1194
INDUSTRIAL AND ENGINEERING CHEMISTRY ACKNOWLEDGMENT
The authors gratefully acknowledge assistance received from many of their co-workers. I n the Fermentation Division, 2. Louise Smith and Max D. Reeves performed the analyses for 2,3butanediol, acetoin, and ethyl alcohol; Harold P. Vind and Lucille B. Czapla made the analyses for sugar; and Robert G . Benedict maintained and supplied pure bacterial cultures. Spectroscopic analyses of the starches were made by Eugene H. Melvin and Robert J. Bertc of the Analytical and Physical Chemical Division; and chemical analyses of the starches were made by Fontaine R. Earle and his assistants, of the same division. The piloeplant cooks referred to were conducted by James M, Van Lanen, Fred W. Tanner, Jr., and Virgil E. Sohns of the Agricultural Motor Fuels Division. The strain Aerobacter aerogenes NRRLB199, which was used in the experiments, was kindly made available by C. H. Werkman of Iowa State College in 1941. The commercial starch-converter liquors were supplied by courtesy of the Corn Products Refining Company.
ZEIN FIBERS
Vol. 37, No. 12
LITERATURE CITED (1) Assoc. of Official Agr. Chem., Methods of Analysis, 5th ed., p. 369 (1940). (2) Boeseken and Cohen, Rec. tmw. chim., 47,839 (1928). (3) Bortels, Bwchem. Z.,182,301 (1927). 23,996 (1931). (4) Churchill, IND.ENG.CKEM., (5) Collins, Lamar, and Lohr, U. 9. Geol. Survey, Water Supply Puper 658,61 (1932). (6) Goering, Iowa State Coll. J. Sci., 16,65 (1941). (7) Herrick, Hellbach, and May, IND.ENO.CHBIM., 27, 681 (1935). ( 8 ) Hill and Isaace, Brit. Patent 483,939 (1938); U. S. Pateit 2,224,912(1941). (9) Johnson, IND.ENO.CHEM.,ANAL.ED., 16,626 (1944). (10)Permutit Co., personal communication. (11) S h d e r and Hartmann, J. Biol. Chem., 45,365 (1921). (12) Silverman, Iowa Stat6 Coll. J . Sci., 17,120 (1942). 36,404 (1944). (13) Slotter and Langford, IND. ENO.CHEM., (14) Ward, Pettijohn, Lockwood, tLnd Coghill, J . Am. Chem. SOC.,66. 541 (1944); PRBSENTED (by title) before the Diviiion of Agricultural and Food ChemCHEMICAL SCCIICTY in New ietw at the 108th ,Meeting of the AMERICAN
York, N.Y.
e
Preparation by Wet Spinntng C. B. CROSTON, C. D. EVANS,AND A. K. SMITH Northern Regional Research Labomtory, A new method for pmducing textile fibers from zein and the equipment used in the process are described. Zein fibere which have high tensile strength and wool-like properties are produced from alkaline dispersions of zein. The spinning dispersions are formulated to give a product of high viscosity by the use of denaturing agents, such as alcohol or urea, or by aging the solutions. The filaments are coagulated in an acid bath in which salts may or may not be used. The coagulated filaments are given a mild formaldehyde precure prior to stretching and drying. The influence of the extent of precure on the load-elongation properties of the fiber and on fiber strength is determined. Shrinkage and water resistance of the fibers are controlled by acetylation followed by a strong formaldehyde treatment.
P
ROCESSES for making protein fibers from zein have been patented by Swallen (6) and Meigs (4)and reviewed briefly by Harold (3). The Swallen process involves the use of organic solvents such as ethyl alcohol in the preparation of the spinning solution. This process also uses 20% or more of plasticizer and 2 t o 5% of formaldehyde in the spinning dispersion, and extrudes the solutions through standard spinnerettes. The extruded filaments are solidified by evaporation of the solvent (dry spinning) or by a liquid coagulant (wet spinning). Meigs described a wet spinning process for a wide variety of proteins-namely, the “globulins, prolamines, and phosphoproteins”-and includes zein among the proteins he u8es for examples. The Meigs patent contains rather startling claims for stretching the fibers 300 to 20000J0, regardless of the type of protein used. Zein has been mentioned in other patents as one of the proteins from which fibers can be spun or regenerated. However, the absence in the literature of any information on their strength or other properties makes impossible their practical evaluation or aompaxison with other fibers.
