Table of Contents

Solutions classified as perfumes generally contain 10–25% of perfume concentrate. Perfume-containing products, such as cologne, contain 2–6% of concentrate. Perfume concentrate may be incorporated into aftershave lotions, aerosol sprays, and bath oils, and enhance the scent of soaps, talcs, face powders, and deodorants. In addition, industrial perfumes mask undesirable odors in paints and cleaning materials (1).

Figure 1. Major components of synthetic jasmine.

Jasmine, one of the most popular scents, is widely used in fragrances. About 83% of women’s and 33% of men’s fragrances contain jasmine (5). The scent of synthetic jasmine depends on four components shown in Figure 1: benzyl acetate (1), indole (2), cis-jasmone (3), and methyl jasmonate (4).

Because essential oils and other natural ingredients are scarce and expensive, synthetic materials that duplicate natural scents are the major source of today’s fragrances. Techniques such as IR and UV spectroscopy, gas chromatography, mass spectrometry, thin-layer chromatography, and optical rotatory dispersion are used to determine the components of perfume (6). Recent technological developments have aided the perfume industry—for example, an electronic “nose” has been used to discriminate among perfumes (7; see also Carbosiloxane polymers for chemical sensors in this issue). The headspace measurement of evaporation rates of perfumes applied to skin is an indicator of the performance of perfume ingredients (8).

Practicing the art of perfumery can be a time-consuming process. The preparation of a perfume formulation may require a specific blending of more than 100 ingredients. The slightest changes in the relative proportions of perfume components can result in a different product. Using the traditional approach, the optimization of each of these variables could require a seemingly infinite number of experiments.

Automated process research (APR) technology couples automated synthesis equipment and statistical design of experiments, which results in a much more rapid optimization of reaction conditions (9–11). This technology has several applications, including chemical scaleup, chromatography and purification, materials science, and the optimization of personal care and consumer products. APR technology was applied to the optimization of the relative proportions of four jasmine perfume components to produce the most desirable fragrance.

Figure 2. An overview of the APR mixture-based designs. In the first experimental design, each component was evaluated in 12.5% increments from 0 to 100%. The resultant subfamily of most pleasing formulations was further examined in the second design. After two rounds, the optimized formulation contains the indicated percentage of each component.

Screening experiment

Stock solutions (10 vol% each) of 1, 2, 3, and 4 were prepared in food-grade ethanol. Design Expert software (Stat-Ease Inc., Minneapolis, MN) was used for experimental design.

In the first experimental plan, an eight-level, four-factor distance-based design for mixtures was implemented to screen each component from 0 to 100% relative to the other components. Varying amounts of the four components were mixed to a total volume of 1 mL in 96-well microplates. A total of 165 mixtures was prepared, with 12.5% step increments of each component. Glycerol (100 mL) was added as a fixative to each sample. For example, a sample with relative percentage of 1 equal to 100% would consist of 1 mL of 1 stock solution (10% in ethanol) and 100 mL of glycerol. The automated preparation of these perfume samples with varying proportions of components took only 1–2 hours. A portion of each sample was transferred to test strips to evaporate the ethanol. Four individuals identified their favorite fragrances. These fragrances rep resent a subfamily of desirable perfume formulations.

The screening experiment revealed that the average relative percentages of 1, 2, 3, and 4 for the best fragrances were 12.5, 12.5, 25, and 50%, respectively.

Optimization

The subfamily of formulations was further optimized with a user-defined, distance-based design for mixtures. On the basis of the results of the screening experiment, the relative percentages of 1, 2, 3, and 4 were varied from 0 to 25, 0 to 25, 12.5 to 37.5, and 37.5 to 62.5%, respectively. The vertices, centers of edges, thirds of edges, triple blends, constant plane centroids, axial check blends, and overall centroid were included in the statistical design for a total of 94 perfume mixtures. The automated preparation of these perfume samples was run overnight. The same four individuals were requested to identify a common favorite scent.

The favorite fragrance consisted of 8.3, 16.7, 20.8, and 54.2% of 1, 2, 3, and 4, respectively. An overview of the methodology and the relative percentages corresponding to the most pleasing fragrance following two designed APR experiments is shown in Figure 2.

The outlook for our industry

The use of APR technology in chemical process research and development has been patented recently (12). This technique is extremely useful in the highly competitive personal care products industry. Libraries of perfume formulations can be generated to facilitate the rapid testing of a new component. The combination of all ingredients may then be optimized rapidly by several statistically designed experiments. APR can potentially be coupled with other technologies to identify the optimum scent. APR can greatly decrease the amount of time from discovery to market in the personal care industry.

Acknowledgments

The authors thank Angieszka Bischoff, Daria Darczak, Carla Edwards, and Suhail Bahu for evaluating the perfume samples and for reviewing the manuscript.

References

1. Karppinen, S. L. A View of the World of Perfume; http://zebra.biol.sc.edu/smell/shannon/shannon.html (accessed Sept 5, 2000).
2. Bedoukian, P. Z. Perfum. Flavor 1985, 10 (2), 1–28.
3. Ohloff, G. Perfum. Flavor. 1978, 3 (1), 11–22.
4. McAndrew, B. A. Chem. Br. 1982, 18, 864–870.
5. Rouhi, A. M. Chem. Eng. News 1999, 77 (43), 38–46.
6. Bedoukian, P. Z. Perfumery and Flavoring Synthetics, 2nd ed.; Elsevier Biomedical: Amsterdam, 1967.
7. Carrasco, A.; Saby, C.; Bernadet, P. Flavour Fragrance J. 1998, 13, 335–348.
8. Vuilleumier, C.; Flament, I.; Sauvegrain, P.; Firmenich, S. A. Perfum. Flavor. 1995, 20 (2), 1–10.
9. Wagner, R. W.; Feirong, L.; Du, H.; Lindsey, J. S. Org. Process Res. Dev. 1999, 3, 28–37.
10. Emiabata-Smith, D. F.; Crookes, D. L.; Owen, M. R. Org. Process Res. Dev. 1999, 3, 281–288.
11. Zhang, J.; Kirchhoff, E. W.; Zembower, D. E.; Jimenez, N.; Sen, P.; Xu, Z.-Q.; Flavin, M. T. Org. Process Res. Dev., in press.
12. Flavin, M. T.; Seaney, L. M. U.S. Patent 6,044,212, 2000.

John Aikens is director of client communications at MediChem Research Inc. (jaikens@mcr.medichem.com).

Constance Cassidy is a senior technical writer at MediChem Research Inc. (630-257-1500, ext. 1084; ccassidy@mcr.medichem.com).