wnv The University of /Washington Department of Chemistry of A PRESERVATION OF IRON AND STEEL BY MEANS OF PASSIVIFYING FACTORS BY THOMAS GORDON THOMPSON A Thesis Submitted in Partial Fulfilment of the Requirement for the Degree of Doctor of Philosophy SEATTLE, WASHINGTON 1920 The University of Washington Department of Chemistry BY THOMAS GORDON THOMPSON A Thesis Submitted in Partial Fulfilment of the Requirement for the Degree of Doctor of Philosophy SEATTLE, WASHINGTON H. C. PIGOTT PRINTING CONCERN 1920 PRESERVATION OF IRON AND STEEL BY MEANS OF PASSIVIFYING FACTORS. PASSIVITY. When iron is immersed in solutions of chromic acid or potassium dichro- mate 1 for a few hours, treated with concentrated nitric acid or made the anode of an electric circuit in certain electrolytes, it becomes very inert. This phenomenon is known as passivity. Thus, for example, if a piece of iron or steel be placed in a solution of copper sulphate for a brief period, metallic copper will deposit upon the surface. If, however, the iron has been "passivified" by immersion in potassium dichromate or concentrated nitric acid, this reaction will not take place until the passivity has been destroyed. A further test for passivity is found in the fact 2 that the iron will not dissolve in nitric acid having a specific gravity of 1.20. Should iron be placed in concentrated nitric acid, the metal will begin t'o go into solu- tion rapidly and then suddenly all chemical action will cease. Upon applying the above tests it will be found that the iron is in the passive state. If iron be made the positive electrode and placed in certain electrolytes and the amount of current flowing is sufficient, it will be noted that' the iron will not go into solution according to Faraday's law. The iron will be pbssivified, the rate of solution reduced to a minimum, and bubbles of oxygen will arise from the anode. Many explanations of this peculiar phenomenon have been offered, and the subject' has claimed the attention of a vast number of investigators. No really satisfactory explanation, generally acceptable and capable of explaining the passive condition of metals under all circumstances, has, however, ever been advanced. Many investigators are supporters of the so-called oxide and gaseous film theories. A complete bibliography of the literature relating to passivity up to 1907 is given by Heathcote 3 and by Byers. 4 A symposium, "The Passivity of Metals," published by the Faraday Society in 1913, gives outlines of the more recent work upon this subject. A very clear and comprehensive chapter, ''The Passive State of Iron," is given by J. Newton Friend in his interesting book, "The Corrosion of Iron and Steel." That' other metals besides iron may be rendered passive has been pointed out from time to time by various investigators. Thus Dustan and Hill 5 describe means for producing the passive state of zinc, together with a number of other metals. The effect of potassium dichromate upon zinc, when used as a galvanising agent, will be discussed in the experimental portion of this Report. 1 Dunstan and Hill (Jour. Chem. Soc., vol. 99, pp. 1833-53) have demonstrated that many other salts besides the dichromates may be utilised as passivifying agents. 2 Heathcote, Jour. Soc. Chem. Ind., 1907, vol. 26, p. 899. 3 Heathcote, Jour. Soc. Chem. Ind., 1907, vol. 26, p. 899. 4 Jour. Am. Chem. Soc., 1908, vol. 30, p. 1718. 3 Jour. Chem. Soc., vol. 99, pp. 1853-67. 425704 6 PASSIVITY AND ITS RELATION TO THE CORROSION OF IRON AND STEEL. Passivifying Agents as Pigments. Although the phenomenon of passivity has been known since 1794, no attempt was made until quite recently to make use of this peculiar fact as a means for preventing or inhibiting the corrosion of iron and steel. In 1895 M. P. Wood ' called attention to the valuable properties possessed by insoluble and basic chromates when used as protective coatings for iron and steel. The use of such materials, when properly applied, has been confirmed by all investigators in this particular field. The work of Gush- man has done much to demonstrate the value of chrome pigments. Soluble Passivifying Agents in Steam Boilers. The corrosion of steam boilers takes place under conditions very dif- ferent from those encountered in the ordinary corrosion of iron and steel. The gases dissolved in the boiler water at atmospheric temperatures are rapidly removed, to a very large extent, upon heating, thereby decreasing the relative corrosivity of the water. However, other factors are present that will stimulate corrosion within the boiler. Briefly, these are the lack of homogeneity of the boiler plates, contact of the plates with dissimilar metals, the effect of strains, and the influence of dissolved salts. The nature of these dissolved salts determine to a great extent the corrosive action that goes on within the boiler. From the work of Grave ' and Moseley 3 it might be concluded that the chemical composition of the boiler plates, their lack of homogeneity, and the effect of strains would have very little to do with the ease with which the plates could be passivified. The use of soluble chromates, such as potassium dichromate, as a means for preventing corrosion in boilers, was first recommended by Cushman. 4 He suggested the use of 5 Ibs. of potassium dichromate to 1500 gallons of water. When this was tried on a commercial basis, it was found that the potassium dichromate had practically no influence in limiting corrosion. Archbutt 5 found that when potassium dichromate was added to the con- centrated boiler water, corrosion was not prevented, "especially in locomotive boilers where metallic copper is in contact with the boiler." Under such conditions it would be impossible to maintain the passive state. The exact nature of the dissolved salts was not stated. Friend and Brown 6 have studied the effect' of potassium dichromate and potassium chromate in varying concentrations in a number of solutions of sodium chloride ranging from fairly dilute to saturated. Tests were run for a few days 1 Trans. Am. Soc. Mech. Eng-., vol. 16, p. 671. 2 Z. physik Chem., vol. 77, pp. 513-76. 3 Jour. Am. Chem. Soc., 1915, vol. 37, pp. 2326-34. 4 United States Department of Agriculture, Office of Public Roads, Bulletin No. 30. 5 Engineering, vol. 93, p. 824. 6 Journal of the Iron and Steel Institute, 1911, No. I, p. 125. at room temperature and at 95 C. With the potassium dichromate in sodium chloride corrosion took place readily both in the cool and hot solu- tions; and it was found that corrosion increased with the increase in concentration of the dichromate. However, with potassium chromate, increase in concentration of this salt in the presence of the sodium chloride caused decrease in the rate of corrosion. The point at which corrosion was prac- tically eliminated corresponded with the presence of an amount of dichromate which would have cost so much on a commercial basis as to make its application prohibitive. The tests of Friend and Brown at 95 were con- ducted for just three days, and during that time the solutions were heated only twelve hours a day, the solutions being allowed to cool to room temperature over night. While this treatment stimulated corrosive action, the results do not represent the true action of sodium chloride and potassium dichromate at 95 C. As the great majority of boiler waters are practically free from chlorides, it seems probable that corrosion would be inhibited in waters containing sulphates alone if the water were treated with proper amounts of potassium dichromate. No quantitative data has yet been secured with mixed sol itions of varying strengths of sulphates and potassium dichromate. Cushman 1 has shown qualitatively that with the increase of copper sulphate in solutions containing uniformly small amounts of potassium dichromate, the rate at which the copper was deposited was increased. t Anodic Passivity. The possibility of making a boiler the anode in an electric circuit was suggested by Byers and Voris. 2 They were able to keep iron passive when it was used as an anode and placed in an electrolyte consisting of sodium chloride and potassium dichromate, there being forty times as much potas- sium dichromate as sodium chloride. The electrolyte was heated so that a pressure of from 20 to 50 Ibs. per square inch was obtained. As there are so many factors that would tend to destroy this passivity in an ordinary steam boiler, it therefore seems very doubtful whether this form of passivity would be of any real practical value. With the increase in concentration of the boiler water, the increase in the amount of the potassium dichromate would have to be forty times that of the sodium chloride, and this very fact would render the cost of operation impossible. No work has yet been done showing the effect of the dichromate with the various sulphates commonly found in boiler waters, when the iron or boiler is made the anode in an electric circuit. PURPOSES OF THIS RESEARCH. Sjnce the corrosion problem is so vast in extent, the author has limited himself to studying the means for preventing corrosion in steam boilers and receptacles for containing water for purely industrial purposes. The object 1 Cushman and Gardner, "Corrosion and Preservation of Iron and Steel," p. 113. 