July, 1946
INDUSTRIAL AND ENGINEERING CHEMISTRY
deterioration, and certain niilitary agcnciea have stressed the point that many of their fahric items must he able t o withstand this condition. ;\lthough quantitative differences do exist among soil burial re.ults determined on identical fabrics under different contlitions with various soil.5, experience in ninny laborafories indicates t h a t certain fabric trcntments are consistently poor in all or essentially all tests and t h a t other treatments are consiqtent ly much.better. Ttie soil suspension test repre~entsa n attempt to rctain the heterogeneous microflora of the soil b u t t o bring the environmental conditions under closer control than in the soil burial test. r\ltlioiigh a heterogeneous microflora is retained in actual practic(,, the writers cannot defend the thesis t h a t soil suspension are merely simplified soil burial te The population of oryani-nis which grow on the fabric in soil suspension tests is qualitatively and quantitatively different from the population found on the same fabric in soil burial teqts in the same soil. l’reviou.-ly reported tests of 2,2’-methyle1icbis(~-chlorophenol), niimber 11 (commercially called “compound G - A ” ) , in parallel v i t h p~ntacliloroplioiiol,salicylanilide, tetrabromo-o-cresol, and otticr phenols ( I ? ) hnd already eGtablirlicd that the former com. pound has flunyicidal activity of a high order. I t is apparent from tile data pre-entetl here that other clovly al-o have high fungicidal activity which, hon. not exceed that of compound 11.. T h e possible uses for tlicse compounds have not becn thoroughly erplored, and it might he found tliat for specific uses certain of them noulti have distinct advantages over conipound 11. Compound 57, for example, niigtit he superior to compound 11 as a s p m y or du-t t o prevent “late hlight” of potatoes or other plant dileai;eh, etc. .-Uttioiigli compound 11 has been used commercially as n fabric preservative on a n extensive scale in the 1a.t three year.sedata. ACKNOWLEDGMENT
T h e authors are indebted in particular t o W. S. Gump, of Givaudan-Dclawanna, Inc., who synthesized the compounds used in this etudy, and t o E. Iiunz, R. Horsey, and Thomas Wallace of the same concern for advice and interest during the course of the work. LITERATURE CITED
t c 3 - b
(1) Britton, E. C . , and Bryner, F.. U. S. Patent 1,969,963 (1934). (2) Dean, J. D., Strickland, W. B., and Berard, W. N., Am. Dyes h f R e p t r . , 34, 195-201 (1945). (3) Furry. h I . S.,and Zametkin, M., Ibid.,32, 395-8 (1943). (4) Greathouse, G. A., Klemme, D . E., and Barker, H. D., IND. E s o . CHEST.,. &SAL. ED.,14, 614-20 (1942). (5) Gump, W. S..U. S. Patent 2.250.480 (1941). if3 Ibid.. 2.353.724 *-, -~ , - , - - ~ , (1944). ~~
( 7 ) I b i d . , 2,354,012 t19a4j. (8) Klarmann, E . , and Gates, L., Ibid.. 1,967,825 (1934). (9) Klarmann, E., and Yon TVoxern, J., J . Am. Chem. Soc., 51, 60510 (1929). (10) Kunz, E. C., Luthy, hl., and G u m p , ‘A‘. S., U. S. Patent 2,353,736 (1944). (11) Marsh, P. B., Greathouse, G. .1., Bollenbacher, K., and Butler, AI. L., ISD.ESG.CHEM.,36, 176-81 (1944). (12) hfarsh, P. B., Greathouse, G. A., Butler, M. L., and Bollenbacher, K.. U.S. Dept. Agr., Tech. BuZL 892, 1-22 (1945). (13) Thoin, C.. Humfeld, H., and Holinan, H . P., Am. Duestuff X e p t r . , 23, 581-6 (1934). (14) T r a u b . E. F., Sewhall. C. A., and Fuller, J. R.. Surgery, Gynecol. O L s t ~ t . ,79, 205-16 (1944).