U. S . Department of Agriculture, Peoria, I l l .
Zein is the only prolamine which has attained any substantial industrial importance. Its solubility in organic solvents has determined t o a large extent its commercial utilization for f3ms and coatings, and at present its largest use is aa a replacement for shellac. Likewise, the fiber spinning methods described in the literature use organic solvents in the preparation of spinning solutions. This investigation was directed toward the development of a practical process for making zein fibers by a wet spinning method and toward evaluation of the fibers. The equipment used in this work differs from that for spinning viscose rayon and other protein fibers only in details of construction and in arrangement. Data are presented on the preparation of zein-spinning dispersions, the effect of treating the fiber with formaldehyde prior to stretching (precure), and the effect of precure on the wet and dry strengths of the fibers. The use of trditional organic solvents for zein has been temporarily abandoned in the present investigation, and an aqueous alkaline dispersion for wet spinning has been developed. The object in making the change is to avoid the extra cost which arises in recovering organic solvents. It has been knownfor some time that zein is soluble in alkaline solution, but only recently the details of this type of dispersion have been investigated and ita limitations defined by Ofelt and Evans (6). In their studies on the solubility of zein in sodium and potassium hydroxide and quaternary ammonium hydroxide, Ofelt and Evans showed that a highly concentrated zein solution can be prepared in the pH range 11.3 to 12.7. A complete description of this dispersion procedure is being prepared for publication. AQUEOUS SPINNING DISPERSION
High-viscosity zein solutions of 13 to 16.5% aoncentration me required for spinning fibers, and the upper limits in this range
December, 1945
INDUSTRIAL AND ENGINEERING CHEMISTRY
1198
Figure 1 (Right). Fiber Spinning Assembly through First U n i t Operation A. Compreaaor
B. Prp.mnre remmal containing C.
D.
win solution
Metaring pump CaPdle filter
E. Coagulating bath F. Flat-murfad reel C. Prwuring bath
Figure 2 (Below). View of Processing Equipment A B. Stretchlngreel. C: Stretching bath D. Dryer
give the most satisfactory spinnerette operation. The most desirable viscosity for the spinning dispersions is at the incipient gel s t a g e t h a t is, at the point where the protein mass slowly flows together when cut, Properly aged dispersions are nontacky and appear to be a t the point of changing from viscous to plastic flow With this type of dispersion, there will be no formation of “bubbles” on the face of the spinnerette to interfere with continuous operation. Dispersions of high viscosity were prepared by using (a) an initially high zein concentration, (a) a dispersion of low concentrafion and low viscosity with aging, and ( 0 ) denaturing or gelling agents, such as urea or alcohol, or surface denaturation induced by air bubbles which were stirred into the dispersion. Some dispersions were aged at 25’ C.for a~long as 7 days with no evidence of ammonia formation or putrefaction. The dispersions of lower concentration and of initial low viscosity were cleared of air bubbles on standing, although dispersion with initial high viscosity required centrifuging to remove the air bubbles. The ectud viecosity, after the dispersion had passed through the spinning
assembly, was a matter of conjecture because of the temperature changes and of the notable thixotropic property of these protein dispersions. Urea and alcohol were the most frequently used denaturing agents, and their addition to alkaline dispersionsof low zein concentration decreased the required aging time from 7 to 2 days. In addition, the use of urea produced a fiber of improved softness. This result was probably due to the formation of a urea-formaldehyde resin in the fiber during the formaldehyde curing operation. The solvent action of urea wm necessary to form a clear spinning dispersion when quaternary ammonium hydroxides were used as the zein dispersing agents. If dispersions were spun without aging or- without chemical denaturation, their spinning properties were improved when a large amount of air ap nitrogen was incorporated into them by vigorous stirring. The air was removed prior to spirUring the dispersion. From this behavior it was concluded that surface denaturation had an effect on the spinning properties of the zein solution similar to that of chemical denaturation with urea or alcohol. Heat also served as a denaturing agent, but the diapersions were very sensitive to heat treatments; at about 40’ C. gelatior occurred rapidly and made this method impractical. Likewise, inorganic salts caused rapid denaturation and insolubilization which were difficult to control. Trace quantities of salts, however, may provide adequate means of obtaining the required amount of denaturation. Furthermore, the addition of formaldehyde to the spinning dispersions made the spun filsments very elastic. The spinning dispersion containing formaldehyde and the fibers freshly coagulated from these dispersions exhibited considerable tacakinesa. The tacky filaments had a tendency t o "plaster" or stick together during subsequent handling; the maximum concentration of formaldehyde which