2 Jour. Am. Chem. Soc., 1912, vol. 34, pp. 1368-79. of this research is to study the effect of various passivifying agents upon the rate of corrosion of different grades of iron in solutions of salts com- monly occurring in boiler waters. The experimental work may be divided into four parts : First : To ascertain the effect of varying amounts of potassium di- chromate in solutions of different concentrations of sodium chloride, sodium sulphate, calcium sulphate, magnesium chloride, and magnesium sulphate at room temperature and at the temperature of boiling water. Since the publication of the suggestion of the use of potassium di- chromate as a means for preventing boiler corrosion, many text-books have been published which mention the use of potassium dichromate as a preventive for corrosion. It is the aim of the author to present sufficient data to establish definitely the truth or fallacy of this statement. Second : To determine the efficiency of di-sodium phosphate as an inhibitive agent, the samples of iron and steel being placed in the above- mentioned solutions. Third : To secure data regarding the action of zinc or galvanised iron in various solutions of sodium chloride and sodium sulphate containing varying amounts of potassium dichromate. Fourth : To determine the practicability of keeping iron passive with potassium dichromate in the presence of the salts characteristic of boiler waters, when the iron is made the anode in an electric circuit and heated to a pressure varying from 120 Ibs. to 150 Ibs. per square inch. EXPERIMENTAL. The Nature and Preparation of the Samples of Iron and Steel. The test-pieces used in all the following experiments consisted of a commercially pure iron and a low carbon steel, both of which had been rolled into sheets. The author is indebted to the American Rolling-Mill Company of Middletown, Ohio, who kindly supplied him with samples of their manufacture. The following shows the composition of the test-pieces: Carbon Commercially Pure Iron. Per Cent. 0.012 Low Carbon Steel. Per Cent. 0.112 Phosphorus 0.004 0.045 Silicon 0.001 0.003 Sulphur 0.017 0.052 Maneanese 0.015 0.388 For the corrosion tests, plates having a surface of 42 square centimetres were cut -from the sheets, the plates being 8 cm. X 2 l / 2 cm. After cutting, a small hole was punched at the top of each plate so that it could be attached to a thread and suspended in solution. The plates were then numbered and pickled in a solution of dilute sulphuric acid until all foreign matter was removed by washing rapidly and thoroughly in water and then treating with alcohol, ether, and finally with gasoline. After the latter had evaporated from the surfaces of the plates by holding them over a source of indirect heat, they were placed in desiccators containing calcium chloride and eventually weighed. Preparation of Solutions. Normal solutions of sodium chloride, sodium sulphate, magnesium chloride, and magnesium sulphate, and a saturated solution of calcium sulphate at 20 C. were made. From the first' four solutions, N/10, X/100, and X/1000 solutions were prepared. A solution of potassium dichromate containing 100 grammes of the salt per litre and a normal solution of di-sodium phosphate were also made. Small amounts of these solutions were measured into the bottles in which the individual corrosion tests were to be run, so that when the other solutions were introduced they would also contain either a thousandth or a hundredth part of an equivalent weight of potassium dichromate or sodium phosphate. Amounts of these two latter substances were weighed out and placed in the various solutions so that they gave strengths of either N/10 or N/5. In mentioning the normality of a potassium dichromate solution no reference is meant to its oxidizing power. A normal solution of the dichromate is considered here as containing half the gramme molecular weight. Apparatus. In running the corrosion tests at room temperature, one plate was placed in each bottle. The bottles were of 300 cubic centimetres capacity, and were arranged in such a manner as to permit' the passage of a slow stream of air through the solution. Five of these bottles were placed in series without the slight increase in pressure in the first few bottles having any effect upon the rate of corrosion. The outlet' tube was so far above the surface of the solution that any danger of particles of the liquid being carried over into the next bottle was practically eliminated. The apparatus used for determining the rate of corrosion at the tem- perature of boiling water consisted of three large tin-plated tubs, in which the bottles containing the individual test-plates were placed. These tubs were heated by means of gas-burners, the water being supplied by means of a constant leveling device shown. Two of these devices were utilized for each tub, so that an accident to one would prevent the water from boiling away during the night. The bottles containing the plates were stoppered with wooden plugs, which had previously been steamed and then wrapped with tinfoil. These plugs were loosely fitted in order that the volume of gases could regulate itself without difficulty upon heating. A wooden rack supported the bottles within the tub. A cover, with a few small holes for the escape of steam, fitted securely over the top of the apparatus. 10 Conditions Under Which the Tests were Run. In order to promote corrosion, should it occur, as rapidly as possible at room temperature, air was slowly bubbled through the solutions. The air was thoroughly washed in a series of wash bottles and permitted to pass through the solutions ten hours every day. This produced two important factors, namely, keeping the solutions saturated with oxygen and insuring the homogeneity of the liquids by constant stirring. All the tests were performed in a laboratory free from direct' sunlight. The bottles used to contain the solutions and test-plates for treatment at 100 C. were subjected to the action of boiling water for twenty hours a day for two weeks, before they were utilised for running the corrosion tests. During this time they were removed each day from the tub, emptied, cooled, and washed, then refilled and returned to the tub. The test-plates were not suspended in the solutions but were laid diagonally across the lower end of the bottle, with just tire four corners touching the sides. Cleaning the Test-Plates. The rust which had collected on the test-plates run at room temperature was removed by treatment with either ammonium citrate or dilute sulphuric acid. Ammonium citrate, after treatment for a little time, removed the rust from the plates immersed in solutions containing no potassium di- chromate or sodium phosphate, or very small amounts of these substances. Sulphuric acid had to be resorted to in order to remove the rust formed on some of the plates which had been in solutions containing N/10 and N/5 dichromate. In cases of this nature the rust was always in the form of hard minute spots. With the solutions containing the larger amounts of sodium phosphate, the corrosive material was of a peculiar nature, it being a dull green colour which always occurred in the form of streaks or large spots on the metal. These streaks and spots could be easily removed by gently rubbing with the fingers. After cleaning and washing the plates were dried in the manner previously described. With the tests run at the temperature of boiling water it was impossible to clean the plates. Long immersions in ammonium citrate for two and three days had no desirable effect. On the other hand, if the plates were treated with sulphuric acid, part of the very thin scaly material covering the plate would be removed while the remainder would still be intact, and before this could be removed portions of the exposed metal would begin to dissolve. After a number of various attempts to clean these plates, it was finally decided, after washing with water and alcohol and drying, to weigh them directly. In the great majority of cases, no test for iron was obtained in the solutions, and thus weighing, without cleaning, probably gave fairly accurate results. The spots occurring on the plates after immersion in the hot solutions containing potassium dichromate were invariably of black or brownish-black colour. Those in solutions containing larger amounts of sodium phosphate were of the characteristic dull green colour. 11 ACTION OP SOLUTIONS OF CHLORIDES AND SULPHATES UPON IRON AND STEEL AT ROOM TEMPERATURE IN THE PRESENCE OF POTASSIUM DICHROMATE. The following results show the effect of various amounts of potassium dichromate upon the rate of corrosion of iron and steel when introduced into solutions of different concentrations of chlorides and sulphates. In some cases the addition of dichromate prevented corrosion completely, especially with the more concentrated solutions of this substance, while in others its inhibiting character was very striking. When placed in solutions free from dichromate, the low carbon steel showed a slightly greater tendency toward corrosion than did the commer- cially pure iron. Test-plates of the low carbon steel and the commercially pure iron were placed in distilled water and subjected to the same treatment as accorded to the samples in the various solutions. The results thus obtained were utilized as a standard in computing the relative corrosion. This was deter- mined by taking the loss in weight of the samples in the distilled water and dividing into the loss in weight by the test-plates in the solutions*. Loss in weight in X solution. Relative corrosion = Loss in weight in distilled water. COMMERCIALLY PURE IRON. ROOM TEMPERATURE. Table I. Action of Water. No. A B C D Original Weight. 18.4197 19.6435 16.6748 19.2308 Loss in Weight. 0.1995 0.2021 0.1969 0.1956 Remarks. Badly corroded. Table II. Potassium Dichromate. Original Loss in No. Normality. Weight. Weight. 81 1/1000 19.8732 0.0010 82 1/100 18.8425 0.0004* 83 1/10 18.9613 0.0000 84 1/5 18.0086 0.0001 Relative Cor'sion. Remarks. Slight corrosion on edges. No corrosion. Metal bright. No. A B C D LOW CARBON STEEL. ROOM TEMPERATURE. Table III. Action of Water. Original Weight. 7.3013 7.8240 8.0327 7.9241 Loss in Weight. 0.2232 0.2113 0.2135 0.2203 Remarks. Badly corroded. * Gain in weight. 12 Table IV. Potassium Bichromate. No. Normality. Original Loss in Relative Weight Weight. Cor'sion. Remarks. 329 330 331 332 1/1000 1/100 1/10 1/5 8.4997 0.0005 Slight corrosion. 8.1320 0.0001 No corrosion. 8.1652 0.0003 Metal bright. 8.0663 0.0004 COMMERCIALLY PURE IRON. ROOM TEMPERATURE Table V. Sodium Phosphate. No. Normality. Original Loss in Relative Weight. Weight. Cor'sion. Remarks. 1 2 3 4 1/1000 1/100 1/10 N 18.2439 0.2298 115 Covered with rust. 19.8746 0.2203 110 20.1324 0.1897 93 20.4938 0.1204 59 t *-':. Table VI. Sodium Chloride. No. Normality. Original Loss in Relative Weight. Weight. Cor'sion. Remarks. 