564
T H F ,/OCILVAL OF I N D U S T R I A L A N D E S G I X E E R I N G C H E N I S T R Y .
The above total corresponds t o the following molecular formula of enamel:
Aug.,
1912
was found to be decomposed to a depth of 2 cm. Quantities of free sulfur were deposited upon the walls, and the calcium compounds of the cement were found to 2.513 Si02 0.497 Nap0 0.262 BzOs have been converted near the surface into crystals of gypsum. Large quantities of hydrogen sulfid were 0 039 MgO given off when the dosing chamber discharged, causing RESEARCH LABORATORY, complaint of foul odors by neighbors and residents a t LISK BIANUFACTUKING CO , LTD., CANANDAIGUA, N. Y the hospital. At a later date a similar condition was observed in the septic tank receiving the sewage of TEE PRODUCTION OF EXCESSIVE HYDROGEN SULFID IN the city of Grinnell, Iowa. SEW AGE DISPOSAL PLANTS AND CONSEQUENT DISI n both these cases various measures have been INTEGRATION OF THE CONCRETE.' tried t o prevent the formation of excessive odors and By WILLIAMM. BARRAND R. E. BUCHANAN. decomposition of the concrete. A water-softening Received February 15, 1912. plant was installed at the Knoxville insritution, thinkI t is well known t o those who are familiar with the ing this might improve conditions, b u t , except for operation of the septic tank that in the chambers of making the water more desirable for domestic pursuch tanks, which operate in the absence of light and poses, nothing was achieved. Frequent cleaning of air, the bacteria produce, in the decomposition of the the tank and disinfecting a t the time of cleaning did organic matter of the sewage, quantities of reducing not materially improve conditions. gases. Among the decomposition products is observed During the progress of this investigation our a t more or less hydrogen sulfid gas, resulting in most tention was called to the septic tank a t the State cases from the decomposition of organic sulfur comHospital a t M t . Pleasant, Iowa, and before visiting pounds existing in the sewage. I n most tanks of this the place, a n analysis of the water supply caused us character the amount of hydrogen sulfid thus liberated 9 t o predict the condition existing there. A visit is not sufficiently large t o cause complaint from excesshowed the septic tank to be in much the same consive odors or t o account for the conditions in certain dition as those first studied. tanks observed by the writers. The object of this investigation was to determine Attention was first called to the disintegration of the cause of the conditions above noted and to detercement mortar in concrete resulting from sewer gases, mine whether bacteria were in any measure responsible by 0lnistead.a The case observed by him occurred for the changes, or whether they were purely of a in the outfall sewer for Los Angeles, built in 1895. chemical nature. The various observations made by This resulted from the holding back of the sewage in us, together with analyses of the water supplies and inverted siphons in this sewer, forming, in principle, the sewage, seemed to point to some definite relationa septic tank. Beyond this point the neat cement ship existing between the compounds of sulfur in the mortar was badly disintegrated and it became neces- water supply and the formation of excessive hydrogen sary t o reconstruct the walls and roof of this portion sulfid, with the consequent formation of sulfuric acid of the srwer. It was observed that the disintegration on the walls of the tank. resulted from the formation of sulfuric acid upon the This was used for a working hypothesis for the folwalls and roof, which attacked the cement, giving lowing reasons: calcium sulfate as the final product. It was also First.-The water supplies in the cases observed observed that the amount of sulfuric acid formed were found t o have a n unusually high content of appeared t o be too large t o have been produced by sulfates. the organic sulfur compounds in the sewage proper. Second.-The odor of hydrogen sulfid was, in all It was also suggested that the only possible sources cases, very pronounced about the dosing chamber. of this excessive amount of sulfur were the oil wells Third.-Free sulfur was found floating on the surand the sulfur which is always present in small quanti- face of sewage which had been standing for some time, ties in the normal sewage. The report does not indi- and quantities of free sulfur were deposited on the cate that the source of these sulfur compounds was walls in various parts of the septic tank. definitely located. Fourth.-An abundance of sulfur bacteria was The case of a septic tank in England where a con- evident upon microscopic examination. dition similar t o the one cited above existed has been F i f t h .-The disintegrating concrete showed much reported by D u n n , ~but no attempt is made to show greater amounts of sulfate than the unaffected conany relation between this condition and the character crete, and a s already stated showed crystals of gypsum of the water supply. on microscopic examination. As early as November, 1908, one of us visited the A large proportion of proteins contain a considerable sewage disposal plant of the Inebriate Hospital at percentage of sulfur. When such compounds are Knoxville, Iowa, and observed the following condi- decomposed in the absence of oxygen, the sulfur tions: The concrete composing the roof and walls makes its appearance for the most part in combinaabove the high water line in the dosing chamber was tion with hydrogen as hydrogen sulfid. Inasmuch rapidly disintegrating. The under surface of the roof as proteins and the products of hydrolytic cleavage From a Bulletin of the Eng. Expt. Sta., Iowa State College. are generally t o be found in sewage, it is to be expected Eng. News, 44, 317. that a certain amount of hydrogen sulfid will be de8 Cement Eng. News. 22, 10.