1196
Vol. 37, No. 12
INDUSTRIAL AND ENGINEERING CHEMISTRY
After coigulation, the filaments were pulled through the bath and gathered into a tow on a large solid-faced reel, around which the tow was passed several times before falling into a mild formaldehyde precuring bath. Cold water waa sprayed over the fibers on the reel t o remove most of the acids and the salts adhering t o t.hem. The container holding the precuring mixture was rotated on a turntable a t such a rate that the tow entered faster than the peripheral speed of the container. The tow entered the container or “curing pot” near the edge, and settled to the bottom in zigzags which formed a doughnut-shaped cake. After the fiber had precured for the desired time, the continuous tow was withdrawn from the curing pot in the reverse direction from which it had entered, and was stretched and dried. Stretching was in a 6-foot tank of water between two large reels driven at variable speeds (Figure2) at a temperature between 45” and 60” C. The maximum amount of stretch before excessive breakage of filaments occurred was between 200 Figure 3. Continuous Fiber Dryer a n d Wind-up Bobbin to 350%. The limit to which the fiber could be stretched depended on the nature of the spinning could be used under these conditions was thus limited to about dispersion, the amount of formaldehyde precure, and the tempers2% of the weight of zein. However, the plastering difficulty ture of the stretching bath. could be overcome by the use of salts in the coagulating bath or From the stretching reel the wet tow was fed into the top of by the incorporation of denaturing agents, as mentioned above, the dryer shown in Figure 3. In the dryer the tow was pwed in the spinning dispersions, in which case the concentration of between four pairs of free-moving glass cylindrical reels in such a formaldehyde could be increased. fashion that about 120feet of the continuous tow waa exposed to a Table I gives formulas for preparing satisfactory spinning disvariable-velocity air stream, which entered at the bottom of the persions. The dry powder zein was first quickly and thoroughly dryer and was maintained at about 65” C: An aluminum wind-up wetted with part of the ice cold water by manual stirring. The bobbin, driven by the same motor which operated the stretching remaining cold water was added by mechanical stirring, and after reel, pulled the fiber through the dryer. The speeds of the a smooth suspension was attained, the dilute alkali was slowly stretching reel and bobbin, however, could bevarieddependently or added through a dropping funnel. If formaldehyde was to be independently, thus making possible an additional stretching of the used in the dispersion, it was added with the alkali. A high local fiber in the dryer. The tow waa in the dryer for about 5 minutes. concentration of either alkali or formaldehyde precipitated the During the preliminary work the fiber waa stretched without protein, but solid urea could be added at any stage in the disprecure by running it directly from the reel, immediately after persing process without harmful effects. The alcohol had to be the coagulating bath, through the tank of water at about 30” C. added after the zein had been dispersed in the plkali. Before t o a second reel. Considerable stretch was realized between the spinning, the dispersions were deaerated by centrifuging or by two reels, but the elongation obtained was lost when the fiber was allowing them to stand until free of bubbles; to remove the last removed from the reel for further treatment. The a m e n t s some amount of air, the dispersions were subjected to partial vacuum. times plastered seriously when fiber was allowed to accumulate on the reel. The continuous stretching and drying procedure prevented shrinkage of the fiber a t this stage of it,sproduction. EQUIPMENT AND PROCESS FOR SPINNING Figure 1 is a photograph of the apparatus used for spinning the fibers; Figures 2 and 3 show the equipment for continuous stretching and drying of the fibers. The zein dispersions were forced by air pressure, which could be controlled within a range of 0-200 pounds per square inch, through the metering pump, candle filter, and spinnerette, and into the acid coagulating bath. The spinnerettes were 1.5 inches in diameter and contained 100 to 3000 holes, with hole diameters in the range 0.003 to 0.006 inch. The coagulating bath composition most frequently used consisted by weight of 4.7% sulfuric acid, 3.1% acetic acid, and 4.4% zinc sulfate. The salt was not always necewary, and with many formulations either of the acids, in the proper concentratioqgave satisfactoryresults. When the spinning formulations resulted in tacky fibers, the salt of the coagulating bath hardened the surface of the fibers and prevented them from plastering. The alkaline zein dispersions behaved differently from other protein dispersions in that they did not provide sufficient lubrication for the metering pump, which would often freeze and shear the drive pin. This difficulty was overcome by cooling the pump with a stream of cold water.