5 6 8 8 1/1000 1/100 1/10 N 20.4132 0.2133 105 Covered with rust. 19.5453 0.2175 107 19.7463 0.2556 128 19.7077 0.1561 78 Table VII. Magnesium Sulphate. No. Normality. Original Loss in Relative Weight. Weight. Cor'sion. Remarks. 9 10 11 12 1/1000 1/100 1/10 N 19.0452 0.2028 103 Covered with rust. 18.7993 0.2468 122 20.4995 0.1686 83 19.5231 0.1214 55 Table VIII. Magnesium Chloride. No. Normality. Original Loss in Relative Weight. Weight. Cor'sion. Remarks. 13 14 15 16 1/1000 1/100 1/10 N 19.8018 0.2283 113 Covered with rust. 20.3402 0.1828 90 20.2853 0.1258 62 19.2753 0.1176 58 LOW CARBON STEEL. ROOM TEMPERATURE. Table IX. Sodium Sulphate. No. Normality. Original Loss in Relative Weight. Weight. Cor'sion. Remarks. 249 250 251 252 1/1000 1/100 1/10 N 8.0917 0.2597 118 Covered with rust. 7.8634 0.2267 104 7.4396 0.1648 75 7.5003 0.1124 52 13 Table X. Sodium Chloride. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks, 253 1/1000 7.9239 0.2285 105 Covered with rust. 254 1/100 7.9880 0.2332 108 ** 255 1/10 7.7302 0.2505 115 " 256 N 7.4951 0.1786 82 Table XI. Magnesium Sulphate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 257 1/1000 7.0276 0.2640 122 Covered with rust. 258 1/100 7.8348 0.2832 130 " 259 1/10 8.0051 0.1562 72 " 260 N 8.1704 0.1137 52 Table XII. Magnesium Chloride. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 261 1/1000 8.1027 0.2529 116 Covered with rust. 262 1/100 7.6020 0.2670 122 " 263 1/10 7.9926 0.2218 102 " 264 N 7.4304 0.1893 87 COMMERCIALLY PURE IRON. ROOM TEMPERATURE. Table XIII. Sodium Sulphate with N/1000 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 17 1/1000 17.7931 0.0190 10 Corrosion around edges. 18 1/100 17.4386 0.0165 8 " " " 19 1/10 18.8439 0.0077 4 Few small rust spots. 20 N 17.5677 0.0132 7 " large " Table XIV. Sodium Sulphate with N/100 Potassium Dichromat't. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 21 1/1000 18.8235 0.0000 No rust. 22 1/100 17.9828 0.0005 " " 23 1/10 17.8812 0.0021 1 Few minute rust spots. 24 N 18.5079 0.0046 2 Rust around edges. Table XV. Sodium Sulphate with N/10 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 25 1/1000 17.7866 0.0001 * Metal bright. No rust. 26 1/100 18.6974 0.0005 ' " " " " 27 1/10 18.5133 0.0006 1 Several black spots. 28 N 18.1652 0.0043 2 " " " Gain in weight. 14 Table XVI. Sodium Sulphate with N/5 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 29 1/1000 18.6887 0.0001 1 Metal bright. 30 1/100 18.9888 0.0002 1 " " 31 1/10 18.4944 0.0003 Several minuti 32 N 18.5613 0.0045 2 Many minute r No rust. LOW CARBON STEEL. ROOM TEMPERATURE. . Table XVII. Sodium Sulphate with N/1000 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 265 1/1000 8.0241 0.0112 5 Few rust spots. 266 1/100 7.0198 0.0225 10 " " " 267 1/10 8.4727 0.0090 4 " 268 N 7.9607 0.0092 4 " " " Table XVIII. Sodium Sulphate with N/100 Potassium Bichromate. Original Loss in Relative No. Normality. Weight Weight. Cor'sion. Remarks. 269 1/1000 6.7617 0.0001 No corrosion. 270 1/100 6.7436 0.0033 2 Few rust spots 271 1/10 7.7283 0.0032 2 " " " 272 N 7.9914 0.0045 2 " " " Table XIX. Sodium Sulphate with N/10 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. 272 1/1000 6.6671 0.0001 274 1/100 6.2877 0.0000 275 1/10 6.7019 0.0067 3 276 N 8.4086 0.0049 2 No Remarks, corrosion. Few rust spots. Table XX. Sodium Sulphate with N/5 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. 277 1/1000 6.9826 0.0003 278 1/100 6.9373 0.0001 279 1/10 6.9747 0.0010 280 N 6.7125 0.0027 Remarks. Cor'sion. No corrosion. " 1 Few rust spots. COMMERCIALLY PURE IRON. ROOM TEMPERATURE. Table XXI. Magnesium Sulphate with N/1000 Potassium Bichromate. Original Loss in Relative on. Remarks. Slight rusting on edge Many small rust spots. 1 Gain in weight. No. Normality. Weight. Weight. Cor's 33 1/1000 18.6374 0.0011 1 34 1/100 17.7247 0.0031 2 35 1/10 18.5333 0.0080 4 36 N 18.8740 0.0119 6 15 Table XXII. Magnesium Sulphate with N/100 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 37 1/1000 18.8471 0.0002 ' No rust. 38 1/100 18.7850 0.0022 1 Several black spots. 49 1/10 18.2635 0.0107 5 Many small spots. 40 N 18.1070 0.0264 13 Covered with spots. Table XXIII. Magnesium Sulphate with N/10 Potassium Bichromate Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 41 1/1000 18.6751 0.0000 Metal bright. 42 1/100 18.7850 0.0003 l " " 43 1/10 18.7790 0.0008 1 Minute rust spots. 44 N 19.8571 0.0280 14 Covered with spots. Table XXIV. Magnesium Sulphate with N/5 Potassium Bichromate. No. Normality. Weight. Weight. Cor'sion. Remarks. Original Loss in Relative 45 1/1000 19.1785 0.0000 Metal bright. 46 1/100 19.5526 0.0003 1 " " 47 1/10 19.9281 0.0010 1 Few minute rust spots. 48 N 18.2517 0.0238 12 Covered with rust spots. LOW CARBON STEEL. ROOM TEMPERATURE. Table XXV. Magnesium Sulphate with N/1000 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. 281 1/1000 6.9384 0.0038 2 282 1/100 7.0442 0.0127 6 283 1/10 7.6038 0.0073 4 284 N 7.6373 0.0124 6 Remarks. Few spots over surface. Many small spots. 'Few small rust spots. Many small spots. Table XXVI. Magnesium Sulphate with N/100 Potassium Bichromate. Original Loss in Relative Cor'sion. Remarks. No corrosion. 3 Few rust spots. No. Normality. Weight. Weight. 285 1/1000 7.7890 0.0000 286 1/100 7.7853 0.0050 287 1/10 7.9115 0.0094 288 N 7.6104 0.0218 11 Many rust spots. Table XXVII. Magnesium Sulphate with N/10 Potassium Bichromate. Original Loss in Relative FO. Normality. Weight. Weight. Cor'sion. Remarks. 289 1/1000 7.5233 0.0002 No corrosion. 290 - 1/100 7.8860 0.0001 " " 291 1/10 8.1318 0.0004 " " 292 N 7.6104 0.0210 11 Covered with si 1 Gain in weight. 16 Table XXVIII. Magnesium Sulphate with N/5 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 293 1/1000 7.4765 0.0002 Metal bright. 294 1/100 7.2624 0.0002 295 1/10 8.1285 0.0007 No corrosion. 296 N 8.3027 0.0241 12 Covered with rusty spots. COMMERCIALLY PURE IRON. ROOM TEMPERATURE. Table XXIX. Saturated Solution of Calcium Sulphate (20 C.), Original Loss in Relative No. Normality. 1 Weight. Weight. Cor'sion. 85 1/1000 18.3412 0.0098 5 86 1/100 18.0028 0.0007 - 87 1/10 18.7949 0.0003 - 88 1/5 18.7370 0.0002 - Remarks. Several large streaks. Black spots. No corrosion. LOW CARBON STEEL. ROOM TEMPERATURE. Table XXX. Saturated Solution of Calcium Sulphate (20 C.). Original Loss in Relative No. Normality. 1 Weight. Weight. Cor'sion. Remarks. 365 1/1000 8.0091 0.0088 4 Minute spots. 366 1/100 7.7364 0.0001 No corrosion. 367 1/10 7.1023 0.0004 " " 368 1/5 7.4972 0.0005 " " COMMERCIALLY PURE IRON. ROOM TEMPERATURE. Table XXXI. Sodium Chloride with N/1000 Potassium Bichromate. No. 49 50 51 52 Table No. 53 54 55 56 Original Loss in Relative Normality. Weight. Weight. Cor'sion. 1/1000 18.2921 0.0452 23 1/100 18.3231 0.0100 5 1/10 18.5411 0.0122 6 N 18.2830 0.0173 9 XXXII. Sodium Chloride with N/100 Original Loss in Relative Normality. Weight. Weight. Cor'sion. 1/1000 17.3569 0.0004 l 1/100 17.7729 0.0021 . 1 1/10 18.8033 0.0183 9 N 18.1411 0.0329 17 Remarks. Covered with rust. Narrow long black streaks. Covered with thin layer of rust. Potassium Bichromate. Remarks. No corrosion. Several rust spots. 'Covered with thin layer of rust Covered with rust spots. 1 Normality of potassium dichromate. 2 Gain in weight. 17 Table XXXIII. Sodium Chloride with X/10 Potassium Bichromate. No. 57 58 Normality. 1/1000 1/100 Original Weight. 17.9824 18.9423 Loss in Weight 0.0004 0.0003 Relative . Cor'sion. 1 1 Remarks. No corrosion. 59 1/10 18.8458 1.0307 15 Covered with layer of rust. 60 N 18.7504 0.0633 32 Table XXXIV. Sodium Chloride with N/5 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight . Cor'sion. Remarks. 61 1/1000 17.5235 0.0001 1 Metal bright. 62 1/100 18.4591 0.0004 1 " " 63 1/10 17.9659 0.0423 21 Many small rust spots. 64 N 18.2212 0.0530 27 Badly corroded. LOW CARBON STEEL. ROOM TEMPERATURE. Table XXXV. Sodium Chloride with N/1000 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 297 1/1000 8.0472 0.0374 19 Rust spots in streaks. 298 1/100 8.2619 0.0185 9 Many rust spots. 299 1/10 7.7362 0.0156 8 Large rust spots. 300 N 8.5746 0.0190 10 Table XXXVI. Sodium Chloride with N/100 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 301 1/1000 8.6156 0.0001 No corrosion. 302 1/100 7.5995 0.0002 Several minute spots. 303 1/10 8.5122 0.0174 9 Many small rust spots. 304 N 8.1990 0.0375 19 Covered with large rust spots. Table XXXVII. Sodium Chloride with N/10 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 305 1/1000 8.4199 0.0007 No corrosion. 306 1/100 8.5453 0.0016 1 Few minute rust spots. 307 1/10 7.9314 0.0336 17 Covered with small spots. 308 N 8.7001 0.0652 33 Table XXXVIII.- Sodium Chloride with N/5 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 309 1/1000 8.0314 0.0005 No corrosion. 310 1/100 6.6100 0.0035 2 Few minute spots. 311 1/10 7.5867 0.0409 20 Many large rust spots. 312 N 8.1445 0.0515 26 1 Gain in weight. 18 COMMERCIALLY PURE IRON. ROOM TEMPERATURE. Table XXXIX. Magnesium Chloride with N/1000 Potassium Bichromate. Remarks. Corrosion on edges. Few minute rust spots. Covered with minute rust spots. Many small rust spots. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. 65 1/1000 17.5235 0.0018 1 66 1/100 18.4563 0.0024 1 67 1/10 18.6983 0.0164 8 68 N 18.8969 0.0128 Table XL. Magnesium Chloride with N/100 Potassium Bichromate. Original Loss in No. Normality. Weight. Weight. 69 1/1000 17.0475 0.0002 70 1/100 17.3322 0.0009 ' 71 1/10 17.3590 0.0151 72 N 17.8621 0.0246 Relative Cor'sion. Remarks. Several minute spots. No corrosion. 8 Covered with thin layer ru.st. 12 " Table XLI. Magnesium Chloride with N/10 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. 'Cor'sion. Remarks. 73 1/1000 19.5877 0.0004 * Metal bright. 74 1/100 18.1777 0.0003 1 " " 75 1/10 18.0645 0.0209 10 Covered with rust spots. 76 X 18.1682 0.0528 26 small lumpy Table XLII. Magnesium Chloride with N/5 Potassium Bichromate. No. Normality. 77 1/1000 78 1/100 19 1/10 80 N Original Loss in Relative Weight. Weight. Cor'sion. Remarks. 18.9682 0.0004 l Metal bright. 17.4384 0.0002 1 18.3128 0.0212 11 Covered with small spots. 17.9440 0.0540 27 lumpy LOW CARBON STEEL. ROOM TEMPERATURE. Table XLIII. Magnesium Chloride with N/1000 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 313 1/1000 8.7930 0.0020 1 Few rust spots. 