Initial Reaction Rate between Phos - American Chemical Society

the authors described the batch mixing of phosphate rock and concentrated phos- .... were prepared by grinding small quantities in a ball mill und...
0 downloads 0 Views 613KB Size
Initial Reaction Rate between Phosphate Rock and Phosphoric Acid R. L. COPSON. R. H. NEWTON, AND J. D. LINDSAY’ Tennessee Valley Authority, Wilson Dam, Ala.

The reaction between Tennessee brown phosphate rock and concentrated phosphoric acid proceeds as follows: There is first a “fluid stage,” then a “plastic stage,” and finally a “dry stage.” A study was made of the chemical and physical factors affecting the length of the initial fluid stage, using a specially devised apparatus to measure the consistency of the reacting mass. The length of the fluid stage was decreased by increasing the acid concentration (up t o 85 per cent HsP04, where a minimum occurred), by increasing the

temperature, and by decreasing the particle size of the rock. The fluid stage was found to be longer for phosphate rock samples of higher P205 content. When a small portion of the finely ground rock was first dispersed in the acid, the fluid stage when the remainder of the rock was added t o the acid was very much shortened. I t was apparent that the phenomenon of the fluid stage and subsequent setting was due in part t o the chemical reaction and in part t o the absorption of the liquid by the solid phase.

I

The chemical composition of superphosphate prepared with concentrated phosphoric acid was studied by Newton and Copson ( 3 ) . No additional data on the composition of the product are given in the present paper. Subsequent to the initial work on batch mixing, an alternative process was developed (9) in which all of the mixing was accomplished in a high-speed mixer during the fluid stage, and the product was discharged while still fluid. This method was specially suited to a continuous manufacturing process in which the acid and rock were supplied continuously to the high-speed mixer, and the fluid product was discharged onto a conveyor on which it traveled until the dry stage was reached. This process is illustrated schematically in Figure 2. In addition to the advantages of a continuous process, a considerable saving in power is possible by this method since the mixture is not kneaded while passing through the plastic stage.

N A PREVIOUS paper (I) on the manufacture of con-

centrated superphosphate, the authors described the batch mixing of phosphate rock and concentrated phosphoric acid. The concentrated acid, containing 70 to 85 per cent H3P04, and the finely ground rock were charged into a powerful kneading machine and were mixed until the

1.5

n

c

d 3L

1.0

c

r: 0.5

0

0.5

1.0 TIME MINUTES

-

2.0

2.4

FIQURE 1. POWER CURVEFOR BATCHMIXINQ OF GROUND PHOSPHATE ROCKAND PHOSPHORIC ACID(78PERCENT&PO4) IN KNEADINQ MACHINE product was substantially dry. There existed initially a brief “fluid stage” during which the mixture was reasonably fluid and the power requirement was low. After a few seconds the mixture became plastic and the power required for mixing rose sharply. Finally the mixture became fairly dry, and the power curve passed through a maximum. With 76 to 78 per cent H3P04and finely ground rock (80 per cent through 200 mesh) the product after 2 to 3 minutes of mixing was dry enough to be handled on conveying equipment direct to the storage pile. Figure 1 gives a typical curve obtained with a wattmeter in the motor circuit during a >batchmixing operation ( 1 ) . Present address, University of Idaho, Moscow, Idaho.

175

176

INDUSTRIAL AND ENGINEERING CHEMISTRY

VOL. 29, NO.

