PRODUCT AND PROCESS DEVELOPMENT creased the time t o attain 60% conversion by 15% and the Mooney viscosity was increased by 8 points. These results verify a prediction that could be made from the data in Tables V . and VI. None of the contaminants studied influenced the stability of the latex as judged by visual examination.
J. L. Hutson, and R. D. Barnett, 111, of The General Tire & Rubber Co., Baytown, plant supervised the experimental work performed a t Baytown. literature cited (1) Back, A. L., IND.ENG.CHEM.,39, 1339 (1947). (2)
Vinylacetylene and butadiene dimer seriously interfered with GR-S-100, GR-S-101, and X-636 polymerization; acetaldehyde retarded the X-636 system
Of the ten impurities studied, vinylacetylene and butadiene dimer seriously affected polymerization in the GR-S-100, GR-S101, and X-636 systems and acetaldehyde was a powerful retarder in the x-636 system. The amounts of these impurities that caused either a 10% increase in time t o attain 60% conversion or a 10-point variation in Mooney viscosity were so small that large enough fluctuations in the concentrations of these contaminants might conceivably occur in the manufacturing or in the recycling step of the monomer t o cause production of off-specification rubber. The information presented should prove useful in explaining and preventing some polymerization reaction anomalies experienced by copolymer plants. Compensating for the contaminants present in the butadiene is not a practical means of process control. A uniform supply of butadiene charged t o the reactors and containing minimum quantities of objectionable contaminants is the best safeguard against polymerization difficulties from impurities. Acknowledgment
1953, CD-3055. (3) . , Burke, 0. W.. Starr, C. E.. and Tuemmler, F. D., “Light Hydrocarbon Analysis,” p. 18, Reinhold, New York, 1951. (4) Cram. L. H.. Rubber Chem. and Technol.. 19. 1092 (1946). (55 Frani, R. L., Blegen, J. R., Inskeep, G. E:, and Smith, P. V., IND.ENG.CHEM.,39, 893 (1947). (6) Fryling, C. F., and Mitchell, L. A., Phillips Petroleum Co.,
private Oommunication to Federal Facilities Corp., Office of Synthetic Rubber, Dec. 18, 1845, CR-955. (7) Fryling, C. F., and Pritchard, J. E., Phillips Petroleum Co., private communication to Federal Facilities Corp., Office of Synthetic Rubber, Oct. 24, 1950, CR-2495. (8) Harrison, S. A., and Meincke, E. R., And. Chem., 20, 47 (1948) (9) Hobson, R. W., and D’Ianni, J. D., IND.ENG.CHEM.,42, 1572 I
(1950). (10) Houston. R. J.. Anal. Chern.. 2 0 . 4 9 (1948).
(llj Howland, L. H., Reynolds, J. A‘., and Brown, R.
(12) (13)
(14) (15) (16) (17)
The work discussed herein was performed as a part of the research project sponsored by the Federal Facilities Corp., Office of Synthetic Rubber, in connection with the Government Synthetic Rubber Program. A. L. Hollis of the Office of Synthetic Rubber and W. B. Reynolds of Phillips Petroleum Co. participated in outlining the program and the methods of investigation. T. J. Kennedy of Phillips Chemical Co., Copolymer Section, supervised the additional experimental work with acetylenes cited in the body of the report. G. D. Hanson,
Boyer, T. W., Kentucky Synthetic Rubber Corp., Progress Rept. Federal Facilities Corp. for period ending Sept. 30,
(18) (19)
W., IND.ENG.
CHEM.,45, 2738 (1953). Kolthoff, I. M., and Dale, W. J., J . Polymer Sci., 3, 400 (1948). Kostas, G. J., and Faull, J. H., The General Tire & Rubber Co., private communication to Federal Facilities Corp., Office of Synthetic Rubber; McCleary, D. C., U. S. Rubber Co., private communication to Federal Facilities Corp.; Fryling, C. F., IND. ENQ.CHEM.,40, 928 (1948). Phillips Petroleum Co., progress rept. to Federal Facilities Corp., Office of Synthetic Rubber, December 1950, CD-2010. Ibid., Jan. 31, 1953, CR-3212. Phillips Chemical Co., Copolymer Section, unpublished data, January 1950. Robey, R. F., Wiese, H. K., and Morrell, C. E., IND.ENG. CHEM.,36, 3 (1944). Schiller, J. C., and Seyfried, W. D., “Product Quality Studies of Butadiene,” RUR SRlO and RUR SR99 (Technical Reoort. Humble Oil and Refining Co.). Seotember‘l9. 1944. St. John, W. M., Uraneck,C. A:, and Fryling, C. F., J . Polymer
Sci., 7, 159 (1951). (20) Whitby, G. S., “Synthetic Rubber,” pp. 260 and 683-4, Wiley, New York, 1954. (21) Whitby, G. S., Wellman, X., Flouts, V. W., and Stephens, H. L., ~ N D ENG. . CHEM.,42, 445 (1950). RECEIVED for review October 14, 1954.
ACCEPTED April 27, 1955.
END OF PRODUCT AND PROCESS DEVELOPMENT SECTION
Dimensional Stabilization of Textile Fabrics N. A. MATLIN AND A. C. NUESSLE Textile Applications Laboratory, R o h m & Haas Co., Philadelphia 37, Pa.
HE most important end use of textile fabrics is in the manufacture of clothing. Only a few decades ago i t was common practice t o buy garments several sizes too large, in the hope that after a few washings they would shrink t o an approximate fit. Today, however, both common sense and fashion decree a reasonable immediate fit. One of the first major advances in the dimensional stabilization of textiles was Sanforizing, a mechanical preshrinking process in which cotton fabrics are compressed in the warp direction t o such a degree t h a t on subsequent washing they shrink little or not a t all. The utility of the process was so evident that “Sanforized” is now a household word. More recently, chemical September 1955
processes have been developed which extend the range of fiber and fabric types t h a t can be stabilized. Most of these treatments minimize t h e tendency of the fabric to stretch as well aa t o shrink. This dual resistance is termed dimensional stability. The present discussion centers almost entirely on chemical processes that impart such stability to fabrics. The primary emphasis throughout, however, is on the control of shrinkage. The mechanics of shrinkage will vary with the type of fabric, and especially of t h e fiber. Cotton and rayon shrink because of fiber swelling, wool because of felting, the hydrophobic synthetics because of thermal disorganization and mechanical distortion. All fabrics shrink when strains developed in manufacturing are
INDUSTRIAL AND ENGINEERING CHEMISTRY
1729
INDUSTRIAL AND ENGINEERING CHEMISTRY
1730
released, but wool and acetate rayon particularly exhibit this effect. CELLULOSICS
Vol. 47,No. 9
amounts being required for equal stabilization. Table I1 illustrates the effect of formaldehyde-urea ratio. Urea-formaldehyde Yesins used i n the textile industry have formaldehyde-urea ratios of 1.25 t o 2.5. Melamine-formaldehyde resins are generally used a t formaldehyde-melamine ratios greater than 2.0. As the formaldehyde ratio is high, it is not as critical a variab1e:as with the urea resins.
Cotton and rayon fibers are composed almost entirely of cellulose. They are very hydrophilic and can imbibe large amounts of water. Although a completely satisfactory cause-and-effect relationship between fiber swelling and fabric shrinkage has never been developed, one early theory explained shrinkage on the basis of yarn and fabric geometry. Because a cellulose Table 11. Effect of Formaldehyde-Urea Ratio on fiber increases in both length and diameter when wet, the yarn Stabilization of Rayon Challis also increases in diameter but decreases in length. The yarn Warp Shrinkage, becomes shorter because the increase in fiber length is not enough Resin Solids F/U Ratio % t o compensate for the greater distance the fibers must travel 0 ... 9.2 along the swollen yarn helix. The fabric shrinks for two reasons: 5 1.1 4.1 10 1.1 3.6 this actual shortening of the yarns and the greater distance neces15 1.1 2.8 sarily traversed by the yarns about themselves (1%). This ex5 1.3 3.5 10 planation is certainly incomplete. A stretched rayon or cotton 1.3 2.8 15 1.3 0.8 fiber will actually shrink on wetting as locked-in strains are 5 2.0 2.5 released. On the other hand, a completely relaxed rayon fabric 10 2.0 1.7 15 2.0 0.8 will stretch on wetting, shrinking only when it is dried t o 40% Catalyst, 4% (NHdzHPOd,based on resin solids. Cured 10 minutes at or lower moisture content. Nevertheless an empirical relation300° F. Washed 40 minutes a t 200' F. in 0.1% soap and rinsed. ship between swelling and shrinkage is known t o exist. Table I shows roughly such a correlation for fabrics made of several common fibers, and Figure 1, based on work of Gagliardi and Methoxy- derivatives of the simple methglolureas and methylNuessle (IS), shows a correlation for a rayon fabric treated with various thermosetting resins and reactants. Thus, the hyolmelamines may be prepared by reaction with methanol. Examples are: pothesis of shrinkage due to swelling can be very useful, even though - other factors undoubtedly . operCHaOCHz\ /X\ /CHaOCHa ate-e.g., the tendency of yarns to b e CH~OCH~NHCO~HCH~OCH~ come more highly crimped (in one or both ,N-C C--N\ I I1 Dimethoxymethylurea directions) when fabrics, even loosely CH~OCH/' N, \CH~OCH~ woven ones, are tumble-washed.
h\
Table I.
