79.5 74.5 Z-Gly-L-Ala L-Phe-OMe <3 5 10 12 78.0 71 .O Z-Gly-L-Phe L-Ala-OMe <3 9 17 17 7 7 . 5 7 0 . 5 Z-Gly-Gly-L-Phe L-Ala-OMe 10 15 25 20 81.5 72.0 For-L-Phe L-Ala-OMe <3 <3 <3 <3 7 9 . 5 72.5 Ac-L-Ala L-Phe-OMe <3 <3 35 27 79.0 7 1 . 5 a For = formyl; all other abbreviated designations of compounds follow IUPAC-IUB rules. * The optical purity of the starting materialswas verified by gas-liquid partition chromatography : N-acetyl and N-benzoyl compounds were converted to the menthyl ester derivatives, while the N-formyl compounds were hydrolyzed and analyzed as the N-trifluoroacetyl-L-prolyl peptide esters; the methyl ester compounds were assayed similarly.' c The limit of measurement was generally 3 z, although in some cases a more accurate value was obtained by duplicate procedures. All spectra were determined on a Varian A-60 spectrometer with the center of gravity of the chemical shift given in hertz downfield from tetramethylsilane ( J = 7.2 i 0.3 Hz). The compounds were dissolved in deuteriochloroform (deuterioethanol for the tetrapeptide).nuclear magnetic resonance (nmr) spectra.14* The methyl doublet signal in an L-L (or D-D) compound was at lower field than the equivalent signal for the D-L (or L-D) analog due to deshielding and offered a potentially convenient means for the quantitative analysis of such mixtures. We have now employed this technique to examine the influence of several coupling agents and N-acyl protecting groups on the extent of racemization during peptide synthesis.In a typical experiment, N-acetyl-L-phenylalanine (0.207 g, 1 mmole), carbonyldiimidazole (0.162 g, 1 mmole), and methylene chloride or acetonitrile (5 ml) was stirred at -5" for 1 hr. The hydrochloride of L-alanine methyl ester (0.139 g, 1 mmole) and triethylamine (0.14 ml, 1 mmole) were added and the solution was kept at 0" for 15 hr. After suitable washings, the solution was dried and evaporated; the solid residue was redissolved in deuteriochloroform and the solution was used for nmr analysis.