1254
INDUSTRIAL A N D EATGI,VEERISG CHEMISTRY
Vol. 18, Yo. 12
Fatty Oils as Substitutes for Ethyl Alcohol in Citrus Flavors1 By H. A. Schuette and B. P. Domogalla UNIVZRSITY O F WISCOISIK,~ I A D I S O N WIS. ,
In a study of substitutes for ethyl alcohol in the preparation of orange and lemon flavors, data have been obtained on the use of fatty oils which have considerable analytical importance. For the assay of such solutions with a saccharimeter it is recommended that a 100-mm. tube be used and that the appropriate factors for lemon and orange oils be 1.7 and 2.7, respectively, or at least values of that order of magnitude. The use of a preservative (ethyl alcohol) having been eliminated, it became a matter of more than passing interest to determine the stability of such fatty oil solutions. Tests were made on flavors in which olive, corn,
peanut, and cottonseed oils were used as vehicles. Acid numbers of such solutions showed in 54 months an increase from 25 to 55 per cent for the lemon flavors, and 19 to practically 100 per cent for the orange flavors. Rancidity had developed to such a degree, even after 14 months, as to mitigate against the commercial use of such flavors, although reports on initial baking tests-the only field where such flavors appear to have any use-were on the whole satisfactory. With age and increasing acidity there is a reduction in the optical rotation to the end that all such flavor solutions show a content of essential oil lower than when originally made.
EW problems in the manufacture and sale of flavoring extracts have arisen since the adoption of the Eighteenth Amendment. This is particularly true of the citrus extractsJ2orange and lemon, which are alcoholic solutions of the essential oil in question. While it is probably true that a proper construction of the enforcement acts of most states places no prohibition upon the sale for legitimate purposes of flavoring extracts made with alcohol, yet manufacturers have appreciated the wisdom of making a search for a solvent which will satisfactorily replace ethyl alcohol. Illoreover, such a n investigation has an academic interest. The various factors which determine the value of a solvent for the industry have been discussed a t length by De Groote3 and by hIass.4 Briefly summarized, such a solvent must be odorless, tasteless, colorless, harmless, and entirely acceptable from a bromatological standpoint; and must possess in addition those properties by virtue of which a flavoring extract may be prepared which is brilliant in appearance, is freely miscible in all proportions with water or sirup, diffuses readily through the food, possesses body and strength, has stability until entirely consumed and a permanence a t oven temperatures without undue loss of flavor or transformation foreign to its original character, shows no tendency to become rancid with age, and is not appreciably affected by ordinary changes in temperature. For the accomplishment of these ends there have been suggested the acetic acid esters of glycerolJ5t,he higher alcohols and more particularly the isopropyl form;6 glycolsJ7of which the ethylene and propylene types are representatives; esters of ethyl alcohol;8 and the edible fatty 0ils.g The essential oils have also been emulsified3 with a gum, such as arabic or tragacanth, and the emulsion in turn diluted with a sirup. Mono- and diacetin have but a limited solvent action,l0
if any, upon orange or lemon oils, and triacetin falls into a class with the fatty oils. The higher aliphatic alcohols, with the possible exception of isopropanol, are unsuitable on account of toxicity or taste. More information must be gathered with respect to the physiological action and toxicity of isopropanol before it may be used as a solvent for essential oils that are intended for alimentary purposes.ll The esters of ethyl alcohol possess characteristic odors which mask that of the essential oil. Emulsified essential oils offer a partial solution of the problem, but their tendency to separation and decomposition makes their use an economic loss. The fatty oils as a class are solvents for the citrus oils, at least in the concentration in which these oils are used in alcoholic solution. I n this instance the solvent or vehicle, which is neither odorless nor tasteless, has a recognized food value. However, the odor or taste of a prime fatty oil is hardly such as t o condemn a flavor for that purpose. Vegetable and animal oils so employed become rancid with age, concerning which quantitative data will be submitted. The field of usefulness of such mixtures is restricted, the only place where they find application being in the baking industry. To study the effect of solvent upon the rotation of lemon and orange oil solutions when fatty oils replace ethyl alcohol; to investigate over a long period of time the desirability of using such solutions as food flavors from the standpoint of stability, since the use of alcohol as preservative has been dispensed with; and to note what effect, if any, age has upon the analytical constants originally determined, is the threefold object which prompted the investigation the results of m-hich are herein summarized.