Relationship between Fiber Swelling and Fabric Shrinkage
Fiber Rayon Cotton Acetate Nylon
Swelling Very high High Low Very low
Fabric Shrinkage High Moderate Low Negligible
Many substances may be used t o reduce the swelling of cellulose fibers with concomitant reduction of fabric shrinkage. Some of the more important types are: Urea-formaldehyde resins Melamine-formaldehyde resins Cyclic urea-formaldehyde resins Thiourea-formaldehyde resins Formamide-formaldehyde resins
Phenol-formaldehyde resins Ketone-formaldehyde resins Formaldehyde Glyoxal Higher dialdehydes
Urea- and Melamine-Formaldehyde Groups. Certain of these types represent entire families of compounds with individual members of varying composition and utility. The urea- and melamineformaldehydes are two such groups. Some variations in composition, apparently minor but actually exerting considerable influence on the properties of the specific substances, are formaldehyde ratio, methoxyl content, and degree of precondensation. Because urea has four replaceable hydrogens, as many as four molecules of formaldehyde can theoretically combine with it. Actually only monomethylol- and dimethylolurea are well defined. T h e typical urea resin paste is a mixture of t h e two. Triand tetramethylolurea are unstable and tend t o liberate formaldehyde. Of the six methylolmelamines, at least four are known. T h e stabilizing efficiency of the urea resins increases sharply as t h e formaldehyde-urea ratio is raised from 1 to 2. Monomethylolurea is less effective than the dimethyl01 compound, larger
NHCH~OH Tetramethoxymethylmonomethylolmelamine The degree of methoxylation has little effect upon stabilizing action. Methoxylated resins possess greater storage stability, however, as well as better bath life Because of the enhanced stability of the methoxylated compounds, they sometimes require slightly more catalyst or a stronger one, or a more severe cure than the corresponding unalliylated types. Methoxylation also facilitates the preparation of precondensates of higher formaldehyde content and those of higher molecular weight. Urea- and melamine-formaldehydes may be precondensed over a fairly wide range of molecular weights during manufacture They are sold a s monomers, dimers, and linear (or B-stage) polymers. The smaller molecules penetrate rayon and cotton fibers; the larger ones are confined t o the surface. While both types stabilize cellulose, the location of the resin controls such properties as fabric handle or feel and crease recovery. Thermosetting surface resins stiffen fabric and do not appreciably enhance crease recovery. With proper control of formaldehyde ratio, methoxyl content, and degree of precondensation a wide range of compositions becomes available. Particular resins will vary in reactivity, penetrating ability, and over-all efficiency. I n many instances a certain urea resin will thus resemble more closely its melamine analog than it will other urea types. Thermosetting resins and reactants can reduce swelling by three means. The substance may form a surface coating, i t may polymerize within t h e cellulose and cross-link adjacent polysaccharide chains through hydrogen bonds, or it may form interchain covalent cross links. I n practice the three processes are not mutually exclusive. Some surface resin is almost always deposited when the textile is treated with a n internally polymerizing substance. Undoubtedly, there is some covalent cross linking as well.
*
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
September 1955
A precondensed linear polymer gives surface resin. The polymer fills fiber crevices and forms a sheath which is bonded to the surface of the fiber. Such resins are available as sirups of about 50% nondiffusible content-half the resin penetrating the fiber, half being confined t o its surface. The surface resin stiffens the fabric, as yarns are bonded t o yarns and fibers to fibers. Such immobilization undoubtedly contributes t o the stabilizing effect. Conversion of certain resin formers-especially the melamine-formaldehydes-to colloidal dispersions increases surface deposition (19,46, 6 5 ) . T h e colloid is formed by adding to 2.5 a suitable amount of a carboxylic acid ( K i of 1.4 X X 10-4) to the resin and allowing the mixture to stand. The well known bluish cast is taken as indication of conversion. Most melamine types can be treated in this way, but only a few ureas will form satisfactory colloids. Surface deposition is generally regarded as a less efficient means of stabilization than internal deposition. The difference is ordinarily small, however, even if the same large concentration of catalyst (required for the formation of a suitable colloid) is also used for the application in monomer form. This is illustrated in Table 111, which compares a 10% solids application of a methylated melamine-formaldehyde resin in colloidal form, in solution using the same quantity of acid as in colloidal form but with no waiting period to convert the colloidal state, and in solution but using the normal type and amount of catalyst. The resistance t o flat abrasion is better in the fabric treated with the external resin. This stems from the fact that the surface resin has little effect on the viscoelastic properties of the fibers. I n fact, in some instances, and b y some methods of test, the resistance t o flat abrasion can exceed that of the untreated control (52).
1731
zation, even though formaldehyde itself will not react with the cellulose under these conditions. These effects are shown in Table IV. Whenever a formaldehyde resin is cured onto cellulose, some covalent cross linking is likely t o occur. How much takes place depends on the relative availability and reactivity of the resin The -OH is generally present -N€Iand the cellulose -OH. in greater amount, but this is compensated for by the greater reactivity of the -NH--. For example, mere exposure of dimethylolmelamine to moist air, even in the absence of catalyst, is enough to initiate polymerization. When dimethylolmelamine is cured onto cellulose, therefore, some polymerization will almost certainly occur in addition t o cross linking with cellulose hydroxyl groups. With a typical urea-formaldehyde paste of formaldehyde ratio of 1.5 or less, there is ample opportunity for both polymerization and cross linking with the cellulose. With monomers of higher formaldehyde ratio, however, cross linking is favored. This is a possible reason for the higher efficiency of dimethylolurea when compared with compounds of lower formaldehyde content.
-I 8p
/
10
d
u W
7
a
Table 111. Effect on Shrinkage of Phase State of Resin Standing Time of Catalyst Bath, Hours 7 . 5 % HOAc 1.5 7.5% H0.4~ 0 0.5% NHiCl 0
Phase Colloidal Solution Solution
Abrasion Shrink- Resistance, Diffusible age, TBL Solids, % Warp, % Cycles < 10 2.9 2560 80
2.1
1180
2.3 1650 7.2 5090 10% (solids) methylated melamine-formaldehyde resin applied to rayon challis. Washed 60 minutes at 180' F., including rinse.
...
...
...
80
...
0
Tahle IV.
Stabilization of Spun Rayon under Conditions Favoring Resin Formation Warp Shrinkage,
Agent
15% monomethylolurea Control
15% UF paste
7%
Catalyst 1 % NaHCOs
Cure 10 minutes 300' F.
0 . 5 % xH4C1
1 dav a t room temw. 10 days a t room temp. 10 minutes a t 300' F.
Control Washed BO minutes a t 180' F., including rinse.
at
4.2 8.9 6.9 2.1 1.8
8.3
Internal resin is assumed t o be present in the amorphous regions of the fibers, binding the cellulose chains b y means of hydrogen bonds. While covalent reaction cannot be eliminated with certainty, conditions can be adjusted to suppress it strongly. Yet under such conditions excellent stabilization is obtained. For example, monomethylolurea cured on fabric with an alkaline catalyst such as sodium bicarbonate should tend t o polymerize rather than cross link, as reaction between cellulose and formaldehyde is not favored a t high pH. Furthermore, i€ the monomet hylolurea were to react with the cellulose, i t would not e a d y form a cross link, because i t is monofunctional. I n spite of this, the fabric is stabilized t o a reasonable degree. Similarly, treatment of cellulose with a urea-formaldehyde monomer a t room temperature under mild acid catalysis gives good stabili-
10
20
30
40
WATE-R
SO
60
70
80
IMBIBITION
90 100
(%)
Figure 1. Empirical correlation between water imbibition and fabric shrinkage
I n evaluating a chemically stabilized fabric, resin efficiency is not the only thing to be considered. I n fact, the shrinkage of a treated fabric depends not just on resin effect, but on the degree of stretch of the fabric during treatment as well. This was demonstrated in this laboratory some years ago (51). An 18inch length of rayon challis, previously dried under tension, was washed and allowed to dry in a relaxed state. It shrank to 16 inches. Five other 18-inch strips were padded through a solution of a urea-formaldehyde monomer and were dried and cured while being held under varying degrees of tension. On laundering, the amount of shrinkage was found to be highly dependent on the amount of tension a t the time of treatment. The fabric under greatest tension exhibited the greatest shrinkage. The fabric under least tension, however, actually stretched during laundering. This is shown in Figure 2. Another important consideration is the durability of the stabilizing effect. Because of their marginal resistance to hot alkaline washing, simple urea-formaldehydes have found wider use in imparting crease recovery to spun rayon fabrics which will be dry cleaned than in stabilizing such fabrics. The durability of such finishes may be enhanced by changes in chemical structure: alteration of formaldehyde ratio, cyclization, etc.