PRECURING AND STRETCHING
The mild formaldehyde treatment introduced after coagulanon and before stretching of the fibers has been designated “precure”. The precure was found to be a necessary treatment of the fibers before they could be effectively stretched. Noncured filamenta were easily broken when handled at room temperature, and stretched indefinitely at elevated temperatures without showing
TABLE I. SPINNING DISPERSIONS OF ZEIN (% BY W ~ I G H T ) Formula No.
Commercial Zein. %
Water, % 69; 0 62.5 59.0 57.0 59.5 68.0 60.0 58.0 65.0 81.4
* Most frequently uaed disperaion.
1716 21.0 24.0 24.0
22.0
17.0 22.0 24.0 17.66
..
Formaldehude. . . % .0.5
...
0.5 1.0 0.5 0.5
Urea, % ._
... .... .. 1.5
... ...
2.5 1.5 3.0 1.5
...
i:i.
0.5
of ethyl alcohol. P l u s 2 : o ~of tetraethanol ammonium hydroxide.
b Plua 4 0 0
0.4 A’ NaOH. %
A$;& 168 24 2 3 7 72 4 2 72 48
December, 1945
INDUSTRIAL AND ENGINEERING CHEMISTRY
1197
a satisfactory indehyde and 0.2% ammonium chloride added as catalyst. The treatment of fibers by this low concentration of formaldehyde is crease in strength. very mild and eliminates the time of treatment as a critical The extent of the precure, aa deterfactor. Good fibers could be recovered from this bath after mined by the conseveral days of treatment at room temperature. In another series of experiments the relation between the load c e n t r a t i o n of 3"Diam. and elongation of wet fibers for different curing treatments was formaldehyde and Glass Tube1 by the time and studied. The wet samples of fibers were precured in the regular curing bath (described above) at room temperature for 0, 4.5, t e m p e r a t u r e of 24,48, and 96 hours. The load-elongation data were obtained in treatment,greatly water a t 45' C. with equipment illustrated in Figure 4. I n makinfluenced t h e ing the elongation measurements, increment weights were added character of the 3 3*Diarn. every 5 minutes to a 500-filament tow of approximately 4000 fibers. A lowTubing deniers, and the increase in length was noted for each added cured sample weight. The loading was continued until the fiber broke, and could be given a high percentage of the resulting data are plotted in Figure 5. Fiber A used in obstretch which retaining curves 1 and 2, and fiber B used for the other four curves sulted in flexible were from different spinning dispersions. The load-elongation data are only relative, depending on the fibers of small rate of loading; nevertheless, they show clearly an increase in diameter having tensile strength with increased precure and elongation. The large s h r i n k a g e curve for each precured fiber is linear up to a certain load, at values in water. Fiber A highly c u r e d which point there is a rather sharp break to another linear porB u n d l e 7 tion of greater slope. Except for the transition portion, any sample would 1 4 curve can be represented quite well by two intersecting straight tolerate relatively less s t r e t c h i n g lines. As the amount of precure was increased, the slope of 45.C. both of these lines increased, and the point of intersection appears and would proat a higher load value and lower percentage elongation. Comduce coarser fibers parison of curve 2, for the 18-hour precure of fiber A, with the of higher tensile curves of fiber B indicates that the spinning conditions had a strength (Figure pronounced effect, on the slope of the curve and the properties 5). However, exof the fiber. In contrast to the precured samples, the loadcessive curing reelongation curve for the noncured sample is a straight line with sulted in brittle Figure 4. Schematic Diagram of more than 600% stretch, and the fibers exhibit very low tensile fibers. Apparatus for Measurin the Loadstrength and very small increase in strength with elongation. The effect of Elongation Values of &in Fibers In another series of experiments the increase in the dry and wet time, concentre, strengths of the fiber with increasing stretch was measured. The tion, a n d temfibers were stretched in water a t 45" C. (with equipment illusperature of the formaldehyde precure treatment on the proptrated in Figure k). After stretching, the wet fiber bundles erties of the fiber was investigated in a series of experiwere fastened to a frame for drying. The strengths were determenta in which these factors were varied. Temperatures of 25" and 46" C. were used with concentrations of 2.6 to 40% formaldehyde in the preI I I cure treatment; then the bundles were washed I I I I I I I I 1 I in warm water and stretched by the direct application of a given weight per filament on the end of tbe suspended bundle. In these experiments the elongation varied inversely with the amount of cure, and both the amount of stretch and amount of cure influenced the strength of the fibers, As would be espected, increasing the temperature or concentration of formaldehyde increased the rate of C U I ~ . From these tests the following generalizations could be made: A t 25' C. a treatment of 10% formaldehyde for 7.5 minutes gave a cure almost equivalent to that of 2.6% formaldehyde for 1 hour. Likewise, a treatment of 20% formaldehyde for 16 minutes corresponds to that of 5% formaldehyde for 2 hours. The tensile strength of fibers from all four sets of conditions was about 1.3 grams per denier; however, the stretch for the first two sets of conditions waa about 300%, but for the last two it waa Elongotion in percent only 170%. A precure of 2.5% formaldehyde Figure 5. Effect of Amount of Formaldehyde Precure on Loadat 46" C. for 30 minutes was similar to that of Elongation Values of Zein Fibers 10% formaldehyde at 25 O C. for 30 minutes, 1 FiberAnithno eaum 4. Fiber B after 24-hour p m a ~ The precuring solution used most in the presa: FW A d* i a L pr6. Nbsr B after &hour p r e c ~ 8. Flber B after 4.5-hour preoum 6. Fiber B after %-hour prsowe ent work consisted by weight of 1.3% formal-
3 '4" i
S
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
1198
VoI. 37, No. I2
with the more easily demonstrated birefringence of the fibers, shows that a considerable degree of orientation has been obtained.
1
0
1 1 1 1
Figure 6.
50
I
l
l
I
I
I
I
100 150 200 250 Elongotion in percent
I
I 300
I
I J
350
Relation of Tensile Strength of Fibers to Elongation 1. Fiber B after 4.5-hour precure 2. Fiber B after 24-hour precure 3. Fiber B after 48-hour precure
mined at 70’ F. and 50% relative humidity on the Scott IP-2 Serigraph by the bundle method. The resulting strength test data (plotted in Figure 6) show a linear relation between elongation and strength. The plotted data show no indication of a change in the slope of the strength curves to correspond to the break in the load-elongation curves of Figure 5. The strongest fibers were obtained at the maximum stretch before breaking. The dried fibers, when immersed in water a t 45’ C., quickly returned to their original unstretched length. The load-elongation curves in Figure 5 are the same shape as the corresponding curves for wool and rubber given by Astbury ( I ) except that the Hooke law region, as described by him, is missing. A series of curves similar to those in Figure 5, but having in addition an initial linear part of high slope similar to Astbury’s, were recorded on the chart of a Scott IP-2 Serigraph. The recordings were made by using single wet tows of fiber B at about 21’ and 35’ C. The rate of loading was much greater than that described above. Astbury explains the 100% extensibility of wool (100% relative humidity) and the loadelongation curves by the transformation from the a- to the 8keratin structure. However, i t is difficult to explain the high extensibility and recoverability of wet zein fibers by the same theory. The elasticity of wet zein fibers is somewhat like that of rubber, and it may be explained by the tendency of the molecules to assume the most random distribution (E). If the break or change in slope of the load-elongation curve was due to plastic flow or to molecular slippage, tensile strength probably would not be a linear function of elongation. A detailed study of the x-ray diffraction diagrams of zein fibers has not been carried out at this laboratory, although sufficient evidence has been secured to indicate that some fibering occurs as evidenced by the arcing in the x-ray pattern of the more highly stretched fibers. Diffraction diagrams of the oriented fibers are similar t o the so-called @-pattern obtained from the stretching of hair and wool. The evidence from the x-ray studies, along
PROPERTIES OF ZEIN FIBERS
The process for the production of zein fibers is not yet fully developed. However, a preliminary survey of fiber properties is presented to indicate the excellent possibilities for the establishment of a commercial fiber from zein. Fibers obtained in the manner described above are high in wet and dry strengths. Dry strengths of 1.87 and wet strengths of 0.75 gram per denier have been obtained. I n comparison, the better known spun protein fibers have dry tensile strengths of 0.6 to 0.8 gram per denier and wet strengths of about one third the dry values; the dry tensile strength for a good grade of wool is about 1.3 grams per denier, The zein fibers have high resilience and are very flexible, although no plasticizer is used. Fibers finer than 0.6 denier ( 7 . 5 in ~ dimeter) have been produced by applying a high stretch in a heated coagulating bath. As stated earlier, the fibers prepared by the above method shrink to their precured length when immersed in hot water. This shrinkage can be greatly reduced by treating the fiber with strong formaldehyde while it is held under tension. Even so, there is some loss in strength from the maximum values, and the fibers from this “post cure” measure 1 to 1.25 grams per denier. Preliminary studies on dyeing the zein fibers showed that, after postcuring with strong formaldehyde, the fiber does not withstand the severe acid conditions normaIIy used in dyeing and processing wool. I n this respect zein fibers are similar to casein fibers, and like casein, the dyeing properties, in general, are greatly improved by acetylation. The work on acetylation will be reported in a future publication. It is sufficient to say here that acetyl contents above 2’% are necessary to give satisfactory resistance to acid-dye bath conditions. Acetylation followed by curing with formaldehyde has given a satisfactory control of shrinkage, and in some cases, the shrinkage has been lowered to 4% as determined by boiling for 15 minutes in water. From a study of the acid and alkaline peptization curves of zein, the fibers wc-xld be expected t o offer considerable resistance to the usual alkaline conditions encountered in the use of soaps and scouring powders. Acetylated and properly cured fibers have shown excellent resistance to boiling in buffered solutions of pH 8 to 9 and, in some cases, have shown no loss in strength upon boiling for 2 hours at this alkalinity. Fiber texture, feel, and appearance were not impaired by this boiling treatment. ACKNOWLEDGMENT
The authors acknowledge the assistance of N. Cyril Schieltz of this laboratory for the x-ray analysis of zein fibers. LITERATURE CITED
(1) Astbury, W. T.. “Fundamentals of Fibre Structure”, London, Oxford Univ. Press, 1933. (2) Guth, Eugene, “Surfaoe Chemistry”, p. 103, Lanoastar, Pa., Eoience Press Printing Go., 1943. (3) Harold, B.A., Am. Dyeatuff Reptr., 29, 53-7 (1940). (4) Meiga, F. M.,U.8.Patent 2,211,961(1940). (6) Ofelt, C.O., and Evans, C . D., in preparation. (6) S w d e n , L. C . , U. 8.Patent 2,166,929(1939).
Properties of Granular and Monocrystal-. line Ammonium Nitrate-Correction An unfortunate error has been foupd in this article by W. H. Ross, J. Y. Yee, and S. E. Hendncks in the November issue. On page l O q the pictures were inadveftently revemed. i n other words, as printed, Figure 1 is at the nght and Figure 2 is at the left.