Quantitative analysis of the volatile constituents of lemon peel oil. Effects of silica gel chromatography on the composition of its hydrocarbon and oxygenated fractions

Abstract

Fifty-one constituents accounting for -99.7 5% of total volatiles common to both Sicilian and California commercial lemon peel oils have been identified and quantified in a single GC capillary run. The quantification was performed by using both internal standard and appropriate response factors, which are necessary for accurate volatile analysis and simultaneous determination of nonvolatiles. Direct comparisons to results obtained by using area percent or internal standard only show the advantage of using response factors. Fractionation of the oils by open-column silica chromatography facilitated identification of constituents and improved the accuracy of the analysis. Yields of the fractions recovered from the column were usually greater than 96%. The reactivity of high-quality silica gel with lemon hydrocarbons was investigated, and while quantitative effects are small, several oxygenates are formed on the column. Use of a cold (3 C) column significantly reduces the amount but not the number of oxygenates formed. Lemon oil is an essential oil of considerable commercial importance. In the mid 1980s the United States led in lemon oil production with -25% of total world output (2300 metric tons) followed closely by Italy and Argentina (Lawrence, 1985). In spite of the large domestic production, more lemon oil than any other essential oil except orange is imported into the United States. These imports amount to e700 metric tons and are third in dollar value only to distilled lime and patchouli oils (Pisano, 1986). Considering the economic value of lemon oil, it is surprising that many trace constituents that probably contribute to lemon flavor remain unidentified, and in addition, only 37 constituents have been thoroughly quantified (Staroscik and Wilson, 1982a,b). An understanding of the relationship between taste and composition of lemon oil requires that a relatively complete quantitative data base be developed. Some results on the correlation between certain components of cold pressed lemon oil and odor, but not taste, have been published (Drawert and Christoph, 1984; Schieberle and Grosch, 1988). The most accurate GC analyses of essential oils and flavor volatiles are obtained by using both an internal standard and response factors (RFs). Shaw (1979) has argued convincingly for the adoption of this approach, but it is still not utilized enough. Of some 20 substantive papers concerning the analysis of lemon oil published since 1960, only one group (Staroscik and Wilson, 1982a,b) used both an internal standard and response factors; another group (Selzer and Christoph, 1981) employed an internal standard alone.