AUTOXIDATION OF FURFURAL A. P. DUSLOP, PAUL R . STOUT, A N D S.i3IUEL SFADESH The Quaker Oats Company, Chicago, I l l .
Color and acid formation in furfural at room temperature are shown to be due to autoxidation, the course of which is different from that of benzaldehyde. It occurs at a considerably lower rate and the reaction is interrupted a t a much lolier lebel of acidity. A mixture of acids is formed from furfural, not merely acetic acid or solely furoic acid as had been speculated previously. The changes may be prevented by storing furfural in an oxygen-free atmosphere, or effectively inhibited by adding to furfural a small quantity of a variety of baric substances such as tertiary amines or alkali metal soaps or phenolic antioxidants. ’Water causes a partial inhibition of the autoxidation.
I
S SOR1\IhL storage, furfural slon-ly dwkens in color; n change accompanied by formntion of ncid. The magnitude of these changes with reference t o the amount of furfural is small even niter relntively long periods of time. Severtheless, it was considered desirable t o find some means of preventing or retarding the reactions responsible, mid to accomplish this it \v:is necessary t o have some knowledge of the cniise. Previous to this Ivork, it was known that the color bodies in on distillation, and aged furfural were left behind a s a resi~lut~ t h a t tlie yield of rccovered furfural ~ r i gtiner:llly s high. In addition, stabilizers such as hydroquinone and pyrocntectiol h:id been found to inhihit color formntion. There hnJ been considerable
speculation regarding the nature of the formed acid. For example, acetic acid -xas postulated, a view originating in plant data since acetic acid and also acetaldehyde are by-products of the furfural manufacturing process. Another hypothesis attributed t h e development of acidity t o t h e formation of furoio acid, a belief based largely on the kno-xn behavior of benznldehyde. This was not unreasonable, since i t is true t h a t furfural parallels benzaldehyde in a number of well known reactions. Some of these reactions are illustrated in Figure 1, nhere R denotes the fury1 or phenyl radical. Under the influence of sodium cyanide, furfural condenses with itself t o form furoin, the furan analog of benzoin. Both aldehydes undergo the Cannizznro reaction in the presence of strong alkalies to give the eorresponding alcohol and ncid in equimolecular proportions. With ammonia, complex nldimines are formed: hydrofuramide from furfural, and liydrobenzamitfe from benzaldehyde. Under a p propriate conditions both can be oxidized t o the carboxy derivatives or reduced t o furfuryl or benzyl alcohols, respectively. The usual aldehyde tlcrivativcs can be prep:u.ed from either, as illustrated in this case by tlie oximes. O X l G E N IN COLOR A S D ACID FORIIATION
Preliminary studies soon showed tlint, n.1ic.n technical furfural was stored with free access t o nir a t room temperature, it darkened i n color and devclopcd :icidity much more rnpidly than
706
INDUSTRIAL AND ENGINEERING CHEMISTRY
R-CHzOH
J
R-COOH
R-CHOH-CO-R
R-CH20H + R-COONa
Vol. 38, No. 7
The results in Figure 3 sho\v that technical furfural, on exposure t o air at room temperature, exhibits an induction period which is n o t apparent in the autoxidation of pure furfural. I n each case, however, the rate of acid formation reaches a maximum and then the curves slope off quite abruptly, although a t different levels. At this stage of the investigation it was reasoned that the formed acid had its origin in a readily oxidizable impurity, a theory which would explain why the autoxidation:of pure furfural practically ceases a t a relatively low level of acidity. I n others words, it was indicated that, in preparing the sample of pure furfura!, the concentration of acid precursor had been decreased.
Figure 1. Reactions of Furfural
similar samples in tightly stoppercd flasks. Difficulty was experienced, however, in checking the rates of acid formation in supposedly duplicate experiments. Table I demonstrates that this was due t o variations in the surfaceivolume ratio. I n this experiment differing volumes of distilled furfural were exposed to air in Petri dishes of equal area, and the increase in acidity TWS mcnsured. The results show that the rate of acid formation is a function of the surface/volume ratio. Recognition of this fact made it possible to obtain reliable and reproducible results. The surface/volume ratios employed in our studies were as much as a hundredfold greater than that obtaining in normal storage of furfural-that is, in 55 gallon drums. By accelerating the rate in this manner, we were able t o obtain results in a reasonable time.
