2096
INDUSTRIAL AND ENGINEERING CHEMISTRY
proceeds more rapidly. In the presence of moisture, silica is easily attacked by the hydrogen fluoride formed. It is presumed that under some conditions, even the silicone plastics and greases will be affected. Rubber is vulcanized by the the sulfur compounds and becomes hard and brittle; organic greases become discolored and hard, although pure paraffins have been found to stand up somewhat better. However, with the present availability of fluorocarbon plastics and greases, the choice of suitable materials for use with SF,, especially where there is a good possibility for breakdown, should not be too difficult. TOXIC EFFECTS
The toxic properties of fluorine are well established.
As far as
SF, and SzF2 are concerned, the rapid hydrolysis of these compounds in moist air to give hydrogen fluoride and sulfur dioxide indicates that any hazard would be due to these gases. Physiological tests made to date show no effects other than those which would be expected on this basis. Fortunately, all these compounds possess strong characteristic odors which render them easily detectable. However, there is some question as to whether the threshold of smell is below the toxic limit.
Vol. 45, No. 9
ACKNOWLEDGMENT
The authors would like to express their appreciation to H. V. Wadlow for microchemical analyses, to C. J. Calbick for electron diffraction studies, and to K. H. Storks, D. M. Dodd, and M. H. Read for infrared spectroscopy. LITERATURE CITED
Ahearn, 8 . J., and Hannay, K.B., J . Chem. Phys., 21, 119 (1953). Camilli, G., Gordon, G . S., and Plump, R. E., Trans. Am. Inst. Elec. Engrs., 71, 111,348 (1952). Dibeler, V. H., and Mohler, F. L., J . Research Natl. Bur. Standards, 40, 25 (1948). Dubnikov, L. &I., and Zorin, N. I.,Zhur. ObshcheE Khim., 17, 185 (1947). Edelson, Da\-id, J . Am. Chem. Soc., 74, 282 (1952). Lapemann, R. T., and Jones, E. A , . J. Chem. Phgs., 19, 534 (1951) .
Mcifee, K. B., unpublished work. Schumb, W. C., Trump, J. G., and Priest, G. L., IND.ENQ. CHEhf., 41, 1348 (1949). Trauta, A I . , and Ehrinann, K., J . prakt. Chem., 142, 79 (1935). RECEIVED for review February 3, 1953. ACCEPTED June 11, 1953 Presented before t h e Conference on Electrical Insulation, National Research Council, Lenox, %lase.,October 1952, and before the winter meeting of the Aiiierican Institute of Electrical Engineers, January 1953.
ine
ercetin from
ar
E. F. KURTII Oregon Forest Products Laboratory, Corvallis, Ore.
HE flavonol, quercetin, is now extracted commercially from Howering buckxheat and the Chinese scholar tree, Xophom japonica, as the rhamnoglucoside, rutin, It may be produced by the hydrolysis of the rhamnoside, quercitrin, obtained from the bark of black oak ( 1 ) . Quercetin is a yellow, crystalline powder, almost insoluble in water, and its melting point has been variously reported to be between 312" and 317' C. Extensive literature has appeared sho\ving that it is nontoxic and has antioxidant and physiological propert,ies (2, 4, 7 , 9). A410ngwith other flavonoids it possesses vitamin P activity (8). Dihydroquercetin and quercetin are closely related coinpounds. The former is a white crystdline flavanone t,hat has two more hydrogen atoms in the molecule than quercetin. Quercetin is, therefore, an oxidized derivative of dihydroquercetin. The flavanone is a substantial constituent of the barks of certain trees, such as Douglas fir and Jeffrey pine. It has been claimed that dihydroquercetin, which is present also in the heartwood of Douglas fir, interferes with the pulping of this species with calcium bisulfite liquor. Therefore, the action of bisulfites on dihydroquercetin was explored and a process for the conversion of this abundant, raw material to quercetin developed. In this connection, it has been found that dihydroquercetin was converted to pure quercet,in rapidly and in high yields by the simple process of refluxing it with an aqueous solution of an alkali metal bisulfite or ammonium bisulfite ( 6 ) . Chemically pure quercetin in the form of yellow crystals melting a t 316" to 317' C. separated promptly from the hot aqueous solut'ion in the order of about a 90% yield, by weight. When calcium bisulfite liquor was reacted similarly with dihydroquercetin in an open vessel, a finely divided, insoluble crust of a calcium-quercetin complex was deposited that adhered tenaciously to the sides of the vessel. In the conventional calcium bisulfite pulping renction, therefore, digestion of wood chips may be hindered by t8heimpervious deposit of this complex and by the removal of the calcium ions from the reacting liquor. The difficulty encountered in pulping Douglas fir heartwood chips with calcium bisulfite is not experienced when ammonium bisulfite is used. This distinc-