314 1/100 8.4156 0.0038 2 " " " 315 1/10 8.5263 0.0143 7 Rust around edges. 316 N 7.4669 0.0160 8 Many minute spots. 1 Gain in weight. 19 Table XLIV. Magnesium Chloride with N/100 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight . 'Cor'sion. Remarks. 317 1/1000 8.7930 0.0020 1 Minute spots. 318 1/100 8.3371 0.0057 3 Several large black spots 319 1/10 7.7715 0.1042 7 Covered with minute spots. 320 N 8.2331 0.0550 27 " *' Table XLV. Magnesium Chloride with N/10 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight . Cor'sion. Remarks. 321 1/1000 7.7367 0.0004 No corrosion. 322 1/100 8.0481 0.0004 " " 323 1/10 8.0177 0.0193 10 Covered with spots. 324 N 8.5306 0.0632 31 Badly corroded. Table XLVI. Magnesium Chloride with N/5 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight . Cor'sion. Remarks. 325 1/1000 7.9113 0.0003 No corrosion. 326 1/100 8.1286 0.0001 " " 327 1/10 7.9591 0.0259 13 Covered with rust spots. 328 N 8.3227 0.0756 38 Badly corroded. SUMMARY OF RESULTS. ROOM TEMPERATURE. Solutions of Sodium Sulphate and Potassium Bichromate. 1. The differences in chemical composition and physical properties had no influence upon the rate of corrosion. 2. Only the solutions which contained very dilute amounts of potassium dichromate showed any marked corrosion. 3. N/10 arid N/5 potassium dichromate rendered the iron and steel passive in the more dilute solutions of the sodium sulphate. 1 4. In all other cases the potassium dichromate decidedly inhibited corrosion. Solutions of Magnesium Sulphate and Potassium Bichromate. 1. The differences in chemical composition and physical properties of the iron and steel had no influence upon the rate of corrosion. 2. Tenth normal and normal solutions of magnesium sulphate con- taining potassium dichromate in all proportions used, were covered with rustf spots of varying sizes. While this was generally true in the case of sodium sulphate solutions, the rate of corrosion with the magnesium sulphate solutions was much greater. 1 The passivification of the test-plates was determined by dipping them into nitric acid of 1.20 specific gravity. 20 3. N/10 and N/5 potassium dichrornate passivified the test-plates of both iron and steel in N/100 magnesium sulphate. This was also true with N/5 potassium dichromate in N/100 magnesium sulphate. With the latter solution containing N/10 potassium dichromate, the passive state was dem- onstrated twice with four test-plates of iron and twice with three test-plates of the steel. To determine the passive state of the iron and steel, other plates than those used to secure the loss in weight were run under identical conditions. At the end of two weeks they were washed carefully, and then immersed in the nitric acid of 1.20 sp. gr. 4. In all cases potassium dichromate inhibited corrosion to a marked extent. Saturated Solution of Calcium Sulphate (20 C.) and Potassium Dichromate. 1. With the exception of N/1000 potassium dichromate, the addition of this salt to the saturated solution of calcium sulphate produced results similar to those obtained in solutions containing only the dichromate. Solutions of Sodium Chloride and Potassium Dichromate. 1. There was no difference in the behavior of the iron and steel. 2. The most peculiar action was obtained with N/1000 solution of both the sodium chloride and potassium dichromate. With this solution the rate of corrosion was many times higher than that obtained with magnesium chloride or with the sulphate solutions. With the iron, four tests gave a loss in weight of 0.0452, 0.0397, 0.0190, and 0.0432 gram. 3. The passive state was demonstrated in many cases with N/1000 and N/100 sodium chloride containing N/10 and N/5 potassium dichromate. 4. The inhibiting action of the dichromate is less in the sodium chloride solutions than in the corresponding sulphate solutions. Solutions of Magnesium Chloride and Potassium Dichromate. 1. The behavior of the iron and steel appears to be the same in many cases. 2. Test-plates were passivified in N/1000 and N/100 solutions of the chloride with N/10 and N/5 potassium dichromate. 3. In the more concentrated solutions of the chloride, an increase in either salt produced an increase in corrosion. 4. The inhibiting action of the potassium dichromate is less than in the corresponding sulphate solutions. 5. The action of many of the magnesium chloride solutions were sim- ilar to those of sodium chloride. Normal solutions of the former, however, were exceptions, corrosion being much greater. 21 ACTION OF SOLUTIONS OF CHLORIDES AND SULPHATES UPON IRON AND STEEL AT THE TEMPERATURE OF BOILING WATER IN THE PRESENCE OF POTASSIUM DICHROMATE. In the tests run at room temperature potassium dichromate reduced corrosion to a marked extent in every case, in some instances preventing all chemical action upon the iron and steel. Thus tanks used for storing non-potable waters at atmospheric temperatures could be well preserved by the addition of small quantities of potassium dichromate. However, with waters in boilers the addition of this* compound will, in many cases, serve as a stimulative agent rather than as a preventive for corrosion. This will be seen by studying the data obtained when the iron and steel were placed in solutions free from dichromate at boiling temperature for two weeks and comparing them with the results from those boiled in solutions in which this salt' was present. In many of the solutions the presence of the dichromate was occasioned by a slight increase in corrosion. This was especially true in the chloride solutions. In fact, from the general appearances of the test- plates, this increase in corrosion seemed greater than that actually repre- sented by the following figures. Plates of the iron and steel were introduced into distilled water and subjected to the same treatment as accorded to the samples in the various ; solutions. At the end of two weeks the plates were removed, dried, and weighed. The plates were all slightly tarnished. The gain in weight of the plates served as a standard for calculating the relative corrosion. The test-plates immersed in the salt solutions free from dichromate corroded to a certain extent, a small brown precipitate settling on the bottom of the containing vessel. It was thus impossible to weigh the plates directly and determine the rate of corrosion by gain in weight. The plates, however, were very easily cleaned with ammonium citrate and the loss in weight determined. The equivalent amount of ferric oxide represented by the loss in weight was calculated. While this does not accurately represent the actual amount of corrosive matter produced it is fairly approximate, so that it can be reasonably compared with the data obtained from plates where cleaning was impossible. N/5 dichromate produced a decrease in the rate of corrosion in nearly every case. In some instances, especially with solutions of sodium sulphate containing N/10 dichromate, there is a decided increase in corrosion, while in others the dichromate appears to have little or no effect. In general the addition of dichromate to boiler waters would be of no practical use, as an alteration in the concentrations of the salts would produce a difference in action. This would be especially true if chlorides were present. Under such circumstances the effect of the sulphates would be acted upon by the chlorides and the final result would be a stimulation of corrosion. The presence of dichromate in solutions of the chlorides caused a decided increase in the rate of corrosion, the magnesium chloride being more violent in its behavior than the sodium chloride. Chemical interaction 22 undoubtedly takes place between the chlorides, potassium dichromate, and water, with the production of small amounts of hydrochloric acid, thus causing the great increase in corrosive action. Friend ' has suggested the following chemical equation to show the production of the hydrochloric acid : 2NaCl+K 2 Cr 2 O+ H*0 h K.Cr 0*+2 Because of experimental difficulties the results reported for the action of the stronger solutions of the chlorides with N/10 and N/5 dichromate upon the iron and steel are low : COMMERCIALLY PURE IRON. TEMPERATURE OF BOILING WATER. Table XLVIL Action of Water. No. A B C D E Original Weight. 19.1064 18.8200 18.8349 19.1126 18.7442 Gain in Weight. 0.0055 0.0048 0.0043 0.0063 0.0057 Remarks. Tarnished. Table XL VIII. Potassium Dichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 189 1/1000 19.1515 0.0023 190 1/100 19.7182 0.0015 191 1/10 18.2635 0.0013 192 1/5 19.6910 0.0010 -- Remarks. Tarnished Slightly tarnished. LOW CARBON STEEL. TEMPERATURE OF BOILING WATER. No. A B C D Table XLXIX. Action of Water. Original Weight. 7.0429 7.0774 7.4627 6.9830 Gain in Weight. 0.0074 0.0074 0.0081 0.0090 Remarks. Minute spots. Table L. Potassium Dichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 445 1/1000 6.2030 0.0017 446 1/100 7.4011 0.0007 447 1/10 6.6794 O.OOlo 448 1/5 7.0553 0.0015 __ Remarks. Slightly tarnished. 23 COMMERCIALLY PURE IRON. TEMPERATURE OF BOILING WATER. Table LI. Sodium Sulphate. No. Normality. Original Weight. Loss in Weight. Loss as Ferric Oxide. E 193 194 195 196 1/1000 1/100 1/10 N 19.2936 18.6913 18.4841 16.6731 0.0039 0.0043 0.0057 0.0062 0.0056 0.0061 0.0081 0.0089 Table LII. Magnesium Sulphate. No. Normality. Original Weight. 'Loss in Weight. Loss as Ferric Oxide. F 197 198 199 200 1/1000 1/100 1/10 N 17.4416 19.6567 18.6475 18.9933 0.0043 0.0049 0.0062 0.0071 0.0061 0.0070 0.0089 0.0101 Table LIII. Sodium Chloride. No. Normality. Original Weight. Loss in Weight. Loss as Ferric Oxide. I 201 202 203 204 1/1000 1/100 1/10 N 18.8491 16.3931 17.9781 16.9963 0.0043 0.0051 0.0063 0.0054 0.0061 0.0073 0.0090 0.0077 Table LIV. Magnesium Chloride. No. Normality. Original Weight. Loss in Weight. Loss as Ferric Oxide. I 205 206 207 208 1/1000 1/100 1/10 N 17.7218 18.4160 18.5508 17.0679 0.0049 0.0058 0.0096 0.0107 0.0070 0.0083 0.0137 0.0153 Remarks. Remarks. Remarks. Remarks. LOW CARBON STEEL. TEMPERATURE OF BOILING WATER. Table LV. Sodium Sulphate. Loss as t Original 'Loss in Ferric No. Normality. Weight. Weight. Oxide. 