INDUSTRIAL AND ENGINEERING CHEMISTRY
1732
There are also a number of side effects t o be considered. Fabrics treated with simple urea-formaldehydes rarely discolor. T h e cost of processing is low, as plain urea resins often sell for less than 15 cents per pound. A cardinal disadvantage of the straight ureas is their tendency to pick up chlorine from bleaching liquors, retaining it as chloramides ( 4 7 ) . On subsequent ironing of the fabric, hydrochloric acid is released and the fabric is tendered. The degree of damage depends not only on the amount of chlorine initially retained and on the ironing temperature, b u t also on the type of fiber. Cotton is more prone than rayon t o suffer such degradation.
& I
\
\
Vol. 47, No. 9
tached-cyclic ureas act predominantly as cross linkers, the cellulose chains being bonded through the methylol or methylated methylol groups. Some polymerization is likely t o occur, however, through either methylene or ether bonds. A few of these compounds, particularly the dimethylolhydantoin derivative, are believed t o act as formaldehyde carriers, much of the nitrogenous portion of the molecule being readily removed on washing. All are relatively nonvolatile and so may be used on both cotton and rayon without danger of over- or undertreatment. Most of them do not yellow on application. They have good durability t o laundering. Retention of chlorine from bleach baths should be negligible with the cyclic ureas. Actually each one does pick up a small amount of chlorine, probably on nitrogens which have become unblocked through loss of formaldehyde. The extent of damage t o a fabric on chlorination and ironing is a complex function of the chlorine concentration, fabric-bath ratio, ironing temperature, buffering power of the resin, etc. (47). At least one cyclic urea is widely used, b u t opinions vary as t o whether it may be regarded as sufficiently chlorine-resistant t o withstand all practical conditions of chlorination, or merely the less severe ones. I n common with a few other types-notably the melaminescertain cyclic ureas have been found very effective in producing durable glazed and embossed finishes. I n such finishes the fabric is pressed or shaped into a new form by calendering. T h e function of the resin is to make the fabric reslstant t o washing, during which i t would otherwise revert t o its original uncalendered form. Thus, the resin primarily acts t o prevent fiber swelling. I n the case of embossing, the resin not only prevents fiber swelling, b u t also inhibits longitudinal deformation. Thiourea- and Formamide-Formaldehyde Resins. Thiourea resins have found little use in stabilization, probably because of their relatively high cost and the lack of distinct and unique advantages. Also, they are degraded by chlorine. A fabric stabilized with a thiourea-formaldehyde would begin t o shrink after several chlorinations. At present they are used only t o give nylon a crisp, flame-retardant finish. Dimethylolformamide was used in Germany during World reduce the water imbibition War I1 t o “hydrophobize”-Le., of-rayon clothing. It has never been widely exploited in this country. Formaldehyde. The reaction of cellulose with formaldehyde was discovered in 1903, yet it was not until 1926 t h a t it was recognized that cross linking of adjacent cellulose chains resulted. Because of t h e difficulty of controlling the cure, the substance has only recently found wide use in stabilization. Studies of t h e reaction of formaldehyde with viscose rayon ( 2 4 )have shown t h a t both methylene and polyoxymethylene links are formed. Strongly acid conditions are needed t o catalyze the reaction Oxalic acid or a mixture of sodium sulfate and sulfuric acid a t p H less than 2 is commonly used. I n commercial processes the formaldehyde is used in conjunction with a water-soluble polymer such as poly(viny1 alcohol) or hydroxyethylcellulose, presumably to help control the volatility of t h e aldehyde. I n the laboratory
yY 15
I
2
3
4
5
F U L L S A N F O R I Z E WASHES Figure 2. Washed dimensions as a function of cured length 15% Urea-formaldehyde paste
While melamines are somewhat more durable to laundering, and are free from discoloration during application, they cost more. They yellow on chlorination, though this can be minimized b y the judicious choice of a curing catalyst. Strength loss, on ironing following chlorination, is usually lower with melamines than with ureas, not only because chloromelamines are more difficult t o decompose than the urea analogs, b u t because the more basic melamines buffer the hydrochloric acid and protect the cellulose. Where chlorination is no problem-for example, with colored fabrics which would never be bleached-either urea or melamine resins may be used t o obtain stabilization, crease recovery, or durable glazed or embossed effects. Cyclic Urea Resins. Cyclic urea-formaldehyde resins are enjoying rapidly increasing use as cellulose modifiers. Typical examples are: CHSOCH2N-CO-NCHzOCHa
I
I
CHzOCHg Dimethoxymethyluron
HOCH2NCONCH20H
I
CHa-C-CO
1
I
CHa Dimethyloldimethylhydantoin
C H ~ O C H Z N C O N C H ~ O C H ~ CHaOCHzNCONCH2OH
&H--dW CH30CH2hOdCH20CH3 Tetramethoxymethylglycoluril
dH2--bH2 Methylolmethoxymethylethyleneurea
Though they may be classed as being derived from urea, they are not necessarily manufactured from this substance. Because the nitrogens are blocked-Le., have no active hydrogens at-
Table V.
Effects of Adjuncts in Reaction of Formaldehyde with Rayon
Warp Shrinkage, % Abrasion ResistTreatment 1 wash 5 washes ance, TBL Cycles Complete 1.4 2.2 1230 Without HEC 1.4 2.0 1210 . ... 63 Without P\TazSOa Control 8.0 9.7 3000 Formulation. 1.0% hydroxyethyloellulose (HEC) 3.7% formaldehyde 0 . 3 % softener 5.0% NanSOa HzSOa to pH 1.3 Padded,.dried,, and cured 5 minutes a t 300° F. Washed 60 180° F., including rinse.
. .
Fabric Color White White Brown White
minutes at
September 1955
*
INDUSTRIAL AND ENGINEERING CHEMISTRY
good results can be obtained with or without such polymers. T h e data in Table V suggest t h a t the effect of hydroxyethylcellulose is neglible. Sodium sulfate, however, must be present if acid damage is t o be prevented. At first sight this is puzzling, since t h e p H was 1.3 in each case and less sulfuric acid was required when the salt was absent. It is probable, however, t h a t the salt exerts a protective action under the anhydrous conditions prevailing a t the elevated curing temperature. Regardless of the exact bath formulation-and there are many -a properly applied formaldehyde treatment will give a stabilized effect having excellent durability t o repeated laundering. While it is questionable whether poly(viny1 alcohol) or a like substance is of use in controlling the cure, a real benefit results from its use in improving fabric handle or feel. As the alcohol is fixed by reaction with part of the formaldehyde, the added stiffness is fast t o washing. Formaldehyde alone is a t a disadvantage in the treatment of spun rayon suitings, because the treated fabric has a thin feel, as contrasted with the desirable full, n~ooly handle produced b y internally polymerizing resins. Such effect can be approached b u t not equaled by using a softener and a polymer such as poly(viny1 alcohol) with the formaldehyde. Glyoxal and Other Dialdehydes. Glyoxal produces a n effect similar t o t h a t of formaldehyde. I t s lower volatility is no longer thought t o be as advantageous as was once believed, however. I n common with higher dialdehydes, b u t unlike formaldehyde, it yellows somewhat during curing. This tendency is minimized b y using glyoxal in conjunction with ureaformaldehyde resins. Hand, too, is naturally improved b y addition of the resin. The low solubility of higher dialdehydes in water, their increased propensity t o yellow, and their greater cost have prevented their industrial application. Because formaldehyde and the dialdehydes must be catalyzed with strong acid, they have not been found useful for cotton. Rayon is not ordinarily subject t o acid damage under normal textile processing conditions, b u t cotton is extremely sensitive t o overcatalysis. T o date, no method has been found of making cotton react with aldehydes without incurring a n excessive loss of strength. B y contrast, the N-methylol resins produce a more moderate strength loss which has been found acceptable in both rayon and cotton. Nonnitrogenous Aldehyde Derivatives. Phenol-formaldehyde resins, one of the earliest classes t o be applied t o textiles, were specified in the basic patents for t h e creaseproofing of rayon fabric (20). While a phenol-formaldehyde can readily be cured onto cellulose with alkaline catalysis to obtain dimensional stability and other effects, its marked tendency t o yellow has precluded its use for such purposes. Resorcinol-formaldehydes and similar types suffer from the same disadvantage. Ketoneformaldehydes, such as unsym-dimethylolacetone, form infusible materials when cured at high p H ( 1 6 ) . They possibly react with cellulose as well (10). Like the phenol-formaldehydes, the ketone compounds tend to yellow, b u t t o a much lesser degree. An afterwash or mild bleach will often reduce or eliminate the discoloration. Accordingly, certain types have enjoyed limited commercial use. Like most other resins and reactants, they can be used with substances such as hydroxyethylcellulose. One patent, for example, claims a reaction mixture of acetone, formaldehyde, and starch ( 2 2 ) . Other nonnitrogenous aldehyde derivatives are beginning t o appear on the market. With mild catalysis-magnesium or calcium chloride-they stabilize effectively. One such material of undisclosed composition has found some use in the stabilization of rayon. It is much too early, however, t o tell whether these substances possess concrete advantages not obtainable with the simpler formaldehyde treatments. Nonaldehydic Treatments. Ethyleneureas manufactured from ethyleneimine and diisocyanates were used t o some extent in wartime Germany, primarily t o reduce the hydrophilic nature of