Aug.,
1912
T H E J O U R N A L OF I1VDiYSTRIA4L A S D EAYGIA\-EERI.\-G C H E X I S T R Y .
veloped in such sewage as it passes through an anaerobic septic tank. The quantity so produced is not large and ordinarily little of i t escapes into the air. Even in open septic tanks the heavy membrane or scum which usually forms contains many oxidizing bacteria, which may effectually prevent t h e ' escape of hydrogen sulfid. I n the instances here discussed the amount of hydrogen sulfid evolved cannot be entirely accounted for by the decomposition of such protein substances, for the analyses of the raw sewage show it to be no higher in albuminoid ammonia, that is to say proteins and their decomposition products, than sewage from other disposal plants in which there is no such evolution of this gas.
565
The preceding table shows the amount of sulfur in the sewage calculated as SO, and the amount in the effluents, showing how it has been removed by reduction to hydrogen sulfid. The investigations of various European workers, Particularly BeyerinckI and van Delden,2have demonstrated that certain anaerobic bacteria in the presence of organic matter reduce sulfates to sulfids. Each of these investigators isolated specific organisms, which, when inoculated into sterile solutions containing sulfates and a little organic matter, brought about a very considerable evolution of hydrogen sulfid. I t was therefore thought possible that the large amount of this gas given off in the septic tank and the dosing
ANALYSESO F RAW SEWAGE.(Figures show parts per million.) Nitrogen Nitrogen as as free albuminoid Sample from ammonia. ammonia. Chlorides. 35.0 Hospital for inebriates, Knoxville, Iowa. . . . . . 21 .OO 6.50 ... Hospital for inebriates, Knoxville, Iowa. , , . , , 32.60 4.60 54.0 6 .OO Grinnell, Ia., city plant.. . . . . . . . . . . . . . . . . . . . 6.50 Grinnell, Ia., city plant. .................... 11.50 6.00 122.0 33.0 Knoxville, I a . , city plant'. . . . . . . . . . . . . . . . . . . 25 .OO 6 .OO Ames, Ia.,city plant'. ..................... 18.20 8.20 1 Excessive hydrogen suliid was not observed at these plants.
...
All other plants examined, where the water supply had a smaller sulfate content than that shown for the city of Knoxville in the next table, gave no excessive odor of hydrogen sulfid. Chemical analyses of the raw sewage and effluents were made to determine the amount of sulfates in solution, and the total sulfates in the raw sewage. These results indicate, first, that the organic matter in normal sewage does not contain enough sulfur to produce the quantities of hydrogen sulfid observed ; second, t h a t where the sulfates in the original water supply are high, there is a marked reduction in the amount of mineral sulfates found in the effluent, showing t h a t a portion of these mineral sulfates must have been reduced and removed from solution. I n three cases where excessive hydrogen sulfid was observed these relations were very marked. Several other septic tanks were observed where the sulfates in the water supply ranged from 16 to 143 parts per million as a t the Knoxville city plant, and in all such ANALYSESOF R AW SEWAGE.
Total
sulfur Source of sample. Knoxville, Ia ..... Knoxville, I a , . Knoxville, Ia. Knoxville, I a .
Inebriate Hospital, Inebriate Hospital, Inebriate Hospital, Inebriate Hospital, Grinnell, I a : ...................... Grinnell, I a . . ...................... Grinnell. I a . ....................... State Hospital, Mt. Pleasanc, I a . . . . . . State Hospital, Mt. Pleasant, I a . . . . . . State Hospital, 31t. Pleasant, I a . . . . . . Knoxville. I a . ..................... Knoxville, I a . . .................... Knoxville, l a . . ....................
.. .
.... ....
as SO4. Water supply 1664 Softened water I566 Raw sewage 1552 Tank effluent .. Water supply 476 Raw sewage 452 Tank effluent 381 W a t e r supply 900 Raw Sewage 871 Tank effluent 716 Water supply 143 R a w sewage 163 Tank effluent
Soluble sulfur as SO4. 1664 1566 1533 965 476 427 352 900 850 690 143 147 145
cases very little hydrogen sulfid could be detected a t the tank. The Knoxville city plant shows the highest sulfates in the raw sewage of any plant observed where excessive hydrogen sulfid did not appear, and i t is doubtful whether this amount of sulfates might not cause trouble in some cases.
