Sept., 1916
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
per kg., it follows t h a t 5 5 4 - 2 2 9 = 3 2 j mg. per kg. must he in a n insoluble form. As stated above, this amount must he well within the truth. If a n y considerable amount of t h e tin of the liquor is insoluble, t h e insoluble tin of t h e drained solids must he considerably more t h a n 3 2 j mg. per kg. The remaining figures in t h e column marked “Insoluble t i n in drained solids” were all calculated in the same manner, using average figures for t h e weight and water content of drained solids and liquor as t h e individual determinations were not made. The column headed “Insoluble tin in drained solids” is of particular interest although, as stated above, it is believed t h a t t h e figures in this column are somewhat less than t h e t r u t h ; t h e relatiye amount reported as insoluble tin in samples of different ages shows t h a t the process of hydrolyzation is a continuous one. The figures given in t h e last column headed “Soluble tin in total sample” were obtained by adding together the figures in t h e columns headed “Milligrams t i n per kilogram liquor” and “Soluble tin in drained solids,” after calculating these respective figures t o milligrams of tin per kilogram of t i n of original sample. I t is interesting t o note t h a t t h e figures in this column are almost identical with t h e figures in t h e column headed “Milligrams tin per kilogram liquor.” I t is probable, therefore, t h a t a n approximate idea of the amount of soluble tin in a sample of canned food can be obtained by determining t h e amount of t i n in t h e liquor of t h e food, although t h e figure so obtained will obviously include any insoluble t i n which may be present in t h e liquor in a finely divided state a n d also any tin t h a t may be present in colloidal form. Recent workers agree t h a t t h e ideas of t h e earlier writers on t h e toxicity of t i n were erroneous. I t is now known t h a t t h e toxicity of soluble tin compounds is a t least very much less than it was formerly supposed t o be. I t is evident, however, t h a t t h e results obtained in t h e study of soluble t i n salts cannot be used a s a criterion on which t o judge t h e toxicity of tin in canned foods. Whatever t h e insoluble combination in which t i n occurs in canned foods it is in all probability less likely t o be absorbed from the intestinal tract t h a n soluble t i n compounds. The same is true, perhaps t o a less extent, of t i n in colloidal form. A t any rate, t h e need of experimental work on the toxicity of tin a s it occurs in canned foods is obvious, NA~ONA CANNHRS’ L ASSOE~ATION.WASHINGTON
THB EFFECT OF CURING ON THE AROMATIC CONSTITUENTS OF VANILLA BEANS1 B,.
FRANK
me*r
Received March 13. 1916
The fruit of vanilla has long been used a s a flavoring agent. The plant is native t o Mexico, where it is a t present extensively cultivated for t h e production of t h e vanilla beans of commerce. It has been introduced into a number of other tropical and subtropical countries where it is also profitably cultivated. Among the other sources are South America (Guadaloupe), I
Published by permission of the Secretary of Agriculture.
815
Tahiti, Reunion, Madagascar, Comores, Seychelles, and Mauritius. Vanilla beans as found on t h e market are described by t h e United States Pharmacopoeia’ a s t h e “cured, full grown, but immature, fruit of Vanilla planifolia, Andrews. (Family orchidaceae.)” Vanilla is used for flavoring purposes in the form of t h e so-called extract or tincture of vanilla, which is prepared by extracting the coarsely comminuted beans with a hydroalcoholic menstruum varying in strength from 4 j t o 6 j per cent. The aroma of vanilla, t o which t h e flavor is attributed, does not preexist in the beans but is formed by a chemical reaction induced during the drying or curing process t o which t h e beans are subjected after harvesting. Behrens2 states t h a t ripe vanilla beans have little or no odor but become strong in odor by curing,
C O Y I B R C I A L VANILLABBANS
due t o the splitting up of a n existing glucoside with the formation of t h e compound vanillin. The above statement regarding t h e odor of vanilla was confirmed by Busse: who states t h a t vanillin is formed from a nonodorous body partaking of t h e nature of a glucoside. In a n extensive research with vanilla, Lecomte‘ found t h a t the fruit of vanilla contained, besides the glucoside coniferin, two ferments, one an oxidase and t h e other a hydrolyzing agent. The latter conUnited s t a t ~ Pharmacopoeia. ~ Eishfh Decennial Revision. 1900, +97. “Ueber dap Vorkommen des Vanillins in der vmille.” Der Tiomzpllonarr, 3 (1899). 299. 1 W. Bur& “ V a n i l l c C h e m i e der vanille Prueht. Arbeiten ~ U Bdem Kaiserliehen Gcsundheitnarnte,” Berlin. XV. 1898 bir 1899, p . 101. a H. Lccomte. ‘,Sur la formation du parfurn de la vanille.” Cornpi. vend.. I33 (1901). 745. %