I
0
I
I
10 40 60 TIME, DAYS
I
80
I 0
FigurB 3. Acid Formation i n Furfural, with Free Access to Air
Figure 2. Effect of Storage Atmosphere on Furfural
The results in Table I indicate that the development of acidity in furfural is brought about by autoxidation. Conversely, then, by providing a n oxygen-free atmosphere it should be possible t o prevent these changes. Accordingly, duplicate samples of freshly distilled furfural were stored, one in contact with air and the other in a n atmosphere of nitrogen. Over a period of 9 weeks the air sample (Figure 2) deepened in color and exhibited an increase in acidity of 0.24 equivalent per litcr. I n the same period the nitrogen sample showed only a slight change in color and no measurable acid formation. Apparently then, color and acidity are interrelated, and oxygen is necessary for their development. AUTOXIDATION OF FURFURAL
Up t o this point it was not clear whether t h e formed acid had its origin in furfural or in a n impurity in the technical product Therefore comparative tests were made t o determine the relative rates of acid formation on storing iechnical furfural and pure furfural at room temperature with free access t o air.
To check the validity of this theory, a sample of technical furfural was allowed t o oxidize as before until the acidity curve sloped off; then i t could be assumed t h a t all of t h e postulated impurity had been oxidized. The formed acids were removed, and the furfural was purified by distillation and then exposed t o air a t room temperature. N o diminution in the tendency t o form acid could be detected, a result entirely a t variance with the “impurity” theory. On the contrary, the experiment established furfural itself as the acid precursor. Inhibition of autoxidation is strikingly shown by the curve a t the bottom of Figure 3. This result was obtained by washing technical furfural a i t h dilute, aqueous sodium carbonate, dehydrating under vacuum, and exposing the product t o air as in the previous experiments. Similar results have been obtained by addition of small quantities of tertiary amines t o technical furfural. CHARACTER O F ACID FORMED
As mentioned previously, acid formation in furfural has been attributed to acetic acid, according to one hypothesis, and to furoic acid, according t o another. During this investigation the data obtained precluded the possibility t h a t the formed acid was solely acetic or solely furoic acid. Autoxidized furfural is very dark in color, the intensity varying directly with the concentration of formed acids. On titrating aqueous solutions of such.furfura1, an indicator effect has been npted.’ Neither furoic acid nor acetic acid, added to furfural in the concentrations considered here, impart such a n indicator effect. At various intervals during the autoxidaFion of technical furfural (room temperature, free access to air), the distribution ratio
I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY
July, 1946
707
HC- C H
-
3 5% 0
FURFURAL
HC C - 0
io
c
W 0'0'
0
HC-CH
U
wc c-coon
3
z
'0' 7JROIC ACID
-2
-ka
-4
-.
t
I1
0
HC=CH
- I
a
11
It
1
1
0
0
W=CH 1
1 HC=CH)
1
H C C=O 0 OH p-FORMYLACRYLIC ACI
HC=CH I
/
C=O
HC C-O-
WR ACIDS
1 -H,O
TIME, DAYS
Figure 4. Distribution. of Acids between Furfural and Water during Formation
of the formed acid between furfural and water was measured. The results are shown in Figure 4. The distribution ratio of th(, small initial acidity in technical furfural \vas 1.1; as autoxidation proceeded, the acid distribution ratio decreased, went through a minimum value of 1.5, and then showed a n upward trend. The distribution ratio of furoic acid between furfural and water is 5.8; hence it, is evident t h a t the formed acid is not solely furoic. .4lthough the distribution curve in the early stages of autoxidation approaches a value reasonably close to that for acetic acid, 1.1, i t then shows a n upward trend. This indicate8 that, whereas acetic acid may be one of the products of the autoxidation of furfural, it is not the sole acid formed. These conclusions are further substantiated by the results shown in Table 11. It is evident t h a t acetic acid is almost entirely removed from furfural in a 607, overhead distillate. On the other hand, furoic acid remains almost completely in the residue in a similar diztillation. Autoxidized furfural exhibits an entirely different distillation characteristic, since it is apparent t h a t a mixture of volatile and nonvolatile acids is present.