.. .. .. .. .. ..
N
Presented before the Industrial and Sanitary Chemistry Group, Midwest Regional Meeting of the American Chemical Society, Madison, Wis., M a y 27 t o 29, 1926. 2 I n this paper the word “extract” refers to the solution in ethyl alcohol of the essential oil, while for solutions in any other solvent the word “flavor” has been reserved. 8 A m . Perfumer, 15, 5 5 (1920). 4 Tea Coj’ee T r a d e J., 40, 484 (1920). 5 Esselen, U. S . Patent 1,378.099 (June 5 , 1921). 6 Smith and Eoff, U. S. Patent 1,384,680 (July 12, 1921). 7 Z b i d . . U. S. Patent 1,364,681 (July 12, 1921). 8 Paul, J. Assoc. OficiaZ A g r . Chem., 4, 468 (1921). B Thurston, Midland Drrrggzst, 63, 68 (1910). 10 Boyles, Tea Coi7ee T r a d e J., 45, 424 (1923). 1
Materials
CITRCSOILS-Eight samples each of lemon and orange oils of Californian and European origin were selected as fairly representative of the product from which commercial extracts are made. The lemon oils were found on analysis to contain from 4 to 6.25 per cent citral and to possess the following constants: [a]’: +59.96 to +61.20; n25 1.4725 to 1.4744; and specific gravity 0.8543 to 0.9548 at 25’/4O c. Similarly for the orange oils, [a]’,” varied between +94.85 and S95.15; nzswas 1.4710 to 1.4727;and the specific gravities ranged from 0.8432 to 0.8589 a t 2 5 ’ / 4 O C. 11 The published literature on this subject, which is meager, has been brought together b y Grant, J. L a b . Clin. M e d . , 8 , 362 (1923).
December, 1926
INDUSTRIAL AND ENGINEERING CHEMISTRY
FATTY OILS-& solvents for the citrus oils in the preparation of the corresponding flavors there were used sweet almond, corn, cottonseed, neutral lard, olive, peanut, rape, sesame, and soy bean oils. All were "prime"--that is, of minimum acidity. 9 determination of their physical and chemical constants indicated that they were truly named. The corn, olive, cottonseed, and peanut oils were obtained from grocers' stocks while pharmaceutical supply houses furnished the otherf . The soy bean and rape oils were refined in the laboratory by a method of Puscher.I2 Clear, light yellon, and practically odorless products were obtained by treating 500gram portions of the oils twice with 10 grams of a mixture consisting of equal parts of 95 per cent ethyl alcohol and sulfuric acid. After standing 24 hours, the oil was separated from the acid and the resulting deposit of impurities. Removal of the alcohol was followed by neutralization of the acid and drying. All oils were tested for optical activity. On11 two of this group gave positive results, which is contrary to the reports of Peter13 and of Rakusin,'4 who state that almond, cottonseed, peanut, and rape oils are optically active. When viewed in a 100-mm. saccharimeter tube the olive oil gave a rotation of +0.65' V. a t 20" C. For the sesame oil under similar conditions a value of +3.3" V. was found. CITRUSFLhvoRs--The conditions described by the definitions of lemon and orange extracts were simulated. Five per cent solutions by volume of the essential oil in the respective fatty oil were made. The variety of the former a t hand and the number of different solvents selected as vehicles admitted of the preparation of seventy-two solutions of each essential oil, a number believed to be large enough to give representative average data. None of such solutions possessed the marked odor of the particular essential oil used, as do the time-honored extracts. This is to be expected in view of the differences in vapor pressures of the solvents. Differences in viscosity contributed to the lack of brilliancy of the flavors in contrast to that of the extracts. The effect of temperature changes upon these fatty oil solutions of the citrus oils was observed under conditions which it is reasonable to assume would obtain in actual practice. To that end samples in the conventional 2-ounce panel bottles of the trade were kept under thermostatically controlled conditions for 10 hours a t 0" C., then for 24 hours a t 15.5' C., and finally for a like period a t 25" C. The peanut