1733
rayon. Stabilization was obtained as well. KO catalyst is required for such a substance t o bond adjacent cellulose chains. Of the nonnitrogenous, nonaldehydic compounds suggested for use in dimensional stabilization, ones such as 1,3-dichloro-k propanol have been tried ( 4 5 ) . At high p H the chloroalcohol cross links the cellulose. Divinyl sulfone undergoes a similar reaction (69) I n contrast with the dry cure given most urea and melamine resins, both the chloroalcohol and the divinyl sulfone react with cellulose in aqueous medium. High cost, the absence of any strong favorable points, and a general lack of practicality have precluded use of these substances in this country. Miscellaneous Treatments. All the treatments so far discussed have operated through reduction of fiber swelling, the setting of t h e crimp of the fibers being purely ancillary. Cotton or rayon can be stabilized b y two other means, however: by the release of strains plus the adjustment of crimp within the fabric, and by fiber immobilization. T h e Definizing process ( 6 ) makes use of the first. Rayon fabric is padded through 30 to 40% aqueous sodium hydroxide, which contains a proteinaceous colloid t o repress disruption of the rayon. After a 15-second lag in a relatively relaxed state, the fabric is run into dilute sulfuric acid ' to neutralize the alkali, then washed and dried. Time, temperature, and concentrations all must be rigorously controlled. Properly carried out, the process can give adequate stabilization without chemical add-on or reduction in water imbibition. Shrinkage is probably reduced b y the release of fiber strains and by setting the crimp of the fibers within the larger structure of the fabric, although the harshness resulting from overtreatment indicates t h a t some interfiber and interyarn bonding are involved. A somewhat different, though related, phenomenon is the shrinkage reduction accompanying cotton mercerization. Cotton, when mercerized under tension, will have a wider washed width than unmercerized cotton. Slkali-soluble hydroxyethylcellulose has been suggested for use in stabilizing cotton. T h e chief effects are fiber immobilization and concomitant stiffening. ' The substance seems t o have more utility in preventing stretching than in reducing shrinkage. Many similar bonding processes have been patented. Wallach ( 6 7 ) wove a fabric from cotton blended with 10% ethylcellulose fiber, exposed it t o a solvent for t h e latter, and dried the goods. Mantell and Heim (58)treated cotton and rayon piece goods with a n aqueous solution of urea, sodium hydroxide and zinc or stannic oxide, and dried the fabric in a stretched state. Even though these processes possess some limited value (and it is known t h a t thermosetting resins act partly through fiber immobilization) thermoplastic resins have little value in fiberbonding cellulose. This is due partly t o the easily deformable nature of the thermoplastics and partly to their poor adhesion to wet cellulose fibers ( 4 8 ) . Water-repellent treatments which might tend to reduce shrinkage by decreasing water imbibition of the fibers are effective only in cold water. A hot wash swells the fibers as though the repellent were not present. Various aspects of the shrinkage control of cellulose fabrics are discussed by Wardell ( 6 8 ) and Woodruff (69). WOOL
R70ol is the most complex of all textile fibers and the one which has enjoyed most study. I n considering why wool shrinks, and what can be done about it, i t is helpful t o understand the morphology of the fiber. The cortex, a cylindrical mass of desiccated spindle cells, composes the bulk of the fiber. Each cortical cell lies in one of two helically twisted segments, sometimes called the ortho- and the paracortexes. The two portions of the cortex differ from each other chemically and probably physically. Forming a sheath about the cortex is a n irregular, scaly cuticle believed t o be composed of three layers. Outermost is the thin, chemically unreactive epicuticle. Below this appears t o be a
1734
INDUSTRIAL AND ENGINEERING CHEMISTRY
TI P
O Figure 3.
G
7
Steps in fiber migration of wool
thicker, more reactive exocuticle, and perhaps an even lower layer called the endocuticle. The unique structure of the fiber is responsible for the main (Swelling and recause of shrinking of wool textiles-felting. laxation shrinkage are of little or only indirect importance.) A number of theories have been proposed to account for felting. .Many contain a germ of the truth, b u t Speakman’s has suffered the test of time and criticism with least change. I n the idealized diagram of Figure 3 a wool fiber is pictured lying on two fibers perpendicular to it. The fiber is attached a t its tip end t o a sort of semimovable anchor-for example, a mass of tangled fibers. A tensile force is imposed on the fiber, stretching it. The load is
Table VI.
Treatments Used to Control Shrinkage of Wool Textiles
Method Aqueous halogenations Nonaqueous halogenations Nonhalogen oxidations
Gross Action Attrition Attrition Attrition
Alkali treatments Reductions
Attrition Attrition
Enzyme treatments
Attrition
Thermosetting resins
Addition
Amide and urethane resins
Addition
Vinyl and acrylic resins
Addition
Specific Action Cuticle alteration Cuticle alteration Cuticle alteration and change in viscoelastic properties Cuticle alteration Change in viscoelastic properties and cuticle alteration Change in viscoelastic properties and cuticle alteration Fiber immobilization and cuticle alteration Fiber immobilization and cuticle alteration Fiber immobilization and cuticle alteration
Vol. 47, No. 9
other proposed factors can only implement or supplement fiber migration. They do not supplant it. The wool chemist’s task is a t first glance easy. One has merely to reduce or destroy the difference in coefficients of friction of the cuticle or to alter the elongation or energy absorption capacity of the fiber, thus reducing extension or recovery. Very few commercial processes attempt to alter the flow properties of the individual fibers. Some change the viscoelastic nature of the entire fiber assembly-the fabric-by immobilizing the fibers and yarns within a relatively rigid network. The vast majority effect some change in the cuticle. Table VI lists the important processes which enjoy or have enjoyed some commercial vogue. Aqueous Halogenations. Gaseous chlorine and fluorine, solutions of bromine and bromates] and dry powders containing oxygen compounds of chlorine have all been proposed for use in shrinkproofing wool. Today solutions of chlorites, hypochlorites, and various chloramides are used almost exclusively in nonsolvent halogenations. Some of them have gained wide commercial use: the Schollerize, Harriset, h’egafel, Kroy, and Protonize, to name a few. Each operates on the same principle, controlled attrition of the fiber cuticle. Their differences are in their secondary effects: speed of treatment, ease of control, damage t o the fiber, etc. T h e outermost layer of the wool fiber, the epicuticle, is a smooth featureless structure. I t s topography is uniform, permitting the long-period effect of the individual scales to be easily sensed. The exocuticle, however, possesses an irregular surface. It is striated and furrowed, and pocked irregularly with pits and craters. With the epicuticle damaged or removed, the wool fiber still possesses its long-period scale. effect, but superimposed upon i t is the strongly featured microgeography of the exocuticle which tends in its roughness and scarification to be uniform along the fiber. While the gross features of the scales accent-indeed, are the cause of-any unidirectional effect, the detail of the uncovered exocuticle might be expected to mask such orientation. Chlorination decreases the directional frictional effect. [D.F.E. = (ut - u,)/(ut u?),where u t and upare the coefficients of linear friction of the fiber in the tip and root directions, respectively.] The directional friction effect is decreased not by decreasing ut but by increasing ZL,. This effect is shown in Table VII, from Harris and Frishman (26). While i t is tempting to suppose that chlorination works by allowing the exocuticle to exert its “leveling” effect, this view is inconsistent with known observations, among which is the fact that while the epicuticle comprises 0.1 to 0.2% of the total fiber weight, a loss of 3 t o 5% is commonly suffered in chlorination. The chlorinating agent can diffuse through the epicuticle to attack the exocuticle or even lower structures. The release of osmotically active degradation products leads to the formation of Allwoerden bubbles. Osmotic pressure or simple mechanical agitation ruptures the bubbles and “lifts off” the epicuticle. The chlorine-vulnerable group in wool keratin is the cystine link. I n chlorination, -S-Sbonds are oxidized to sulfoxides, sulfones, and ultimately to cysteic acid Alexander and his colleagues ( 8 ) believe t h a t acid chlorination acts through a mechanism in which cystine is cyclized and finally released as free cysteic acid:
+
then removed, allowing the fiber to contract. Because the scales of the fiber prevent its tipward contraction, the fiber necessarily contracts in the root direction, dragging its semimovable anchor with it. I n felting, this three-step process occurs over and over again, drawing the fibers into a thick, tangled mat or felt. The process of fiber migration postulates three essentials: a unidirectionally oriented cuticle, the ability of the fiber to stretch, and its capacity to rkcover. While such fiber ,migraCla Hf tion is the true cause of -NH-CH-CO-XH-CHR+ -NH-CH-CO-N-CHRHzNCH( C0OH)CHzSOaH felting, other factors are AH2 AH2-----SOa I contributory. It should $OHbe understood, however, SI that twisting, curling, I -NH-CH(COOH)NH-CHRinterlocking, a tendency &Hi----80~ of the cuticle t o gelatiHZNCH( COOH) NH-CHR-COOH --f HzNCH( COOH) CHaSOyH nize or of the whole fiber I I t o soften, and all the dHS -,O,
September 1955 Table VII.
INDUSTRIAL AND ENGINEERING CHEMISTRY Effect o f C h l o r i n a t i o n on Frictional Characteristics of Wool
Treatment Control NaClO 5 % available C1 7% available C1 10% available C1
U
ur
0.39
0.15
D.F.E. 0.34
0.35 0.44 0.47
0.22 0.36 0.44
0.28 0.10 0.03
Whatever the case, the -S-Slink is oxidized and the degree of cross linking of the cuticle decreases. This lowering of “molecular weight” causes the expected increase in swellability. I n addition, the amino acid residues within each polypeptide chain become more susceptible to hydrolysis. Rupture of the individual chains follows, with further decrease of interlink chain weight. It is the old story of protein hydrolysis with the apparently slight but actually exceedingly important difference t h a t the hydrolysis-Le., the attrition-is nonuniform over the available surface of the protein. The condition leading to this nonuniformity of reaction may be visualized as follows. Wool fibers vary from point t o point along their length in dyeing and rheological properties. This is natural, since the chemical and physical characteristics of the fibers reflect the changes-large and small-in the total metabolism of the sheep. There is every reason t o believe t h a t this nonuniformity is mirrored in the swelling properties of the fibers. It is known t h a t the rate-controlling step in the reaction of wool with a chlorine solution is diffusion, probably through the fiber itself (3’). Nonuniformity of swelling would thus lead t o nonuniform chlorination. The attrition would be localized a t “random” points on each scale, exerting no preferential attack on the tips rather than the bodies of the scales. The ut would not be expected to decrease. Rather, as a result of what might be termed this random attack, the entire fiber would be roughened. The difference between the u’s decreases and the directional frictional effect mould drop. Theory here agrees perfectly with observed behavior. Because the basic residues of keratin are thought t o be scattered randomly along the polypeptide chains while the acidic one would groups are clustered around the disulfide links (44), expect greater chain damage to occur in alkaline than in acid chlorination. Main chain hydrolvsis depends t o a great degree, however, on the accessibility of the peptide bonds, the acidic residues being subject to rapid hydrolysis only if they are readily accessible. According to Middlebrook (&), however, proline residues are intermixed with the acidic groups in regions of high disulfide cross linking, causing a “gathering together” of the chains within the wool “molecule,” and reducing the molecular diameter from 68.6 A. (in regions of low proline content) t o 56.7 A. The acidic residues are thus relatively inaccessible. The only manner in which they may be rapidly “opened up” is b y destroying the cystine links, in this case by oxidation with the chlorine solution. The oxidation potential of the chlorine solution drops sharply, however, with increasing pH. T h e primary attack on the cystine residues is thus limited in alkaline solution. I n direct consequence, main chain hydrolysis is also limited. Fiber damage is held t o a minimum, therefore, even though effective shrinkproofing takes place. This, and the development of techniques whereby yellowing of the wool is avoided, are responsible for the newer trend towards alkaline chlorination. Several processes employ halogenating agents other than chlorites or hypochlorites. Chloramides and chlorosulfonamides find some favor. Some of the former are made b y reaction of compounds containing active chlorine with amine resins. Originally the hypochlorite and resin were simply added to the same liquor, or a resin-treated fabric was chlorinated. Such “internal resists,” as they were once called, were used on a rather intuitive
1735
basis, much a s glue was once added to wool stripping liquors. The resin was assumed t o protect the wool. Subsequent work showed t h a t the resin exerted no direct protective action b u t t h a t it did make the reaction of wool and chlorine more uniform and easily controlled. Today this rate-controlling property is the main advantage in the use of such reaction products. Nonaqueous Halogenations. About ten patents cover the use of nitrosyl chloride, chromyl chloride, bromine, chlorine, and organic hypochlorites in solvent solution for treating wool. While the organic hypochlorites are suggested for use both in stabilization and in making wool fabrics resistant t o mustardlike vesicants, the other agents were intended mainly for stabilization alone. None is used today. Of this class, sulfuryl chloride has been the only compound t o gain a wide use. It is used in the Dri-sol method. Though the presence of water is necessary for effect, and though the mechanism of solvent chlorination seems identical with t h a t of aqueous chlorination, there apparently are subtle differences. Poor stabilization results if the wool contains little residual oil or grease. The data of Farnworth and Speakman ( 1 7 ) in Table VI11 illustrate this. T h e results are even more striking when one considers t h a t the addition of oleic acid to solvent-extracted, unshrinkproofed wool greatly increases felting shrinkage. This is clearly shown by values of Peryman and Speakman (61) in Table IX.
Table VIII.
Effect of Oleic Acid on Dri-sol ( S u l f u r y Chloride) Stabilization
Area Oleic Acid Shrinkage, Treatment Fabric Added 70 43 None None Unextracted 3 None SOZClZ Unextracted 44 None None Extracted 25 None s02c12 Extracted -7 s02c12 Extracted 0.5%; -5 sozc12 Extracted 1.0% 25 l.O%b SOZClZ Extracted a Added t o wool after extraction but prior t o Sod% treatment. b ddded t o SOL& bath.
No.
Table IX. Effect of Oleic Acid on S h r i n k a g e of ’ Solvent-Extracted Nonstabilized Wool Oleic Acid Added t o Wool, 7 0
Area Shrinkage,
0.0 0.1 0.2 0.4 0.75
4 10 20 18 23 24
1.5
%
T o account for the effect of the oleic acid as illustrated in Table
VIII, i t was proposed t h a t peroxides present in the wool oil form free radicals and t h a t the following reactions occur:
+
+ + +
+ + + +
R* SOzCL + RCI SO2 C1* Wool-s-s-wool C1’ + wool-sc1 wool-s* Wool-s* sozc12 + wool-sc1 SO2 c1* Rupture of the cystine bonds is accompanied b y attrition of the cuticle and felting is prevented. The intermediate values for samples 4 and 7 in Table VI11 may be due t o two possible causes. First, dibenzyl disulfide is known t o be split b y sulfuryl chloride in the absence of peroxides (6R). Cystine breakdown could be partially due t o such a mechanism. Even in the absence of peroxides, then, disulfide scission could take place, leading t o reduced-but not completely eliminated-shrinkage. Second, and more likely, extraction of the samples was incomplete. Some oil probably remained in the wool, enabling the sulfuryl chloride t o produce a low effect. It is now known t h a t even several days’ extraction of wool with volatile solvents will not entirely eliminate all residual grease. While it is probable t h a t the actual mechanics of stabilization
1736
INDUSTRIAL AND ENGINEERING CHEMISTRY
with sulfuryl chloride are the same as those of aqueous chlorinations, the picture is clouded by the reported observations t h a t while u,. increases and the directional frictional effect decreases on treatment, ut sometimes suffers a slight decrease. The validity of the drop is in question, however, as other workers have found t h a t both u,. and ut increase ( 8 , S $ , S 4 ) , Nonhalogen Oxidations. T h e Sanforlan process, a hybrid consisting of simultaneous treatments of wool with sodium hypochlorite and potassium permanganate, enjoys some use today. Other methods in which hydrogen peroxide, per acids, or other per salts are the active agents are of minor importance b u t do have fair possibilities. Permonosulfuric, peracetic, perbenzoic, and performic acids are claimed to be efficient, as are perborates, percarbonates, and hydrogen peroxide itself, the last when used in conjunction with very dilute solutions of salts of metals possessing more than one oxidation state ( 1 3 ) . This use of hydrogen peroxide-metal salt suggests that t h e reaction is not just a simple oxidation but is a free radical-initiated process involving rupture of cystine bonds. Alkyl hydroperoxides, dialkyl peroxides, aromatic per acid esters, benzoyl peroxide, and substituted benzoyl peroxides are stated to be without effect ( 1 ) . The physical action of such oxidations is still far from clearly known. While hydrogen peroxide attacks the wool fiber, it is doubtful if the epicuticle is touched by such oxidants. Many techniques for isolating this membrane depend upon the digestion of the remainder of the fiber with peracetic acid followed by ammonia, leaving the epicuticle intact. Furthermore, Allwoerden bubbles appear on fibers previously treated with peracetic acid when they are immersed in 1%aqueous sodium bicarbonate (42). The situation is made somewhat complex, however, by the observation t h a t after oxidation or reduction of the wool, the Allwoerden bubbles appear only on t h a t portion of the cuticle covering the orthocortex (21, SO). This could indicate t h a t the epicuticle is indeed oxidizable, but only t h a t part of i t which covers the paracortex. On the other hand, it could mean t h a t the orthocortex was preferentially attacked. Smith and Harris (60) found in treating wool with hydrogen peroxide t h a t while the cystine content decreases with increasing hydrogen peroxide concentration, the total sulfur increases (on aftertreatment with sodium hydroxide). The chemical differences between the ortho- and the paracortexes could explain this behavior. Mercer has found t h a t the paracortex contains more sulfur than the orthocortex. If it contains more blocked cystine or cysteine (see ‘7, 56, 6 6 ) , the observations of Smith and Harris would indicate that hydrogen peroxide preferentially oxidizes the orthocortex. The ortho-para ratio decreases and the fibers assume less of a n elastic and more of a viscous nature, a condition unfavorable t o felting. Steele (63) has found t h a t with increasing oxidation of the fiber cystine, both the elastic modulus and the elastic recovery decrease. Clark and Buhrke (11) showed qualitatively t h a t the elastic nature of wool decreases with increasing keratinization. If one may assume that increasing keratinization results in decreasing ortho-para ratio, their findings agree with Steele’s, and would support the view t h a t oxidation renders wool shrink-resistant by lowering fiber recovery. Anderson’s observations ( 4 ) are a t variance with this view, however. Alkali Treatments. Almost every possible organic and inorganic base has been suggested for use in stabilizing wool. They have been used in dilute and concentrated solutions, in aqueous, and in solvent systems. The only usable methods are those based on the work of Hall and IVood, and of Freney and Lipson, who used sodium or potassium hydroxide or a n alkali metal alkoxide, dispersed or dissolved in organic solvents containing only small quantities of water (1 to 5%). ‘ Alcoholic potassium hydroxide feltproofs wool b y action upon the disulfide bonds. A pretreatment v i t h potassium cyanide, which transforms cystine t o lanthionine, will permit alcoholic alkali t o exert only a poor stabilizing effect on a wool fabric [Table X, based on work of Farnworth and Speakman (IS)].
Table X.
Vol. 47, No. 9
Effect of Alcoholic Allrali on Wool Shrinkage Area Shrinkage, Treatment
%
Sone
42
Alcoholic KOH K C N followed by alcoholic KOH
23 2 22
KCN
It is doubtful whether the action of the alkali is restricted t o the sulfur groups, however. Aqueous alkali has a nonspecific action on proteins. I n Freney-Lipson systems, even though there is no “free” water, instantaneous dilution could occur when the solution comes in contact with wool of 8 t o 14% regain. Peptide cleavage could readily take place. If extensive hydrolysis took place, one would expect gross gelatinization and attrition. Yet alcoholic potassium hydroxide does not gelatinize horn keratin, but sulfuryl chloride, chlorine, and bromine do ($4). Furthermore, there are excellent photomicrographs showing t h a t the wool cuticle suffers only slight visible alteration on treatment with nonaqueous alkali ( 1 5 ) . It has been suggested t h a t alcoholic alkali acts by attrition of the bases of t h e scales, allowing each one t o move freely under stress, as about a hinge. This fits well with the apparent lack of physical damage t o the scales and with the poor abrasion resistance of alkali-treated wools, yet it is completely a t odds with data from frictional tests. If the scales were loosened a t their bases, ut and possibly u7 should be lower than for untreated wool, Both u’s increase, however ($1, 32). Lindberg and his coworkers thus claim t h a t alcoholic potassium hydroxide damages or even removes the epicuticle. Obviously, more work is needed before the action of alkali processes can be explained unambiguously. Reductions and Enzyme Treatments. Enzyme treatments of wool are practically, if not theoretically, the same as ordiuary reductions, since the enzymes are always used in a strongly reducing bath. Both processes depend for effect mainly upon alteration of the stress-strain-time characteristics of the fiber. Some degree of cuticle damage always occurs as well. T h e epicuticle is unharmed by simple reduction and by enzymatic hydrolysis. I n fact, both processes are utilized in its isolation. The lower lavers of the scales are vulnerable t o attack, however. Because only one pinhole in the epicuticle is sufficient t o admit B reagent to the underlying strata, and diffusion can occur through the epicuticle, damage t o the underlayers can readily take place. This view is supported by electron photomicrographs illustrating that controlled enzymatic digestion removes part of the cuticle (16). The main force of the reduction is probably directed a t the orthocortex. Mercer ( 4 1 ) showed t h a t supercontraction initiated by pepsin is the result of the coiling of what he then called the “microfibrils” of wool. I n several later papers he disclosed t h a t the orthocortex is preferentially dissolved on treatment of whole fibers with proteolytic enzymes, the convoluted semicylinder of the paracortex being left as a residue. While the Wool Industries Research Association’s papain-bisulfite process is naturally not allowed to proceed t o such a n extent, controlled partial degradation of the orthocortex must result. As there is excellent reason t o believe that t h e paracortex contains most or all of t h e blocked cystine and cysteine of wool, and is therefore more highly cross linked than the orthocortex, one can expect a change on enzyme treatment toward a less elastic state as the ortho-para ratio decreases. Simple reductions, not entailing the use of enzymes, probably operate on the same principles of decrease of the ortho-para ratio and cuticle attrition. One notable exception is the Harristrip process, a method nut intended for stabilizing wool at all, but giving good r e s u h nevertheless. I n the Harristrip wool is treated with a buffered solution of a strong reducing agent such as sodium hydrosulfite, in which is emulsified ethylene dibromide or
September 1955
INDUSTRIAL AND ENGINEERING CHEMISTRY
a similar substance. .Reduction of cystine occurs but, here again, probably mainly in the orthocortex. Recoupling t o give dithioether links follows immediately. The thioether links are extremely stable to hydrolysis. Thus the orthocortex is probably made as resistant as-perhaps even more than-the paracortex. Because felting depends on fiber elongation and this in turn is dependent on the ease with which interchain bonds are ruptured by water and heat, the Harristrip process reduces or eliminates the felting of wool, probably by making elongation i n the felting medium more difficult. The physical characteristica of the fiber, as evidenced under normal use conditions, are not harmed, however (50). It has been noted in this laboratory t h a t a contraction of 10 t o 15% in area often occurs during processing by Harristrip. This was probably due to poor emulsification of the ethylene dibromide, allowing some supercontraction t o take place before the dithioether bridges were formed. The Harristrip has the advantage over almost all other feltproofing processes in t h a t it makes the wool highly resistant t o further supercontraction. Figure 4 illustrates the resistance of Harristripped wool t o felting. Thermosetting, Urethane, and Amide Resins. The initial use of urea- and melamine-formaldehyde resins for wool stabilization stemmed naturally from the original Tootal Broadhurst Lee patents. It is common philosophy t h a t what is good for one use might be good for another. Textile finishers therefore applied such resins to wool, hoping t h a t something would happen. Something did: The resins gave a certain degree of washability t o the wool. Since those early days, certain amino resins, especially melamines, have been found t o be eminently suited for shrinkproofing wool. Methylated melamine-formaldehydes are the outstanding examples. There is little doubt that after aqueous chlorination methods, the Lanaset treatment represents the most widespread American means of stabilizing wool. Amino-formaldehyde resins can operate by two possible methods. T h e first could be decomposition to formaldehyde and the combination of the liberated formaldehyde with the wool keratin. It has been known for many years that formaldehyde makes wool more resistant t o laundering. At p H 6 t o 8 formaldehyde combines with the e-amino groups of the lysine residues; at higher pH’s with the guanido units of arginine and with the cysteine or even cystine groups. Reaction with the sulfur acids occurs only a t fairly high concentrations. Cyclization of the cysteine to a thiazolidine probably occurs (64). It is well known t h a t formaldehyde is also bound by the primary amide groups of keratin. All these reactions of formaldehyde with the protein groups give rise to no new interchain bonds, however. There is little or no effect on shrinkage. On the other hand, Brown and others (9) have postulated the formation of cross links involving the reaction of formaldehyde with the active hydrogen of arginine and lysine residues. Middlebrook (43) assumed that formaldehyde combines with aspargine to give a hydroxypyrimidine and with free carboxyl groupe of glutamic acid t o form cross links between the latter and residues of arginine. While such bonding may and probably does contribute somewhat t o stabilization and to the harshness which is sometimes encountered in resin-treated goods, the greater effect is produced by essentially nonchemical means Several years ago i t was found (89) t h a t Resloom HP, a melamine-formaldehyde, distributed itself uniformly throughout wool fibers. Another paper ($8) revealed similar behavior using Aerotex M-3 (Lanaset), another melamine resin. The resin was found t o prevent diffusion of dyes into the treated fibers. Two years later, however, Royer and Maresh (67) were able t o show by means of phase microscopy that the melamine resin was heavily concentrated on the surface of the fibers, particularly at the edges of the scales. Another group (64)presented convincing evidence t h a t the rheological properties of wool fibers are not changed upon resin treatment, but t h a t considerable increase in tensile strength is to be
1737
noted in treated yarns. This was thought t o be due t o “spotwelding” or fiber immobilization. While the directional frictional effect of melamine resin-treated fibers varies widely according t o the form in which the fibers were treated-individual fibers, top, or fabric-the consensus is t h a t in actual practice there is little lowering of the directional frictional effect by the resin.
a w
a Q
ae
0
IO
5
.
v
180’F
140°F
HOURS
Figure 48.
15
OF
LAUNDERING
Effect of Harristrip on wool shrinkage
Because they operate by spot-welding, amino resins have some very obvious advantages as well as some clear-cut disadvantages. No weight loss is suffered, as with chlorination; instead, there is a net gain. The yarn and fabric tensile strengths are increased considerably. The application of the resin requires a minimum of supervision, as fine control is not critical. Because of the decrease of elongation and energy absorption capacity of the fabric, however, the wearing qualities of the fabric are sometimes harmed. This occurs i n spite of a sharp increase in abrasion resistance as measured by laboratory tests. Further, the resin-treated wool is sometimes hareh t o the hand. The use of acid colloids of melamine resins is claimed to eliminate the drop in wearing qualities of the treated wool. It also eliminates the need for a high cure-240°, rather than 300” F., is all t h a t is needed. Opinions of the worth of the acid colloid processes are conflicting. Some mills swear by them, others swear a t them, claiming t h a t proper cure of the monomeric resins gives superior results. Whatever the case, however, melamine resins are very effective and are ideally suited for American production methods. Unfortunately, they have not done away with the bad points of the chlorination processes; they have substituted new faults for old. Imperial Chemical Industries has a number of patents covering the use of polyurethanes for wool stabilization. Their value is unknown but may be guessed at, as they have never been used commerciallv. The Commonwealth Scientific and Industrial Research Organization has developed a method of shrinkproofing wool by application of N-methoxymethyl-nylon (26, 27, 35). This substance has excellent How properties b u t is poorly durable. When set on the fiber as nylon itself by acid or heat treatment, the unshrinkable finish is good. Unsubstituted nylon has poor
INDUSTRIAL AND ENGINEERING CHEMISTRY
1738
flow properties and cannot be deposited on the wool effectually. T h e CSIRO method operates b y fiber immobilization, by physjcal alteration of the cuticle, or by both means. It is possible, though distinctly unlikely, that the fiber is masked by a smooth coating of polymer which fills in the "troughs" of the scales. It s much more probable t h a t t h e directional frictional effect is reduced by a n increase in both coefficients of friction. The process is currently being tried in conjunction with nonaqueous alkali treatment (36). Thermoplastic Polymers. Speakman, Barr, and Lipson (37, 61) were t h e first t o investigate the internal polymerization of addition monomers in wool. Elaboration and extension of this work have led to numerous papers by many workers but as yet t o no commercially feasible process. The main effort in the thermoplastic polymer field has been devoted t o external application of fully polymerized resins. A curious result of this widespread work is that the group is now represented by some of the worst and the best of all shrinkproofing treatments. Homopolymers and copolymers of simple alkyl acrylates, vinyl chloride and acetate, acrylonitrile, butadiene, styrene, and other similar substances have all been tried without much effect. Others, containing various functional groups-alkoxy, isocyanate, sulfone, etc.-have enjoyed some academic vogue b u t have not been exploited commercially (14,29,49,53, 68,66). A wide variety of acrylics has been investigated in this laboratory for use in shrinkproofing wool. The majority have been of Iittle worth. The few effective ones operate by increasing u,, and t o a lesser extent ut, thus reducing the directional frictional effect of the fibers. While fiber immobilization does play a role, i t is still not known how important a contribution this makes t o stabilization. The rheological properties of yarns and fabrics treated with a n effective thermoplastic resin are much improved over untreated wool. Tensile strength, elongation at rupture, energy absorption capacity, and all dependent variables are increased, but in such a manner as to suggest simple superposition of the resin characteristics upon those of the fibers. Insufficient work on the fiber eve1 forbids one to draw definite conclusions as t o how the resin affects fiber properties. The data presented in Table X indicate the widely varying results obtainable from a series of thermoplastic copolymers.
Table XI.
Effect of Thermoplastic Resins on Wool Shrinkage
Applied Area Felting Resin Solids, % Shrinkage, % A 3.5 0 B 5.0 0 B 3.5 2 1.5 0 C D 3.5 55 E 3.5 53 F 7.0 0 F 3.5 20 F 1.5 45 60 Control ... 5-hour wash a t 140" F., in water buffered t o pH 5 to 7.
A detailed investigation into the action of resin A revealed an interesting fact, which was confirmed by work with resin C. It is commonly believed t h a t a thermoplastic finish on wool can be durable only if some covalent link exists between the resin and the wool. T o test this hypothesis, a series of wool swatches was subjected t o various chemical treatments designed t o block the functional groups of the keratin. One piece was exhaustively methylated with hydrochloric acid in anhydrous methanol, esterifying the carboxyl groups, and methylating the majority of the peptide links. A second swatch was acetylated with acetic anhydride in both the absence and presence of acid, destroying guanido groups and acetylating the resulting ornithine residues, as well as
Vol. 47, No. 9
the lysine and most of t h e serine, threonine, and tyrosine groups. A third piece was ethylenated by Harristrip using a n excess of ethylene dibromide; all unblocked cystine and cysteine was thus converted t o unreactive form. A fourth piece of wool was treated with Sanger's reagent (l-fluoro-2,4-dinitrobenzene); lysine, histidine, cysteine, tyrosine, and possibly tryptophan residues were converted t o their dinitrophenyl derivatives. Two swatches of wool were given all four treatments and two others were not treated at all. Of these two pairs, one sample each was padded t o give 3.5% applied solids of resin A. The resinpadded wool was dried and cured. None of the other samples received resin treatment. Table X I 1 gives the felting shrinkage of the samples during laundering.
Table XII.
Effect of Group Masking on Wool Shrinkage
Resin A, Pretreatment % Methylation 0 Acetylation 0 Ethylenation 0 Dinitrophenylation 0 All 4 0 All 4 3.5 None 0 None 3.5 Wash buffered to pH 5 to 7. 180" F.
1 hr. 17
10 0 6 0 0 7 0 First
Area Felting Shrinkage, % 5 hr. 10 hr. 15 hr. 49 51 52 40 54 56 0 1 13 27 53 62 0 29 34 0 0 0 27 37 54 0 0 6 5 hours a t 140" F., next 10 hours at
The figures suggest what at first appeared to be a n unacceptable thesis: No covalent cross links need be formed between the wool keratin and the resin t o impart a durable finish. Subsequent investigations revealed t h a t hydrogen bonding plays no part either. The resin seems t o adhere t o the fibers purely b y adhesive forces. It should be understood t h a t layers a few hundred millimicrons thick have properties radically different from those of bulk matter (40). Materials as different as metallic gold, gelatin, and poly(methy1 methacrylate) show the same extraordinary adhesiveness at very small dimensions. Every textile chemist knows, however, t h a t durability of finiah is not necessarily synonymous with retention of resin weight. A finish can be destroyed without removing the resin. There are indications t h a t t o produce a n effective unshrinkable finish an acrylic must possess the proper flow characteristics. I n Table XI, resin F shows the need for exercising care in determining the proper concentration level of resin in a test. Many addition polymers lacking any intrinsic value have been purposelv tested a t high levels t o produce what appears t o be a good effect. Really worth-while ones, however, are able t o act at low levels, producing a washable wool fabric completely acceptable in every respect. The treated fabrics are the equals and often the superiors in many respects of not just untreated wool fabrics, b u t of those made from the newer synthetic fibers. It ie probable t h a t vinyls, acrylics, and similar polymers will find increasing use in wool stabilization. Miscellaneous Treatments. There are many other types of processes intended t o make wool washable. Silicones, anhydrous isocyanates, drying oils, cellulose acetate, and abrasive silicates have all been suggested. T h e silicones will bear watching, as will the English anhydrocarboxyglycine process. This last has the advantage t h a t i t is a n additive process and t h a t the additive itself is chemically similar t o wool ( 5 ) . HY DROPHOBIC S
While cotton, viscose, and wool are capable of imbibing large quantities of water, acetate and especially the newer synthetics absorb relatively little. This behavior is mirrored in their regains at room temperature and 65% relative humidity. Wool absorbs u p t o about 14% moisture, viscose 11%, cotton approximately 8%, acetate about 6y0,and the newer fibers from about 4y0t o less than 0.5%. Swelling is thus a minor factor and shrink-
-
INDUSTRIAL AND ENGINEERING CHEMISTRY
Sepfember 1955
age due t o swelling is virtually nonexistent. Felting is naturally of no importance, since the man-made fibers possess fairly smooth scale-free contours. T h e shrinkage of fabrics made from such fibers is due largely to molecular disorganization, leading to fiber contraction, and t o mechanical dist'ortion. Molecular disorganization is mainly physical in nature, the presence or absence of water having little effect on its course. As few or no covalent cross links are present in the polymers of which the fibers me composed, the orientation of the main chains depends t o a great extent on hydrogen bonding and the weaker van der Waals forces. A4sthe temperature increases, therefore, the polymer chains in the amorphous regions coil and kink to an ever-increasing degree. The phenomenon is seen as a contraction of the fibers, causing the yarns and thus the fabric t o shrink. The second cause of shrinkage is gross mechanical distortion. During laundering any fabric becomes badly creased or wrinkled. Because of the thermoplastic nature of the hydrophobic fibers, textiles made from t,hem when creased a t moderately high temperatures will retain the creases a t lower ones. Thus the creases set in the fabric during the hot wash part of the laundering cycle will persist throughout the cooler rinse periods and drying. Such creases and wrinlrles cause a large diminution in fabric area and are extremely difficult to remove. I n the early 1940's nylon fabrics were somet,imes chemically treated with formaldehyde or a melamine resin to obtain dimensional stability. I t was rapidly discovered, however, t h a t the only effective means of stabilizing such fabrics was by heatsetting. Today most thermoplastic hydrophobics are stabilized by this means. They are heated above their glass transition temperatures and are >hen cooled to the lassy state while under physical constraint. r h e temperature o f setting varies with the fiber and the end use of the fabric. Generally, a temperature 36" to 70" F. higher than the maximum expected during use is chosen, the use of still higher temperatures being without additional advantage. Kylon fabrics are generally set a t about 400" F. Heat-set fabrics are stable to gross distortion and usuallj~ suffer from thermal disorganization only when the setting temperature is closely approached. Such fabrics will wrinkle in processing or laundering, but the wrinkles are easily removed with a n ordinary press or iron., European and American practices in heat-setting usually differ only in mechanical details. Because of lower production with consequent batchwise finishing, the English are able t o heat-set fabrics while maintaining control of both warp and filling dimensions. I n this country production is higher and finishing routines are usually continuous. Control of filling dimensions is therefore difficult. Common practice is t o permit draw-in in the filling, compensating for i t initially by weaving the cloth wider in the loom. Acetate rayon occupies a unique position in the range of textile fibers. It is more hydrophobic than cellulose or wool, yet less so than Dacron or similar materials. It appears t o be so thermoplastic in nature that it does not respond to heat-setting, nor is it susceptible to resin treatment. It is fortunate, therefore, t h a t acetate has litt,le tendency t o shrink under conditions of norm:tl laundering. The fiber is commonly blended with viscose and cotton today and in such cases i t may be treated with urea or melamine resins with great advantage. The treated fabric will be dimensionally stable during mild laundering.
1739
Collins, J . Textile Inst., 30, P46 (1939). Cunliffe, Sharp, and Ashworth (to British Cotton and Wool Dyers Assoc.), Brit. Patent 614,966 (1948). Drechsel and Thomas (to American Cyanamid Co.), U. S. Patent 2,516,836 (1950). Elliott and Manogue, J . SOC.Dyers Colourists, 68, 12 (1952). Ellis, "Chemistry of Synthetic Resins," vol. I, pp. 543-4, Reinhold, New York, 1935. Farnworth and Speakman, Nature, 161, 850 (1948). Ibid.. 163. 798 (1949).
Fluck and Lynn (to'hmerican Cyanamid Co.), U. S. Patent 2,609,307 (1952).
E'oulds, Marsh, and Wood (to Tootal Broadhurst Lee), Ibid., 1,734,516 (1929)
Fraser and Rogers, Biochim. et Biophys. A c t a , 1 2 , 4 8 4 (1953). Gagarine, D. M. (to Dan River Mills), U. S. Patent 2,486,399 (1949).
Gagliardi and Nuessle, Am. D y e s t u f f Reptr., 39, 12 (1950). Gruntfest and Gagliardi, Textile Research J., 18, 643 (1948). Harris and Frishman, Am. D y e s t u f f Reptr., 37, 52 (1948). Jackson, Teztile Research J., 21, 655 (1951). Jackson and Lipson, Ibid., 21, 156 (1951). Kienle, Royer, and McCleary, Am. D y e s t u f f Reptr., 34, 42 (1945).
Kropa ( t o American Cyanamid Co.), U. S. Patent 2,606,892. (1952).
Leveau, Parisot, and Cehe, B u l l .
inst.
T e r t i l e France, 42, 7
(1953).
Lindherg, Textile Research J., 18, 470 (1948). Lindberg and Gralen, I b i d . , 18, 287 (1948). Lipson, N a t u r e , 156, 268 (1945). Lipson and Howard, J . SOC.Dyers Colourists, 62, 29 (1946). Lipson and Jackson (to Commonwealth Sci. Ind. Research Organization), Brit. Patent 692,380 (1953). Ibid., 692,381 (1953).
Lipson and Speakman, N u t u r e , 157, 736 (1946). Mantell, and Heim, (to United Merchants & Manufacturers), Can. Patents 449,551-449,554 (1948). Marshall and Aulabaugh, Textile Research J . , 17, 622 (1947). Meissner and Byrne, J . A p p l . Phys., 23, 1170 (1952). Mercer, Textile Research J., 22, 476 (1952). Mercer and Golden, Ibid., 23, 43 (1953). Middlebrook, Biochem. J . , 44, 17 (1949). Middlebrook, B i o c h i m et Bioph,ys. A c t a , 7 , 547 (1951). Musset, Industrie textile, 69, 103 (1952). Korthern New England Sect., AATCC, Am. D i p s t u f f Reptr., 35, 13 (1946).
Nuessle and Bernard, Ihid., 39, 396 (1950). Nuessle and Kine, IND.ENG.CHEM.,4 5 , 1287 (1953). Nyquist (to American Cyanamid Co.), U. S. Patent 2,565,259 (1951).
Patterson, Geiger, Mizell, and Harris, J . Research N a t l . Bur. Standards, 27, 89 (1941).
Peryman and Speakman, J. Textile Inst., 41, T241 (1950). Philadelphia Sect., AATCC, Am. D y e s t u f R r p t r . , 38,822(1949). Pinkney (to E. I. du Pont de Nemours & Co.) U. S. Patent 2,585,583 (1952).
Ratner and Clarke, J . Am. Chmm. Soc., 59, 200 (1937). Rhode Island Section, A$TCC, Am. D y e s t u f f Reptr., 38, 842 (1949).
Ripa, Textile Research J . , 2 3 , 7 7 6 (1953). Royer and Maresh, Ibid., 17, 477 (1947). Schoene (to U. 5 . Rubber Co.), U. S. Patent 2,579,871 (1951). Schoene and Chambers (to U. S. Rubber Co.), Ibid., 2,524,399 (1950).
Smith and Harris, J . Research N a t l . Bur. Standards, 16, 301 (1936).
LITERATURE CITED
Alexander and Earland (to Wolsey, Ltd.), Brit. Patent 656,938 (1951).
Alexander, Fox, and Hudson, Biochem. J., 49, 129 (1951). Alexander, Gongh, and Hudson, T r a n s . Faraday Soc., 45, 1058, 1109 (1949).
Anderson. Nature. 158. 554 (1946). Baldwin, Barr, and Speakm'an, J. SOC.D y e r s Colourists, 62, 4 (1946) ; (to Imperial Chemical Industries) Brit. Patent 567.501 (1946) . ~, I~
Beer, I,.,and Stevenson, A. S. (to Alrose Chemical Co.), U. S. Patent 2,497,519 (1950). Benesch and Benesch, J . Am. C h e w SOC.,75, 4367 (1953). Bohm, J . Soc. Dyers Colourists, 61, 278 (1945). Brown, Hornstein, and Harris, Textile Research J . , 21, 222 (1951).
Cameron and Morton, .I. SOC.Dyers Colourists, 64, 329 (1948). Clark and Buhrke, Science, 120, 40 (1954).
Speakman and Barr (to Imperial Chemical Industries), Brit. Patent 559,787 (1944). Speakman, Nilssen, and Elliott, N a t u r e , 142, 1035 (1938). Steele, private communication, Rohm & Haas Co., Philadelphia 37, Pa., 1954. Stock and Salley, Textile Research J . , 19, 41 (1949). Thomas and Kropa (to rlmerican Cyanamid Co.), U. S. Patent 2,578,861 (1951).
Van Scott and Flesch, Science, 119, 70 (1954). Wallach, R. N. (to Sylvania Industrial Corp.), U. S. Patent 2,319,834 (1943).
Wardell, Am. D y e s t u f f Reptr., 41, 546 (1952). Woodruff, R u u o n and Synthetic Textiles, 31, KO.5, 71; No. 6, 51 (1950).
RECEIVED for review October 16, 1954. ACCEPTEDMarch 12, 1956. Presented before the Division of Polymer Chemistry, Symposium on Textile Chemicals. at the 126th Meeting of the AMERICANCHEMICALSOCIICTY, New York, N. Y.,1954.