e J. Behren.,
816
T H E J O U R N A L OF I N D U S T R I A L .AND E N G I N E E R I N G C H E M I S T R Y
verts t h e glucoside coniferin into coniferyl alcohol and glucose, t h e former oxidizing this alcohol into vanillin. Later investigations have shown t h a t t h e agreeable aroma of vanilla is not due alone t o t h e compound vanillin b u t is strongly influenced b y other substances of entirely different character. Walbauml states t h a t t h e value of vanilla beans does not depend solely on t h e vanillin content. T h e oleoresinous constituents of t h e beans were examined and found t o contain anisaldehyde a n d anisic alcohol besides other compounds which could not be identified, b u t which possessed a characteristic agreeable odor. The presence of other odorous constituents in vanilla beans is also mentioned b y Iserman; who states t h a t
Vol. 8, No. 9
uents which are known t o impart much of the pleasant flavor t o the cured beans. CURING OF B E A S S
As has been previously stated, the flavor of vanilla beans is brought about b y a curing process with a consequent development of the true flavor upon which t h e efficacy of t h e beans as a flavoring material depends. The curing process varies somewhat according t o t h e source of t h e beans, t h e prime object, namely, t h e evaporation of t h e moisture and the formation of t h e flavoring constituents, being attained in every case. This is accomplished in a general way by a process of heating a n d subsequent sweating, t h e details of which vary in t h e different producing countries. Two general methods are in vogue ior the curing
LABOPATOILY
C u s a o VANILLA BEANS
t h e special aroma and taste is probably due t o vanilla resins. Bussea states t h a t vanillin is present in the living fruits only in insignificant quantities a n d is mainly produced in the process of curing. Vanillin is not the sole criterion of quality and does not represent t h e natural aroma of t h e fruit. Besides vanillin there is present in t h e beans vanillic acid, resin, oil, sugar, mucilage, and tannin, Up t o t h e present time attention has been largely directed towards the vanillin content of the beans, only bare mention being made of the resinous constit3 H. Walhaum. "Dar Vorkommen "on Anisalkohol und Anisaldehyde in der Pruchfen der Tahiti-vanille. Festschrift. Otto Wallaeh. Zur Erinnerung an seine Forschungen auf dem Gebiete der Terpenc in den Jahrcn 188+1909." p . W Y . 2 S. I~crman. "Vanilla Bcanr: Chemistry and Character of Conrtituents." Western Druggist. July, 1910,358-362. H. Burrr. "Studien ueber die Vanille," Apolh. Zeii., 18 (1898). 894.
of vanilla beans, dry and t h e ,t water me1 ,ds. T h e d r y method, in use in Mexico, consists essentially of placing t h e freshly gathered beans in the sun for several hours, after which they are wrapped in woolen cloths t o promote sweating, a n d the process. of heating and sweating is continued for several days until most of t h e moisture has evaporated. T h e final drying is usually completed in t h e shade. I n unfavorable weather a drying oven is employed, in which the beans are sweated at a temperature of 60' C. The curing process is continued until the beans are soft a n d supple and have acquired a chocolate-brown color, the complete length of time required being frequently a s long as 5 mos. T h e hot water method, used in Reunion, consists in placing t h e beans in hot water (90 to I O O O C.) ~
Sept., 1916
T H E JOCRLVAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
for I O t o 30 secs., after which they are wrapped in woolen blankets or placed in a box lined with blankets, where t h e sweating process takes place. Subsequently the beans are exposed t o t h e sun a n d later t h e curing is completed in a drying room until t h e finished beans have t h e proper color a n d texture. This process requires from I t o z mos. for completion. The curing process is difficult a n d precarious, calling for a high degree of skill and experience. Weather conditions also have much t o do with t h e rapidity a n d success of t h e process. T h e exact details Of t h e two methods differ greatly according t o locality. The general method of procedure for both processes is considerably modified in t h e minor details, t h e one object being t o produce finished beans which will have the maximum amount of flavor of superior quality a n d at t h e same time will have had removed
