146 HALES THE INFRA-RED SPECTROSCOPIC ESTIMATION OF [Vol. 75 The Infra-Red Spectroscopic Estimation of 4-Methyl-2 : 6-Ditertiary Butyl Phenol in Mixtures Containing 2-Methyl and 3-Methyl-4 : 6-Ditertiary Butyl Phenols BY J. L. HALES SYNOPSIS-Industrial requirements set the problem of finding a rapid, accurate method for the estimation of 4-methyl-:! : 6-ditertiary butyl phenol in crude and refined products. It was also desirable to obtain a check on the type and amount of impurity in the products. A description is given of an infra-red spectroscopic technique whereby this may be obtained. The limitations inherent in the method because of the presence of certain impurities are indicated. Experimental values are given for estimations on several synthetic mixtures and on a crude product, and the magnitude of the errors introduced by impurities is indicated. INTRoDucTIoN-Weinrichl has shown that although m- and 9-cresol form mixtures which are difficult to separate, the two corresponding ditertiary butyl derivatives have a marked difference in volatility.Should, however, a cresol fraction containing o-cresol in addition to the m- and 9- derivatives be ditertiary butylated, the product will contain, in addition to the 4-methyl-2 : 6-ditertiary butyl phenol (b.p. 191" C./lOO mm.) and 3-methyl-4 : 6- ditertiary butyl phenol (b.p. 211" C./100 mm.), a proportion of 2-methyl-4 : 6-ditertiary butyl phenol (b.p. 194"C./100mm.) which will render more difficult the separation by distillation of the first-mentioned component in a reasonably pure state. The 4-methyl-2 : 6-ditertiary butyl phenol has technological applications as a softener and anti-oxidant, and the present purpose was to work out a rapid infra-red spectroscopic method of analysing mixtures in which it is the preponderant component, but which contain the other two isomerides, and in addition small quantities of monobutyl derivatives of the three cresols.March, 19501 4-METHYL-2 : 6-DITERTIARY BUTYL PHENOL 147 A method is described which gives an accurate value for the quantity of the desired component, and also a check on the amounts of the other two components present.A correction may be applied to allow for errors introduced due to the presence of monobutyl cresols in the crude mixture. EXPERIMENTAL Samples of the following were available- (a) The three individual dibutyl cresols.(b) A crude mixture of all three dibutyl cresols. (c) A fraction representative of the monobutyl cresol content of (b). The spectra of these samples and of their solutions in cyclohexane were obtained on a Hilger D209 double-beam infra-red spectrometer,2 using a rocksalt prism (see Fig. 1).* The region from 700 to 800 cm.-l was found to be the most useful for analytical purposes, since it includes key bands suitable for the estimation of each of the three components. Furthermore, rocksalt has a high dispersion in this region. Since all the aromatic ring frequencies lie in this range, the analysis is liable to interference from aromatic impurities, such as unchanged or monobutylated cresols. cycZoHexane, which is fairly transparent in this region, was chosen as the solvent for this analysis.Standard solutions of various concentrations of the individual components were made up and their spectra were measured. Sample cell thickness was 1-00 mm. for all analytical measurements. Wide slits (1.20 mm., corresponding to spectral slits of 6 cm.-l) were used with the spectro- meter, so as to give a high signal-noise ratio for the thermopile, and to avoid resolving the key band at 775 cm.-l (Fig. 1) into two bands at 771 and 776 cm.-l, which would complicate the analysis. Optical density at the three key frequencies for each of the three components was plotted against concentration (Fig. 2). Scattered radiation was allowed for, using a mica shutter and a l-mm. thick cell filled with cyclohexane, so as to approximate closely to the experimental conditions.A number of synthetic mixtures were prepared, their optical density a t the key frequencies measured, and from the curves illustrated in Fig. 2 the analytical figures were worked out by the method of successive approximation^.^ The results are given in Table I. TABLE I SYNTHETIC MIXTURES This is referred to later in the paper. Ditertiary butyl phenols Found .. Found (corrected) Taken . . .. Found .. Found (corrected) Taken . . .. Found .. Found (corrected) Taken . . .. Found .. Found (corrected) Taken . . .. .. .. . . .. . . . . . . . . . . .. .. .. .. .. .. .. .. . . .. .. ,. .. .. .. r 3 4-methyl-2 : 6-, 2-methyl-4 : 6-, 3-methyl-4 : 6-, g./100 ml. g./100 ml. g./100 ml. .. 