oil and lard oil solutions alone of the series studied responded to temperature changes. At 0' C. the peanut oil solutions partially solidified with a separation of glycerides, whereas the effect upon the lard oil solutions was complete. At 15.5" C. all solutions remained clear except again those containing lard oil, which now had only partially solidified, a condition which did not prevail a t 25" C. Baking tests were very favorable. Representative samples of both flavors were given to several experienced pastry cooks with the request that they use them in cakes of their own choosing but in the same amount as they would have used of a standard extract. The finished cakes were in most instances sampled by noninterested observers, The consensus of opinion mas that these flavors, which had been prepared with a fatty oil base, were just as good, if not better, than the alcoholic extracts. Data of this kind are necessarily subject to the variations in the personal equation. They were collected under the widest latitude of conditions lacking means of exact measurement. Yet, it is interesting 11 Brannt, "A Practical Treatise on Animal and Vegetable Oils," 1896, Vol I, p:480 Baird & Co , Philadelphia 18 Chem - 2 l g . 11, Rep. 267 (1887). 1 4 J Russ. Phys.-Chem. Soc , 37, 442; Chem Zenlr , 76, 11, 623 (1905).
1255
to observe that the reports on the distinctness of the flavor were rather uniform. Thurston'j had made a similar observation in that he stated that the fixed oils hold the aroma and flavor of the volatile oils much better than alcohol does, but that flavors of this type in comparison to the alcoholic extracts do not hare an odor proportionate to their strength. Experimental Procedure
The analyst is indebted to Mitchellle for his systematic study of lemon extracts and the method for determining oil of lemon, He observed that a 1 per cent solution of oil of lemon in ethyl alcohol produced a rotation of 3.4" V. a t 20' C. when viewed through a column 200 mm. long. Under similar conditions oil of orange gives a reading of 5.3' V. Subsequent investigation has indicated that results which are in closer agreement with theory will be obtained if these factors are reduced to 3.2 and 5.2, re~pective1y.l~ Any attempt at the application of Mitchell's polariscopic method to the assay of fatty oil solutions of the citrus oils must take into account the optical activity of solvents of this type. It is true that this property is not a prominent one of the fatty oils, yet some are known to possess it to a minor degree. However, it has a diagnostic value in that it admits of a division of the fatty oils into two groups'*-viz., those whose optical activity is due to the presence of sterols, and in special cases to sesamin or certain aliphatic alcohols; and those, as for example, the members of the castor oil and chaulmoogra groups, whose fatty acids themselves contain 60
I
c 13
20
20
40
50
60
Time in month8
Figure 1-Effect
of Age upon Acidity of L e m o n Oil in F a t t y Oil S o l u t i o n
asymmetric carbon atoms. Bishop,Ig Peter,I3 Thoerner,*O Crossley and Le Sueur,21 Utz,22Sprinkmeyer and Wagner,23 and RakusinL4have made valuable contributions in this direction and have brought together much of the literature on the subject. The adoption in toto of Mitchell's method to the determination of the essential oil content of the flavor solutions in question was not feasible because of the color and opacity of the latter. This difficulty could only be overcome by the use of a 100-mm. observation tube, the adoption of which became the st,andard procedure in this investigation. A m , Pevfumer, 14, 46 (1913). A m . Chem. Soc., 21, 1132 (1899). 17 Assoc. Official Agt. Chem., Methods, 2nd ed., 1925, p. 352. 1s Lewkowitsch, "Chemical Technology and Analysis of Oils, F a t s and Waxes," 6th ed., 1921, Vol. I . p. 352. hfacmillan & Co., London. 19 J . pharm. chim., 16, 300 (1887). 2 0 J . Sac. Ckem. Ind., 14, 42 (189:). 2 1 I h i d . , 17, 992 (1898). 2 2 Phnrm. Zffi., 45, 490 (1900). p a Z . Nakr. Genussm., 10, 347 (1905). 16