curing vanilla beans were made. I t was also desired t o ascertain whether t h e long-continued curing as now practiced is absolutely necessary for t h e production of beans of commerce and whether shorter time of curing noticeably affects t h e percentage of vanillin in t h e cured beans. A comparison of t h e efficacy of t h e two general methods of curing, namely, t h e dry method and the hot water method, and t h e effect of temperature upon t h e curing was sought. Furthermore, it was desired t o study t h e effect of t h e various curing methods upon the formation of the oleoresins of vanilla which are present in all cured beans and which agreeably modify t h e flavor of t h e beans. LAB 0 RAT 0 R Y C U R I N G E X P E R I M E N T S
X small
quantity of green beans was obtained from
TABLE I-TREATMENT, Loss IN WEIGHTA N D CURINGOF VANILLABEANSAT ROOMTEMPERATURE A N D IN THE WATEROVEN Immersion in water three successive times a t temperature given, for periods indicated, a t intervals noted IN WATEROVEN (40-60' C.) BEANS CURED BEANSCURED A T ROOMTEMPERATURE Per cent Curing Per cenl Curing Period Loss in Loss in Period REMARKS Days REMARKS No. Weight Days No. Weight TREATMENT S Beans became uniformly brown in 6 days, 14 Curing progressed slowly for 7 days 2 32.5 None 1 53.0 33 sweating continuous until 8th day when 67.0 24 after which color developed rapidly to 66.4 beans were wrapped in waxed paper and uniform brown: beans well cured in 14 allowed to dry until soft and pliable. days. At end of this time, however, Beans possessed characteristic but mild beans were wrapped in wax paper and allowed to dry. Cured beans were aroma. dark brown in color, soft, pliable and possessed fine aroma. Temp., 40' C Beans all brownish color in 5 days with good 7 39.3 14 Curing proceeded about same as No. 1. 8 49.6 8 After subjecting to sweating process 28 odor developing. After 8 days' sweating in 68.5 46 66.6 for 14 days beans were uniformly Period, 30 sec. water oven beans were wrapped in waxed brown in color. Beans wrapped in Interval. 1 min. paper and allowed to dry slowly until proper waxed paper and allowed to dry until texture was attained. Odor was fine they presented a wrinkled appearance vanilla-like characteristic hut not especially and proper pliability. Beans possessed strong. very fine aroma. Temp., 60' C. 11 Beans became yellowish green after 5 36.5 6 41.4 8 After 5 days beans were uniformly brown in Period, 20 sec. 33 third day. After 11 days all were 66.0 66.2 color with pronounced odor and apparently 33 Interval, 1 min. uniformly brown in color and had curing rapidly. After 8 days' sweating characteristic wrinkled appearance beans were treated as above and dried Beans wrapped in waxed paper and to proper pliability. Odor was strong, allowed t o dry t o proper flexibility. vanilla-like. Beans characterized by strong and exceptionally fine odor. Temp., 80' C. 3 40.7 12 Beans assumed brown color and de4 31.2 8 Uniformly brown after 4 days' sweating, Period, 10 sec. 66.2 24 veloped odor more rapidly than any of 64.0 29 . curing more rapidly than any of above and Interval, '/z min. above. After 12 days' sweating beans becoming soft and pliable. After sweating wrapped in waxed paper and set 8 days beans treated as above. Final aside to dry. After 12 days of drying appearance excellent both as regards color beans possessed a most exquisite and pliability. Odor strong. aroma, deep brown color and fine texture, Temp., 90' C. 9 23.3 10 Beans were all pliable, wrinkled and 10 30.0 3 Beans possessed a uniform dark brown color Period, 10 sec. 66.1 28 brown in 10 days with odor fairly 25 in 2 days with prominent odor developing. 66.3 Interval, 1/1 min. strongly developed. Allowed to dry After 3 days' sweating beans were exposed in air for 18 days. Beans pliable and t o air and dried until proper degree of wrinkled with only faint odor and pliability was reached. Strong but unDale brown color pleasant odor noted. Temp., 100' C. 11 35.2 10 Be'ans- became light brown and soft on 12 44.0 3 Beans uniformly brown but pale after 1 day's 64.0 Period, 10 sec. 18 third day. Taken from towel on 10th 69.0 17 sweating. Odor only fair and color pale Interval, '/a min day and allowed to dry in air for 8 brown after 3 days. Beans allowed to dry days. Odor prominent but not sufficiently, the cured beans being very strong. Color pale brown. light in color with faint odor.
