Chapter 6

With regard to denitration which is both a delicate and disagreeable operation: none of the agents recommended to substitute the sulphydrates have proved available. Of these the author mentions ferrous chloride (6), ferrous chloride in alcohol (7), formaldehyde (8), sulphocarbonates. The different sulphydrates (9) have very different effects. The calcium compound tends to harden and weaken the thread. The ammonia compound requires great care and is costly. The magnesium compound works rapidly and gives the strongest thread. Investigations have established the following point. In practice it is not necessary to combine the saponification of cellulose ester with complete reduction of the nitric acid split off. The latter requires eight molecules of hydrogen sulphide per one molecule tetranitrocellulose, but with precautions four molecules suffice. It is well known that the denitration is nearly complete, traces only of nitric groups surviving. Their reactions with diphenylamine allow a certain identification of artificial silks of this class. Various other inventors, e.g. Du Vivier (10), Cadoret (11), Lehner (12), have attempted the addition of other substances to modify the thread. These have all failed. Lehner, who persisted in his investigations, and with success, only attained this success, however, by leaving out all such extraneous matters. Lehner works with 10 p.ct. solutions; Chardonnet has continually aimed at higher concentration up to 20 p.ct. Lehner has been able very much to reduce his pressures of ejection in consequence; Chardonnet has had to increase up to pressures of 60 k. per cm. and higher. The latter involves very costly distributing apparatus. Lehner made next considerable advance by the discovery of the fact that the addition of sulphuric acid to the collodion caused increase of fluidity (13), which Lehner attributes to molecular change. Chardonnet found similar results from the addition of aldehyde and other reagents (14), but not such as to be employed for the more concentrated collodions. The author next refers to his discoveries (15) that alcoholic solutions of a number of substances, organic and inorganic, freely dissolve the lower cellulose nitrates. The most satisfactory of these substances is chloride of calcium (16). It is noted that acetate of ammonia causes rapid changes in the solution, which appear to be due to a species of hydrolysis. The result is sufficiently remarkable to call for further investigation. The chloride of calcium, it is thought possible, produces a direct combination of the alcohol with a reactive group of the nitrocellulose. The fluidity of this solution using one mol. CaCl_{2} per 1 mol. tetranitrate (17) reaches a maximum in half an hour's heating at 60°-70°C. The fluidity is increased by starting from a cotton which has been previously mercerised. After nitration there is no objection to a chlorine bleach. Chardonnet has found on the other hand that in bleaching before nitration there is a loss of spinning quality in the collodion. The author considers that the new collodion can be used entirely in place of the ordinary ether-alcohol collodion. With regard to the properties of the denitrated products they fix all basic colours without mordant and may be regarded as oxycellulose therefore. The density of the thread is from 1.5 to 1.55. The thread of 100 deniers shows a mean breaking strain of 120 grammes with an elasticity of 8-12 p.ct. The cardinal defect of these fibres is their property of combination with water. Many attempts have been made to confer water-resistance (18), but without success. Strehlenert has proposed the addition of formaldehyde (19), but this is without result (20). In reference to these effects of hydration, the author has made observations on cotton thread, of which the following table represents the numerical results:

Breaking Strain Mean of 20 experiments

Skein of bleached cotton without treatment 825 Skein of bleached cotton without treatment, but wetted 942 Ditto after conversion into hexanitrate, dry 884 The above, wetted 828 The cotton denitrated from above, dry 529 The cotton denitrated as above and wetted 206

The author considers that other patents which have been taken for spinning nitrocellulose are of little practical account (21) and (22). The same conclusion also applies to the process of _Langhans_, who proposes to spin solutions of cellulose in sulphuric acid (23) (24) and mixtures of sulphuric acid and phosphoric acid.

GROUP 2. _Lustra-cellulose._--Thread prepared by spinning solutions of cellulose in cuprammonium.

This product is made by the Vereinigte Glanzstoff-Fabriken, Aachen, according to a series of patents under the names of H. Pauly, M. Fremery and Urban, Consortium mulhousien pour la fabrication de fils brillants, E. Bronnert, and E. Bronnert and Fremery and Urban (1). The first patent in this direction was taken by Despeissis in 1890 (2). It appears this inventor died shortly after taking the patent (3) The matter was later developed by Pauly (4) especially in overcoming the difficulty of preparing a solution of sufficient concentration. (It is to be noted that Pauly's patents rest upon a very slender foundation, being anticipated in every essential detail by the previous patent of Despeissis.) For this very great care is required, especially, first, the condition of low temperature, and, secondly, a regulated proportion of copper and ammonia to cellulose. The solution takes place more rapidly if the cellulose has been previously oxidised. Such cellulose gives an 8 p.ct. solution, and the thread obtained has the character of an oxycellulose, specially seen in its dyeing properties. The best results are obtained, it appears, by the preliminary mercerising treatment and placing the alkali cellulose in contact with copper and ammonia. (All reagents employed in molecular proportions.) The author notes that the so-called hydrocellulose (Girard) (5) is almost insoluble in cuprammonium, as is starch. It is rendered soluble by alkali treatment.