1 S J .
INDUSTRIAL A N D ENGINEERING CHEMISTRY
1256
Table I-Polarization of Five Per c e n t Solutions of L e m o n a n d Orange Oils in Fatty Oils -LEMON OIL-ORANGE OILDeviation Deviation -Rotationfrom -Rotationfrom Obsd. Calcd. theor’al Obsd. Calcd. theor ‘a1 100 mm. 1% oil 100 mm. 1% oil Fatty oil 2!’ C. soln. Found content 20’ C. toln. Found content solvent V. V. % % V. V. ?& % Almond 8 . 4 4 1 . 6 8 8 4 ; 9 6 -0:04 1 3 . 7 0 2 . 7 4 0 5:07 +0:07 Corn 8 . 2 6 1.652 4 . 8 6 - 0 . 1 4 13.40 2.680 4 . 9 6 -0.04 Cottonseed8.46 1 . 6 9 2 4 . 9 7 - 0 . 0 3 13.60 2.720 5 . 0 3 +0.03 Lard oil 8 . 4 2 1.684 4 . 9 5 - 0 . 0 5 1 3 . 4 8 2 . 6 9 6 4 . 9 9 -0.01 Olive 8.07 1.614 4 . 7 5 -0.25 13.04 2.608 4 . 8 3 -0.17 Peanut 8 . 4 7 1 . 6 9 4 4 . 9 8 -0.02 13.69 2.738 5 . 0 7 $0.07 Rape 8 . 5 4 1 . 7 0 8 5 . 0 2 +0.02 13.70 2.740 5 . 0 7 $0.07 Sesame 8 . 5 6 1.712 5 . 0 3 + 0 . 0 3 13.55 2.710 5 . 0 2 +0.02 Soybean 8 . 5 8 1 . 7 1 6 5 . 0 4 + 0 . 0 4 13.71 2.741 5 . 0 8 +0.05 Average 1.7 10.07 2.7 *0.06
The results obtained under these conditions are given in Table I. The observed readings are the averages of those of eight solutions of the essential oil in the fatty oil, corrected by a blank determination on the solvent, assuming that the effect of solvent upon the rotation is additive. The per rent by volume of the essential oil found on analysis is the quotient obtained by dividing the respective observed reading by the average rotation produced by a fatty oil containing in solution 1 per cent of the latter.
Vol. 18, No. 12
commercial standpoint to their ability to preserve the prime condition which was theirs when made. Acidity is an index of such value, not of itself alone, but rather because of the concomitant changes which take place when it develops to the condition of rancidity. Rancid fatty oils have a pronounced “off” odor and taste which is penetrating and pervades other food materials into which they have been incorporated. Olfactory and organoleptic tests made of such flavors over a period of 54 months confirm the analytical data given in Table 11. A liberal estimate in the light of these data places the shelf life of such flavors a t less than 12 months. Table 11-Effect
of Age u p o n t h e Acidity of L e m o n a n d Orange Oils i n Fatty Oil Solutions Initial acid number
--Per cent Increase in Acidity-5 14 54 months months months
Lemon Oil Olive oil Corn oil Peanut oil Cottonseed oil
0.59 0.39 0.43 0.09
Olive oil Corn oil Peanut oil Cottonseed oil
0.87 0.31 0.41 0.08
3.74 5.16 2.80 2.20 Orange Oil 2.76 6.23 4.16 9.58
11.88 21.91 8.39 5,56
25.46 55.41 23.31 54.44
9.19 24.92 9.78 19.75
19.52 64.59 33.98 97.53
TOcarry out these tests composite samples of the flavors in corn, cottonseed, olive, and peanut oils were set aside on the laboratory shelf, where they were fully exposed to the light and were subject t o all temperature changes. At irregular intervals during the test period, acid numbers were determined by the method of the Association of Official Agricultural Chemi~ts.~4These data are graphically represented in Figures 1 and 2. The sharp upward trend of the curves is significant. Effect of Age upon Optical Activity
0
10
20
Figure 2-Effect
SO 40 Time In months
50
6C
of Age u p o n Acidity of Ornnge Oil in Fatty Oil Solution