377 1/1000 6.5609 0.0063 0.0090 378 1/100 6.0070 0.0061 0.0088 379 1/10 7.6734 0.0066 0.0094 380 N 7.1523 0.0067 0.0096 Remarks. 24 Table LVI. Magnesium Sulphate. Original Loss in No. Normality. Weight. Weight. 381 1/1000 6.3835 0.0058 382 1/100 6.7226 0.0058 383 1/10 6.8952 0.0063 384 N 7.2419 0.0071 Table LVIL Soc Original Loss in No. Normality. Weight. Weight. 385 1/1000 6.3420 0.0054 386 1/100 7.4483 0.0062 387 1/10 7.2679 0.0068 388 N 6.7082 0.0059 Table LVIIL Mag: Original Loss in No. Normality. Weight. Weight. 389 1/1000 6.7509 0.0062 390 1/100 7.1248 0.0068 391 1/10 7.5381 0.0092 392 N 7.0032 0.0118 Remarks. Chloride. Loss as Ferric Oxide. 0.0077 0.0089 0.0097 0.0084 Loss as Ferric Oxide. 0.0089 0.0097 0.0131 0.0169 Remarks. Remarks. COMMERCIALLY PURE IRON. TEMPERATURE OF BOILING WATER. Table LIX. Sodium Sulphate with N/1000 Potassium Bichromate. Original No. Normality. Weight. 121 1/1000 18.2543 122 1/100 17.9236 123 1/10 18.0497 124 N 18.5067 Gain in Relative Weight. Cor'sion. Remarks. 0.0081 155 Few streaks and spots. 0.0052 100 0.0042 79 Several spots. 0.0057 107 Few spots. Table LX. Sodium Sulphate with N/100 Potassium Bichromate. No. 125 126 127 128 Table No. 129 130 131 132 Original Gain in Relative Normality. Weight. Weight. 'Cor'sion. 1/1000 18.2348 0.0038 71 1/100 19.5568 0.0049 92 1/10 18.7812 0.0060 113 N 18.0550 0.0076 143 LXL Sodium Sulphate with N/10 Original Gain in Relative Normality. Weight. Weight. Cor'sion. 1/1000 18.9583 0.0036 68 1/1-00 18.7540 0.0086 162 1/10 19.6902 0.0158 299 N 18.8220 0.0077 145 Remarks. No corrosion. Tarnished. Few spots. Streaks and spots. Potassium Bichromate. Remarks. No corrosion. Well tarnished. Few rust spots. Few streaks. 25 Table LXII. Sodium Sulphate with N/5 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion 133 1/1000 18.1663 0.0033 62 134 1/100 19.1555 0.0033 62 135 1/10 18.0014 0.0120 226 136 N 18.0851 0.0124 234 No Remarks, corrosion. Few spots and streaks. LOW CARBON STEEL. TEMPERATURE OF BOILING WATER. {Table LXIII. Sodium Sulphate with N/1000 Potassium Bichromate. No. 393 394 395 396 Original Gain in Relative Normality. 1/1000 Weight. 7.8221 Weight. 0.0067 Cor'sion. 84 1/100 1/10 N 8.7803 9.1039 8.9027 0.0073 0.0095 0.0086 91 119 108 Remarks. Few streaks and spots. Many minute spots. 'Few spots and streaks. Table LXIV. Sodium Sulphate with N/100 Potassium Bichromate. No. Normality. 397 1/1000 398 1/100 399 1/10 400 N Original Gain in Relative Weight. Weight. Cor'sion. Remarks. 8.2508 0.0057 71 Few spots. 7.5331 0.0058 73 " " 8.2940 0.0098 123 Streaks and spots. 8.7572 0.0095 119 " " " Table LXV. Sodium Sulphate with N/10 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 401 1/1000 8.9566 0.0037 46 402 1/100 9.1438 0.0091 114 403 1/10 9.0236 0.0124 155 404 N 9.0028 0.0119 149 Remarks. No corrosion. " Tarnished Few spots and streaks. Table LXVI. Sodium Sulphate with N/5 Potassium Bichromate. Remarks. No corrosion. Tarnished. Few rust spots. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 405 1/1000 9.1142 0.0043 54 406 1/100 9.2546 0.0032 40 407 1/10 8.7928 0.0114 143 408 N 8.2371 0.0097 121 COMMERCIALLY PURE IRON. TEMPERATURE OF BOILING WATER. Table LXVII. Magnesium Sulphate with N/1000 Potassium Bichromate. Remarks. Few spots. 'Black spots and streaks. Black streaks. 4 Original Gain in Relative No. Normality. Weight Weight. Cor'sion. 137 1/1000 19.1782 0.0048 91 138 1/100 19.2714 0.0101 191 139 1/10 18.2537 0.0077 145 140 N 18.5238 0.0110 208 26 Table LXVIII. Magnesium Sulphate with N/100 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 141 1/1000 19.1782 0.0046 87 Dull in color. 142 1/100 18.7740 0.0073 138 Spots over surface. 143 1/10 18.7936 0.0097 183 < 144 N 18.9597 0.0126 238 n tt a Table LXIX. Magnesium Sulphate with N/10 Potassium Dichromat'e Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 145 1/1000 17.1741 0.0006 11 Slightly tarnished. 146 1/100 17.8488 0.0013 23 " " 147 1/10 18.3274 0.0086 162 Few small spots. 148 N 18.3712 0.0106 200 Table LXX. Magnesium Sulphate with N/5 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 149 1/1000 18.5256 0.0003 6 No corrosion. 150 1/100 18.1176 0.0011 21 Slightly tarnished. 151 1/10 19.6333 0.0065 123 Spots and streaks. 152 N 18.8247 0.0082 151 " " " LOW CARBON STEEL. TEMPERATURE OF BOILING WATER. Table LXXI. Magnesium Sulphate with N/1000 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 409 1/1000 8.3184 0.0037 46 Black spots and streaks. 410 1/100 7.0408 0.0160 200 Many spots. 411 1/10 7.9015 0.0097 121 Spots and streaks. 412 N 7.1737 0.0115 144 " * * * ' Table LXXII. Magnesium Sulphate with N/100 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 413 1/1000 8.1879 0.0042 53 Few black spots. 414 1/100 9.2314 0.0118 148 Many small spots. 415 1/10 7.9020 0.0088 110 Spots over surface. 416 N 7.1787 0.0108 145 'Many small spots. Table LXXIII. Magnesium Sulphate with N/10 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 417 1/1000 7.8221 0.0010 13 Slightly tarnished. 418 1/100 8.0716 0.0013 16 " " 419 1/10 8.8831 0.0085 106 'Few spots over surface. 420 N 7.8375 0.0100 125 Many small spots. 27 Table LXXIV. Magnesium Sulphate with N/5 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 421 1/1000 7.2037 0.0014 18 422 1/100 9.1524 0.0009 11 423 1/10 8.2165 0.0057 71 424 N 8.5532 0.0095 119 Remarks. Slightly tarnished. Few small spots. Spots and streaks. COMMERCIALLY PURE IRON. TEMPERATURE OF BOILING WATER. Table LXXV. Sodium Chloride with N/1000 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 153 1/1000 18.1790 0.0075 141 Few black spots. 154 1/100 18.6186 0.0069 130 n a 155 1/10 18.2633 0.0066 124 " " " 156 N 18.9162 0.0062 117 Table LXXVL Sodium Chloride with N/100 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 157 1/1000 18.9220 0.0120 226 Few black spots. 158 1/100 18.1262 0.0208 392 Spots over surface. 159 1/10 17.6083 0.0267 504 Many small spots. 160 N 17.2738 0.0210 396 Table LXXVII. Sodium Chloride with N/10 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 161 1/1000 18.2339 0.0011 20 ' Tarnished slightly. 162 1/100 19.3364 0.0048 91 Well tarnished. 163 1/10 18.2891 0.0282 532 Many small spots. 164 N 18.7213 0.0447 843 Covered with rust. Table LXXVIII . Sodium Chloride with N/5 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight , Cor'sion. Remarks. 165 1/1000 19.2181 0.0018 34 Slightly tarnished. 166 1/100 19.7539 0.0078 147 Few spots. 167 1/10 18.9736 0.0262 494 Many small spots. 168 N 19.3316 0.0453 855 Covered with rust. LOW CARBON STEEL. TEMPERATURE OF BOILING WATER, Table LXXIX. Sodium Chloride with N/1000 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 425 . 1/1000 8.6269 0.0065 81 426 1/100 7.4392 0.0095 119 427 1/10 7.7890 0.0082 102 428 N 7.5234 0.0052 65 Remarks. Few small spots. Many small spots. Several streaks and spots. 28 Table LXXX. Sodium Chloride with N/100 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 429 1/1000 8.8366 0.0139 174 Many minute spots. 430 1/100 7.6972 0.0134 168 " " " 431 1/10 8.1314 0.0167 209 Streaks and spots. 432 N 8.1278 0.0190 238 a tt Table LXXXI.- -Sodium Chloride with N/10 Potassium Bichromate, Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 433 1/1000 7.6622 0.0013 16 Slightly tarnished. 434 1/100 7.6972 0.0052 65 Few spots. 435 1/10 7.8208 0.0254 318 Many minute spots. 436 N 8.7049 0.0434 543 Badly corroded. Table LXXXIL Sodium Chloride with N/5 Potassium Bichromate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 437 1/1000 7.5060 0.0014 18 Slightly tarnished. 438 1/100 7.3744 0.0093 116 (Pew minute spots. 439 1/10 7.6206 0.0254 318 Covered with spots. 440 N 7.7049 0.0468 585 Badly corroded. COMMERCIALLY PURE IRON. TEMPERATURE OF BOILING WATER Table LXXXIII. Magnesium Chloride with N/1000 Potassium Bichromate. Remarks. Spots and streaks. Many small spots. Streaks and spots. Original Gain in Relative wo. Normality. Weight. Weight. Cor'sion. ley 1/1000 18.6425 0.0091 173 170 1/100 19.0886 0.0116 219 171 1/10 19.7181 0.0132 249 172 N 18.0293 0.0231 436 Table LXXXIV. Magnesium Chloride with N/100 Potassium Bichromate. Remarks. Few minute spots. Many spots. Badly corroded. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 173 1/1000 19.1877 0.0057 108 174 1/100 18.9242 0.0149 281 175 1/10 18.6621 0.0229 432 176 N 18.4985 0.0321 606 LXXXV. Magnesium Chloride with N/10 Potassium Bichromate. Remarks. Tarnished. Well tarnished. Many spots and streaks. Badly corroded. WO. Normality. Original Weight. Gain in Weight. Relative Cor'sion. 177 178 179 1/1000 1/100 1/10 17.9358 19.4608 18.3315 0.0010 0.0014 0.0289 19 26 545 180 N 18.9670 0.0479 903 Table LXXXVI. Magnesium Chloride with N/5 Potassium Bichromate. No. Normality. 181 1/1000 182 1/100 183 1/10 184 N Original Weight. 18.4936 19.4126 18.3315 19.1092 Gain in Weight. 0.0011 0.0011 0.0321 0.0744 Relative Cor'sion. 20 20 606 140 Remarks. Slightly tarnished. Many spots and streaks. Badly corroded. LOW CARBON STEEL. TEMPERATURE OF BOILING WATER. Table LXXXVII. Magnesium Chloride with N/1000 Potassium Bichromate. No. Normality. 465 1/1000 466 1/100 467 1/10 468 N Original Gain in Relative Weight. Weight. Cor'sion. 7.2108 0.0122 153 6.9734 0.0122 153 7.0399 0.0180 225 8.0032 0.0227 284 Remarks. Many small spots. Streaks and spots. Many spots and streaks. Table LXXXVIII. Magnesium Chloride with N/100 Potassium Bichromate. No. Normality 469 1/1000 470 1/100 471 1/10 472 N Original Gain in Relative Weight. Weight. (Jor sion. 6.8296 0.0045 56 7.1818 0.0163 204 7.1590 0.0193 241 7.9684 0.0237 284 Remarks. Few spots. Many minute spots. Streaks and spots. Covered with rust. Table LXXXIX. Magnesium Chloride with N/10 Potassium Bichromate. No. Normality. 473 1/1000 474 1/100 475 1/10 476 N Original Weight. 7.3246 7.7206 8.0416 7.3858 Gam in Weight. 0.0020 0.0015 0.0282 0.0459 'Relative Cor'sion. 25 19 353 574 Remarks. Slightly tarnished. Covered with spots. Badly corroded. Table XC. Magnesium Chloride with N/5 Potassium Bichromate. No. Normality. 477 1/1000 478 1/100 479 1/10 480 N Original Weight. Gain in Weight. 