February, 1925
INDUSTRIAL AND ENGINEERING CHEMISTRY
161
Determination of Rubber and Inorganic Materials in Soft Rubber Goods' By R. T. Mease and N. P. Hanna BUREAUO F
STANDARDS, WASHINGTON,
D.
c.
HE method commonly used for the determination of form as to prevent filtration. Experimental work was then rubber in vulcanized rubber goods is an indirect one.2 done with mineral oils of different qualities. When this method is used the analysis of many samIt is common knowledge that certain saturated hydrocarples will indicate more rubber than is actually present. The bons, such as mineral oils, cause rubber goods to disintegrate at error of the method will be quite large in samples containing ordinary temperatures, and that the action is greatlyaccelerated compounding ingredients that decompose or are volatile a t by an increase of temperature. The action is not accompathe temperature required for ignition. Several direct meth- nied by any great swelling of the sample. From this it would ods for the determination of rubber hydrocarbon are discussed appear that a mineral oil when used as a rubber solvent might meet conditions (1) in the literature, but none A method for the determination of inorganic materials and ( 2 ) j and that it may be Of them have been found in rubber goods which is accurate, reasonably rapid, and purified to meet conditions Sufficiently accurate Or apapplicable to a large variety of compounds is proposed, and (3), (4)7 and (5). The to a large enough is a n indirect method for the determination of rubber vent found most witable variety of compounds to be considered standard* For hydrocarbon. It differs from the usual methods for the for the purpose intended many purposes it is desirdetermination of the inorganic materials and consists i n was a mixture Of 75 per cent the removal of the rubber in the compound by solution, by volume of 0% arbitrarily able to determine not Only quantitative separation of the inorganic materials by fildesignated No* 32, and rubber, but also the amount 25 per cent of No. 34. The and nature Of the Gomtration. and their Dartial analysis. oils have the following pounding materials used. properties: Action of Solvents on Vulcanized Rubber Compounds Table I-Properties of Oils Finally Usedu Oil No. 32 Oil No. 34 To reduce the amount of error inherent in the methods in Color ............................. Colorless Colorless common use and in many cases the time required for analysis Acidity.. ......................... 0.01 0.04 ........................... 0.13% 0.10% as well, it would seem better to remove the rubber from the Sulfur.. Viscosity at 20' C. (68' F.) (Saybolt compounding materials and determine the latter in as nearly Universal). ...................... 56 sec. 52 sec. 37.8' C. (100'F.) (Saybolt an unchanged state as possible. This would not only make it Viscosityat Universal), .................:. ... 45 sec. 43 sec. point.. ...................... 132.2' C.(270°F.) 129.4'C.(265'F.) possible to determine the rubber by difference, but would also Flash 176.7 C. (350 F.) 157.2 C. (3 15' F.) Fire point. ........................ permit the examination of the pigments thus obtained. To Specific gravity. ................... 0.8530 0.8405 effect such a separation many solvents have been recoma Tests made in the laboratories of the Bureau of Standards by N. Berrymended. If the separation is to be made quantitative the man and G . E. Graham. solvent used must have the following properties: During the course of experiment it was learned that mix(1) I t must dissolve vulcanized rubber compounds in a reasona- tures of oils generally act better than one alone. It is not necble time and under conditions easily obtainable in the laboratory. (2) Aft,er solution it must permit complete separation of the essary that the oils rigidly meet the above specifications given pigments from the solution of rubber by some simple method in Table I, but before using them in quantitative work they such as centrifuging or filtration. should be tried separately and as mixtures. ( 3 ) It must have no chemical action upon any of the comThe advantage of the mixture of oils that was chosen is its pounding ingredients. (4) It must not burn, oxidize readily, or evaporate at the ability to effect nearly complete solution in a relatively short time and at low temperature. If 0.5 gram of the sample temperature required to effect solution. (5) It must be easily obtained or prepared and at a moderate is heated in this solvent at 150" C., in most cases the rubber price. will be broken down at the end of 3 or 4 hours, 70 to 95 per In the long list of known solvents it is not difficult to find cent of the rubber hydrocarbon present will be in solution, and one or more that will meet all conditions except (2). Solvents the solid can be filtered off by means of an asbestos pad. The whose initial action is such as to cause much swelling of the solvent and temperature used do not change the weight of sample generally do not meet condition ( 2 ) . The action of the inorganic materials present. The separation can be made benzene and xylene is typical of this class. Such solvents as quantitative, and a method was adopted for the determination cresol or tetrachloroethane, which have a chemical action on of inorganic ingredients and used to determine rubber hydromany compounding ingredients, are not applicable when the carbon by differences. latter are to be determined or examined. By applying the foreMethod going rules many solvents can immediately be eliminated from (1) Weigh two portions (a and b) of 0.5 gram of the finely consideration so far as their application to quantitative methods is concerned. The reasons for the peculiar actions of the ground sample. Extract each with a mixture of 32 parts by differenttypes of solvents have not been investigated. Many volume of acetone and 68 parts of chloroform for a minimum of solvents were tried, including cresol, kerosene, diethyl phthal- 8 hours.s If the liquid in the extraction bucket is still colored ate, ethylene chloride, and several others. I n all cases there a t the end of this time, continue the extraction. Remove the remains an insoluble portion, and in some cases it is of such a samples and put each into a 150-cc. lipped assay flask, add 20 to 25 cc. of the mixed oils, cover with a watch glass, and 1 Received April 7, 1024. Published by permission of the Director,
T
O
U. S. Bureau of Standards. * Bur. Standards, CI'IG.S8,4th ed. ;THIS JOURNAL, 16,397 (1924).
8 Bur. Standards, Tech. Pager 162; Rubber Ape and Tire Ncws. 6 , 445 (1920).
162
INDUSTRIAL AND ENGINEERING CHEMISTRY
heat in an air bath a t a temperature of 150' to 155' C. until solution appears complete, which requires about 3 hours and then 15 to 30 minutes more. Solution may be considered complete when the rubber colloid has been broken down and the oil seems quite clear. Remove the flasks from the air bath, cool to about 110' C., and add in a small stream 10 to 1Fj cc. of benzene, while mixing thoroughly; allow to cool and then dilute with sufficient petroleum ether to fill the flasks to within about 2 cm. of the top; mix thoroughly,' cover the flasks to prevent evaporation, and allow the mixture to stand overnight. (2) Treatment of Portion a. Prepare a Gooch crucible with finely divided asbestos that has been treated alternately with strong caustic soda solution and concentrated hydrochloric acid and then washed well with water. Ignite the crucible, cool and weigh; call this weight c. Filter the mixture by decantation through the crucible, using suction; wash well with petroleum ether, followed by warm acetone, and by a warm mixture of equal volumes of acetone and chloroform if the filtrate is dark. Finally, wash with hot alcohol. A portion of the pigments will remain in the flask. Dry the crucible and flask with their contents for 1 hour at a temperature of 105' to 110' C. Cool and weigh. Call the weight of the flask and adhering residue d, and of the crucible and contents e. (3) Remove the acid-soluble materials from the flask and Gooch crucible, as follows: Add to the flask and crucible a few cubic centimeters of boiling alcohol; allow to soak for 2 or 3 minutes and then wash two or three times with boiling water; let the flask cool, add 10 cc. of concentrated hydrochloric acid, and swirl the flask to bring the acid in contact with the pigments. Pour the acid from the flask into the crucible and let it stand until no more bubbles rise through t8he liquid. If carbonates are present there is danger of loss by excessive frothing. This can be prevented by first adding a few drops of the acid to the crucible and sucking it through the pad. After the violent action has ceased, add the remainder of the 10 cc. of acid. When no more gas is evolved, draw the acid through the pad and again wash with 20 cc. of acid, adding a little a t a time; then wash well with hot water and transfer as much as possible of the residue remaining in the flask to the asbestos pad. If by qualitative tests the sample was found to contain antimony, save the filtrate and washings for treatment as described in (4). Dry the flask and crucible for 1hour a t 105' to 110' C., cool, and weigh. Call the weight of the flask f, and that of the crucible containing the organic residue and acid-insoluble materials, h. Burn the organic residue from the asbestos pad by igniting in a furnace a t 700" C., cool, and weigh. Call this weight k. If the sample contains barytes, save the contents of the crucible for treatment as described in (5). (4)Determination of the Sulfur Present as Antimony Trisul$de.* Determine the antimony and express the sulfur present as antimony trisulfidein percentage of the sample and call this S. (5) Determination of Barytes. Transfer the contents of the crucible from (3) to a 50-cc. porcelain crucible and fuse with 5 grams of a mixture of equal weights of sodium carbonate and Cool the crucible, nitrate. Stir well during the fusion. put it into a 250-cc. beaker, cover with distilled water, and heat on the steam bath until all the solid material has been loosened from the crucible; filter and wash the residue back into the beaker and dissolve with 10 per cent hydrochloric acid, using about 10 cc. excess and keeping the solution warm. Filter off the asbestos, wash well with hot water, and precipitate the barium in the solution with 10 cc. of a 10 per cent solution of sulfuric acid, Heat on the steam bath, filter off the precipitate, ignite, and weigh the barium sulfate. Calculate the percentage of sulfur present as barytes and call this value rn. 4
Collier, Levin, and Scherrer, India Rubbcv J . , 6 4 , 580 (1921).
Vol. 17, No. 2
(6) Treatment of Portion b for Sulfur in Compounding Materials. Treat Portion b as described under (2), but omit the weighing of the crucible. After the pigments have been dried, transfer the asbestos pad and pigments into the flask. The last traces of pigments can be removed from the sides of the crucible with wads of absorbent cotton moistened with a little warm water. Add to the flask about 10 cc. of bromine water and swirl the flask to moisten the contents. Add 20 cc. o€ concentrated nitric acid saturated with bromine.5 Allow to stand in the cold for 15 minutes and then heat on the steam bath for 1 hour. Transfer the contents of the flask to a 75-cc. crucible, and evaporate to dryness. Complete the determination of sulfur by the method of the Bureau of Standards.2 Let the value for the percentage of sulfur in fillers be represented by n. The percentage of total inorganic materials corrected equals (e d k) - (c + f + h ) + (m + - ?z
+ +
0.5
To determine the percentage of rubber hydrocarbon subtract from 100 the sum of the percentages of acetone, chloroform, alcoholic alkali extracts, free carbon, glue, total sulfur corrected, and total inorganic materials correctede2 Principal Source of Error and Accuracy of Method
The principle source of error in the proposed method is in determining when solution is complete, especially if the sample contains free carbon, To determine the accuracy of the method, it was tried on samples representing different types of rubber compounds, designated as A-1, A-2, 206-25, 20645, and 207-15. The composition of each according to the compounder is given in Table 11. Table 11-Composition of Experimental C o m p o u n d s (Per cent by weight) A-1 A-2 206-25 206-45 207-15 MATERIAL Rubber 79.25 78.40 40.00 40.00 37.00 Magnesium carbonate 0.0 0.0 12.00 12.00 15.00 0.0 0.0 0.0 0.0 8.00 Barytes Barium carbonate 0 .O 0.0 0.0 0.0 7.00 Zinc oxide 10.00 0.0 26.00 26.00 8.00 Iron oxide 0.0 0.0 0.0 0.0 5.00 Slate dust 0.0 0.0 0.0 0.0 4.00 Sulfur 4.25 3.90 2.00 2.00 1.00 Antimonv 5.00 17.25 0.0 0.0 15.00 Hexametkylenetetramine 0.50 0.39 0.0 0.0 0.0 2.00 2.00 0.0 Paraffin i;oo 0.0 12.00 12.00 0.0 0.0 Whiting 0.0 0.0 0.0 6.00 6.00 Litharge 0.0 15 min. 45 min. Time of cure 1 hr. 1: m. 1 hr. 15 m. 25 min. 154.4' C. (310' F.) Temuerature of cure 137.8 C. (280' F.) Calcdated specific gravity 1,100 1.091 1.588 1.588 1.651
Table I11 contains the results of analysis of the samples listed in Table I1 for the determination of rubber as conpounded by the new method and are illustrative of the results obtained by the method as used by the Bureau of Standards. Table 111-Analysis of Experimental Compounds by New M e t h o d MATsR I A L A-1 A-2 206-25 206-45 207-15 Acetone extract 3.05 2.63 3.26 3.36 1.37 Chloroform extract 0.27 1.44 0.59 0.43 0.36 Total inorganic materials 56.90 66.86 1 2 . 4 4 56.62 15.53 (corr.) 1.95 4.33 1.95 7.30 5.22 Total sulfur (corr.)m 1.609 1.648 2.057 1.607 1.062 Specific gravity 37.36 37.08 37.58 75.93 76.19 Rubber hydrocarbon 39.23 38.81 39.46 79.25 80.0 Rubber as compoundedb 68.04 67.46 67.15 89.54 89.96 Rubber by volume Rubber as compounded 4 0 . 0 0 37.00 4 0 . 0 0 7 8 . 4 0 70.25 (Table 11) 0.48 2.04 3.86 1.07 3.17 Free sulfur Calculated specific gravity 1.588 1.651 1.588 1.091 1.100 (Table 11) a This value does not include sulfur in combination with the compounding materials, but includes sulfur added as such and that added as available sulfur in the ingredients used, such as the free sulfur in commercia1 antimony pigments. b Rubber h7pocarbon plus 5 per cent of its weight is taken as "rubber as compounded except when the sum of the acetone and chloroform extracts is less thLn the figure represented by the arbitrary 5 per cent. I n the latter case, rubber as compounded is considered as the s u m of rubber hydrocarbon, acetone extract corrected, and chloroform extract. I
Bur. Standards, Sci. Paper 174.
February, 1925
IATDUSTRIAL AND ENGIhTEERING CHEMISTRY
In comparing results obtained for rubber as compounded with the value stated by the compounder to be present, the calculated and determined specific gravities should also be compared. Any errors introduced in compounding the samples or in their analyses would be indicated by this comparison. Thus, if a sample has a calculated gravity of 1.5888 and is said to contain 40 per cent of rubber, and if the determined specific gravity is l .607, then the sample contains more inorganic materials than was planned and is therefore
163
lacking in rubber content. The case given is that of 206-25 and is also true of 206-45. A modification of the method given above has been found useful in this and in other laboratories for quantitatively separating rubber compounds from frictioned rubber goods, such as rubberized cloth and asbestos materials. It makes possible an examination of the fibers used in weaving the fabric and an analysis of the rubber compound.
A Remarkable Property of Selenium' By Ernest C. Crocker ARTHUR D. LITTLE, INC., CAMBRIDGE, MASS.
flameproof insulated electric wire, it was discovered quite empirically that selenium had great value. The customary method for flameproofing telephone switchboard wire consists in treating the cotton braid with fusible materials, such as borates, phosphates, tungstates, molybdates, and even silicates, together with some salts which evolve ammonia or other gas, and some water on heating. While this method secures a degree of flameproofness under adjusted test conditions, which pass the best and refuse the poorer, such flameproofing is of very limited real value in service. In particular, while the specifications state that a single wire will not continue to burn in a horizontal position after removal of the igniting flame, a bundle or loose cable of such wires may and usually does burn freely. In trying out numbers of different classes of materials for leads which would indicate different methods of attack of the flameproofing problem, it was noted that selenium, which is itself combustible (though not free burning), imparted remarkable flameproofness. So far as known this is the first instance of a completely combustible material being used as a flameproofing agent. An outstanding feature of the flameproofing ability of selenium is that the larger the bundle of wires the more secure is the flameproofness. Another and perhaps no less valuable property is that when used no electrolyte is applied to the insulation of the wire to cause appreciable conduction during humid weather. Both of these features are highly desirable in flameproofed wire for use on the back of telephone switchboards. There a slight fire, if it spreads, can ruin the service of thousands of subscribers; and surface conductivity of the wire, such as is now unfortunately usual in damp weather, lowers the efficiency of all service, makes for cross talk, and frequently causes the sound of bell ringing in nearby wires. One part by weight of selenium protects 3.6 parts of cotton braid and 10 parts of rubber covering in the presence of 10 parts of copper conductor, constituting the usual No. 22 switchboard wire. Three milligrams of selenium thus protect a centimeter of length of wire, or one pound protects nearly a mile. The protection referred to is freedom from spreading of the flame along the wire held horizontally when the Bunsen burner, candle, or other igniting agent is removed. This much protection is secured by the usual processes only when the wire is practically crusted with salts. On application of the flame much difficulty is experienced in getting the wire to ignite a t all. When ignited, however, and the igniting flame is removed, the flame usually goes out without appreciable spreading. The selenium imparts a strong blue color to the border of the flame and particularly to that
part close to the wire. Just before the flame goes out the luminous part seems to rise upon a bank of blue flame with a separation of 6 mm. or more between the wire and flame a t the moment of extinction. It appears that the concentration of selenium vapor is highest close to the wire, where it forms a blanket which dilutes the otherwise flammable gases which get through it, and serves to insulate and thus prevent the heat of the flame from going down and making more gas out of the cotton and rubber below. It is interesting indeed to find as we do here, that when one heavily treated wire is braided or twisted into a cable with two untreated wires, the resulting cable is completely flameproofed. For some reason not yet apparent selenium combined with the rubber covering of the wire3 by vulcanization (in place of sulfur) is not effective for flameproofing, even though the amount of selenium be considerable. The selenium is most effective when applied to the outside of the cotton braid, but also works very well when applied to the outside of the rubber, under the braid. The complete story of why selenium is so effective in preventing combustion is yet to be told. There is possibly an action analogous to the combustion-retarding effect of tetraethyl lead or diethyl selenide on gasoline vapors. Selenium appears to be unique in possessing flameproofing power of the first magnitude. Arsenic oxide has some of the same action, though of a lower order, and molybdenum compounds a trace. The element tellurium is practically without action, as are also iodine, tin, bismuth, and sulfur. Compounds of selenium such as selenious and selenic acids, sodium selenite, and selenium oxychloride have been found to possess flameproofing power in high degree. In general, however, the use of the element selenium is preferable. It is perhaps best applied in powdered form, with an adhesive, to the rubber covering of the wire before the cotton is braided on. When selenium-treated wire is burned there is a strong odor reminiscent of both sulfur dioxide and garlic. Although unpleasant and noticeable over a wide area, no ill effects due to its inhalation have been noted. At any rate, during fires its fumes are far less objectionable and perhaps less poisonous than those from carbon tetrachloride. Selenium may be used to flameproof paper, scenery, and other diffuse materials, but the amount needed is relatively large. The effect appears to depend on surface action and the surface area is here quite great. Electric wire and probably other comparatively concentrated masses of combustible material are, however, economically protected. This is indeed fortunate, since the usual flameprookg, which works fairly well on the more diffuse materials, is but poorly suited for insulated wire.
Received December 8, 1924. This investigation was conducted by Arthur D . Little, Inc., for the Simplex Wire & Cable Co., Cambridge, Mass. The results are published with the kind permission of Everett Morss, president of the Simplex Company.
a This selenium-vulcanized rubber was made more than seven years ago by C. R. Boggs, superintendent of the Simplex Wire & Cable Company. It had retained the greater part of its original elasticity. See THISJOURNAL, 10, 117 (1918).
URING a recent investigation2 to find new ways to
D
1