SI ACIDS
0
0
Figure 5.
Summary of Furfural Autoxidation
IXffererices bet\vc-en the ttvo altlchydos on autoxidation can be seen further in previously published work. For example, Siverne et al. ( 3 ) demonstrated that under conditions which gave a 40uc yield of perhvnzoic nrid from t)cnznldchyde, no perfuroic
TECH. FURFURAL
COMPARISON WITH AUTOXIDATION OF' BENZALDEHYDE
The autoxidation of furfural appears t o be different from t h a t of benzaldehyde. The latter proceeds rapidly and almost completely, with benzoic acid as virtually t h e sole end product. Furfural, on the other hand, oxidizes much more slowly, and the reaction is interrupted a t a much lower level of acidity. For instance, under our highly accelerated conditions only 7 7 , of the furfural was destroyed before auto-inhibition occurred. I n addition, the end result with furfural is a mixture of acidic products and not one single acid.
TABLE I.
E F F E C T OF SVRFACE/VOLChlE
FORMATIOX
(Distilled furfural exposed 7 d a y s t o air a t room samples, 63.5 sq. cm.) Volume, ml. 10 Acidity, equivnlent/liter 0.006 Initial 0 100 Final 0.001 Increase
TABLE 11.
0
temperature:
surface of
20
40
0 006 0.076 0 070
0 006 0 059 0.0>3
ACIDS IS FE'RFCR.41. YCof Initial Acidity i n :
1'OLATILITY D I S T R I U r T I O S O F
Condition of Furfural Autoxidized Pure acetic acid Pure furoic acid
++
RATIOO N ACID
Initial Acidity, Equivalent/L. 0 26.5 0.270 0.245
Engler flask, atmospheric presqure.
60% dist.0 31.5 07.4 6.6
40% still tlottoms 68 5 2 6 93 4
TIME, HOUGS
Figure 6. Relation of Autoxidation and .4cid Formation
acid was obtained from furfural. 1Iil:is and XTcAlevy (b) showed t h a t the small amount of peroxide formed on autoxidation of furfural \vas not identical ivith authcntlc perfuroic acid synthesized from furoyl peroxide. .4 study is reported clseivhere ( I ) OIL the nature of the products of autoxithtiun of furfural, arid n mcchnnium is postulated to account for t hilir formation. Figure 5, which summarizes t.hat work, shows t h a t the only volatile acid protlucrd is formic acid. KO acetic acid \vas found, and although somc furoic arid waa
INDUSTRIAL AND ENGINEERING CHEMISTRY
108
Vol. 38, No. 7
be 1.17 equivalents of acid per mole of oxygen absorbed. This value is within the range predicted in the preceding section. EFFECT OF WATER
Figure 7.
Influence of Rater on Autoaidalion of Pure Furfural
isolatcd, it is a minor product of aiitoxidation: Of thc nonvolatile acids formed, ,8-forniyl:icrylic acid is i w l a t ~ din rc’lativrly sniall amounts b u t is a n import ant intcrmrdintc. It polymerizes rradily to form t h e WR acids (watcr-soluble rrsin acid.;), and these on dehydration yield the SI acids (solid-indicat or a(>icl$). T h r latter two acid fractionq arc tiark resinous matc.rials, and constitute the polymcr which clc~vclopsin furfural during normal storage. T h e SI acids are rmponsible for the indicator effcct previously noted. T h e differing solubilities and volatilities of the variou5 acids formed explain t h e results shown in Figure 4 and Tahle 11. In addition, this scheme of oxidation predicts that more than 1 but, less than 2 equivalents of acid should be formrd per mole of oxigen consumed. A verification of this prediction is apparcnt’ in t h e results described in the following section. OXYGEN UPTAKE
In order t o corrrlate oxygen uptake \vith acid formation, samples of furfural were shaken in a n oxygen atmosphcrcx at room temperature in a modificd Barcroft-Warburg apparatus t o measure the amount, of osygrn absorbed. Figure 6 s h o w the bcxhavior of technical furfural undcr these conditions. Although t h e autoxidation is considerably acccleratcd in this exprriment, the shape of t h e acidity curve closely parallels t h a t obtained in t h e earlier work using air but no agitation. -4rrlationship between oxygen uptake and acid formation is clearly demonstrated, and a t the end of 427 hours the ratio \vas found t o
1-.
,
I
I
PURE +Na,C03
Inhibition of the autoxidation due t o water was observed in some of our tests, and this effect was investigated more fully with the results shown in Figure 7. The curves show the relative rates of oxygen absorption when working with pure furfural and also with pure furfural plus varying amounts of added water (1.0, 3.0, and 5.07’). The experimental procedure was the same as t h a t described in the preceding section. It is obvious t h a t the degree of inhibition is related to the concentration of water. If water ncre the sole controlling factor, however, the oxygen uptake would be a linear function of timewhich it is not. Since it has been shown t h a t acidity increases with time, it is reasonable t o assume t h a t acid and water are jointly responsible for the inhibition. Such a combination nould impiy t h a t hydrogen ion is the eontrolling factor, a possibility which is not inconsistent with the results obtained in this expcxriment. For example, as the concentration of water decreases, an increasing amount of formed acid will be required t o attain a -hydrogen ion concentration sufficient t o inhibit autoxidation effectively. Such a hypothesis offers a n explanation as t o \vhy these curves slope off at different levels. EFFECT OF INHIBITORS
The technique previously described was used to study t h e effect of various inhihitors on the autoxidation of furfural. T h e results are illustrntcd in Figure 8. T h e induction period in the technical furfural curve is apparent, whereas pure furfural exhibits no such induction period. Tripropylamine as a n example of a tertiary amine and hydroquinone as a n example of a phenolic antioxidant are effective inhibitors for eithcr technical or pure furfural. Sodium carbonate treatment, on the other hand, is effective in inhibiting the autoxidation of technical but not of pure furfural. I n the latter case a short induction period is noted. Apparently treatment of pure furfural wtih sodium carbonate has caused the formation, though t o a lesser degree, of a type of substance which occurs naturally in technical furfural and which causes a long induction period. Such a substance might be a soap of a high-molecular-lyeight, acid. Since furfural, when distilled for purification, would be virtually free of such acids, Drily a minute amount of furfural-soluble soap could be formed on sodium carbonate treatment, and a short induction period would result. Technical furfural, with a greater amount of soap, would have a longer induction period, and sodium carbonate treatment might form enough additional soap t o cause an indefinitely long induction period. Tripropylamine is completc.ly soluble in furfural and is available for soap formation in eithcr pure or technical furfural, and thus is a n effective hibibitor in either. T h e quantit ies of bnse rcquired for inhibition of autoxidation need not be sufficient t o make the furfural alkaline. ACKNOWLEDGMENT
The authors wish t o thank F. X, Peters for encouragement, C. D. Hurd for counwl, and 1Iary L. Leslie for technical assistance. LITERATURE CITED
(1) Dunlop, A . P., and Swadesh, S I to be published. (21 Milas, N. A . . and SIcAlevy, A , , J . A m . Chem. SOC.,56,12f9,1221 I
I
0
I
100 LOO TIME, H O U R S
-
I
I
300
(1931)’ ( 3 ) Sneren, D., Findley,T. W., and Scanlan, J. T., I b i d . , 66,1026
O
(1914).
I
400
Figure 8. Effect of Inhibitors on Furfural Autoxidation
PREJF:UTFD before t h e Organic Chemistry session a t t h e Technical ConferSO( L I E T Y , Evanstnn, Ill., ence of the Chicago Section, A ~ i . n i c a h ’C H E I I I C ~ November 16, 1945.