1-81 2.48 2.42 ..1-67 1.78 1.81 . . 1.72 1.68 1.69 .. 1.75 . . 1-70 . . 1.72 . . 1.55 .. 1.51 . . 1.52 .. 2.10 . . 2.07 . . 2.05 1-30 0.75 0.7 1 0.99 0.5 1 0.44 0.83 0.26 0.23 1.08 0-73 0-64 0.87 0.60 0.55 0.53 0.27 0-22 Corrected figures take into account the overlapping of key bands. It should be noted (Fig. 2) that the key band at 775 cm.-l, for the 4-methyl-2 : 6- ditertiary butyl phenol has an extinction coefficient considerably higher than those of the bands at the key frequencies for the other two components. This enhances the accuracy and reliability of the estimation of this component, particularly as it is the major component * The spectrometer that was used incorporated a D.C. amplifying system, and slight drifts in the latter were the main factor limiting the consistency of the quantitative results.148 100 80 - 60- 40..20 - O -i HALES : THE INFRA-RED SPECTROSCOPIC ESTIMATION OF ? A I I I 1 I I Wol. 75 100 80- 60 - 40- $ 20- U B Wave numbers (cm.-l) Fig. 1. Absorption Spectra. The arrows indicate analytical bands. A : 2-Methyl-4 : 6-ditertiary butyl phenol. C: 4-Methyl-2 : 6-ditertiary butyl phenol. B: 3-Methyl-4 : 6-ditertiary butyl phenol. D. Monobutyl cresol fraction. Concentration (g./lOO ml.) y = 755 crn.-l -0 I Concentration (g./lOO rnl.) Concentration (g./100 mt.) Fig. 2. Analytical curves for ditertiary butyl phenols. A: 2-Dfethyl-4 : 6 ditertiary butyl phenol. B: 3-Methyl-4: 6-ditertiary butyl phenol. C: 4-Methyl-2 : 6-ditertiary butyl phenol.March, 19501 4-METHYL-2 : 6-DITERTIARY BUTYL PHENOL 149 in the present case.In Table I the uncorrected figures illustrate the error introduced when no allowance is made for the overlapping of the key bands, and it is seen that the effect is smallest in the case of the band at 775 cm.-l. These considerations are substantiated by the relative magnitude of the errors in the figures given in the table. MONOBUTYL CRESOL FRACTION-This fraction had several absorption bands near 775 cm.-I (Fig. 1) and the presence of monobutyl cresols in the crude product to be analysed will thus reduce the accuracy of estimation of the dibutyl cresols. A correction for errors due to the presence of monobutyl cresols may be applied using an “internal reference” method. The monobutyl cresol fraction had a fairly intense band at 1081 cm.-l which does not coincide with any bands in the dibutyl cresols, and this can be used to detect and estimate the effect of small quantities of monobutyl cresols in the crude mixture.However, a l-mm. layer of cyczohexane absorbs strongly at 1081 cm.-1, so that any comparison of optical densities must be made on the undiluted components. Although this may introduce deviations from Beer’s law, the correction under consideration need not be known to a very high accuracy. Hence the optical density of the band at 1081 cm.+ was compared with the densities at the three key frequencies for the monobutyl cresol fraction, and from the density of the 1081 cm.-l band in the crude dibutyl cresol mixture, the respective corrections for density at the three key frequencies could be estimated.Their magnitude is shown by the figures in Table 11, which gives the results of two analyses on the crude mixture. TABLE I1 ANALYSIS OF CRUDE DIBUTYL CRESOL MIXTURE Ditertiary butyl phenols r 4-methyl-2 : 6-, 2-methyl-4 : 6-, 3-methyl-4 : 6- g./100 ml. g./lOO ml. g./100 ml. A .I Found .. .. .. . . 1.28 0.40 0.18 Corrected for presence of monobutyl cresols . . .. .. .. 1-24 0.26 0.01 Found .. .. .. .. 1-26 Corrected for presence of monobutyl cresols . . .. .. . . 1.22 0-41 0.27 0.19 0.02 Total crude material was 2 g. per 100 ml. in each case. Thanks are due to the Esso European Laboratories for providing the specimens of pure and crude materials required for this investigation. The work has been carried out as part of the research programme of the Chemical Research Laboratory, and this paper is published with the approval of the Director of the Laboratory. REFERENCES 1. 2. 3. Weinrich, W., Ind. Eng. Chem., 1943, 35, 264. Hales, J. L., J . Sci. Inst., 1949, 26, 359. Fry, D. L., Nusbaum, R. E., and Randall, H. M., J . A$pZied Phys., 1946, 17, 150. D.S.I.R. CHEMICAL RESEARCH LABORATORY TEDDINGTON, MIDDLESEX First submitted, July, 1949 Amended, January. 1960