N o useful purpose obtains in attempting to express these factors in any manner except in one which states their order of magnitude. The application of any factor whose value is expressed with such nicety as to demand the third decimal place seems quite out of keeping with the fact that several inherent variables must be considered in flavor solutions of this type. Such variables, controllable only to the extent that choice of solvent may be a personal one, are the natural variations in percentage composition of the essential oil and of the fatty oil serving as solvent, and the nature of the latter itself. The small mean deviation for each flavor solution seems to justify the tentative acceptance of these factors. These factors suffer no change if sweet almond, rape, and soy bean oils are omitted from the tabulations, because they do not have the prominence in the American diet that the others have and therefore would probably find no place in a manufacturer’s list of edible oil solvents.
Effect of Age upon Stability
Whatever of value solutions of the citrus oils in fatty oils may possess as food flavors is second in importance from a
From the analytical side some interest attaches to the effect of time upon the rotation recorded when the solvent was fresh. A redetermination of the optical activity of samples of four solutions, part of which had served for the acidity studies, indicated an apparent loss of essential oil. The average indicated loss of lemon oil, using the factor 1.7, was 5 per cent; that of the orange oil, calculated with the factor 2.7, 9 per cent. The causes for this loss are somewhat speculative. The answer probably lies in the knowledge that chemical changes have taken place in the solvent with the result that the latter had become very rancid. It is believed that these are the contributing, if not the sole, causes. Table 111-Effect
Solvent Olive oil: Initial After 54 months Corn oil: Initial After 54 months Peanut oil: Initial After 54 months Cottonseed oil: Initial After 54 months
of Age u on Rotation of L e m o n a n d Orange Oils in Patty oil solution -LEMON OIL----ORANGE OIL--Rotation at Essential Rotation at Essential Z O O C. oil content 20‘ C. oil content v. Per cent v. Per cent 8.07 7.75
4.75 4.56
13.04 12,20
4.53 4.53
8.26 7.50
4.86 4.58
13.40 12.00
4.96 4.44
5.47 7.95
4.98 4.67
13.69 12.25
5.07 4.53
8.46
4.97 4.73
13.60 12.40
5.03 4.56
8.05
The results of these tests are given in Table 111. Like those of the preceding table, their value is merely indicative of an inevitable condition which will exist when fatty oils are t o be used as substitutes for ethyl alcohol in citrus flavors. 21
A s s o c O f i c i a l 4 g u CFem , hlethods, 2nd ed , 1926, p 293.
I S D C S T R I A L A S D ENGINEERING C H E M I S T R Y
December, 1926
Acknowledgment
1257
flavors to baking tests, and to the Exchange Lemon Products
Cellulose Xanthate’ By G. W. Blanco Dci PONTRAYONCo., BUFFALO, N. Y.
HE ever increasing popular approral and versatile uses of rayon here and abroad make it a timely subject for discussion. Approximately 160 million pounds of rayon are produced annually, 90 per cent of which is made by the viscose process. Viscose has commanded the attention of numerous investigators for the last quarter of a century and there is today a healthful amount of research contribution2 and patents bearing on miscellaneous phases of this important cellulose derivative. The field is one of complicated chemistry, and it is the purpose now to call attention to some of the salient points in the preparation and properties of cellulose xanthate without necessarily expanding on or adding to our present knowledge of the reactions involved.
T
GENERAL-It seems desirable first to review briefly some of the properties of carbon disulfide, comparing them with those exhibited by carbon dioxide. With caustic the former produces thiocarbonates, which in the presence of water decompose into sodium carbonate and hydrogen sulfide. Thiocarbonic acids are related to carbonic acid as indicated below:
c=o
\OH Carbonic /OH C=S \OH Sulfocarbonic
,/OH C==O ‘\SH Thiocarbonic
/SH
c-0
\SH Dithiocarbonic
/SH C-S \SH Trit.hiocarbonic
/OH
c=;s
\.SH Sulfothiocarbonic
These acids are very unstable. Trithiocarbonic acid decomposes into hydrogen sulfide and carbon disulfide. It may be precipitated by hydrochloric acid as a reddish brown, oily liquid from solutions of its alkali salts, which are the products of interaction between carbon disulfide and alkali ~ u l f i d e s . ~Secondary esters are formed by the action of alkyl halides (iodide) on the sodium s a l t - e . g., /SNa
c==s
\SNa
+
I\ C*HsI,CzHj
+
/SCzHt
- c=s
\SCsHi
A similar reaction starting with sodium ethyl xanthate proves the position of sodium in the xanthate ent’ity. Yanthogenic or sulfothiocarbonic acid represents the parent substance from which xanthates and their derivatives may be prepared. Ethyl xanthic acid, or the ester of dithiocarboxglic acid, is a heavy liquid insoluble in water and decomposing into alcohol and carbon disulfide a t 25’ C. It is both an ester Presented before the Midwest Regional Meeting a n d the Meeting of the Division of Cellulose Chemistry of the American Chemical Society. Madison, Wis., M a y 27 t o 29, 1926. * Margosches, “Die Viskose,” 1906. 3 Richter, “Organic Chemistry,” 3rd ed.. p 433.
+
+
The thiocarbonic acids resemble carbamic acid and its derivatives in general constitution, as will be brought out later.
?:/
/NHz
/OH c=o
Preparation of Xanthates
/OH
and an acid4 and may be prepared by shaking sodium alcoholate with carbon disulfide and treating the mixture with sulfuric acid, thus: /OGH6 CS2 C2H50Na = C=S \SNa /OCzHs /OC& C=S HzSOl = C=S NaHSOa \SNa ‘SH
c=o
C/ O4H
Carbamic acid
Sulfocarbamic, Dithiocarbamic thiocarbamic, or acid xanthogenamic acid /OCzHa /OCZH~
/NHz
c=o
c-0
\NH,
c=s
\NH,
Urea
Urethane
METALLIC XANTHATES-The /OR formula C--C
\SH
\NHz
\SH Thiocarbamic acid
\NH%
Thiourethane
xanthates having the general
were first discovered by Zeise in 1824. They
\SMe
are generally prepared by shaking an alcoholate with carbon disulfide, e. g., /OCzHs C& KOH CsH60H = C=S +H20
+
+
\SK
Potassium ethyl xanthate consists of yellow silky needles. Sodium ethyl xanthate is prepared in a similar manner. CELLULOSE XANTHATE-The first patent on CellUlOSe xanthate was taken out by Cross, Bevan, and Beadle in 1892.5 Cellulose treated with 15 per cent sodium hydroxide liquor and then pressed was treated with carbon disulfide to the extent of 30 to 40 per cent of the material used, in a closed vessel for 3 to 4 hours. The raw xanthate solution was subsequently purified by organic acids, sulfur dioxide, or salt solution. Cellulose xanthate is prepared by treating suitably cellulose material with the required amount or excess of caustic solution of given concentration at definite temperature for a specified length of time and afterwards removing the excess caustic. The alkali cellulose formed is disintegrated and allowed to mature for determined periods a t a constant temperature. It is then treated with carbon disulfide in a closed vessel observing predetermined optimum conditions. 4 Heuser, West, and Esselen. “Textbook of Cellulose Chemistry,” 1924, p. 64. 6 British Patent 8700 (18921