'Relative Cor'sion. 7.2677 0.0005 6 6.7138 0.0019 24 8.3392 0.0419 524 7.3858 0.0879 1099 Remarks. Slightly tarnished. Tarnished. Badly corroded. SUMMARY OF RESULTS. (TEMPERATURE OF BOILING WATER.) Solutions of Sodium Sulphate with Potassium Bichromate. 1. With very dilute solutions of the sulphate the addition of dichromate causes a slight inhibition in corrosion. 2. In the more concentrated solution the addition of dichromate stim- ulates corrosion. 30 3. In many cases the corrosion of the test-plates appeared to be greater in comparison with those in solutions free from dichromate than is actually represented by the gain in weight. 4. Generally the low carbon steel was more easily attacked than the commercially pure iron. Solutions of Magnesium Sulphate with Potassium Dichromate. 1. With the exception of N/1000 magnesium sulphate the addition of dichromate in small amounts stimulated corrosion. 2. N/10 and N/5 dichromate inhibited corrosion in varying degrees. 3. Generally the low carbon steel was more easily attacked than the commercially pure iron. 4. The action of the magnesium sulphate solutions were less than those of sodium sulphate. Solutions of Sodium Chloride with Potassium Dichromate. 1. N/1000 dichromate had practically no effect upon the corrosion of the plates, although there was a general tendency towards inhibition. 2. N/100 dichromate stimulated corrosion in every case. 3. With the more dilute solutions of sodium chloride, N/10 and N/5 dichromate inhibited corrosion. 4. N/10 and N/5 dichromate in the more concentrated solutions of sodium chloride stimulated corrosion to a marked extent. 5. Commercially pure iron showed a slightly greater increase in the rate of corrosion than did the low carbon steel. 6. Sodium chloride and potassium dichromate interacted with water to produce traces of hydrochloric acid. Solutions of Magnesium Chloride with Potassium Dichromate. 1. In all cases the action of these solutions upon the iron and steel was greater than the corresponding solutions of sodium chloride. 2. N/10 and N/5 dichromate in the dilute solutions of magnesium chloride inhibited corrosion. 3. In all other solutions increase in either the dichromate or the chloride produced conditions that strongly stimulated corrosion. 4. Magnesium chloride and potassium dichromate interact with water to produce hydrochloric acid. 31 ACTION OF DI-SODIUM PHOSPHATE IN THE PRESENCE OF SODIUM SULPHATE AND SODIUM CHLORIDE. When an iron strip is made the anode in an electric circuit and placed in a solution of di-sodium phosphate, it is passivified. It has been shown by Heyn and Bauer that iron placed in this solution will not corrode as readily as in solutions of commoner salts, and that a point or "limiting concentration " is reached where all corrosive action ceases. As this phos- phate is used to a certain extent as a boiler water softener, it was decided to study its effect in solutions of sodium chloride and sodium sulphate upon the iron and steel used in other experiments. The test's were conducted in the same manner as previously described and the data obtained follows: COMMERCIALLY PURE IRON. ROOM TEMPERATURE. Table XCI. Di-sodium Phosphate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion 241 1/1000 18.7146 0.1494 75 242 1/100 17.1147 0.0190 9 243 1/10 18.4924 0.0005 1 244 N 18.0340 0.0001 1 Remarks. Badly corroded. Green and brown No corrosion. rust. LOW CARBON STEEL. ROOM TEMPERATURE. Table XCII. Di-sodium Phosphate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. 369 1/1000 6.4191 0.1339 61 370 1/100 6.9566 0.0232 10 371 1/10 7.3431 0.0004 1 372 N 6.8778 0.0000 Remarks. 'Badly corroded. Greenish brown rust. No corrosion. COMMERCIALLY PURE IRON. ROOM TEMPERATURE. Table XCIII. Sodium Sulphate with N/1000 Di-sodium Phosphate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 89 1/1000 17.6301 0.1893 95 Badly corroded. 90 1/100 17.3585 0.1792 85 91 1/10 18.3143 0.1850 93 " " 92 N 18.9224 0.1109 55 " " Table XCIV Sodium Sulphate with N/100 Di-sodium Phosphate. Original Loss in Relative Normality. Weight. Weight. Cor'sion. 1/1000 18.7378 0.0489 24 1/100 19.5230 0.0505 25 1/10 19.6386 0.0613 13 Xo. 93 . 94 X 0.0861 33 Remarks \ Covered with a light brown layer mottled with [ greenish streaks. Badly corroded. 1 Gain in weight. 32 Table XCV. Sodium Sulphate with N/10 Di-sodium Phosphate. No. Normality. Original Weight. Loss in Weight Relative . Cor'sion. Remarks. 97 98 99 100 1/1000 1/100 1/10 N 19.2083 18.5526 18.9841 19.6082 0.0002 0.0061 0.0195 0.0309 1 3 10 15 No corrosion. Small lumpy green spots. Large " " Covered with brown rust. Table XCVL Sodium Sulphate with N/5 Di-sodium Phosphate. No. Normality. Original Weight. Loss in Weight Relative , Cor'sion. Remarks. 101 102 103 104 1/1000 1/100 1/10 N 20.0196 19.3779 19.9065 18.7244 0.0004 0.0000 0.0032 0.0266 1 1 13 No corrosion. Small green festoons on ed Greenish-brown deposits. LOW CARBON STEEL. ROOM TEMPERATURE. Table XC VII. Sodium Sulphate with N/1000 Di-sodium Phosphate. Original Loss in Relative No. Normality. 'Weight. Weight. Cor'sion. Remarks. 209 1/1000 7.2392 0.2221 Ill Badly corroded. 210 1/100 6.7040 0.1597 80 211 1/10 7.7197 0.1862 93 < 212 N- 7.2695 0.1428 71 Table XC VIII. Sodium Sulphate with N/100 Di-sodium Phosphate. Original Loss in Relative No. Normality. 'Weight. Weight. Cor'sion. 213 1/1000 7.0472 0.0285 14 214 1/100 7.7363 0.0304 15 215 1/10 7.5623 0.0733 36 216 N 7.0303 0.0842 42 Remarks. Few small spots. Brown spots and streaks. Badly corroded. Table XCIX. Sodium Sulphate with N/10 Di-sodium Phosphate. Original Loss in Relative No. Normality. 'Weight. Weight. Cor'sion 217 1/1000 7.7705 0.0003 1 218 1/100 7.3620 0.0182 9 219 1/10 6.6200 0.0183 9 220 N 7.7727 0.0427 21 Remarks. No corrosion. 'Few minute green spots. Well corroded. Table C. Sodium Sulphate with N/5 Di-sodium Phosphate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 221 1/1000 7.3201 0.0001 1 No corrosion. 222 1/100 7.6792 0.0002 ' " " 223 1/10 7.5185 0.0093 5 Few minute green 224 N 6.9973 0.0399 20 " large " spots Gain in weight. COMMERCIALLY PURE IRON. ROOM TEMPERATURE. Table CL Sodium Chloride with N/1000 Di-sodium Phosphate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 105 1/1000 18.1240 0.1424 71 Badly corroded 106 1/100 18.9518 0.1376 69 " " 107 1/10 19.0994 0.1260 63 108 N 18.0001 0.1336 67 " " Table CIL Sodium Chloride with N/100 Di-sodium Phosphate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 109 1/1000 18.2028 0.0917 46 Badly corroded 110 1/100 19.1043 0.0962 48 " " 111 1/10 18.7914 0.0208 10 Green streaks. 112 N 18.9768 0.0138 6 " " Table CIII Sodium Chloride with N/10 Di-sodium Phosphate. Remarks. No corrosion. Few green spots. Large green spots. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. 113 1/1000 18.8776 0.0003 1 114 1/100 18.9418 0.0063 3 115 1/10 17.2943 0.0350 17 116 X 19.1335 0.0243 12 Table CIV. Sodium Chloride with N/5 Di-sodium Phosphate. No. Normality. 117 1/1000 118 1/100 119 1/10 120 N Original Loss in Relative Weight. Weight. Cor'sion. Remarks. 17.3865 0.0009 ' No corrosion. 18.1178 0.0009 1 " " 20.0233 0.0316 15 Green spots and 18.3372 0.0339 17 ii streaks. LOW CARBON STEEL. ROOM TEMPERATURE. Table CV. Sodium Chloride with N/1000 Di-sodium Phosphate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 225 1/1000 6.6189 0.1643 76 Badly corroded 226 1/100 6.6663 0.1682 77 " " 227 1/10 6.5467 0.1497 69 " " 228 N 7.4723 0.1477 68 n Table C VI. Sodium Chloride with N/100 Di-sodium -Phosphate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. 229 1/1000 230 - 1/100 231 1/10 232 N 6.7869 7.6427 7.0475 6.8192 0.0877 0.0724 0.0273 0.0159 40 29 13 7 Remarks. Badly corroded. Green streaks and spots. : Gain in weight. 34 Table CVIL Sodium Chloride with N/10 Di-sodium Phosphate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 233 1/1000 7.3540 0.0004 1 No corrosion. 234 1/100 6.5625 0.0048 2 Few minute green spots. 235 1/10 7.0781 0.0344 15 Green streaks and spots. 236 N 7.5965 0.0169 8 Table CVIII. Sodium Chloride with N/5 Di-sodium Phosphate. Original Gain in (Relative No. Normality. Weight. Weight . Cor'sion. Remarks. 237 1/1000 6.3236 0.0001 1 No corrosion. 238 1/100 6.1297 0.0011 1 " " 239 1/10 6.6065 0.0250 11 Green streaks and spots, 240 N 7.5117 0.0271 13 " " " " COMMERCIALLY PURE IRON. TEMPERATURE OF BOILING WATER. No. Normality. 245 1/1000 246 1/100 247 1/10 248 N Table CIX. Di-sodium Phosphate. Original Gain in Relative Weight. Weight. Cor'sion. Remarks. 17.4359 0.0143 269 Greenish brown streaks. 17.9364 0.0102 194 " " " 16.8934 0.0058 108 Green spots. 17.0009 0.0024 46 Tarnished. LOW CARBON STEEL. TEMPERATURE OF BOILING WATER. Table CX. Di-sodium Phosphate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 373 1/1000 6.5418 0.0125 156 374 1/100 7.0031 0.0110 138 375 1/10 6.8985 0.0054 68 376 N 6.5718 0.0019 24 Remarks. Green and brown spots. Small green spots. Well tarnished. COMMERCIALLY PURE IRON. BOILING TEMPERATURE. Table CXI. Sodium Sulphate with N/1000 Di-sodium Phosphate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 333 1/1000 19.2281 0.0170 321 334 1/100 19.4353 0.0142 268 335 1/10 18.1332 0.0125 236 336 N 19.3526 0.0104 196 Remarks. Spots over surface. Green spots. Well tarnished. Table CXIL Sodium Sulphate with N/100 Di-sodium Phosphate. Original Gain in (Relative No. Normality. Weight. Weight. Cor'sion. Remarks. Green streaks. Green spots. Whole plate dull green. 337 i/1000 19.1982 0.0156 294 338 1/100 18.3059 0.0140 264 339 1/10 18.9705 0.0155 293 340 N 18.7855 0.0137 268 J Gain in weight. 35 Table CXIIL Sodium Sulphate with N/10 Di-sodium Phosphate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 341 1/1000 19.6320 0.0146 275 342 1/100 19.7485 0.0196 370 343 1/10 19.5355 0.0169 319 344 X 19.2326 0.0117 221 Remarks. Well tarnished. Green spots and streaks. Badly tarnished. Several green spots. Green streaks. Table CXIV. Sodium Sulphate with N/5 Di-sodium Phosphate. Remarks. Tarnished. Green spots. Badly tarnished. Green streaks and spots. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 345 1/1000 18.4341 0.0045 85 346 1/100 19.6283 0.0143 270 347 1/10 18.1347 0.0121 228 348 ' N 19.0593 0.0117 221 LOW CARBON STEEL. BOILING TEMPERATURE. Table CXV. Sodium Sulphate with N/1000 Di-sodium Phosphate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 449 1/1000 7.3269 0.0174 218 450 1/100 7.6484 0.0202 252 451 1/10 8.0451 0.0132 165 452 N 7.4946 0.0134 168 Remarks Green spots and streaks. Faint green streaks. Dull colour. Tarnished. TaMe CXVI. Sodium Sulphate with N/100 Di-sodium Phosphate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 453 1/1000 7.7013 0.0181 226 454 1/100 7.5766 0.0200 250 455 1/10 8.0451 0.0188 225 456 N 7.4946 0.0134 193 Remarks. Green spots and streaks. Green streaks. Many small green spots. Table CXVII. Sodium Sulphate with N/10 Sodium Phosphate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 457 1/1000 7.0807 0.0174 217 458 1/100 7.4831 0.0176 220 459 1/10 7.7700 0.0191 239 460 N 7.9405 0.0142 178 Remarks. Many green spots. Table CXVIII. Sodium Sulphate with N/5 Di-sodium Phosphate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 461 1/1000 7.0911 0.0067 84 462 1/100 7.6702 0.0182 228 463 1/10 7.9916 0.0136 170 464 N 7.8441 0.0122 153 Remarks. Tarnished. Small spots. Large " Large green spots. 36 COMMERCIALLY PURE IRON. TEMPERATURE OF BOILING WATER. Table CXIX. Sodium Chloride with N/1000 Di-sodium Phosphate. No. Normality. Original Weight. Gain in Weight. Relative Cor'sion. Remarks. 349 1/1000 19.0371 0.0136 257 Many green lumpy spots. 350 1/100 18.8365 0.0139 262 1 t t. t 351 1/10 16.4222 0.0098 185 Greenish coating. 352 N 18.5260 0.0063 118 Well tarnished. Table CXX. Sodium Chloride with N/100 Di-sodium Phosphate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 353 1/1000 18.9083 0.0096 181 Green streaks and spots. 354 1/100 17.0315 0.0130 245 " " " " 355 1/10 18.1345 0.0100 190 Small green spots. 356 N 17.5985 0.0115 216 Green coating. Table CXXI. Sodium Chloride with N/10 Di-sodium Phosphate Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 357 1/1000 18.1934 0.0068 128 Few green spots and lumps. 358 1/100 17.6379 0.0110 207 Green spots and streaks. 359 1/10 18.3824 0.0169 318 Green lumps. 360 N 18.2926 0.0159 300 'Green coating and lumps. Table CXXIL- -Sodium Chloride with N/5 Di-sodium Phosphate Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 361 1/1000 18.4672 0.0020 38 Well tarnished. 362 1/100 18.1511 0.0044 83 Minute green spots. 363 1/10 18.1273 0.0135 257 Few green lumps. 364 N 18.2734 0.0151 288 Green lumps. LOW CARBON STEEL. TEMPERATURE OF BOILING WATER. Table CXXIII. Sodium Chloride with N/1000 Di-sodium Phosphate. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 481 1/1000 6.7084 0.0169 210 482 1/100 6.7072 0.0153 191 483 1/10 8.0876 0.0123 154 484 N 8.2663 0.0071 89 Remarks. Green coating with spots. Green spots and streaks. Green coating. Tarnished. Table CXXIV. Sodium Chloride with N/100 Di-sodium Phosphate. Remarks. Dark green coating. Green coating and spots. Green-brown coating. Original Gain in Relative No. Normality. Weight. Weight. Cor'sion. 529 1/1000 6.2168 0.0091 114 530 1/100 6.9987 0.0107 134 531 1/10 8.1169 0.0159 199 532 N 8.2188 0.0147 184 37 Table CXXV. Sodium Chloride with N/10 Di-sodium Phosphate. Original No. Normality. Weight. 533 1/1000 7.1474 534 1/100 6.7447 535 1/10 6.8138 536 N 7.4259 Gain in Relative Weight. Cor'sion. 0.0087 109 0.0131 164 0.0149 186 0.0164 205 Remarks. Tarnished. Green colour. Spots. Green coating and spots. Table CXXVI. Sodium Chloride with N/5 Di-sodium Phosphate. No. Normality. 537 1/1000 538 1/100 539 1/10 540 N Original Weight. 7.5677 6.6022 7.9037 7.0790 Gain in Weight. 0.0016 0.0073 0.0135 0.0162 Relative Cor'sion. 20 91 169 203 Remarks. Well tarnished. Pew green spots. Green lumpy spots. Room Temperature. Sodium Sulphate. 1. In all cases di-sodium phosphate inhibits corro- sion but not to the extent manifested by potassium di-chromate. 2. With N/1000 di-sodium phosphate, the rate of corrosion is only slightly reduced, but in solutions containing greater amounts of this salt, the degree of inhibition is much more marked. 3. The concentration of the sodium sulphate remaining constant, an increase in the amount of phosphate produces a decrease in the rate of corrosion. However, if the sulphate varies and the concentration of the phosphate remains constant, an increase in corrosive action follows. 4. With solutions containing N/100 or more of phosphate, the action of the electrolytes upon the iron is shown by the formation of dark green or greenish brown spots and streaks. 5. There is a noticeable difference in the behavior of the two types of test-plates. 6. When corrosion did not take place, the test-plates were found to be in the active condition, as they were readily attacked by 1.20 nitric acid. Sodium Chloride. The above summary describing the action of sodium sulphate and di-sodium phosphate also applies in a general way to conditions produced by solutions of sodium chloride and the phosphate. However, with the normal solutions an increase of phosphate from N/10 to N/5 will give a slight increase in the rate of corrosion. Further, if the phosphate is kept constant and the amount of sodium chloride is varied a decrease in corrosion sometimes results, especially with N/100 phosphate. * Temperature of Boiling Water. 1. It was found that if di-sodium phosphate be introduced into solu- tions of sodium chloride or sodium sulphate, an increase in the rate of corrosion of the test-plates was obtained. 38 2. In N/5 di-sodium phosphate there is a general tendency for the rate of corrosion to decrease. 3. The low carbon steel showed a higher rate of corrosion than did the commercially pure iron. ACTION OF POTASSIUM BICHROMATE ON GALVANIZED IRON IN THE PRESENCE OF SODIUM CHLORIDE AND SODIUM SULPHATE. The use of zinc was suggested some years ago as a means for preventing corrosion in boilers. Slabs of zinc were placed in the boiler in contact with the iron or steel, which was preserved at the expense of the zinc. This method for preventing corrosion was adopted on board ships and made particular use of by tramp steamers, the results being highly satisfactory. However, the great objection to the method is the rapid manner in which the zinc will go into solution under boiler conditions. As Dunstan and Hill l have shown that zinc can be passivified by means of potassium dichromate, and in view of the fact that this salt will greatly reduce the rate of corrosion of iron and steel in the presence of large amounts of common boiler salts at room temperature, as shown in previous experiments, it was thought that possibly the rate of solution of the zinc would be like- wise reduced in the presence of potassium dichromate without hindering it's action toward the iron. Accordingly solutions of varying concentrations of sodium chloride and sodium sulphate containing potassium dichromate in different amounts were made up, and test-plates of galvanized iron, similar in size to those previously used, were introduced into the solutions. Tests were run both at room temperature and at the temperature of boiling water for two weeks. The conditions under which the tests were performed were the same as given on p. 10. In cleaning the plates after they had stood in the solutions great diffi- culty was experienced. After trying several means for cleaning them, it was decided to simply wash them thoroughly with water and then with alcohol and dry. It was found to be almost impossible to secure very good checks on the tests. This is attributed to improper means for cleaning, formation of zinc chromate, and to the great variation in the size of the crystals of zinc on the galvanized iron. The rate of solution of the plates in water was taken as a standard. Tests were run with solutions containing varying amounts of potassium dichromate. The plates, after standing for a brief period, especially in the stronger solutions, became very bright; but at the end of the two weeks it was observed that much more zinc had gone into solution than when the plates were immersed only in water. This is easily explained by the fact that the addition of the potassium dichromate would increase the con- ductivity of the solutions and thus cause an increase in galvanic action. A glance at the following results will show that the presence of potas- sium dichromate in either solutions of sodium chloride or sodium sulphate 1 Transactions of the Chemical Society, vol. 99, p. 1861. 39 will stimulate the solution of the zinc rather than inhibit it. If the amount of potassium dichromate is kept constant and the concentration of the sodium chloride or sodium sulphate is increased, the rate of the destruction of the zinc is increased. Further, if the concentration of the potassium dichromate is varied and that of the sodium chloride or sodium sulphate kept constant, the rate at which the zinc is attacked is far more vigorous. The plates immersed in the more concentrated solutions of potassium dichromate and sodium sulphate or sodium chloride were all of a yellow colour, the intensity of which depended upon the strength of the solutions. After these plates had been thoroughly washed with water, they were treated with hydrochloric acid to dissolve the zinc together with the yellow coating. The resulting solutions were tested for chromium. In every case this element was identified, much more chromium being found in the solu- tions obtained from the plates having the deep yellow color than those possessing a lighter color. Thus the yellow deposit on the galvanized iron was due to the formation of zinc chromate. With the tests run at the temperature of boiling water no quantitative results were obtained. In every case the plates were badly attacked, and in some instances flakes of metallic zinc had fallen from the plates. In many others all the zinc had dissolved and the solutions had attacked the iron. This was especially true of the sodium chloride solutions. GALVANIZED IRON. ROOM TEMPERATURE. Table CXXVII. Action of Water. Original Loss in No. Weight. Weight. Remarks. A 5.9566 0.0023 No visible action. B 6.1762 0.0029 C 6.2771 0.0038 D 5.9928 0.0032 E .0821 0.0026 Table CXXVIII. Potassium Dichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 525 1/1000 5.8927 0.0039 130 No visible action. 