Ha r o l d A. Id d l e s a n d Pa u l J. Ro b b i n s, University of New Hampshire, Durham, N. H.
T
H E standard method for the determination of pentoses and pentosans in plant materials (1, 5, 6) involves the distillation of the sample with 12 per cent hydrochloric acid until no more furfural is produced, as indicated by the absence of a pink coloration with aniline acetate.
The distillate is treated with phloroglucinol to produce a precipitate of indefinite composition which may be deter
mined gravimetrically. Because of the empirical nature of this method and the length of time consumed in its opera
tion, several alternative methods have been suggested. The use of thiobarbituric acid as a precipitant has been studied by Dox and Plaisance (2); the potentiometric titration of the furfural in acidified potassium bromide solution with potassium bromate has been proposed by Pervier and Gort- ner (7); and the use of excess bromine to react with the furfural and subsequent estimation of the unused bromine was studied by Powell and W hittaker (8) and recently used by Kline and Acree (4).
The primary object of the work recorded in this paper was the determination of the quantity of furfural produced from six species of American hardwoods. Preliminary to these determinations, a comparison of the four above-men
tioned methods for the estimation of furfural was made on pure furfural samples. And further, distillations of pure furfural by means of the standard method with 12 per cent hydrochloric acid and also with 12 per cent hydrochloric acid and steam were made to contrast the relative amount of destruction in these procedures. On the basis of these comparisons the furfural produced by acid-steam distilla
tion of the hardwood samples has been determined by the volumetric potassium bromate-bromide method of Powell and W hittaker (8).
De t e r m i n a t i o n o f Fu r f u r a l
In comparing the quantitative methods of estimation, the furfural wras purified by distilling technical furfural at a pressure of 16 to 20 mm. four or five times until an almost colorless distillate was obtained. Five-gram samples were immediately dissolved in water and made up to 500 cc. from which 5-cc. aliquot portions were used in the following de
terminations :
P h l o r o g l u c i n o l o k A. O. A. C. M e t h o d . A quantity of phloroglucinol about double the amount necessary for the furfural present was dissolved in 12 per cent hydrochloric acid and added to the sample with stirring. The solution turned yellow and then a green .precipitate appeared which grew darker until it became nearly black. After making the solution up to 400 cc.
with 12 per cent hydrochloric acid and allowing it to stand over
night, the precipitate was collected on a tared Gooch, washed with 150 cc. of water, and dried a t 100° C. for 4 hours. After weighing the precipitate, the amount of furfural recovered was calculated by means of Krober’s (5, 6) formulas.
T h i o b a r b i t u r i c A c i d M e t h o d . To the sample of pure furfural, made up to 200 cc. with 12 per cent hydrochloric acid, was added the precipitant consisting of thiobarbituric acid, slightly in excess of the equivalent amount, also dissolved in 12 per cent hydrochloric acid. The final volume was r^ade up to 400 cc. with 12 per cent hydrochloric acid and allowed to stand overnight to insure complete precipitation. The pre
cipitate, which filtered slowly, was collected on a tared 'jooch and dried at 100° C. The weight of furfural was calculated on the basis of a reaction ratio of one molecule of furfural to one of thiobarbituric acid.
V o l u m e t r i c P o t a s s i u m B r o m a t e - B r o m i d e M e t h o d . Fol
lowing the method suggested by Powell and Whittaker and since used by others, 25-cc. portions of approximately 0.1 iV potassium bromate-bromide were pipetted into four glass-stoppered bottles.
Into two of these the samples made up to 200 cc. with 12 per cent hydrochloric acid were added and into the other two, 200 cc.
of 12 per cent hydrochloric acid were added for blank runs.
After standing for one hour in the dark, 10 cc. of 10 per cent potassium iodide were added and the liberated iodine was titrated with 0.1 N sodium thiosulfate. The number of cubic centimeters required by the sample subtracted from the number required for the blank was a measure of the furfural present.
In an attem pt to employ the potentiometric method pro
posed by Pervier and Gortner great difficulty was found in determining the end point, and erratic and often high results were obtained. This observation has been made' also by Youngburg and Pucher (12). Consequently it was not used in further work.
A comparison of the averages of results for the three meth
ods given in Table I (100.17, 102.0 and 102.88 per cent) shows th a t the standard phloroglucinol method gave prac
tically theoretical results, while the latter two were slightly above the theoretical which is in close agreement with the results for these methods recorded by the original authors (2, 8). The widest variations in the errors observed in the different methods of analyzing weighed portions of furfural were about 2 per cent with phloroglucinol, 1 per cent with thiobarbituric acid, and 1.2 per cent with potassium bro
mate-bromide. If in contrasting these three methods the advantages of greater simplicity, economy of time, and equal constancy of results are considered, then the volumetric method seems to be superior to either of the gravimetric methods.
