H), and 1.52 (s, 3 H).
12.
Benzene was used as received from Aldrich Chemical Company, Inc. Caution! Benzene has been identified as a carcinogen; OSHA has issued emergency standards on its use. All procedures involving benzene should be carried out in a well-ventilated hood, and glove protection is required.
13.
Paraformaldehyde was used as received form Aldrich Chemical Company, Inc.
14.
p-Toluenesulfonic acid monohydrate was used as received from Aldrich Chemical Company, Inc.
15.
Excess paraformaldehyde may separate from the distillate. If this occurs, the liquid should be filtered prior to crystallization.
16.
In the absence of seeding, crystallization may take several weeks. It is preferable to separate a small sample of the precursor thiomenthol from its stereoisomers by HPLC (3% ethyl acetate in hexane as eluant) and prepare a small amount of pure oxathiane from this material. Alternatively, a small amount of the product may be purified by GLC on a 5% FFAP column. The melting point of pure material is 37–38°C. The checkers, who did not have seeding crystals, found that the early crops of crystals melted when the flask was allowed to warm to ambient temperature. Consequently, the cold supernatant liquid was withdrawn from the crystals with a Pasteur pipette while the flask was maintained at ca. 0°C (ice–water bath). The crystals were subsequently recrystallized several times in the same flask without filtration. By this technique, white crystals melting at 32–35°C were obtained; this material is spectrally pure and suitable for asymmetric synthesis. The supernatant liquid was also concentrated, as the submitters described, to obtain additional crops using this technique.
17.
Spectral data; IR (film), cm−1: 2970–2870, 1455, 1440, 1388, 1370, 1355, 1305, 1155, 1095, 1066, 985, 955, 900, 830, and 710; 13C NMR (50 MHz, CDCl3) δ: 21.8, 22.1, 24.4, 29.4, 31.3, 34.7, 41.8 (2 C), 51.5, 67.1, 76.7 ppm; 1H NMR (200 MHz, CDCl3) δ partial: 0.92 (d, 3 H, J = 6.5), 1.27 (s, 3 H), 1.43 (s, 3 H), 3.35 (td, 1 H, J = 10.5, 4.2, HCO), 4.69 (d, 1 H, J = 11.5, SCH2O), 5.03 (d, 1 H, J = 11.5, SCH2O).