526 1/100 6.1605 0.0021 70 Metal brighter. 527 1/10 6.3310 0.0064 213 Metal bright. 528 1/5 6.4807 0.0065 216 GALVANIZED IRON. ROOM TEMPERATURE, Table CXXIX. Sodium Sulphate with N/1000 Potassium Dichromate. * Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 485 ' 1/1000 5.7442 0.0059 197 No visible change. 486 1/100 5.4649 0.0064 213 487 1/10 5.7579 0.0088 293 488 N 6.1893 0.0114 380 40 Table CXXX. Sodium Sulphate with N/100 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight . Cor'sion. Remarks. 489 1/1000 5.5231 0.0159 530 No visible change 490 1/100 5.8883 0.0320 1063 " " " 491 1/10 5.5643 0.0383 1277 Light yellow in color. 492 N 5.7842 0.0775 2583 Table CXXXL Sodium Sulphate with N/10 Potassium Bichromate, Original Loss in Relative No. Normality. Weight. Weight . Cor'sion. Remarks. 493 1/1000 5.7247 0.0114 380 No visible change 494 1/100 5.5725 0.0345 1150 Light yellow. 495 1/10 5.7728 0.0528 1760 Deep yellow. 496 N 5.4051 0.0649 2163 Table CXXX1L Sodium Sulphate with N/5 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight . Cor'sion. Remarks. 497 1/1000 6.0362 0.0171 570 No visible change 498 1/100 5.2548 0.0389 1297 Deep yellow. 499 1/10 6.0078 0.0950 3167 " " 500 N 5.7815 0.0986 3287 " " GALVAXIZEB IRON. ROOM TEMPERATURE. Table CXXXIIL Sodium Chloride with N/1000 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight . Cor'sion. Remarks 501 1/1000 5.8586 0.0052 173 No visible change. 502 1/100 5.8087 0.0097 323 " " ' 503 1/10 5.3215 0.0061 203 " " ' 504 N 5.1719 0.0097 323 " " " 505 3 N 5.5820 0.0145 483 " " " 506 5 N 6.1767 0.0210 700 Table CXXX IV.- -Sodium Chloride with X/100 Potassium Bichromate. Original Loss in Relative No. Normality. Weight. Weight . Cor'sion. Remarks 507 1/1000 5.9378 0.0062 207 No visible change. 508 1/100 6.0623 0.0248 827 a a ; 509 1/10 6.1338 0.0398 1327 " " '' I 510 N 6.3237 0.0568 1893 " " " 511 3 N 6.0456 0.0675 2250 Light yellow in color. 512 5 N 6.0240 0.0691 2333 " " ' ( " 41 Table CXXXV. Sodium Chloride with N/10 Potassium Dichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. 513 1/1000 6.3159 0.0055 183 514 1/100 5.8117 0.0270 900 515 1/10 5.8750 0.0694 2313 516 N 6.1337 0.0987 3290 517 3 N 6.0204 0.0981 3270 518 5 N H.9321 0.113a 3767 Remarks. No visible change. Light yellow in color. Deep yellow. Table CXXXVI. Sodium Chloride with N/5 Potassium Dichromate. Original Loss in Relative No. Normality. Weight. Weight. Cor'sion. Remarks. 519 1/1000 5.9914 0.0018 60 No visible change. 520 1/100 5.9276 0.0282 940 " " " 521 1/10 6.0575 0.0713 2377 Light yellow. 522 N 6.1311 0.0909 3030 Deep yellow. 523 3 N 5.9955 0.0987 3290 " " 524 5 N 6.2547 0.1006 3353 " " THE EFFECT OF THE ELECTRIC CURRENT AND POTASSIUM DICHROMATE IN VARIOUS SOLUTIONS UPON AN IRON ANODE. A study of the behavior of an iron anode in various solutions containing potassium dichromate and heated to high temperatures was undertaken. A small boiler containing a pressure gauge, a safety valve and a common spark plug, which was screwed into the top of the boiler, was constructed. The entire boiler was made the cathode in an electric circuit and a small strip of iron suspended from the spark plug was used as the anode. The iron strips which were used as anodes were made from commer- cially pure iron and had the following dimensions : 10 cm. X 0.50 cm. X 0.10 cm. The total area was 12.10 square centimetres. A small hole was drilled into the top of each strip so that it could be hooked on to A. When the plug was screwed into the top of the boiler, the iron strip would be com- pletely covered by the solution. The hole in the strips decreased the area exposed to 11.97 square centimetres. The boiler was heated by placing in a large paraffin bath. N/10 solutions of sodium sulphate, sodium chloride, magnesium sulphate, and magnesium chloride were used. To these solutions potassium dichro- mate was added until they were also equivalent to N/10 and N/5 solutions of this substance. In running the experiments the solutions, which would be at room tem- perature, were introduced into the cool boiler. The iron strip was then introduced into the solution, and the plug from which it was suspended screwed securely into the top of the boiler. The proper electrical connec- tions were made, so that a current of two and a half amperes flowed through 42 the system. The boiler was then immediately placed in the paraffin bath and heated rapidly until a pressure of 120 Ibs. was recorded upon the gauge. At the end of half an hour the boiler was removed from the bath, and when cool enough the current was stopped and the iron strip removed. In every case it was found that the iron would be in the active condition, as a large part of it had gone in solution during the half hour treatment. Tests were then run starting with warm solutions containing normal (147 grammes per litre) potassium dichromate in N/10 solutions of the four salts previously used. Upon repeating the experiments as described above, the same results were obtained, the increase in the amount of potassium dichromate producing no effect. A warm solution was used at the start to dissolve all the potassium dichromate. Thus it may be concluded that the possibility of keeping iron passive under boiler conditions with the aid of an electric current and potassium- dichromate is practically impossible. GENERAL CONCLUSIONS. 1. At atmospheric conditions the introduction of varying amounts of potassium dichromate into solutions of different concentrations of sodium sulphate, sodium chloride, magnesium sulphate and magnesium chloride, and saturated calcium sulphate, inhibit corrosion t'o a remarkable extent. 2. The differences in the chemical composition and physical properties of the iron and steel employed appeared to have practically no effect when placed in solutions of chlorides and sulphates containing potassium dichro- mate at room temperature. Differences in behavior were noted at the temperature of boiling water. 3. At atmospheric conditions a few of the test-plates were rendered passive by the dichromate contained in certain of the more dilute solutions. 4. The rate of corrosion was reduced in some cases upon the addition of dichromate to the dilute solutions at the temperature of boiling water. 5. In N/10 and normal solutions of the chlorides and sulphates the addition of dichromate to the hot solutions was decidedly injurious to the iron and steel. 6. The commercial application of potassium dichromate to boiler waters is not' practicable because of cost and uncertainty of results. 7. Di-sodium phosphate limits the rate of corrosion of the iron and steel at room temperatures when added to various solutions of sodium sulphate and sodium chloride. 8. Di-sodium phosphate does not produce the passive state when iron or steel is immersed in its solutions. 9. Slight differences in the action of the iron and steel in solutions containing di-sodium phosphate were noted. 43 . 10. Iron and steel in solutions of sodium sulphate and sodium chloride containing di-sodium phosphate and heated to the temperature of boiling water, are more readily acted upon than if the phosphate were not present. Even small amounts of this salt increase the rate of corrosion. 11. The rate of solution of zinc occurring on galvanized iron is increased by the addition of potassium dichromate to solutions of sodium sulphate and sodium chloride. 12. Iron cannot be rendered passive under boiler conditions in N/10 solutions of chlorides and sulphates when varying amounts of potassium dichromate are added and the iron made the anode in an electric circuit. The author of this paper is indebted to the Iron and Steel Institute of Great Britain for financial assistance given through the Carnqgie Scholarship Fund. 44 VITA. The author of this paper was born in Rose Bank, Staten Island, N. Y. He received his early education in the grammar schools of Brooklyn, N. Y., and graduated from the Commercial High School of that city in 1906. Immediately after graduation he accepted a position in the chemical labora- tories of the American Brass Co. of Torrington, Conn., and was employed there until September. 1911. He then entered Clark College, Worcester, Mass., and received the degree of Bachelor of Arts from that institution in 1914. While a student at Clark some time was spent in the chemical laboratories of the American Steel and Wire Co. as assistant chemist. In the fall of 1914 he entered the University of Washington as teaching fellow in chemistry. The degree of Master of Science was granted in the summer of 1915. During this year papers on the "Total Amino Nitrogen in the Seedlings of the Alaska Pea" and the "Tannin Content of Pacific Coast Conifers", the latter with Dr. H. K. Benson, were published. In the same year a Carnegie Research Scholarship was received from the Iron and Steel Institute of Great Britain. The years 1916 and part of 1917 were spent in graduate work at the University of Washington. In June of the latter year, the examinations for First Lieutenant in the Ordnance Department were passed but no call into active service was received until December. Being among the first drafted in July of that year, he served a month at Camp Lewis as a private and two months with the Gas Defense Service, working in the laboratories of the National Carbon Company, of Cleve- land, Ohio. As First Lieutenant he worked in the Geophysical Lab- oratory of Washington, D. C., the laboratories of Ohio State University, Johns Hopkins University, and of Edgewood Arsenal. All of the work performed in these laboratories dealt with the chemistry of war gases. In July, 1918, he was transferred to the Chemical Warfare Service with the rank of captain. UNIVEESITY OF CALIFORNIA LIBRARY, BERKELEY THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW Books not returned on time are subject to a fine of oOc per volume after the third day overdue, increasing to $1.00 per volume after the sixth day. Books not in demand may be renewed if application is made before expiration of loan period. JBL i.0. Due end of SUMMEF subject to recall a SEP JUN 7, 50m-8,'26 Gaylord Bros Makers Syracuse, N Y. PAT. JAN. 21. 1908 UNIVERSITY OF CALIFORNIA LIBRARY