GROUP 3. _Lustra-cellulose_ prepared by spinning a solution of cellulose in concentrated chloride of zinc.

This solution has been known for a long time and used for making filaments for incandescent lamps. The cellulose threads, however, have very little tenacity. This is no doubt due to the conditions necessary for forming the solution, the prolonged digestion causing powerful hydrolysis (1). Neither the process of Wynne and Powell (2) nor that of Dreaper and Tompkins (3), who have endeavoured to bring the matter to a practical issue, are calculated to produce a thread taking a place as a textile. The author has described in his American patent (4) a method of effecting the solution in the cold, viz. again by first mercerising the cellulose and washing away the caustic soda. This product dissolves in the cold and the solution remains unaltered if kept at low temperature. Experiments are being continued with these modifications of the process, and the author anticipates successful results. The modifications having the effect of maintaining the high molecular weight of the cellulose, it would appear that these investigations confirm the theory of Cross and Bevan that the tenacity of a film or thread of structureless regenerated cellulose is directly proportional to the molecular weight of the cellulose, i.e. to its degree of molecular aggregation (5).

GROUP 4. 'Viscose' silks obtained by spinning solutions of xanthate of cellulose.

In 1892, Cross and Bevan patented the preparation of a new and curious compound of cellulose, the thiocarbonate (1) (2) (3). Great hopes were based upon this product at the time of its discovery. It was expected to yield a considerable industrial and financial profit and also to contribute to the scientific study of cellulose. The later patents of C. H. Stearn (4) describe the application of viscose to the spinning of artificial silk. The viscose is projected into solutions of chloride of ammonium and washed in a succession of saline solutions to remove the residual sulphur impurities. The author remarks that though it has a certain interest to have succeeded in making a thread from this compound and thus adding another to the processes existing for this purpose, he is not of opinion that it shows any advance on the lustra-cellulose (2) and (3). He also considers that the bisulphide of carbon, which must be regarded as a noxious compound, is a serious bar to the industrial use of the process, and for economic work he considers that the regeneration of ammonia from the precipitating liquors is necessary and would be as objectionable as the denitration baths in the collodion process. The final product not being on the market he does not pronounce a finally unfavourable opinion.

The author and the Vereinigte Glanzstoff-Fabriken after long investigation have decided to make nothing but the lustra-cellulose (2) and (3). A new factory at Niedermorschweiler, near Mulhouse, is projected for this last production.

BIBLIOGRAPHY

_Introduction_

(1) Bull. de la Soc. industr. de Mulhouse, 1900.

(2) Réaumur, Mémoire pour servir à l'histoire des insectes, 1874, 1, p. 154.

(3) English Pat. No. 283, Feb. 6, 1855.

(4) Swinburne, Electrician, 18, 28, 1887, p. 256.

(5) Weston (Swinburne), Electrician, 18, 1887, p. 287. Eng. Pat. No. 22866, Sept. 12, 1882.

(6) German Pat. No. 3029. English Pat. No. 161780, April 28, 1884 (Swan).

(7) Wynne-Powell, English Pat. No. 16805, Dec. 22, 1884.

_Group I_

(1) German Pat No. 38368, Dec. 20, 1885. German Pat. No. 46125, March 4, 1888. German Pat. No. 56331, Feb. 6, 1890. German Pat. No. 81599, Oct. 11, 1893. German Pat. No. 56655, April 23, 1890. French Pat. No. 231230, June 30, 1893.

(2) Industrie textile, 1899, 1892. Wyss-Noef, Zeitschrift für angewandte Chemie, 1899, 30, 33. La Nature, Jan. 1, 1898, No. 1283. Revue générale des sciences, June 30, 1898.

(3) German Pat. No. 46125, March 4, 1888. German Pat. No. 56655, April 23, 1890.

(4) Swan, English Pat. 161780, June 28, 1884. See also Béchamp, Dict. de Chimie de Wurtz.

(5) German Pat. No. 81599, Oct 11, 1893.

(6) Béchamp, art. Cellulose, Dict. de Chimie de Wurtz, p. 781.