from them sufficient moisture t o prevent molding when t h e beans are finally packed for t h e market. As a general ride from j t o 7 lbs. of t h e fresh beans are required t o produce I lb. of cured beans. I t is stated by Ellis' t h a t t h e temperature of t h e water into which t h e pods are dipped before curing has much t o do with t h e successful curing of vanilla. Treatment with hot water supposedly kills t h e pods and prevents dehiscence. I t is, however, essential t h a t t h e temperature be not sufficiently high t o destroy t h e action of t h e oxidase which is present in t h e pericarp. A destruction of t h e oxidase would seriously interfere with the formation of t h e odorous constituents of t h e cured beans. The object of this investigation was twofold. A number of modifications of t h e process now used in 1
F. R. Ellis, "Vanilla," The Chemist and Druggzst, 67 (1905). 593.
Mexico through a prominent manufacturer of vanilla extract. The small ends of the beans were very slightly yellow but were practically devoid of odor. The curing experiments were conducted with twelve lots of beans, one-half of which were cured a t room temperature and t h e other half in a water oven a t somewhat higher temperature. Each of t h e six lots cured a t room temperature were given different treatment prior t o the sweating and drying. The treatment consisted in dipping the green beans into water a t various temperatures up t o t h e boiling point. Each lot was kept wrapped in a soft towel and placed on t h e laboratory shelf t o undergo the sweating and drying process. The duplicate lots of beans cured in the water oven at a temperature of 40-j~' C. received exactly the same treatment as those cured at room temperature.
818
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
A description of t h e treatment, loss in weight, and time of curing of t h e green beans a t room temperature and in t h e water oven is given in Table I. The general treatment in the curing a t room temperature and in t h e water oven was similar t o t h a t employed in t h e commercial curing of the beans. The processes of sweating and drying were applied, however, under very different conditions t h a n those obtaining a t localities where the beans are cured commercially. During t h e sweating process the beans were wrapped in soft towels and later in waxed paper t o complete the drying. The treatment of t h e two sets of samples was identical in t h a t one sample in each set vias cured without previous treatment, while t h e remaining five samples were subjected t o the action of hot water a t increasing temperatures up t o t h e boiling point. The subsequent treatment of each sample was identical in every case. It will be noted t h a t t h e time of sweating or t h e length of time required to produce t h e brown color a n d t h e characteristic odor of t h e cured beans de-
Vol. 8, No. 9
waxed paper, placed in tight containers, a n d allowed t o stand for a period of 2 mos. After this time the samples were carefully examined and compared with regard t o average weight and length, also color, odor, and general appearance, with several commercial samples of cured beans, and the results tabulated, as shown in Table 11. The average weight and length of t h e laboratory cured beans compare very favorably with t h e commercially cured samples. I n color and odor several of t h e laboratory samples were superior t o the commercial samples. The presence of crystals of vanillin on t h e surface of the beans usually occurs when beans with high content of this constituent are kept in a confined space for a period of time. Crystals of vanillin were observed t o be most prevalent in t h e samples cured a t room temperature, a n entire absence of crystals being noted in t h e samples cured in the water oven with one exception, t h a t of Sample 2 , on which a few crystals were observed. A decided red-brown color was apparent in the cross section of t h e beans cured a t room tempera-
TABLE11-GENERAL COMPARISON O F LABORATORY CUREDA N D COMMERCIALLY CURED \'AXILLA BEANS CUREDBEANS Av. Av. DESCRIPTION OF Wt. Length BROWN CUT BEANS(CROSS SECTION) ODOR DESCRIPTION O F WHOLE BEANS COLOR SAXPLE G. Mm. Uniformly red-brown, inner portion Fragrant, vanilla-like Uniformly colored, soft, pliable. A few Chocolate 4 . 1 19 1 redder than surface small vanillin crystals on surface Reddish brown, inner portion redder Fine, fragrant Soft and pliable, wrinkled. No crystals 5 . 1 19.5 Pale 7 than surface Reddish brown. inner nortion redder Fine, mild Pliable and wrinkled. A few crystals on Chocolate 4 . 9 18 5 than surface surface Dark brown, somewhat lighter than Nearly covered with vanillin crystals Especially fine 5 . 1 19.5 Dark 3 surface Uniformly brown No crystals Light chocolate Very mild 4.4 19 9 Not uniformly colored. A few crystals a t Lighter but uniformly brown Lacking fineness 5 . 2 19.5 Reddish 11 ends Inner portion lighter brown than surA few crystals a t one end Good, strong 5 . 2 19.5 Dark 2 face. Charred odor Uniformly chocolate-brown TJniformly colored. No crystals Mild, fragrant Chocolate 5 . 1 20 8 Uniformly pale brown Prune-like ko crystals Pale 5 . 6 20 6 Inner portion paler than surface Good, but lacking bouquet No crystals Chocolate 4.7 19 4 Lighter in cross section Not uniformly colored. No crystals Very mild 4.8 18.5 Chocolate 10 Uniformly brown Not uniformly colored. No crystals Prune-like Pale 4 . 8 18 12 Much lighter brown than surface Pliable, sticky. KO crystals Strong, not agreeable Dark 5 . 1 23 Mex. A Odor somewhat tobacco-like Uniformly brown. Odor somewhat. Conspicuous pale, longitudinal stripes Strong 3 . 2 18 Dark Tahiti tobacco-like Inner portion reddish brown Not pliable. A few crystals Mild, sweet Very pale S. Amer. 3 . 9 20 Inner portion lighter brown than Very pliable. No crystals Pleasant Almost black Bourbon 2 . 7 17 surface Uniformly pale brown Nearly dry. N o crystals Very faint 2.5 16 Pale Old (a) ( a ) Beans several years old.