tion between the two pulping processes may be caused by the difference in the end product of the two bisulfites and dihydroquercetin. The converaion of dihydroquercetin, which is readily soluble in aqueous sodium, potassium, or ammonium bisulfite solutions, to insoluble quercetin is unique in that the bisulfite reducing agents unexpectedly remove two hydrogen atoms from each molecule of dihydroquercetin. Although the mechanism of the conversion is not clear, the bisulfites apparently act in a catalytic capacity or are reduced to some lower oxidation state, for after the conversion of an initial quantity of dihydroquercetin to quercetin, the residual liquor may be used and reused, with only slightly re luced effectiveness in the conversion of further quantities of dihydroquercetin. In addition, the reaction appears applicable only to 3-hydroxy flavanoids that have a carbonyl group in the four-position, Under the same conditions, catechin from white fir bark, which has the same chemical structure as dihydroquercetin, except for the absence of the carbonyl group, is not affected by bisulfites. EXPERIBI ENTAL
The catalytic character of the bisulfite reagent is illustrated bv four siccessive runs made with the same bisulfite liquor. First, 5 grams of dihydroquercetin and 20 grams of sodium bisulfite were dissolved in 100 ml. water. The solution was refluxed a t atmospheric pressure for 15 minutes, a t the end of which time quercetin was separated by filtration from the hot solution in a yield of 50% by weight. After the residual liquor was refluxed for an additional 25 minutes, an additional 35ql, yield of quercetin was separated, making a total yield of 85%, based on the dry weight of dihydroquercetin. The product was bright yellow in color and melted a t 316' to 317" C. with sublimation. It gave an acetate derivative, melting a t 193' t o 194' C. A mixed melting point determination with an authentic sample of quercetin gave no depression in melting point. To the liquor remaining from the above procedure, another 5 grams of dihydroquercetin was added The resulting mixture
.
I N D U S T R I A Z . A N D E N G IN E E R I N G C H E M I S T R Y
September 1953
was refluxed for 25 minutes and then for an additional 75 minutes, the quercetin being filtered off a t the end of each reaction period. The yields obtained were 58 and 39%, respectively, making a total yield of 93%. Still another 5 grams of dihydroquercetin was added t o the liquor, and the mixture was refluxed first for a period of 35 minutes and then for a period of 105 minutes. The quercetin was removed a t the end of each reaction period. The first period yielded 52% and the second 25%, making a total yield of 7775, Finally, an additional 5 grams of dihydroquercetin was added t o the liquor. After refluxing for successive periods of 35 and 100 minutes, 32 and 41% yields of quercetin, respectively, were obtained, a total yield of 73%. When the residual liquor was acidified with sulfuric acid, unreacted dihydroquercetin was obtained . Increasing the concentration of either dihydroquercetin or bisulfite solution increased the rate of the reaction and also the overall yield of quercetin. Herbert Hergert of this laboratory found that long boiling of dihydroquercetin with dilute sodium bisulfite solution caused some conversion to 4, 6, 3', 4'-tetrahydroxy-2benzyl-3-coumaranone, melting point 262' to 263' C. Geissman and Lischner ( 3 ) obtained this compound by the reduction of quercetin with sodium hydrosulfite (Na2S204). Sulfurous acid, bisulfates, bicarbonates, acid phosphates, and acetates gave no conversion to quercetin. A summary of the yields of quercetin produced when mixtures of dihydroquercetin and bisulfite solutions were refluxed at atmospheric pressure for different periods of time is given in Table I.
tin, which remained as a solid residue after removal of the solvent, was then converted to quercetin. The crude product (5 grams) was dissolved in 100 ml. hot water together with 20 grams sodium bisulfite. The resulting mixture was refluxed a t atmospheric pressure for a period of 30 minutes, after which pure quercetin, melting a t 316" to 317" C., was separated by filtration. A portion of the hot water extract was evaporated to a dry powder. The powder was extracted with warm methyl ethyl ketone and the solvent recovered. Ten grams of the residual reddish-brown dihydroquercetin product that was contaminated with some tannin was refluxed with 100 ml. sodium bisulfite solution containing 10 grams of sodium bisulfite. After 1 hour, substantially 90% of the dihydroquercetin had been converted to quercetin, which separated out as yellow crystals melting a t 316" t o 317" c. Sound Douglas fir bark was ground in a Rietz disintegrator and screened. The fraction retained on a 30-mesh screen was recovered. The material passing through the screen, which consisted largely of bast fibers and a brown amorphous powder, was discarded. The oversize fraction was reground in an attritiontype disk mill and rescreened on the same screen. The part then retained on the screen was largely cork particles and contained 19.0% dihydroquercetin and only 2.7% tannin by weight. The cork particles were extracted with hot water and the extract evaporated to a thin, reddish-brown sirup. The dihydroquercetin in the extract was converted to quercetin by adding 15 grams sodium bisulfite to 100 mi. of the sirup and refluxing for 1 hour. The quercetin was recovered by filtration as bright yellow crystals melting a t 316' to 317' C.
OF DIHYDROQUERCETIN TO QUERCETIN TABLEI. CONVERSION WITH BISULFITE SOLUTIONS AT 100' C.
Bisulfite Solution (100 ml.) 5 7 , NaHSOa
Dihydroquercetin Added, Grams 5.0
10% NaHSOa
5.0
15% SaHSOs
5.0
20% XaHSOa
5.0
10% NH4HSOs
15% N H d " s
5.0 '
Saturated Mg(OH)2
+ SO2
5.0
Reaction Time, Hours
Yield Quercetin,
1,i 17.(1 1 . tI 17.0 0.5 1.25 0.25 0.75 1.5 17.0 < -1 . 2 5
32 57 59
I(
5.0
u
6.0
%
66
50 74 50 a5
32
7Q
.I
56 85 14
2097
SUMMARY
Dihydroquercetin, present in the heartwood and the bark of the Douglas fir and in the bark of the Jeffrey pine, is readily converted in high yield t o quercetin by treatment with hot aqueous bisulfite liquors. Pure crystals of quercetin separate from either hot sodium, potassium, or ammonium bisulfite liquors. Calcium bisulfite liquor with dihydroquercetin produces a finely divided, insoluble calcium-quercetin complex. The reaction affords a simple means for producing quercetin from hot-water extracts of certain plant materials. It also offers an explanation for the difficulty experienced in pulping Douglas fir heartwood with calcium bisulfite liquor. ACKNOWLEDGMENT
Although dihydroquercetin in its pure crystalline form may be employed as a starting material, the crude product, with tannin and other materials extracted from the barks together with the dihydroquercetin, may also be used. Thus, quercetin has been obtained in high yields by refluxing with bisulfite solutions the crude ethyl ether-soluble fraction of the hot water extract of barks, the methyl ethyl ketone extract from the dried hot water extract of barks, or the hot water extract from Douglas fir cork. The tannins and like coloring matters have been found to be completely soluble in hot solutions of the soluble bisulfites, whereas the quercetin is insoluble. This affords a simple means for producing pure quereetin from plant sources. Quercetin was prepared by grinding Douglas fir bark to pass a S/a-inch screen. The ground bark then was extracted countercurrently with water a t a temperature of about 95' C. in wooden leachers, until substantially all the dihydroquercetin and tannin had been leached out ( 5 ) . The resulting hot-water solution containing dihydroquercetin, tannin, sugars, and other water-soluble constituents of the bark was then evaporated to a sirup in alongtube, circulating-type vacuum evaporator. The sirup was extracted with diethyl ether in a conventional liquid-liquid extractor to remove the dihydroquercetin, after which the solvent was distilled off and recycled. The crude crystalline dihydroquerce-
The assistance of H. L. Hergert in obtaining the data in Table
I is acknowledged. LITERATURE CITED
(1) Booth, A. N., and De Eds, F., J . Am. Pharm. Assoc., 40, 384-5
(1951). (2) Couch, J. F., Xrewson, C. F., and Naghski, J., "Chemistry, PKarmacology, and Clinical Application of Rutin," Eastern Regional Research Laboratory, Philadelphia 18,Pa. (February
1951). (3) Geissman, T. A,, and Lischner, H., J . Am. Chem. SOC.,74, 3001-4 (1952). (4)Griffith, J. Q., Couch, J. F., and Lindaner, M . A , , Proc. SOC. Ezptl. Biol. Med., 55, 228-9 (1944). ( 5 ) Kurth, E. F., and Chan, F. L., J . Am. Leather Chemists' Assoc., 48, 20-33 (1953). (6) Kurth, E. F., and Chan, F. L., J . Am. Oil Chemists' Soc., 28, No. 10, 433-6 (1951). (7) Richardson, G. A., El-Rafey, M. S., and Long, 34. L., J . Dairy Sei., 30, No. 6, 397413 (1947). (8) Wilson, R. H., and De Eds, F., J . Pharmacal. E%&. Therap., 95, No. 3, 39!&406 (1949). (9) Wilson, R. H., Mortarotti, T. G., and Doxtader, E. K., Proc. SOC.Ezptl. B i d . Med., 64, 324-7 (1947). RECEIVED for review March 12, 1953. ACCEPTED June 8, 1953. Presented before the Division of Cellulose Chemistry a t the 123rd Meeting of the AMERICAN CHEMICAL SOCIETY,Los Angeles, Calif.