(7) Chardonnet, addit. March 3, 1897, to the French Pat. 231230, May 30, 1893.

(8) Knofler, French Pat. 247855, June 1, 1895. German Pat. 88556, March 28, 1894.

(9) Béchamp, art. Cellulose, Dict. de Chimie de Wurtz. Blondeau, Ann. Chim. et Phys. (3), 1863, 68, p. 462.

(10) Revue industrielle, 1890, p. 194. German Pat. 52977, March 7, 1889.

(11) French Pat. 256854, June 2, 1896.

(12) German Pat. 55949, Nov. 9, 1889. German Pat. 58508, Sept. 16, 1890. German Pat. 82555, Nov. 15, 1894.

(13) German Pat. 58508, Sept. 16, 1900.

(14) French Pat. 231230, June 30, 1893.

(15) German Pat. 93009, Nov. 19, 1895. French Pat. 254703, March 12, 1896. English Pat. 6858, March 28, 1896.

(16) American Pat. 573132, Dec. 15, 1896.

(17) This proportion is the most advantageous, and furnishes the best liquid collodions that can be spun.

(18) French Pat. 259422, Sept. 3, 1896.

(19) English Pat. 22540, 1896.

(20) Application for German Pat. not granted, 4933 IV. 296, Mar. 16, 1897.

(21) German Pat. 96208, Feb. 10, 1897. Addit. Pat. 101844 and 102573, Dec. 10, 1897.

(22) Oberle et Newbold, French Pat. 25828, July 22, 1896. Granquist, Engl. applic. 2379, Nov. 28, 1899.

(23) German Pat. 72572, June 17, 1891.

(24) Voy. Stern, Ber., 28, ch. 462.

_Group II_

(1) German Pat. 98642, Dec. 1, 1897 (Pauly). French Pat. 286692, March 10, 1899, and addition of October 14, 1899 (Fremery and Urban). French Pat. 286726, March 11, 1899, and addition of December 4, 1899. German Pat. 111313, March 16, 1899 (Fremery and Urban). English Pat. 18884, Sept. 19, 1899 (Bronnert). English Pat. 13331, June 27, 1899 (Consort. mulhousien).

(2) French Pat. 203741, Feb. 12, 1890.

(3) The actual lapse of this patent is due to the death of Despeissis shortly after it was taken.

(4) Without questioning the good faith of Pauly, it is nevertheless a fact that the original patent remains as a document, and therefore that the value of the Pauly patents is very questionable.

(5) Girard, Ann. Chim. et Phys, 1881 (5), 24, p. 337-384.

_Group III_

(1) Cross and Bevan, Cellulose, 1895, p. 8.

(2) English Pat. 16805, Dec. 22, 1884.

(3) English Pat. 17901, July 30, 1897.

(4) Bronnert, American Pat. 646799, April 3, 1900.

(5) Cross and Bevan, Cellulose, 1895, p. 12.

_Group IV_

(1) English Pat. 8700, 1892. German Pat. 70999, Jan. 13, 1893.

(2) English Pat. 4713, 1896. German Pat. 92590, Nov. 21, 1896.

(3) Comptes rendus (loc. cit.). Berichte, c. 9, 65a.

(4) English Pat. 1020, 1898. German Pat. 108511, Oct. 18, 1898.

~Artificial Silk--Lustra-cellulose.~

C. F. CROSS and E. J. BEVAN (J. Soc. Chem. Ind., 1896, 317).

The object of this paper is mainly to correct current statements as to the artificial or 'cellulose silks' being explosive or highly inflammable (ibid., 1895, 720). A specimen of the 'Lehner' silk was found to retain only 0.19 p.ct. total nitrogen, showing that the denitration is sufficiently complete to dispose of any suggestion of high inflammability.

The product yielded traces only of furfural; on boiling with a 1 p.ct. solution of sodium hydrate, the loss of weight was 9.14 p.ct.; but the solution had no reducing action on Fehling's solution. The product in denitration had therefore reverted completely to a cellulose (hydrate), no oxy-derivative being present.

* * * * *

The authors enter a protest against the term 'artificial silk' as applied to these products, and suggest 'lustra-cellulose.'

DIE KÜNSTLICHE SEIDE-IHRE HERSTELLUNG, EIGENSCHAFTEN UND VERWENDUNG.

CARL SÜVERN, Berlin, 1900, J. Springer.

