42 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS and organic analysis were still in their infancy. The chemistry of the products was not known sufficiently and so they were accepted for what they were: valuable perfumery ingredients. To-day, with modern methods of synthesis and analysis, we can produce these well-studied products free from impurities, but we must realise that by doing so we are in many cases introducing entirely new aromatics which cannot indiscriminately replace the older, impure versions. When a new chemical is made for the first time and is found to have outstanding qualities as a perfumery chemical it may be accepted by the perfumer long before its chemistry is sufficiently known. The original method used for its preparation may not be economical, because in developing a new perfumery chemical the chemist will concentrate in the first place on prepar- ing the chemical by hook or by crook rather than on obtaining it in a good yield. But when the demand for this aromatic is created he will have to try to find the most economical process for its preparation. Then it may well happen that he will finally devise a process which, although producing the desired compound in a good yield, will not produce at the same time a by- product formed in the original reaction used for the preparation of this aromatic in the first place. If the odour character of the aromatic in question was determined by this by-product formed in thi• original reaction, but not obtained in the more economical process, then the odour quality of the aromatic prepared by the latter will be different from that made by the first method. In view of what has been said so far one may well ask which aromatics should be produced in as pure a state as possible, and which of them should be left to contain certain impurities. As a rough guide one could say that most isolates should be as pure as one can make them. There will surely be no difference of opinion on samples of pure citral, pure citronellal, pure geraniol, pure eugenol, to name only a few isolates. Not so with synthetics. Of course, a large number of synthetics will not be acceptable unless they are free from impurities. Such products are probably those the chemistry of which was well known at the time when they were introduced into perfumery, and which consequently were intro- duced into perfumery in the pure state. Most esters can be named here fatty aldehydes, coumarin, beliotropin, to name only a few more. But the evaluation of other synthetics is, in many cases, a matter of personal preference. I have quoted the example of amylcinnamic aidehyde, and have already mentioned the different types of peach aldehyde and coconut aldehyde as well as the ionones and methyl ionones containing vary- ing proportions of the different isomers. As an additional example take the case of vanillin. Until not so many years ago vanillin was only acceptable when it had a faint yellow colour in

SOME ASPECTS OF PERFUMERY CHEMICALS 43 fact, it was generally accepted that pure vanillin was a faintly yellow product. Later, the intricate chemistry of phenolic aldehydes became better known, and with this knowledge, and with the help of modem techniques and plant made from entirely new materials of construction, the manu- facturers were in the position of preparing vanillin in its pure, white form. But it took quite a number of years before flavour chemists and perfumers became used to white vanillin and realised that its odour was not inferior, but at least equal to, if not better than, that of the yellow material. Well then, the question which aromatics should be prepared in a pure state and which should be left to contain certain impurities can only be answered as follows: . . . purify as far as possible those products which the perfumer requires in the highest degree of purity do not attempt to separate reaction mixtures which are valued by the perfumer on account of the very fact that they are mixtures. But in actual fact the answer is not as simple as that: remember per- fumers are artists and, as I have said, the evaluation of certain products is a matter of personal preference. What one perfumer regards as a good quality aromatic because of the absence of any foreign note, another perfumer may regard as poor because of this very absence. What one perfumer regards as good ionone because of an excess of beta ionone, another may regard as poor because of this excess. This makes the industry of aromatics so different from the industry of other types of fine chemicals. Take, for example, pharmaceutical chemicals. The purer they are the higher they are valued, and their quality is determined by analysis alone. Not so in the industry of aromatics. It frequently happens that one customer will be delighted to have a certain quality of a certain product, and another customer will threaten to break off diplomatic relations if such a poor quality product is supplied to him again. What, then, is the manufacturer to do in order to satisfy all his customers?' The answer is that he has to offer more than one quality of one given aromatic. This has, of course, quite another aspect, too, which I have not mentioned so far, namely, the economic aspect. As you are well aware, there are many methods available for the purifica- tion of chemicals. A liquid product may be converted to a solid derivative which is repeatedly purified by crystallisation to remove all foreign materials. The liquid product is regenerated from the solid derivative, fractionated in vacuum, and possibly redistilled. A solid product may have to be crystallised from a number of solvents. Liquid-liquid extraction and a variety of other methods could be enumerated. But all these methods are costly, so that the further a product ha•, to be purified, the more expensive it becomes. I did say that the perfumer does not always consider the chemically pure material to be the most suitable for his purpose. And so quite often it is.