creases with t h e increase in temperature of t h e water in which t h e green beans were immersed. Indications, therefore, point t o increased activity of t h e oxidases which supposedly bring about t h e changes. Whether this increased activity is favorable or detrimental t o the production of color and odor will be apparent later. After t h e sweating process, the time required for drying t o the proper appearance and texture varied somewhat according t o t h e previous treatment with hot water. The beans were dried until t h e proper pliableness was obtained, or until the loss of moisture was from 64 t o 69 per cent. I n a general way t h e drying increased in rapidity as t h e temperature to which the beans were originally subjected was increased. The formation of t h e brown color as well as t h e vanilla odor was seemingly most rapid in t h e samples subjected t o t h e hottest water. These changes were also most marked and rapid in t h e samples sweated in t h e water o r e n at a temperature ranging from 40 to hoa c. The various lots of cured beans were wrapped in
ture, with a decreasing tendency as the temperature of t h e water increased into which the beans were immersed. Physical tests being insufficient and inadequate to determine accurately t h e quality of beans cured by t h e various modifications of t h e methods, a chemical examination was necessary. S o t only was it desired t o ascertain t h e respective qualities of the beans with regard t o vanillin content, but also the oleoresin content, which is considered an important adjunct to t h e flavor of t h e beans. PREPARATIOX AND P R O P E R T I E S O F EXTRACTS
In order more advantageously t o judge the quality of the cured beans and t o make a thorough comparison of t h e various samples with t h e commercial beans, extracts were made from the samples according to the directions in the United States Pharmacopoeial for t h e preparation of tincture of vanilla. The resulting extracts afforded a n excellent means for com1
United States Pharmacopoeia, Eighth Decennial Revision, 1900, 484,
Sept., 1916
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
paring t h e general physical a n d chemical properties of t h e individual samples with those prepared from commercial vanilla beans. These comparisons are shown in Table 111. The general physical properties of t h e extracts from t h e samples cured at room temperature were superior t o those cured in t h e water oven a t somewhat higher temperature. The colors of t h e former were distinctly more red and brownish red t h a n t h e latter. All t h e laboratory extracts were more attractive in color t h a n t h e extracts from t h e commercial beans. The deeper color of t h e extracts from t h e laboratory cured beans was perhaps due t o t h e presence of vanilla resins which possess a color of deep red t o brown. The odor and taste of t h e extracts from t h e beans cured a t room temperature were superior t o those from t h e oven cured beans. T h e presence of a distinct flowery or fruity odor a n d taste was apparent. Extracts from t h e commercial beans seemed lacking in this respect. I n vanillin content less difference is noted between t h e extracts from t h e commercial beans a n d t h e laboratory cured beans. The extracts from t h e beans cured a t room temperature assayed considerably higher in
819
ples were next in order, followed by t h e Tahiti and t h e old sample. I n order partly t o obtain more expert judgment a n d partly t o verify t h e writer’s judgment of t h e s a m ples, a sample of each extract was submitted t o a prominent manufacturer of flavoring extracts, who, without having any knowledge of t h e identity of any of t h e 1 8 samples submitted, made t h e following report on t h e color, odor, and taste: COLOR-South
American, the best. ODOR-NOS. 3 , 5, I and 7 ranked highest; Tahiti, No. 1 1 , and South American, next; Mexican A, Nos. 9,8,and 2 , Bourbon, 4, 6, In, and 1 2 , hfexican B and old sample follow in order. TASTE-~~OS. I , 2 and 3 ranked first; South American, Nos. 1 1 , 7, 9, and 8, were next in rank, followed by Tahiti, Mexican A, No. 2 , Bourbon, Nos. 4, 6 and I O ; No. 12, Mexican B, and old sample were stated to contain a foreign taste. According t o the report t h e color of t h e extract from t h e South American beans was considered best. This is in close accordance with t h e colors as described in Table 111. Extracts I and 3 possessed about t h e same color as t h e South American, namely, a bright wine-red, while Xos. j and 7 were slightly darker. The remainder of t h e laboratory extracts possessed a n admixture of brown, this color being more prominent in some t h a n in others