Nitrites. none trace high
...
trace
...
Solids on Solids on Oxygen consumed. E:vaporation. ignition. 3040 2454 53.31 114.7 3902 2624 17.18 950 776 10.74 1114 894 63.60 800 538 ... 552 25 1
Nitrates. none trace none
...
none
...
chamber might be due in large part a t least to a reduction of the sulfates of the water supply. The chemical analyses of these waters show them to be exceptionally high in sulfates. Tests were accordingly made to determine whether or not such sulfates under the septic conditions of this plant would be transformed into sulfids. The following solutions were prepared in two liter lots, and j o o cc. placed in each of four liter flasks, the solutions in each case being covered by a layer of neutral paraffine oil. This furnished anaerobic conditions for bacterial growth, inasmuch as paraffine oil permits little or no oxygen t o penetrate it. After sterilization three of each set of the flasks were inoculated, one being retained as a check. The material used in inoculation was secured from different parts of the septic tank and the dosing chamber of the plant at the Inebriate Hospital a t Knoxville. After inoculation the flasks were kept in the laboratory a t a temperature of about 1 8 C.~ COMPOSITIONOF MEDIA USED. Sol. 1. 1000 cc. 30g. KzHPOa . . . . . . . . . . . 0 . 5 g . 1.og. hIgSO4.. . . . . . . . . . . . 8 . o g. NazSOI. . . . . . . . . . . . . . . FeS04. . . . . . . . . . . . . 0 . l g . Material.
Sol. 2. Sol. 3. Sol. 4. Sol. 5. Sol. 6. 1000 cc. 1000 cc. 1000 cc. 1000 cc. 1000 cc.
...
...
...
O.5g. Log. 8 .o g.
0.5g.
Log.
0.5g 1.og.
...
2.0g.
2.og. 0.lg.
4.0g.lO.Og. 0.lg. 0.lg.
0.lg.
...
0.lg.
...
...
...
...
0.5s. 0.5g. Log. Log.
...
The iron in each of the above solutions was added as a n indicator. The formation of sulfid was shown by the deposition of black iron sulfid. Within two or three days the solutions 2 , 3 , 4 , 5 and 6 , almost without exception, began to turn gray and within two weeks a n abundant precipitate of iron sulfid was visible in all. I t is evident, then, that bacteria having the power to reduce sulfates to sulfids mere present in numbers in the sewage in question. The quantity of sulfate daily entering the plant justifies Cent. f u r Baht.. 2nd abl., 1, 1. 49, 104 (1895)
Ibid.,11, 81, 113 (1903).
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y .
566
the conclusion that much of the hydrogen sulfid developed in the tank is formed b y the deoxidation of such sulfates. Stained mounts were prepared at intervals from each of the flasks described above. At first no bacterial type or organism appeared t o predominate ; large and small bacilli, cocci, and a few spirilla were present. At the end of the two weeks most of the flasks showed a decided predominance of spirilla. Those containing the highest percentage of hydrogen sulfid were practically pure cultures of this organism. A medium of the same composition as solution 5 described above, with the addition of 1.5 per cent. agar agar, was used for the purpose of isolation. Different dilutions were plated and the medium covered with paraffine to exclude oxygen. I n three or four days black specks appeared in the plates, the areas of discoloration often reaching I O mm. by the end of the week. Near the center of such areas could be observed small opaque colonies, generally considerably less than I mm. in diameter. These colonies were found t o consist of the characteristic spirillum noted above. When grown in test tubes containing solution 5 under anaerobic conditions, pure cultures were found to produce a rapid reduction of sulfates t o sulfids. I t may be concluded t h a t in the particular sewage of the Knoxville Hospital, this spirillum alone is capable of producing hydrogen sulfid in quantities sufficient t o account for the amount present in the dosing chamber and septic tank. Zelinsky,' in 1893, described a motile elongated bacillus (Bacillus Hydrosulfuroum ponticum) obtained at considerable depth from the ooze of the Black Sea. This organism under anaerobic conditions was found t o reduce sulfates and sulfites t o sulfids. Beyerinckz isolated a specific organism (Spirillum desulfuricans) from canal water. Inoculation of this organism into media containing sulfates resulted in the rapid development of sulfids. Morphologically the organism was from one-half t o a full spiral, rather rarely the latter. I n the absence of oxygen it is actively motile. I t is probable t h a t the organism isolated from the sewage disposal plant a t the Knoxville Hospital is the same as the one described by Beyerinck. Van Deldens made a further study of the S. desulfuricans of Beyerinck and also described the form Microspira aestuarii from near the mouth of one of t h e Holland rivers. The ooze at the bottom of the sea in such places was found t o be blackened b y sulfids. He made a n extensive study of the reduction of sulfur compounds by these organisms in both pure and mixed cultures. M. aestuarii he found t o be the more efficient of the two as a reducing agent, producing in the course of seven weeks as much as a gram of hydrogen sulfid per liter of water. Morphologically this organism resembles the former very closely. Examination of the dosing chambers of the tanks observed showed t h a t the moisture on the walls was strongly acid with sulfuric acid. A considerable porProc. Rus. Phys. Chem. Soc., Fasc. 5. cit., 1896.
* LOC.
LOC.
c i t . , 1903.
Aug.,
1912
tion of this acid was doubtless produced by atmospheric oxidation of the hydrogen sulfid, which of itself might account for the formation of the sulfuric acid and the sulfates found in the concrete. The bacterial study, however, shows another agent which aids materially in the oxidation of the hydrogen sulfid. The work of Winogradsky, Engelman, Miyoshi, and others has shown that certain groups of bacteria under aerobic conditions may utilize hydrogen sulfid b y direct or indirect oxidation as a source of energy. The first evidence of such oxidation is frequently the accumulation within the bacterial cell of granules of free sulfur. These may be recognized microscopically b y their relatively high refractive index, b y their solubility in carbon disulfid and by the characteristic crystals which form when such solutions are allowed t o dry upon a glass slide. These organisms are found later t o further oxidize this free sulfur, the final product being sulfuric acid. The organic food requirements of these bacteria seem to be very small; indeed, many apparently utilize the carbon dioxide of the atmosphere and very simple compounds of nitrogen. I n examination of the septic tanks in question one of these oxidizing sulfur organisms (Beggiatoa) was found growing abundantly in some places on the surface qf the scum. Disintegrating concrete, moreover, was covered by a luxuriant growth of moulds and bacteria. A comparison of the chemical analysis of sound concrete, and t h a t which has undergone partial disintegrations, hows a notable increase in the amount of sulfate present in the latter. It should be noted further t h a t the concrete below the water line of the septic tank and dosing chamber had disintegrated little or none a t all. I t seemed evident, therefore, t h a t hydrogen sulfid alone could not be the immediate cause of such disintegration, for it is constantly present in the sewage in contact with those walls. Inasmuch a s bacteria, however, seemed t o thrive in the moisture generally present on the under side of the roof and on the walls, it seemed probable t h a t much of the hydrogen sulfid arising from the sewage is absorbed and then converted into sulfuric acid through the agency of bacteria. ANALYSIS OF DISINTEGRATED CEMENT WALLIN SEPTIC TANKAT GRINNELL, IOWA,SHOWING VARIATION I N SULFUR CONTEXT O F DIFFERENT PARTSOF CONCRETE Free Sulfates sulfur. (SO4). First or outer' layer, '/4-inch thick.. . . . 19.7% 1.2% 2 1 .o Second layer, immediately beneath first. , 7.2 6 .75 Third layer, one inch below surface.. , , . 0 27
. .. . .. .. . .
One hundred cm. flasks containing various sulfids were inoculated with material from the disintegrated concrete and examined from time to time. In some of these flasks Beggiatoa filaments developed, showing their characteristic sulfur granules within the cells. However, active oxidation of the sulfids t o sulfates occurred in the other flasks in the absence of Beggiatoa, apparently through the agency of other bacteria which swarmed in such cultures. The following solution was prepared t o determine definitely whether or not such bacteria were actually 1
This layer consisted chiefly of a deposit of free sulfur.
,
Aug., 1 9 1 2
T H E J O U R S A L OF I - V D C S T R I A L 4.YD ELYGIaYEERINGC H E X I S T R Y .
capable of oxidizing sulfids to sulfates. Freshly precipitated, chemically pure iron sulfid and calcium carbonate were added in equal amounts in I O O cc. flasks. To each of these flasks was then added 5 cc. of Knop’s solution and after sterilization five were inoculated, the remainder being retained as checks. Within a week the inoculated solutions became cloudy, the uninoculated remaining clear. Stained amounts showed the presence of a capsulated rod showing polar staining in all inoculated flasks. This organism was present in all these flasks in practically pure cultures Comparative analyses of the contents of the check and of these flasks showed t h a t in those in which bacteria had been developed, between two and three times as much sulfate was present as in the check. The presence of this organism in the solution undoubtedly was the cause of the rapid oxidation of the sulfids which took place. However, the organism has not as yet been isolated in pure cultures and quantitative studies made of its oxidizing power. The data a t hand are sufficient to convince us that bacteria may be responsible, in part a t least, for the oxidation of sulfids to sulfates under the conditions as they exist in the sewage disposal plants here discussed. As rapidly as sulfuric acid is produced by bacteria working in and on the concrete or by atmospheric oxidation, it reacts with the calcium compounds of the cement, forming calcium sulfate. This takes up water of crystallization and swells considerably, the result being a cracking and pulverizing action and a consequent disintegration. ’ Careful search has failed to reveal any account of bacterial disintegration of concrete through the action of sulfids or sulfates. R. Grieg Smith1 investigated the disintegration of concrete in the open canals that convey the city water supply of Sidney, New South Wales. He was led to study the action of nitrifying bacteria on concrete through the work of Stutzer and Hartleb, who suggest that such forms might assist in the disintegration of the cement of water reservoirs. He concludes, however, that “there is considerable room for dou%t regarding the action of microijrganisms on cement. There is more reason to believe that the action is purely chemical.” His conclusions would seem to agree with those of Tannatt and Burke,2 who found disintegration of concrete tile to take place when iri contact with soil rich in sulfates. That even in alkali regions bacteria may have something t o do with concrete disintegration is evidenced from the work here recorded and the statement of Schreiner, Oswald and Failyers t o the effect that hydrogen sulfid is produced in the drainage waters from gypsum alkali lands and from iron pyrites. SUMMARY.
I n three of the sewage disposal plants observed, viz: Grinnell, State Inebriate Hospital, and State Hospital a t M t . Pleasant, large quantities of hydrogen sulfid are evolved. This sulfid is produced in two I.
Cent. f . Bakt., and A b f . , 8, 3 7 7 (1902). “The Effect of Alkali on Portland Cement,” Bull. Mont. Agr. Exp. Sta., 1908. 3 Bull. L?. S. Dept. 9gr., Bur. of Soils, 1906
567
ways: ( a ) by the bacterial decomposition of sulfur containing proteins and related compounds, and ( b ) by the reduction of sulfates which are contained in unusual amounts in the water supplies used. Of the two the second seems to be the more important. 2 . The organism chiefly responsible for the reduction of sulfates and evolution of hydrogen sulfid is a spiral bacterium closely related to, if not identical with, Spirillum desuljuricans. 3. The hydrogen sulfid which escapes as gas from the sewage, particularly in the dosing chamber, is partially dissolved in the moisture on the under side of the roof and concrete walls. Here i t is oxidized t o sulfuric acid partly by atmospheric oxidation and partly by bacterial action. The sulfuric acid acts upon the calcium compounds of the concrete forming calcium sulfate, and breaking down the concrete. There is often deposited an excess of free sulfur which the oxidizing agencies have not converted to sulfate. 4. The exact species of the organisms responsible . for the conversion of hydrogen sulfid into sulfuric acid have not been determined. That bacteria, however, in mixed cultures may bring aboub this change has been demonstrated. Beggiatoa and a bacillus have been isolated and are probably the most important organisms in bringing about this change. EXGINEERIXG EXPERIMENT STATION, STATECOLLEGE, AMES,IOWA.
MARKING PORCELAIN AND SILICA CRUCIBLES, ETC.l B y P. A. YODER. Received April 4, 1912. I N T R O D U CTI 0N
.
I n the analytical laboratory there often is occasion t o put upon crucibles, etc., permanent distinguishing marks which will stand ignition and mild treatment with acids and alkalis. The blue pencil or the brush and china paints which are frequently used for this purpose, the markipgs being burned in, often give results so crude as to be quite unsatisfactory, especially for articles that are before one’s eyes daily for years. Moreover, the blue-pencil marks and many china colors lack permanence, and when applied t o silica wares do not adhere satisfactorily. The writer has worked out two methods, one for marking with platinum and the other for the application of china paints, both by the use of rubber type. T H E PLATINUM PROCESS.
The crucibles are cleaned by heating for half a n hour with nitric acid, one part concentrated acid t o one part of water. A sizing is prepared consisting of a hot 5 per cent. solution of gelatin. The parts of the crucibles t o be marked are dipped into this sizing and set aside to drain and dry. When the gelatin is dry, the desired number is stamped on with a solution of platinic chlorid containing 1 2 to 1 5 per cent. of platinum-‘. e , about 32 t o 40 per cent. of the hydrated, crystallized chloroplatinic acid. The pad holding the solution may be made of six or eight folds of smooth linen or muslin and need not be much larger than the type used. This pad is nearly 1 Czrculuf 93, Bureau ol Chemistry, United States Department o Agriculture.
