A Method for the Manufacture of Phosphate Fertilizer Citation Ball, Alpheus M. and Fahs, John Louis (1926) A Method for the Manufacture of Phosphate Fertilizer. Bachelor's thesis, California Institute of Technology. doi:10.7907/2X6M-P768. https://resolver.caltech.edu/CaltechETD:etd-12092004-110048 Abstract No abstract. Item Type: Thesis (Bachelor's thesis) Subject Keywords: Chemical Engineering Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemical Engineering Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: None, None Defense Date: 1 January 1926 Record Number: CaltechETD:etd-12092004-110048 Persistent URL: https://resolver.caltech.edu/CaltechETD:etd-12092004-110048 DOI: 10.7907/2X6M-P768 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 4892 Collection: CaltechTHESIS Deposited By: Imported from ETD-db Deposited On: 10 Dec 2004 Last Modified: 16 Aug 2023 22:03 Thesis Files Preview PDF (Ball_am_1926.pdf) - Final Version See Usage Policy. 1MB Repository Staff Only: item control page

A Ma?THOD FOR THE MAUUPACTURS OF PHOSPHATES FERTILIZER Thesis by Alpheus WM. Ball John L. Fahs in partial fulfillment of the requirements for the degree of Hachelor of Science in Chemical Engineering. CALIPURVIA INSTIDUTS UF VYECHNOLOCY Pasadena, California 1926 A SULFITE PROCESS FOR THE MANUFACTURE OF PHOSPHATE FERTILIZER. Among the foods that must be supplied to plants by fertilization of the soil one of the most important is phosphate. Manure and commercial fertilizers supply phosphate to some degree, but the standard method for introducing phosphates into the soil in quantity is by the use of superphosphate. This product is made by treating phosphate rock, which consists chiefly of the - insoluble calcium phosphate, Ca, (PO,) with concentrated 2’ sulfuric acid. Superphosphate is a mixture of CaHPO,, Ca(H,PO,)o, CaSO,, some free phosphoric acid and some 4 uncomverted calaium phosphate. In addition to these substances superphosphate always contrins all of the impurities that were present in the original rock and acid. The chief advantage of superphosphate is its low cost of production. The large producers of suvper- "phosphate are often smelters of sulfide ores and get their sulfuric acid as a by-product of the smelting industry. Phosphate beds also frequently lie close to the smelting plants so that transportation of the raw materials to the superphosphate works is no great problan. Manufacturing costs are low, for the process eonsists essentially of grinding the rock, mixing it with sulfuric acid, allowing it to stand in a brick lined pit until it stops reacting and cools, then pulverizing it and sacking it for market. | The disadvantages of the superphosphate are (1) that it is subject to reversion and (2) that it contains a lot of inactive material that runs up the freight bill. These disadvantages are done away with by a "double superphosphate” method used in Europe. A quantity of phosphate rock is decomposed with enough dilute sul- furic acid to set free sll of the phosphate as HPO, and to 34 precipitete all of the calcium as CaSO,. The material is filtered and the phosphoric acid solution is concentrated by evaporating in lead pans. Ground phosphate rock in the proper quantity to form monocaleium phosphate is added to the phosphoric acid. The fertilizer obtained in this way contains no gypsum or other sulfate and is more concentrated that superphosphate. A step further toward concentration is the introduction of a cation with fertilizing value in plece of the calcium in combination with the phosphate. This has been done by the American Cyandmide Company in their product "Ammo-Phos", which contains twenty percent of ammonium and twenty percent of phosphate. “The object of the present research is to investigate the possibility of duplicating "Ammo-Phos" or of making a product of a similar nature according to the idea of Mr. Henry Blumenberg, as covered by several U. &. patents held by him. Mr. Blomenberg's idea was to treat a mixture of a solution of ammonium sulfate, calcium nitrate and ground phosphate rock with sulfur dioxide and an excess -of air, precipitating the calcium as Caso, and obtaining a solution of ammonium phosphate. After filtering, the remaining SO, was to be driven off by heating, the solution cooled and neutralized with ammonia to seperete out iron, aluminum snd any remaining calcium, then evaporated until the ammonium phosphate crystallized out on eooling. METHODS OF TESTING The solid matter in the treaters was tested at intervals for phosphate content. ‘When any treater was discharged a test for citric acid soluble phosphate also was made, as eitric acid soluble phosphate is classed commercially as soluble. The difference between the two values thus obtained is the net insoluble phosphate spoken of below. Samples of 500 ce were drawn from the treaters and filtered by suction. The precipitate was dried on "porous plates upon a sand bath until all the water had been removed, then it was powdered and allowed to cool in @ desicgator. 5.000 grams were weighed out and placed in a Kjeldahl flask with a small quantity of a 50-50 mixture of concentrated nitric and sulfuric acids. The flask was carefully heated so that the charge did not regurgitate, with the addition of more of the mixed acid 4f necessary, until the solid matter in the bottom of the flask was white, indicating that the digestion was gomplete. The contents of the flask were cooled and ailutea to 250 ec in a volumetric flask. 10 ce of this diiuted solution were pipetted into a clean beaker, and diluted to about 100 ce. 20 ce each of concentrated nitrie acid and ammonia were added and the solution was heated to 70°C. 10-40 ce of ammonium molybdate solution were added, according to the amount of phosphate believed to be present, and the solution was allowed to stand on the water bath for about half an hour at 60°. ‘the precipitate was filtered at and washed with a hot 1% ‘solution of KNO, until it was free of acid. The pre- Gipitate was removed to a beaker and a 0.2 normal we! solution of NaOH was run in until the precipitate had dissolved. The excess of NaOH was titrated back with 0.2 normal HCL. For the citric acid soluble test, 5.000 grams of the powdered solid were placed in a beaker and about 300 ce of a 2% citric acid solution were added. The whole was heated at 60° for an hour, then filtered and diluted to 500 cc in a volumetric flask. 50 ce were pipetted out and treated as above to determine the phosphate content. The HCl used was 0.2082 normal ani the Na0# 0.2015. by standardizing against Na HPO,, it was found that 1 ee of the base was equivalent to 0.306% of phos~ phate by the first test. APPARATUS ms 5 mT BLES r Ie ry 1 + hy 1 i Fig L. The decomposition apparatus used (see fig. 1) was designed by Mr. Blumenberg. It consisted of three iead-lined cylindrical iron chambers with conical bottoms (@) and ome wooden chamber (D) equipped with a stirring device. Whe lead-lined chambers had a diameter of 20 inches and a height of 24. The wooden one was 14 by 48. The sulfur burner (B) used to generate SO, was a cast iron chamber lined With a refractory meterial and designed to be fed from the top with roll sulfur through a double valve mechanism. Air was pumped through the apparatus by a Nash Hytor pump (A) of size $C driven by a 2 hp Century single phase motor, which also operated the stirring mechanism of the #4 treater. DECOMPOSITION Air compressed by the pump was sent through the sulfur burner, where part of it was converted to 5SQ,, then into the first treeter. he gases were sent down a lead pipe nearly to the bottom of the cone, then allowed to bubble up through the liguor and out through &@ pipe leading to the second, third and fourth treaters “in Series. The waste gases were allowed to escape after passing the fourth treater. In each chember was a charge ‘gonsisting originally of ten gallons of water end six pounds each of ore, ammonium sulfate and calcium nitrate. This charge was recommended by tir. Blumenberg as the one that had given the best results in his preliminary work. The treatment was carried on until the amount of net wo Solids. zs in o Per cent > Fyds loo-F 60 Time in hours. Fic 2. TypicaL Decomeosition Curve. ccs. of sludge Hours. Fic. 3. Setttinc Curve For Ca504 SLupGE. o diusoluble phosphate in the charge was brought tom to 1% of the total insoluble matter. This was found to yequire about forty hours {see fig. 2). ‘In order to -seoure uniform treatment and obtain maximum absorption efficiency for the apparatus used, each new charge was started in #4 treater and moved forward one step every ten hours. Sulfur was burned at the rate of one pound per hour, hence the sulfur consumption amounted to ten pounds per charge. | The following table shows more clearly what goes into the charge: From; (pound equivalents) Ore (28% P, 0,) Ca. so NH PO 26 0.071 4 4 o:d71 Ammonium sulfate 0.091 0.091 Caleium nitrate 0.053 Although the actual reactions involved in the process were not investigated in this work, they may be postulated to be as follows: LL The ammonium sulphate and calcium nitrate, both being soluble, react metathetically and precipitate calcium sulphate: (NH,), so 4 * Ca(NO,), = 2 MH, NO, + CaSO, As is evident from the above table, there is an excess of ammonium sulphate above the amount necessary to complete this reaction. 2. The sulfur dioxide, dissolving in the water to form gulfurous acid, attacks the phosphate rock and brings it into solution: Cag (PO,) 5 + 6H,S0g = 3Ca(HSO,) » 3. The excess of ammonium sulfate precipitates the ealcium thus dissolved: Ca(HS0,), + (NH), 80, = auy, HSO,, + Caso, There is, however, not enough sulfate added to react with all of the calcium of the ore in addition to the geloium of the Ca(NO,)., must remain in solution as long as the solution is so thet part of the calcium kept saturated with 50,° If the caleium nitrate were not added, enough sulphate would be present in the charge to more than take care of the calcium from the ore; and sinee the nitrate ion apparently plays no part in the process and is not recovered in the final product, it would seem thet the calcium nitrate might well be “omitted from the charge. Mr. Blumenberg states, however, that previous work indicated that when it is omitted, the rate of decomposition is materially lowered. %x- periments to cheek this point whould be worth while. The solution obtained from the digestion, there- fore, may be considered to be a mixture of ammonium, galeinn, nitrate, phosphate, sulfite and sulfate ions, together with aluminum and iron, and a trace of vana~ dius, coming from the impurities of the ore. From this solution it is desired to obtain monammonium phosphate. At this point it might be pointed out that, if oxidation of the SO, to sulfate be disregarded, all of the calcium of the phosphate rock. is precipitated by sulfate coming from sulfuric acid (since the ammonium sulphate is prepared from sulfuric acid), and the only fumetion of the SO, is to acidify the solution. Therefore 2 the alleged economy of the process, apart from all other eonsiderations, is to be questioned on this one score alone. FILTRATI OW Thée next step after decomposition is the filtration of the solid material. It was thought that the filtration could be made easier by allowing the solids to settle and decanting the supernatant liquid. As an experiment, a trough was constructed 3 inches high, 4 inches wide and 24 inehes long, with cross-pieces hslf an inch high placed every 4 inches along the bottom. The suspension from the treaters was made to flow along this trough, but With the slowest stream obtainable no appregciable settling took place. To determine more aceurately the rate 10 ef settling, four 100 ce eylinders were filled with the sludge and allowed to stand for several hours. Readings were taken of the top of the sludge layer - at intervals to prepare the accompanying plot (fig.3). ‘In the actual filtering work an eight ineh gweetland filter leaf operating on a vacuum of 15 inches of mercury wes used. Five runs were made using aifferent cloths and treatments: (1) Standard Sweetland filter cloth. (2) Six ounee canvas, close weave. (3) Same as (2), seams waxed. (4) Same as (1), seams waxed, precoated with filter-cel. (5) Same as (4), but no filter-cel. The rate of eake formation was