T a b l e I
Fu r f u r a l Pr e c i p i t a n t Fu r f u r a l Fu r f u r a l
Sa m p l e Ü 8 E D Us e d Pr e c i p i t a t e Fo u n d Re c o v e r e i
Gram Gram ' Gram Gram %
P H L O R O G L U C IN O L M ETH O D
1 0.0574 0 .3 9 0 0 .1 0 5 4 0 .0 5 7 3 9 9 .8 2
2 0.0 5 7 4 0 .3 8 9 0 .1 0 5 0 0.0 5 7 1 9 9 . 4S
3 0.0574 0.3 9 0 0.1 0 7 4 0.0 5 8 4 101.74
4 0.0 5 7 4 0.3 9 0 0.1051 0 .0 5 7 2 9 9 .0 5
Av. 100.17 T H IO B A R B IT U R IC A C ID M ETH O D
1 0.0 5 7 5 0.1 3 2 0 .1 3 6 7 0.0 5 9 1 102.78
2 0.0575 0.1 3 2 0.1 3 5 2 0.0 5 8 5 101.73
3 0.0 5 7 5 0.1 3 2 0.1 3 5 3 0.0 5 8 5 101.73
4 0 .0 5 7 5 0.1 3 2 0.1 3 5 6 0.0 5 8 6 101.91
A v. 102.04 V O L U M ETR IC PO TA SSIU M B R O M A T E --B R O M ID E M ETH O D
Fu r f u r a l 0.1 N Fu r f u r a l FüR F U R A L
Sa m p l e Ü S E D NaîSiO» Fo u n d Re c o v e r e d
Gram Cc. Gram %
1 0.0 5 0 8 2 1 .6 9 0.0521 102.55
2 0.0 5 0 8 2 1 .6 7 0.0 5 2 0 102.36
3 0.0 5 0 8 21 .8 2 0.0524 103.15
4 0.0 5 0 8 21 .6 7 0.0 5 2 0 102.36
5 0.050S 21 .7 7 0.0 5 2 3 102.95
6 0.0 5 0 8 2 1 .7 7 0.0 5 2 3 102.95
7 0.0 5 0 8 2 1 .8 4 0.0524 103.15
8 0.0 5 0 8 21 .9 2 0.0 5 2 6 103.54
Av. 102.88
Co m p a r i s o n o f Di s t i l l a t i o n Pr o c e d u r e s
The standard procedure for the distillation of pentoses produces low results, which may be attributed to the destruc
56 A N A L Y T I C A L E D I T I O N Vol. 5, No. 1 tion of the furfural by the acid upon long exposure. To over
come this difficulty Jolies (S) and Pervier and Gortner (7) have used steam distillation on xylose and arabinose and re
port practically complete recovery of the furfural based on the pentose used, whereas Kline and Acree (4) find th a t steam comparable conditions, furfural samples were subjected to the two types of distillation and the recovery measured by the same method of analysis in both cases. Samples of about 5 grams of freshly distilled furfural were distilled by the A. 0 . A. C. method and the distillate was made up to 500 cc. from which 5-cc. aliquots were analyzed by the volu
metric potassium bromate-bromide method. In all cases the solution being distilled turned dark and a brownish precipitate separated out as the distillation proceeded. Sam
ples were also distilled from 12 per cent hydrochloric acid using the steam distillation method, in which a slow stream of steam was led into the distilling flask while it was maintained at 103° to 105° C. When no further test for furfural was ob
tained, the distillate was analyzed volumetrically. The re
sults appearing in Table II (98.83 per cent recovery when ordinary distillation is employed and 101.54 per cent when steam distillation is used) indicate an average furfural re
covery in the second method 2.71 per cent higher than in
the first with 12 per cent hydrochloric acid alone. The greatest variations in furfural recovered by distilling pure furfural were 5.5 per cent with 12 per cent hydrochloric acid available data being the results of investigations carried on a t the U. S. Forest Products Laboratory (10) by using no further test for furfural was obtained w ith aniline acetate.
The distillate, which was protected by passing it into the receiver through a small filter, was then made up to 500 cc., the acidity of the solution was determined, and sufficient concentrated hydrochloric acid was added to bring the con
centration up to 12 per cent hydrochloric acid. The furfural content was then determined by the volumetric method by using 200-cc. aliquots and an accompanying check by the phloroglucinol precipitation method was run on a 300- cc. aliquot.
The results recorded in Table III show a close agreement between the two methods of determination, the volumetric being slightly higher here as in the determination of pure furfural. In the analysis of six woods the volumetric bro
mate-bromide titration gave from 0.2 to 0.86 per cent higher furfural yields than the phloroglucinol method, or a difference of 1.5 to 5.0 per cent of the furfural itself. The widest variations in several determinations of furfural from each of the six woods were, respectively, 2.3, 2.2, 1.5, 1.2, 0.77, and 2.3 per cent of the furfural by the bromate-bromide method, and with phloroglucinol they were approximately 3.7, 0.8, 4.4, and 2.4 per cent with four of these woods.
In the precipitation method other compounds such as methyl furfural and hydroxymethyl furfural enter into reaction and are determined with the furfural. Undoubtedly these com
pounds may have an effect in the volumetric procedure, although this possible effect has not been determined.
Co n c l u s i o n s
The volumetric method of determination of furfural as proposed by Powell and W hittaker gave consistent results which compare favorably with the gravimetric methods with either phloroglucinol or thiobarbituric acid. The widest variations observed in the different methods of analyzing weighed portions of furfural were about 2 per cent with phloroglucinol, 1 per cent with thiobarbituric acid, and 1.2 per cent with potassium bromate-bromide. By the volumet
ric procedure the whole determination may be made in one day whereas the A. O. A. C. method requires a longer time.
The steam distillation of furfural from a 12 per cent hydro
chloric acid solution led to a recovery wiiich was 2.71 per cent higher than in the ordinary distillation with 12 per cent hydrochloric acid. This would indicate less destruction of the furfural during the time of distillation.
Six species of hardwoods have been distilled with steam and 12 per cent hydrochloric acid to yield furfural which has been determined volumetrically. This method gave
January 15, 1933 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 57 furfural by the bromate-bromide method, and with phloro
glucinol they were approximately 3.7, 0.8, 4.4, and 2.4 per