BEANS BARISON OF VANILLAEXTRACTS PROM L A BORP,TORY CUREDA N D COMMERCIAL COMI TABLE111-PHYSICAL A N D CHEMICAL PER CENT -PHYSICAL PROPERTIES OF EXTRACTSTASTE VANILLIN ODOR SAMPLE COLOR Very agreeable, fruity, lasting 0.255 Pleasant, fruity, very fragrant 1 Wine-red Agreeable, vanilla-like , 0.253 Pleasant, fruity Deep wine-red 7 Mild, characteristic, pleasant 0.210 Pleasant, less fruity Very deep wine-red 5 Very pleasant, fruity 0.317 Very pleasant, flowery Wine-red 3 Agreeable, not fruity 0.212 Pleasant, vanilla-like Wine-red, slight brownish tint 9 Agreeable, slight 0.212 Mild, pleasant Wine-red, slight brownish tint 11 0.215 Mild and pleasant Pleasant. aereeable Wine-red 2 Strong, lasting 0.252 Brown with reddish tint 8 =---Characteristic, pleasant 0.228 6 Mild, lacking fragrance Light wine-red, brownish tint Pleasant, lasting 0.201 Pleasant, fragrance lacking Brown with only trace of red 4 Characteristic, less agreeable 0,239 Mild, characteristic Wine-red, distinct brownish tint 10 Pleasant, slightly prune-like 0.225 Mild, agreeable Wine-red, distinct brownish tint 12 Agreeable, not fruity Pleasant, characteristic 0.259 Mex. A Light brown Strong, harsh, not fruity Strong, vanilla-like, lacking fragrance 0.291 Mex. B Wine-red with brownish tint Strong, unpleasant, licorice-like Strong, extract-like, lacking fragrance 0.200 Very pale brown Tahiti Strong, agreeable, vanilla-like Pleasant, agreeable, vanilla-like 0.268 Wine-red S . Amer. Characteristic, not fruity Pleasant, agreeable, vanilla-like 0.248 Wine-red, with brownish tint Bourbon Mild, not agreeable Mild, vanilla-like 0.196 Brown Old
__
-.
~~~
general average t h a n from t h e beans cured a t higher temperatures. Considerable variation in vanillin content exists, however, in samples prepared from t h e beans which had received t h e hot water treatment before curing. It is very probable t h a t t h e effect of this treatment as regards composition of t h e beans will be more pronounced in other constituents than in t h e vanillin. T o t h e writer t h e laboratory cured beans a n d also t h e extracts appeared to possess a better odor a n d taste t h a n a n y of t h e commercial beans or extracts. The color of t h e extracts in every case also distinctly excelled in color those from t h e beans cured at higher temperatures. Judging from a purely physical’standpoint of color, odor, a n d taste, t h e extracts from Samples I and 3 were considered t h e finest, followed b y Samples 5 , 7, 9 a n d 11. Samples 2 , 8, 6, 4, I O and 1 2 were all lower in color estimation a n d slightly inferior in odor a n d taste. Of t h e extracts from t h e commercial beans, t h e South American sample was distinctly superior t o t h e others in every respect and more nearly corresponded t o Samples I a n d 3 of t h e laboratory cured beans. The Bourbon and Mexican Sam-
Not only was t h e rating of t h e various samples b y t h e manufacturer with respect t o odor and taste in close accordance with t h e writer’s selections, b u t i t is a singular fact t h a t three of t h e laboratory cured samples were stated t o excel in point of odor and taste the extracts from t h e commercial beans and in nearly all cases t h e beans cured at room temperature were superior t o those cured in t h e water oven. I n order t o ascertain whether t h e various modifications of t h e curing process produced any effect upon t h e composition of t h e beans, a n analysis of the residues remaining from t h e preparation of the extracts was examined. EXTRACTION O F T H E R E S I N S
Since it is generally conceded t h a t resins or oleoresins tend strongly t o modify t h e agreeable odor of vanilla beans, it is very probable t h a t the superior aroma and flavor of several of the cured samples of beans were due t o t h e presence of these odorous resinoys substances. These substances are for t h e most part soluble in alcohol, t o a less extent, however, in hydroalcoholic solutions. I n preparing the extracts desciribed 65 per cent alcohol was used, Not all of the resinous constituents of t h e beans were dis-
820
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
solved by alcohol of this strength but it is assumed t h a t approximately t h e same percentage of these constituents was extracted from each sample. Provided t h a t more resinous constituents were present in some samples t h a n in others, and granting t h a t t h e 6; per cent alcohol extracted only a small portion of these constituents, the residues from t h e preparation of t h e extracts still contained much of t h e unextracted resinous substances. These residues were therefore subjected t o extraction with 94 per cent alcohol in order t o obtain t h e residual resins. T h e residues remaining from the alcohol extraction were further extracted with ether. T h e percentages of alcohol and ether extracts. a n d descriptions of each, are given in Table I V . The quantity of alcohol-soluble constituents remaining in t h e beans after t h e preparation of t h e extracts is considerable. While a number of constituents are no doubt present in these extracts, a large portion probably consists of vanilla resins. together with some vanillin. The percentage of alcohol extracts from t h e samples of beans cured in t h e laboratory a t room temperature is uniformly higher t h a n from those cured in T A B L E Iv-.kCOHOI,