found to vary only with the cloth used; that is, (1), (4), and (5) agreed on one rate, and (2) and (3) agreed on another very similer rate. Runs (1) and (2) gave mudéy filtrates, (3), (4), and (5) gave clear ones. For (5), however, on continued operation the filtrate be- came mud dy after about five gallons had been filtered. The same would probably hold true for (3) but the change would not come so quickly. (4) ig then the only really satisfactory procedure. REMOVAL UF 50,- During the vacuum filtration, as might be expected, the solution lost 505. resulting in an acidity kness (inches) ene Oia Ee eR Cake « ‘Itering Time (minutes) 4 6 to te z e A. CAS04 SLUDGE, With STANDARD SWEETLAND FILTER CLOTH. tehness ¢ inches) tering Time (minutes) 2 4 6 8 10 42 B. CASQg4 SLUDGE, with 6-02. CANVAS. —— Filtering Time (minutes). Cake thickness Cinches) 1 2 C. MEUTRALIZED SLUDGE, witn 6-0z. CANVAS. Fic. 4. FirtrRATioNn Curves. il deorease from 0.219 N. to 0.178 N. It was thought that this acidity might be cut down much further by . heating, and a test rum was made to determine the point. ‘Ten gallons of filtered solution were heated ix en enamel lined steam-jacketed kettle. As no power was available 40 operate the mechanical stirrer, the top of the kettle was removed to permit operation as an open kettle. When the temperature of the liquid reached about 60°, there began to form a white precipitate which did not redisgolve upon cooling. This was un~ aoubtedly calcium sulfate, which is present at sat- uration for the hydrated form (gypsum) in the original solution, and which partially preseipitates as the much less soluble anhydrous form at that temperature (see fig. 54). This phenomenon was exhibited whenever the . s0lution was heated or evaporated; and since the an-~ hydrite does not readily revert to the more soluble hydrate, but remains as an insoluble precipitate even after remaining for a month in contact with solution at room temperature, it proved to be a serious factor in subsequent evaporation and crystallization. fhe heating was extended over a period of two hours. Samples of liquid were taken at fifteen-~ 3 “Milliequivalents per Liter- A. SoLusivity or CaSQ,. (Melcher, J-A.C.S. 32,50; 1908). Temperature Centigrade . $ 2 Per cant NHgHz POs — 4o 80 tao 160 200 B Sovusiity of NHgH2, PO, (Buchanan + Winner, J-L*E.C.,12,450;1920). Temperature Cé@ntigrade 20 40 60 80 190 Fis. 5. piaute intervals and titrated for acidity against WaOH, using methyl orange, with the following results: Table l. time. Temp. Vol.of sol. Acidity Total equivalents oe ef acid present (hours). (°C) (liters) (normality) Before filtra- pion 22 87,9 0.219 8.30 After filtra- $ion 22 37.9 £178 6.74 0.25 69 37.9 .162 6.13 0.75 98 36.4 £170 6.17 1.00 98 34.8 172 5.98 1.25 98 32.9 178 5.87 1.50 98 31.4 £185 5.83 1.75 98 29.9 £191 5.72 2.00 98 28.0 203 5.68 It ean be seen that it is impossible to reduce the amount of acid appreciably below a certain anount by this treatment. The acidity of the solution igs due therefore not to SO, but rather to nonvolatile acids, probably sulfurie or phosphoric. It is obviously - dmefficient to heat, cool, then reheat the solution ~ 13 $e evaporate it. Likewise it is useless to attempt to neutralize the solution with ammonia while it is hot. The economieal thing to do is to eliminate either the first heating or the neutralization. | The solubility of monanmmonium phosvhate was investigated {See fig. 5b.). On account of the steep- mess of the slope it is possible to obtain crystals by boiling down to saturation at the boiling point and then allowing the solution to cool. Samples of the partislly concentrated liquor ebteined above were taken. One was neutralized with eoncentrated ammonia, tlie other was not. These two samples were boiled down to a volume, 30 ce, approaching saturation in NH, Hy PO, filtered hot and allowed to erystallize. The results of this experiment are show in the following table: A B Val. sample 500 ee 300 eo Vol. ammonia meee ne 1l ce Evap. to 30 ca 30 ce wt. erystals 13.1335¢ 3.596¢ wt. for analysis - 5035 - 5000 ee standard base 4.33 12.85 