~ARTIFICIAL SILK--ITS PRODUCTION, PROPERTIES, AND APPLICATIONS.~

This work of some 130 pages is an important monograph on the subject of the preparation of artificial cellulose threads--so far as the technical elements of the problems involved are discussed and disclosed in the patent literature. The first section, in fact, consists almost exclusively of the several patent specifications in chronological order and ranged under the sub-sections: (a) The Spinning of Nitrocellulose (collodion); (b) The Spinning of other Solutions of Cellulose; (c) The Spinning of Solutions of the Nitrogenous Colloids.

In the second section the author deals with the physical and chemical proportions of the artificial threads.

_Chardonnet 'silk'_ is stated to have a mean diameter of 35µ, but with considerable variations from the mean in the individual fibres; equally wide variations in form are observed in cross-section. The general form is elliptical, but the surface is marked by deep striæ, and the cross-section is therefore of irregular outline. This is due to irregular conditions of evaporation of the solvents, the thread being 'spun' into the air from cylindrical orifices of regulated dimensions. Chardonnet states that when the collodion is spun into alcohol the resultant thread is a perfect cylinder (Compt. rend. 1889, 108, 962). The strength of the fibre is variously stated at from 50-80 p.ct. that of 'boiled off' China tram; the true elasticity is 4-5 p.ct., the elongation under the breaking strain 15-17 p.ct. The sp.gr. is 1.49, i.e. 3-5 p.ct. in excess of boiled off silk.

_Lehner 'silk'_ exhibits the closest similarity to the Chardonnet product. In cross-section it is seen to be more regular in outline, and a round, pseudo-tubular form prevails, due to the conditions of shrinkage and collapse of the fibre in parting with the solvents, and in then dehydrating. The constants for 'breaking strain,' both in the original and moistened condition, for elasticity, &c., are closely approximate to those for the Chardonnet product.

_Pauly 'silk'._--The form of the ultimate fibres is much more regular and the contour of the cross-section is smooth. The product shows more resistance to moisture and to alkaline solutions.

_Viscose 'silk'_ is referred to in terms of a communication appearing in 'Papier-Zeitung,' 1898, 2416.

In the above section the following publications are referred to: Chardonnet, 'Compt. rend.,' 1887, 105, 900; and 1889, 108, 962; Silbermann, 'Die Seide,' 1897, v. 2, 143; Herzog, 'Farber-Zeitung,' 1894/5, 49-50; Thiele, ibid. 1897, 133; O. Schlesinger, 'Papier-Zeitung,' 1895, 1578-81, 1610-12.

_Action of Reagents upon Natural and Artificial Silks._

1. _Potassium hydrate_ in solution of maximum concentration dissolves the silks proper, (a) China silk on slight warming, (b) Tussah silk on boiling. The cellulose 'silks' show swelling with discolouration, but the fibrous character is not destroyed even on boiling.

2. _Potassium hydrate_ 40 p.ct. China silk dissolves completely at 65°-85°; Tussah silk swells considerably at 75° and dissolves at 100°-120°. The cellulose 'silks' are attacked with discolouration; at 140° (boiling-point of the solution) there is progressive solvent action, but the action is incomplete. The Pauly product is most resistant.

3. _Zinc chloride_, 40 p.ct. solution. Both the natural silks and lustra-celluloses are attacked at 100°, and on raising the temperature the further actions are as follows: China silk is completely dissolved at 110-120°; Tussah silk at 130-135°; the collodion products at 140-145°; the Pauly product was again most resistant, dissolving at 180°.

4. _Alkaline cupric oxide_ (glycerin) solution was prepared by dissolving 10 grs. of the sulphate in 100 c.c. water, adding 5 grs. glycerin and 10 c.c. of 40 p.ct. KOH. In this solution the China silk dissolved at the ordinary temperature; Tussah silk and the lustra-celluloses were not appreciably affected.

5. _Cuprammonium solution_ was prepared by dissolving the precipitated cupric hydrate in 24 p.ct. ammonia. In this reagent also the China silk dissolved, and the Tussah silk as well as the lustra-celluloses underwent no appreciable change.

6. _An ammoniacal solution of nickel oxide_ was prepared by dissolving the precipitated hydrated oxide in concentrated ammonia. The China silk was dissolved by this reagent; Tussah silk and the lustra-celluloses entirely resisted its action.

7. _Fehling's solution_ is a solvent of the natural silks, but is without action on the lustra-celluloses.

8. _Chromic acid_--20 p.ct. CrO_{3}--solution dissolves both the natural silks and the lustra-celluloses at the boiling temperature of the solution.

9. _Millon's reagent_, at the boiling solution, colours the natural silks violet: the lustra-celluloses give no reaction.

10. _Concentrated nitric acid_ attacks the natural silks powerfully in the cold; the lustra-celluloses dissolve on heating.