V O ~ 8. . S O .9
E X A M I X A T I O S O F R E S I N S E X T R A C T E D W I T H ALCOHOL
I n order t o obtain further knowledge of t h e alcohol extracts obtained from t h e residues, t h e chemical constants, acid. ester, and saponification numbers were determined. The results are shown in Table IV. Considerable variation exists in t h e acid, ester, and saponification values of t h e alcohol extracts obtained from t h e residues of t h e laboratory cured beans. The general average of these constants is higher t h a n t h a t of the commercial beans, as was also t h e percentage of extract. The content of free acid constituents of the resinous extracts is high in nearly all instances, signifying the presence of considerable resin acids. The ester values are likewise high, showing t h e presence of saponifiable resinous compounds. The resin acids and t h e combined saponifiable resins seem t o have been formed t o a greater extent in the beans cured in the laboratory t h a n in t h e commercially cured samples. Provided these substances tend t o add fragrance t o t h e odor of the vanilla bean and therefore also t o the extract made from the beans, it would seem t h a t methods which would insure a high percentage of these resinous substances should be preferred. Doubtless a certain percentage of these
6 6 D ETHER EXTRACTS F R O M \'ANILLA
RESIDUES
SAMPLE 1
7 5 3 9 11 2 8 6 4 10 12
Mex. A Mex. B Tahiti S. Amer. Bourbon Old (a) Fatty taste as well as odor
t h e water oven. Likewise those extracts were considerably deeper a n d richer in color. It would seem, therefore, t h a t curing a t room temperature is more conducive t o t h e formation of vanilla resins and coloring matter. T h e highest percentage of alcohol extract (vanilla resins) in t h e commercial beans was in t h e South American sample. This sample also produced a vanilla extract with t h e deepest color. The residues from t h e Mexican and Bourbon beans also contained a high percentage of alcohol-soluble substances, a n d likewise a highly colored extract. These facts are significant of t h e effect of curing upon t h e color of prepared extracts a n d also of t h e comparatively large quantity of flavoring resins which remain in the residues. The ether extracts were found t o consist largely of f a t t y oil and were nearly devoid of odor. Some of t h e less soluble resinous constituents are doubtless present in these extracts. T h e percentages of ether extract seem t o bear no direct relationship as regards treatment during curing. Some variation is apparent, however, among the samples from different sources.
substances are extracted b y t h e hydroalcoholic menstruum used in preparing t h e extracts of commerce, t h e quantity in solution depending largely upon t h e strength of t h e alcohol employed. COh-CLUSIOh-S
From t h e results reported herein i t m a y be said in general t h a t t h e curing process as a t present commercially applied t o vanilla beans is unnecessarily long and extended, requiring on a n average several months for the transformation of the green t o t h e cured beans. In t h e laboratory experiment conducted it has been shown t h a t thg beans can be cured in a much shorter time b y means of a much less tedious process. It has further been demonstrated t h a t the process is simplified t o the advantage rather t h a n t o the detriment of t h e aromatic constituents. I n order t o produce uniformity in the composition of t h e beans, and thereby insure more uniform extracts, the importation and curing of the green beans are suggested. While t h e amount of vanillin in t h e beans was not increased appreciably. as compared with commercial beans, it may be stated with assurance t h a t t h e beans
Sept., 1916
T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
821
cured in t h e laboratory were in most cases superior in vanilla resins a n d coloring matter. This is significant since vanilla resins a n d coloring matter are considered important adjuncts t o t h e quality of vanilla beans. The superior flavor of t h e extracts prepared from t h e laboratory cured beans may therefore be ascribed t o t h e resinous constituents. A considerable ‘proportion of t h e vanilla resins are left unextracted when t h e menstruum is less t h a n 6 5 per cent alcoholic strength. Curing t h e green beans a t room temperature either without previous treatment or after treatment with water up t o 90’ C. for a short period of time apparently produces beans of t h e best quality, as judged by t h e flavoring extracts prepared from t h e samples.
in sour milk, human gastric juice a n d in muscle. I t has been found in t h e vegetable kingdom in t h e peduncles of Solanum du1camara.l There are four lactic acids, three having t h e formula CH3.CHOH.COOH, and one t h e formula CH20H.CH2COOH. Inactive or ethylidene lactic acid (&-hydroxy propionic acid) may be prepared by treating cr-chlorpropionic acid with potassium hydroxide: it forms a zinc lactate t h a t crystallizes with three molecules of water of crystallization. Sarco- or dextro-lactic acid (CH3CHOH.COOH) is found in meat: t h e zinc lactate made from this acid crystallizes with two molecules of water of crystallization. Laevo-lactic acid (CHsCHOH.COOH) differs from t h e dextro acid only in its optical powers: its zinc BUREAUOF P L A N T IXDUSTRY, WASHINGTON lactate crystallizes with two molecules of water of crystallization. Beta hydroxy propionic acid (hydracrylic or ethylene A STUDY OF THE STEREO-ISOMERISM OF A lactic acid) has the formula C H 2 0 H C H z C O O H . FERMENTATION LACTIC ACID When heated i t liberates water and forms an unsatBy STAXLEY JUDSOX THOMAS urated acid, acrylic acid, CH2CHCOOH, Received December 21, 1915 Thus we have three lactic acids having the same Investigations of t h e isomerism of lactic acid soon structural formula but different properties. This followed Pasteur’s famous work with tartaric acid. is known as stereo-isomerism, physical isomerism, or I n 1890, Lewkowitsch’ showed t h a t ammonium optical isomerism. Ethylene lactic acid, CHzOHCH2lactate when exposed t o t h e action of Penicillium COOH, is inactive under all circumstances. Ethylglaucum for several weeks became optically active. idene lactic acid, CH3CHOHCOOH, on t h e other I n 1892, Professor T. Purdie proved, by direct anahand, may exhibit optical activity. On close exlytical evidence, t h a t commercial lactic acid can be amination t h e latter is found t o contain a carbon resolved into two oppositely active isomeric acids.2 atom in its molecule which holds four other atoms T h e results of experiments of various writers on t h e or groups of atoms, all different from each other. kind of lactic acid formed in naturally soured milk Optically active substances, without exception, condo not agree. Some found inactive acid, some dextro-, tain one or more such carbon atoms in their moleand some laevo-rotatory acid. Dr. P. G. Heinemann cules, and t h e property of turning t h e plane of polarcleared u p most of this conflicting testimony by showization bears a close relation t o t h e structure of a subing t h a t different bacteria produce different kinds of stance. Optical activity is, therefore, due t o t h e aci.d. As a summary t o a n excellent paper on t h e presence of a carbon atom in a molecule which holds subject, he says: “Racemic lactic acid is t h e result four dissimilar groups with its four bonds. Such a of t h e formation of pure dextro acid and pure laevo carbon atom is known as a n asymmetric carbon atom, acid by a t least two different species of microorganFurthermore, a n optically active substance may show isms. Racemic (inactive) lactic acid is not known dextro-rotation, t h a t is, t u r n t h e plane of polarizat o be t h e product of one species tion t o the right; or laevo-rotation, t h a t is, turn t h e The object of this paper is t o discuss t h e kind of plane of polarization t o the left. A closer examinalactic acid present in Matzoon, a n Armenian arti- tion of t h e lactic acid molecule may show why t h e ficially soured milk, and t o give t h e results of a n in- acid exhibits both dextro-rotation and laevo-rotation. vestigation of its optical powers. Let us now examine the lactic acid molecule. If Matzoon is a beverage prepared from cows’ milk we attempt t o draw a picture of the molecule in space with t h e aid of a n appropriate ferment. From this we may have: ferment, or “starter,” t h e author isolated a bacillus, a streptococcus, and a yeast. The bacillus was determined as a member of t h e B. bulgaricus group, corresponding more closely t o B. caucasicus (Flugge) t h a n t o any other member. The streptococcus isolated was Str. kejir (Kuntze), a member of t h e Str. Zebenis4 group. The yeast was determined as Saccharomyces kejir: this organism is unable to ferment milk or lac(alJextr0 L actic Acid &) iaevo Lactic ,&id tose, excepting a slight production of alcohol. 1 1 3
4
Ber., 16, 2720. J . Chem. Soc., 1892, 754. J . Bid. Chem.. 2, 612. Freudenreich, Landw. Jahrb. d . Schweis., 10 (1896). 1.
t h a t of a n object and its image in a mirror. If b be superimposed upon a, we will find t h a t the O H group 1
Standard Dispensatory.