Purfty® basis of NH HPO 21.4% 63.6% 4 2° °4 When the ammonia was added a rather heavy Sluminum - iron precipitate came down. As is seen from the above table it is necessary to remove this in order to obtain crystals containing much phosphate. Conseqguent- 14 Ly it seems best to eliminate the first heating, and to neutralize the filtrate from the first filtration. The ‘peutralized solution is also easier to filter when hot peeause there is not so much precipitated matter to renove and the solution seems less viscous. This leaves the process in a state where a filtration is made, a reagent is added to the filtrate and then another filtration is made immediately. If the sludge from the treater could be neutralized be- fore filtering one operation of the filtering apparatus eould be avoided. The liquor filtered would be neutral or alkaline instead of acid, so that less expensive filtering apparatus could be used. The one disadvantage of this procedure would be that more ammonia would have to be used to take eare of the 50, that otherwise would be taken off by the vacuum pump of any suction filter that should be used in the unneutralized sludge. This igs not such a disadvantage, however, because the de- preeiation of the suction apparatus would be much less in the absenee of this corrosive gas. Ten galloms of sludge out of the treaters were meutralized with 2.3 liters of concentrated ammonia and filtered with the No. 4 filter setup. ‘The filtrate was “15 | olear, put it was obvious that the filtration “gyole must pe short, for the rate of flow through ¢he cake decreased very rapidly with the time. CRYSTALLIZATION - Mime was not available to investigate erystallization on a larger seale. the only erystallization work done was that in connection with the neutralization problem above. YIELD The filtered, unneutralized liquor tested 4.39% P,0,- From the solubility curve of mon- ammonium phosphate it can be calculated that the maximum yield in one crop of crystals is about 1.2 pounds of monammonium phosphate from each batch treated. The initial substances going into the process are: phosphate rock 6 lb. ammonium phosphate 6 1b. ealcium nitrate 6 lb. sulfur 10 ib. ammonium hydroxide, conc. 5 lb. To the cost of these materials must be added the operation, depreciation and interest in an ex- pensive plant, consisting of at least four treaters, & sulfur burner, filter presses, evaporation system, erystal driers, ete. Hven assuming that the calcium 16 nitrate is omitted, thet the sulfur consumption 4g reduced to a tenth of its present amount and that the ammonia is recoverable (necessitating an ammonia recovery plant), the process does not seem practicable. Labor in such a plant would have to be skilled to ea greater extent than in most fer- tilizer plants, as most of the apparatus is specialized and requires an intimate knowledge for its efficient operation. CONCLUSIONS: The method suggested by Mr. Blumenberg can be improved by: (a) The use of more efficient treaters, which will reduce sulfur consumption. (b) The omission of calcium nitrate from the original charge. | (eo) Neutralization of the charge before filter- ing, omitting the heating to remove sulfur dioxide. Even With these improvements, it seems inevitable that the process eannot hope to compete vith other methods of treating phosphate rock. SUMMARY: A new method for the treatment of phosphate rock has been investigated in its most important 17 phases (digestion, neutralization, filtration and yield). ‘he major difficulties in each of these steps have been considered with the result that certein changes in the process have been recommended. Further work on the problem, if undertaken, should inelude studies of the problems of evapora- tion and crystallization, as well as the feasibil- ity of recovering the small smount of vanadium present in the oré. the investigators acknowledge the cordial cooperation of The Stockholders! Syndicate of Ios Angeles, who instigated the research and pro- vided the apparatus, and of Dr. 7. U. Lacey of California institute of Yechnology, under whose direction the work was performed.