11. _Iodine solution_ (I in KI) colours the China silk a deep brown, Tussah a pale brown; the celluloses from collodion are coloured at first brown, then blue. The Pauly product, on the other hand, does not react.

12. _Diphenylamine sulphate._--A solution of the base in concentrated sulphuric acid colours the natural silks a brown; the collodion 'silks' give a strong blue reaction due to the presence of residual nitro-groups. The Pauly product is not affected.

13. _Brucin sulphate_ in presence of concentrated sulphuric acid colours the natural silks only slightly (brown); the collodion 'silks' give a strong red colouration. The Pauly product again is without reaction.

14. _Water._--The natural silks do not soften in the mouth as do the lustra-celluloses.

15. _Water of condition_ was determined by drying at 100°; the following percentages resulted (a). The percentages of water (b) taken up from the atmosphere after forty-three hours' exposure were:

(a) (b) China (raw) silk 7.97 2.24 Tussah silk 8.26 5.00

Lustra-celluloses:

Chardonnet (Besançon) 10.37 5.64 " Spreitenbach 11.17 5.77 Lehner 10.71 5.97 Pauly 10.04 6.94

16. _Behaviour on heating at 200°._--After two hours' heating at this temperature the following changes were noted:

China silk Much discoloured (brown). Tussah silk Scarcely affected.

Lustra-celluloses:

Chardonnet Converted into a blue-black charcoal, retaining the Lehner form ofthe fibres.

Pauly A bright yellow-brown colouration, without carbonisation.

17. The _losses of weight_ accompanying these changes and calculated per 100 parts of fibre dried at 100° were:

China silk 3.18 Tussah silk 2.95

Lustra-celluloses:

Chardonnet 33.70 Lehner 26.56 Pauly 1.61

18. _Inorganic constituents._--Determinations of the total ash gave for the first five of the above, numbers varying from 1.0 to 1.7 p.ct. The only noteworthy point in the comparison was the exceptionally small ash of the Pauly product, viz. 0.096 p.ct.

19. _Total nitrogen._--The natural silks contain the 16-17 p.ct. N characteristic of the proteids. The lustra-celluloses contain 0.05-0.15 p.ct. N which in those spun from collodion is present in the form of nitric groups.

The points of chemical differentiation which are established by the above scheme of comparative investigation are summed up in tabular form.

_Methods of dyeing._--The lustra-celluloses are briefly discussed. The specific relationship of these forms of cellulose to the colouring matters are in the main those of cotton, but they manifest in the dye-bath the somewhat intensified attraction which characterises mercerised cotton, or more generally the cellulose hydrates.

_Industrial applications_ of the lustra-celluloses are briefly noticed in the concluding section of the book.

FOOTNOTES:

[3] With these products it is easy to observe that they have a definite fusion point 5°-10° below the temperature of explosion.

SECTION III. DECOMPOSITIONS OF CELLULOSE SUCH AS THROW LIGHT ON THE PROBLEM OF ITS CONSTITUTION

UEBER CELLULOSE.

G. BUMCKE und R. WOLFFENSTEIN (Berl. Ber., 1899, 2493).

(p. 54) _Theoretical Preface._--The purpose of these investigations is the closer characterisation of the products known as 'oxycellulose' and 'hydracellulose,' which are empirical aggregates obtained by various processes of oxidation and hydrolysis; these processes act concurrently in the production of the oxycelluloses. The action of hydrogen peroxide was specially investigated. An oxycellulose resulted possessing strongly marked aldehydic characteristics. The authors commit themselves to an explanation of this paradoxical result, i.e. the production of a body of strongly 'reducing' properties by the action of an oxidising agent upon the inert cellulose molecule (? aggregate) as due to the _hydrolytic_ action of the peroxide: following Wurster (Ber. 22, 145), who similarly explained the production of reducing sugars from cane sugar by the action of the peroxide.

The product in question is accordingly termed _hydralcellulose_. By the action of alkalis this is resolved into two bodies of alcoholic (cellulose) and acid ('acid cellulose') characteristics respectively. The latter in drying passes into a lactone. The acid product is also obtained from cellulose by the action of alkaline lye (boiling 30 p.ct. NaOH) and by solution in Schweizer's reagent.

It is considered probable that the cellulose nitrates are hydrocellulose derivatives, and experimental evidence in favour of this conclusion is supplied by the results of 'nitrating' the celluloses and their oxy- and hydro- derivatives. Identical products were obtained.

_Experimental investigations._--The filter paper employed as 'original cellulose,' giving the following numbers on analysis: