Document x1Q5E2X7MnKMDXVwg11D4aQME

September, 1925 IXDc'STRI9L A S D ENGINEERI;VG CHE-VISTRY 909 The Corrosion of Certain Metals b y Carbon Tetrachloride' By F. H. Rhodes and J. T. Carty CORNELIU, NIVERSITYIT, HACAN, . Y Of the various metals tested, nickel is the most resis- tant to cold carbon tetrachloride, either wet or dry, and to the dry vapor of carbon tetrachloride; and tin is the most resistant to the action of the wet vapor. Some interesting phenomena in connection with the corrosion of aluminium and brass are observed. M 08T of the work on the corrosion of metals has been concerned primarily with corrosion by moist air, by water, or by aqueous solutions of electrolytes, and but comparatively little attention has been paid to the action of nonaqueous liquids on metals. That this latter phase of the general problem of corrosion has attracted so little attention is quite natural, since moqt of the common nonaqueous liquids are without pronounced action on most metals under ordinary conditions. Carbon tetrachloride, however, has been found to attack certain metals to a considerable extent, so that the study of the action of this liquid on metals might,be expected to afford interesting information on the general subject of corrosion. Furthermore, such an investigation should furnish data that would be of value in the selection of materials for use in the construction of containers for this liquid and of apparatus used in extraction processes in which carbon tetrachloride is the solvent. Such data should also bear upon the suggested application of carbon tet'rachloride as a liquid for use in transformers arid elect'rical switches. Previous Work Some work on the action of carbon tetrachloride on metals has already been done, but unfortunat)ely the published data are rather fragmentary and inconclusive and to some extent contradictory. Crocker2 discusses the use of carbon tetrachloride as a solvent in commercial extraction processes and points out that this solvent may corrode the apparatus used in the extraction plant. He makes t'he rather confusing statement that- Steel, cast iron, and wrought iron are only slightly affected by carbon tetrachloride; but the presence of moisture causes decomposition, especially where the solution is in rapid circulation, and if, with rapid circulation there is an increase of heat, the decomposition takes the form of oxide of iron and a ferric chloride. This is quite marked if the moisture is excessive; but to exclude all moisture is practically impossible.* * * Lead, lead-antimony alloys, and tin resist carbon tetrachloride under varying conditions. Copper and its alloys and bronzes, German silver, and gun metal stand the action of carbon tetrachloride, but if the 1 Received May 2 i , 1925. 2 ElecfrorAem. M e / . I n d . , 5, 259 (1007). temperature is high the action is marked. Zinc and aluminium will stand fairly well, nickel about the same as copper; but the presence of moisture, especially if a t a high temperature, is most deteriorating to these metals. Crocker gives no more detailed data and no experimental results to support his statements. Zappi3 found that aluminium is uncorroded even after 4 months' exposur'e to carbon tetrachloride a t room temperature, but that when heated with carbon tetrachloride in a closed tube to a temperature of 100" to 120" C. aluminium reacts rapidly to form aluminium chloride, hexachloroethane, and other higher chlorides of carbon and a resinous product of undetermined composition. Iron powder did not react in the cold and was only slightly attacked in the hot. Sastry' reports that carbon tetrachloride vapor a t its boiling point showed no action on steel, wrought iron, nickel, or aluminium, and only very slight action on copper and lead. As the period of exposure was only 10 hours the results are hardly conclusive. Scope of Present Work In the present investigation we have studied the action, upon a number of metals and alloys, of (a)dry carbon tetrachloride a t room temperature, ( b ) carbon tetrachloride and water a t room temperature, ( c ) the vapor of dry carbon tetrachloride a t its boil- Figure I-Apparatus termining Corrosion for Deby Car- ing point (76Oto 77" C.),and ( d ) bon Tetrachloride Vapor t h i wet vapor from a boiling mixture of water and carbon tetra- chloride at about 67' C. Materials Used The carbon tetrachloride was prepared from presumably pure carbon tetrachloride by agitating first with a 5 per cent 3 A n n soc qutm Argenhno, 2, 217 (1914); C. .4., 9, 3001 (1916). 4 J . SOC.Chem. (London), 35, 94 (1916). ?IfETAI,S Steel Aluminium Copper "Ambrac" Monrl Phosphor bronze Manganese bronze Tobin bronze Brass No. 1 Brass No 2 Brass No. 3 Brass No. 4 Sickel Tin Lead " By difference CU ... ... 99.52 74.21 29.98 95.33 59.20 56.49 90.08 78 90 70 27 59.73 0. ..2.5 0 Zn .. .. .. ... 5 ..6. 0. 39:si 42.64 9 .92a 21. loa 29.730 40. 2 i a Tahle I-Analyses of Metals and Alloys + +Sn ... ... ... .. .. .. 4.47 0.49 0.45 ... 99. . . Mn 0.912 0.00 .. .... 0 ..3. 2. 0.09 ...I ........ o:O66 .. .. .. N.. ..i .. 2O:& 6 6... ....7...5 .. .. .. .. .. .. .. .. .. ... Pb .. .. . , .. .. .. .. .. .. 0.26 Trace .. .. .. ... ... ... 99 Fe 0 :.i.g.f i ... 1.. 6. 1. ... 0 ...4.. ..2 .., ... 0 ..5. 9. ,.. s ... 0:oiz Si 0.2546 0 . 8. 0. . 0.iO ... o:..o..i..g o.i.. ..i..o P .... ........ 0...1..8.. ...... ...... ...... C 0.1...1...4... 0 . 1......4............ 0 .O....iS..s 910 INDUSTRIAL AND ENGINEERING CHEiMISTRY Vol. 17, No. 9 solution of silver nitrate, then with a 2 per cent solution of sodium hydroxide, and finally with distilled water. The mashed material was dried over anhydrous sodium sulfate and redistilled fractionally, the first and last fractions being discarded and only the middle fraction that passed over at constant boiling point being collected for use. The metals used in the corrosion experiments were commercial sheet aluminium, block tin, chemical sheet lead, copper, malleable nickel, mild steel, monel metal, phosphor bronze, Tobin bronze, manganese bronze, "Ambrac" (copper-nickel-zinc alloy), and various brasses ranging from 60 per cent copper to 90 per cent copper.K The block tin and the sheet lead were not analyzed. The analyses of the other metals and alloys are shown in Table I. Results Of the metals exposed to the action of the dry liquid carbon tetrachloride a t room temperature the only one to show appreciable attack was steel, which became covered with a uniformly thin, adherent, brown film. On each of the test pieces of copper, bronze, and brass a very thin, iridescent, brownish film was formed, but no appreciable corrosion or pitting took place. All the other metals and alloys remained bright and unattacked, even after 6 months' exposure. The results obtained in the tests made with liquid carbon tetrachloride and water a t room temperature are presented in Table 11. The period of exposure in these tests was approximately 6 months. Procedure I n the experiments made a t room temperature the test pieces were simply immersed in carbon tetrachloride or in a mixture of carbon tetrachloride and water and allowed to stand. In the tests made with carbon tetrachloride vapor the apparatus used in Figure 1 was used. A glass tube, C, approximately 5 cm. in diameter was connected with a reflux condenser and with the flask containing the boiling carbon tetrachloride or mixture of tetrachloride and water. A return tube, B, with a vapor inlet near the top and with a trapped outlet for the condensed liquid at the bottom was attached to the end of the reflux condenser. The bottom of Table 11-Corrosion b y Carbon Tetrachloride a n d Water at R o o m Temperature METAL Loss in weighta Mg./sq. dm./24 hours REMARUS Aluminium 7.716 Test piece covered with white crust interspersed with black resinous deposit and gelatinous masses of hydrated AlzOa Tin 0.738 Yellow powdery deposit, marked pitting at bound- ary between CClc and water Lead 0.147 Slight white film Monel 0.229 Rather marked pitting, some parts of surface not attacked Steel 6.640 Marked corrosion and slight pitting. metal covered with heavy deposit of Fen03 and Re304 Nickel 0.028 No corrosion, very faint blue or yellow spots on surface of bright metal Copper 0.774 Dark purple film, few localized green deposits Phosphor bronze 3.48 Rather .marked superficial corrosion. metal covered with greenish crust Manganese bronze 1.187 Brown film Tobin bronze 0.340 Brown film, with localized green patches Brass, 60-40 0.342 Slight uniform brown film Brass, 70-30 1.07 Brown film with green spots Brass, 80-20 1.69 Uniform brown film Brass, 90-10 0.618 Iridescent purple film with few green spots "Ambrac" 0.229 Metal covered with grayish green scale and slightly pitted 'Before weighing the test pieces after corrosion they were scrubbed thoroughly to remove any adhering deposit of oxide or basic salt. I Figure 2-Appearance of Test Pieces after Exposure to Wet a n d Dry Vapors the return tube was surrounded by a guard tube, D, so that the refluxed liquid from the condenser did not come in direct contact with the metal test piece, A , in the outer tube. The entire tube was surrounded with a layer of asbestos insulation to minimize the condensation within the tube. The period of exposure in these tests was 100 hours.. 6 The brasses, bronzes, and "Ambrilc" metal were obtained through the courtesy of the American Brass Company division of the Anaconda Copper Co.;the malleable nickel through the courtesy of the International Nickel Company. In every case the water layer acquired an acid reaction although with the test made with the metallic nickel the acid- ity was very faint. With the 6040, 70-30, and 80-20 brasses the aqueous layer assumed a distinct green color. Inasmuch as dry carbon tetrachloride has no appreciable effect upon the metals, it is probable that the corrosive action of wet tetrachloride is due largely to the hydrochloric acid formed by the interaction of carbon tetrachloride and water. The observed differences in the rates of corrosion of the various metals may be due either to differences in susceptibility to the action of hydrochloric acid under the conditions of the experiment or to differencesin the effects of the metals in promoting the decomposition of carbon tetrachloride by water. Nickel, lead, "Ambrac," and the various brasses and bronzes were not affected by the vapor of dry carbon tetra- chloride. Copper became coated with a very thin black film but remained otherwise unattacked. An irregular white film was formed on the surface of the tin, and on the monel test piece there appeared an irregular greenish film, but in neither case was there marked corrosion or pitting of the metal. Steel was more or less uniformly attacked, with the formation of an adherent brown deposit. Aluminium reacted so vigorously with the vapor of carbon tetrachloride that in some cases the test piece was completely dissolved. I"_.... *... Llg./sq. dm./24 hours 307.0 5.62 15.08 94.20 298.5 RBV*.YR &+on tctrat:hloride may therefore he represented by the Test picee etched end covered iri~~uiarly with xe- equat'ioir idar<aitneodurAlzO=La ases 01 hy- 2A1 -t GCClr = 2A1C134-3GClo Slight yello," film None of the other metals that were tested appear 1.0react fimJ3ack slight p i t l i w directly with tho carbon tetrachloride t,n form hexachloro- Slight ikresuliir etcliing &larked eorioslon, test pieci. "t"i"ti"' inctak. The appearatice of t,hc test pieces expused to the wet and dry vapors is showii in Vipres 2 and 3. The reaction of alaminium with tlre \apor of carbon tct.racliloride i5. \.cry iutcrc&ng. Wheii exposcd to the dry vapor metallic F i W m 4 - P h o t o m l u o ~ . r a ~ hacroe.8 Inferface between Attacked and Unstrscked Portlons of 60-40 Brass tile of water vapor alu,ninium is much readily at,tacked by the carbon tetrachloride. The water appears to react with the metallic aluminium or with the aluminium chloride to give aluminium oxide, which forms a protective coating on the surface of the metal and prevents the rapid action of the tetrachloride. Some interesting phenomena were observed in connection with the corrosion of brasses by the wet vapor of carbon tetrachloride. I n all the brasses the zinc was attacked more rapidly than the copper, so that the test pieces soon became covered with a red layer of copper-rich alloy. I n brasses rich in copper the action was conlined to a thin film at the surface of the metal, but with the 60-40 brass a layer of relatively pure copper about 1.6mm. inch) thick was formed during the 100 hours' exposure. This outer layer was of almost uniform thickness and adhered firmly to the inner core of unattacked brass. Upon analysis it was found to contain 92 per cent of copper, although the brass from wliich it was derived contained hut 60 per cent. Metallographic examination shopied that it was not simply a spongy mass of copper left after dissolving the zinc out of the original crystals of the brass, but that it consisted of a dense mass of fine columnar crystals of copper with their axes normal to the original surface of the test piece. The line of demarcation between the brass and the copper did not follow the boundaries of the original crystals of the brass hut out sharply through these crystals. It appears that in the formation of this outer layer the copper was actually dissolved and redeposited. Figure 4 is a photomicrograph taken across the boundary face between brass and copper. Figure 3-Appearance of Teat Pleceli aftor Exposure to W e t and Dry Vapor8 :tluminium is rapidly attacked and disintegrated with the forination of a dark gray powder. This powder has a very peiietrating, resinous odor and "hisses" wlien thrown into \$-titer. Upoir examinatiou it was found to consist of a mixture of aluminium c h l o d e and Irexa[~lilorootfrane,together with small amounts of some black substance and some resinous comprund. Additional quantities of hexaohloroethane Nere recovered from the refluxed liquid in the dis- Benzol Poisoning Progress reports on this subject issued by the Sub-committee on Bene01 of the Committee on Industrial Poisons, National Safety Council, Chemical Section, emphasize the importance o i this work. Investigations disclosed the fact that fiiteen deaths and eighty-three more or less serious illnesses have occurred as a result of industrial benzol poisoning duriiig the last iew years. It was reported that out of a group of eighty-four workers exposed more or less continuously to benzol fumes in varying concentration, thirteen of these, or 15 per cent, showed a white blood cell count so low as to he strikingly suggestive of chronic benzol poisoning.