2

proximately 0.125 inch (0.3 Continuous mixing of om.). this kind has been pracRotation of the container ticed p r e v i o u s l y in the was prevented by means of m a n u f a c t u r e of 16 per a wire; one end of the wire L E V E R FOR passed partly a r o u n d the cent superphosphate from ,RAISING M I X I N G periphery of the turntable phosphate rock and suland was fastened to it, and SHAFT SLIDES furic acid, usually in conVERTlCALLY I the other end was fastened junction with some sort of to a spring. The spring was f a s t e n e d to a clamp "traveling den" in which mounted on a h o r i z o n t a l the mixed product is carrod so that the tension of ried for some hours. When MIXING VESSEL the spring could be adjusted using concentrated phosreadily by slidin the clamp along the rod. 2 s the conphoric acid, both the time sistency of the mixture in in the mixer and the time the container increased, the on the conveyor are very torque transmitted to the much shortened, the forturntable tended to cause t h e spring t o stretch. By mer being only a matter means of a stop arranged on of seconds. The developthe wire, the s p r i n g was ment of a machine capable placed under an initial tenof doing this mixing job sion of 600 grams. A sharp deflection of the recording presented a real problem, mechanism t h u s was obbecause of the difficulty of C C O R D I N O DRUM tained when t h e mixture thoroughly mixing a solid reached a standard consistand a liquid in such a brief ency such that a force exceeding 600 grams was time, and the tendency of transmitted to the spring. the mixture to solidify and The recording mechanism accumulate in any part of EXPERIM E N T A L A PPA R A T US consisted of a chart on a the mixer where it was not revolving drum which rotated so-as to give a chart mechanicallv removed. speed of 0.2 inch (0.5 cm.) er second, and a pen mounted on the I n the present paper a study is reported of those factors wire connecting the turntaEle and the spring. The type of record which determined the length of time during which a mixture obtained by this device is illustrated in Figure 4. The form of of ground phosphate rock and concentrated phosphoric acid these curves corresponds to the first part of the power gra h shown remained sufficiently fluid to be handled in a high-speed in Figure 1, but the upward break in the curve is mucx sharper owing t o the construction and sensitivity of the apparatus. continuous mixer. The length of the fluid stage is obviously related to the rate of the chemical reaction between phosphate A series of experiments made to determine the effect of rock and phosphoric acid, but it is affected as well by physical rate of stirring showed that the length of the fluid stage deconditions such as the concentration of the acid and the creased as the speed of the mixer was increased, but that this particle size of the rock. For example, data are presented effect was small. I n all experiments reported in this paper, which show that with concentrated acid the chemical rethe mixer was operated a t 420 r. p. m. action apparently continues for some time after the mixture has ceased to be fluid, whereas with sufficiently dilute acid In performing an experiment, a weighed amount of acid of the desired concentration was placed in the container and brought the fluid stage may be greatly prolonged even though the to the correct temperature by adjusting the temperature of the chemical reaction may go substantially to completion. For water bath. The mixing blades were lowered into the acid, the purpose of assisting in the design of a continuous mixing and 100 grams of rock added. This amount of rock was used apparatus, it was desired to investigate the rate of change of as standard in all experiments. The addition of rock required less than 1 second, and mixing appeared to be complete 2 seconds the physical condition of the mixture, rather than the chemiafter starting the experiment. The end of the fluid stage was cal reaction rate. marked by the deflection of the recorder pen, as shown in Figure 4. In long experiments, hot or cold water was added to the Apparatus and Procedure water bath in order to maintain an approximately constant temperature. A drawing of the apparatus is shown in Figure 3: The procedure outlined gave results reproducible with The a paratus consisted of a vertical mixer and a device to less than 3 per cent error in all cases except where a mixture record tge toraue transmitted from the rotatinn blades to the container. Thebrass container, of two rock samples of widely 4 inches (10.2 cm.) in diameter differing p a r t i c l e size was and 6 inches (15.2 cm.) deep, used. I n the latter case exwas mounted on a turntable treme care had to be exera n d vertical s h a f t which revolved in ball bearings. The cised to ensure satisfactory container was water-jacketed. uniformity of the rock samCo-axially with the container ples. The majority of the there was mounted a shaft that data represent averages of could be raised or lowered while revolving. The shaft was rotwo or three closely checking tated by a 0.5-horsepower values. motor, through a chain drive I n a d d i t i o n to measureand a sliding sprocket. On the ment of the fluid time as lower end of the shaft were three pairs of blades 0.5 inch o u t l i n e d , some measure(1.27 om.) wide, placed one I 2 m e n t s of the time for the 3 1 2 3 above the other, with a horiproduct to dry and to lose 0 5 0 15 0 5 0 IS zontal angle of 60" between its plastic p r o p e r t i e s comeach air. Clearance between the bfades and the cup was app l e t e l y were made. Since Y

FEBRUARY, 1937

INDUSTRIAL AND ENGINEERING CHEMISTRY

this point was not as definite as the end of the fluid stage and since it was determined by visual examination only, the results were not as exact.

Materials The phosphoric acid used was obtained from the T. V. A. acid plant at Wilson Dam. It was allowed to settle in a lead-lined tank, and the clear acid was decanted. The concentration was adjusted by adding water until the desired specific gravity reading was obtained. 4000

given graphically in Figure 5. The range of acid concentration covered was 60 to 90 per cent H~POI. The length of the fluid stage decreased with increasing acid concentration until a minimum was reached between 80 and 85 per cent HaPO4. The fluid time decreased rapidly with increasing acid temperatures up to 150" F. (66' C.), and thereafter more slowly. TABLE11. ilcid Concn.,

I

I

177

&PO,

EFFECTO F ACID STRENGTH

--

50' F.

(loo C.)

AND TEMPERATURE"

Fluid Stage at Initial Acid Temp. of:-looo F. 150° F. 200' F. 250' F. (38O C.) (66" C.) (93' C.) (121' C.) Seconds

7

72 20 S 6 5 86 52 12 6 4 4 so 41 12 6 4 4 75 57 16 8 5 6 70 81 25 16 11 65 197 81 60 60 .. 60 4080 940 420 .. .. Average composition of rock, 29.4% PnOs,39.0% CaO; average particle diameter, 53 microns. 90

.. ..

10 8

Effect of Particle Size of Rock

I6I0

I

Both the average particle size and the particle size distribution affected the length of the fluid stage. Data showing the former effect in a closely sized material are given in Table IIIA; data for the length of the plastic stage are given in B . These data are plotted in Figures 6 and 7, which show

I

I

70

80

90

P E R C E N T H3W4

F I G U R E 5. EFFECTO F -4CID CONCENTRATION AND TEMPERATURE ON LENGTH OF FLUID

STAGE

Tennessee brown phosphate rock was used throughout this work. Analyses of the samples used in studying the effect of chemical composition are given in Table I. These were prepared by grinding small quantities in a ball mill under as nearly identical conditions as possible. The fineness of the ground samples varied from 75 to 100 per cent through 100 mesh and from 40 to 80 per cent through 200 mesh. For other rock samples not included in Table I, the PzOh and CaO contents are given subsequent tables.

400

3 00

0 0

0

8

200

0 Y

I

COMPOSITION OF PHOSPHATE ROCK TABLEI. CHEMICAL SAMPLES~ Sample

PnOr

No.

%

1 2 3 4 5 6

13.4 22.7 25.2 25.5 28.9 32.1

CaO

%

Fez03

%

BlnO:

%

Si02

%

Fb %

17.4 6.6 7.2 38.4 ... 28.3 4.6 9.1 23.1 2.8 35.5 3.0 2.0 26.3 2.7 34.0 6.1 5.7 19.7 2.7 39.0 5.7 2.8 14.3 3.4 44.7 6.7 0.2 7.8 3.8 33 5. 48.7 4.5 2.8 6.3 3.7 ._ ~. a All samples were dried in an oven at l05O C. for 48 hours before using; analyses were made after drying. b Willard and Winter method (4).

I n studying the effect of particle size, various closely sized samples of practically identical chemical composition were used. Only the average particle diameter is recorded in the tables. For instance, the first sample listed in Table I11 entirely passed a 20-mesh and was retained on a 35-mesh screen (Tyler standard screens). Its average particle diameter is given as the average of these screen openings, or 630 microns. Each sample was t,horoughly mixed to ensure uniformity The quantity of acid used in all experiments was 98 to 100 per cent of that theoretically required to react with 100 grams of rock. It was calculated from the CaO and PzOb content of the rock, as the additional PtOa required to convert all of the CaO to monocalcium phosphate (1).

100

Table I1 summarizes the data illustrating the effect of strength and temperature of the acid. The results are also

800

700

FIGURE 6. EFFECT OF PARTICLE SIZEOF CLOSELY SIZEDSAMPLES ON LENGTH OF FLUID STAQE

that for closely sized materials the fluid stage was proportional to the average particle diameter. Since for a given weight of rock sample the "specific surface" is inversely proportional to the average diameter of the particles, it was concluded that the fluid stage was inversely proportional to TABLE111. EFFECTOF PARTICLE SIZE OF CLOSELYSIZED SAMPLES ------AQ-Av. particle d.iam., microne 630 335 215 165 125

Fluid stage Rec With With'

Ifs!i%

2%

-

Av. particle djam.. mcrom 190 125 90 63 23

Bb

Fluid stage. sec. 44.0 32.6 20.3 15.0 10.0

-

End of plastic stage, min. 35 12 5 0.33 0.25

380 295 203 185 128 99 103 83 81 59 85 52 39 38c 20 19 Average composition of rock, 32.3% PtOs, 45.5% CaO; temperature o acid looo F b hverape'compositjon of rock, 27.8% PzO,,37.9% CaO; acid concentration, 80% &POI; acid temperature, 100' F. Estimated. Q

Effect of Acid Concentration and Temperature

200 M O 400 500 AVERAGE PARTICLE OIAHETER-MICRONS

INDUSTRIAL AND ENGINEERING CHEMISTRY

178

4o

5

20

$

FIGURE8.

EFFECTOF VARYING PRO-

VOL. 29, NO. 2

first, allowing it to become dispersed, and then adding the remainder of the rock to this slurry. The data illustrating this phenomenon are given in Table VI. A further series of experiments (Table VII) showed that t,he particle size of the dispersed rock was of great importance, the fine materials having a much greater accelerating effect. The length of time the dispersion stood before use was found not to affect the length of the fluid stage.

Mechanism

Correlation of these observations with PARTICLE DIAMETER -MICRONS ROCK MIXTURES t h e c h e m i c a l reaction rate was atOF PARTICLE SIZE FIGURE 7. EFFECT tempted by observing the temperature of OF CLOSELY SIZED SAMPLES the mixtures when mixing was carried out under approximately adiabatic conditions in a cylindrical copper vessel supthe specific surface. In Figure 7 , however, the time reported in a Dewar flask. Temperature-time curves for a series quired to reach the end of the plastic stage is shown to be a of acid strengths are shown in Figure 9. These experiments function of a higher power (approximately 2.0) of the averwere made with rock which contained 28.1 per cent P20s age particle diameter. and 38.9 per cent I n Table IV are given data for mixtures of two samples of CaO, and had an phosphate rock of widely differing particle size. A plot of a v e r a g e particle the data (Figure 8) shows that the effect of adding the fine size of 53 microns. material was greater than its proportion would indicate. This T h e i n i t i a l temeffect was more pronounced on the plastic stage than on the p e r a t u r e of t h e fluid stage. These data led to the conchsion that the acid was 77" F. fluid stage and the plastic stage can be varied almost inde(25" C.). pendently by choice of the average particle size and relative Comparison of proportion of both fine and course material. This was of imFigure 9 with portance in operation of the continuous mixer described Figure 5 showed earlier in this paper. t h a t for d i l u t e acids (60 per cent OF PROPORTION OF FINE ROCK^ TABLEI v . EFFECT HaP04) the time End of required to reach Fine Coarse Fluid Plafltic Rock Rock Staae Staae a maximum tem- o - BOo Mi;. SeO. % % perature was less TIME- MINUTES 34.4 35 0 100 than the length of 29.1 20 10 90 the fluid t FIGURE 9. TEMPERATURE-TIME CURVES 26.8 ... 80 20 4 whereas with SHOWING EFFECT OF ACIDSTRENGTH 24.4 30 70 21.8 40 ..* 60 19.3 1.5 50 50 more concentrated 17.5 40 ... 60 15.3 80 20 acid (80 per cent H3P01)the mixture had ceased to be fluid 0:Zb 14.1 100 0 long before the maximum temperature was reached. a Average oomposition of coarse rock 28 4% PgOs 39.07 CaO- average v t i c l e diameter 180 microns: averahe dompositirh of h e r d k 28.1% These observations led to the conclusion that the termiCad: average particle diameter, 53 microns: aoid cohoentra2 0 , 41 6 nation of the fluid stage was largely due to physical phetion,'80% %PO&: acid temperature, 120° F. nomena, such as adsorption. The colloidal nature of the rockTABLEV. EFFECT OF CHEMICAL COMPOSITION OF ROCKQ PoRT1oNS OF

Sample No.

PzOi

% 1 2 3 4

13.4 22.7 25.2 25.5

Fluid Stage See. 0.6 2.8 3.0 5.0

Sample

No.

PrOc %

6 6 7 CadPOdz

28.9 32.1 33.6 45.9

AND

Fluid Stage Sec. 3.4 4.3 13.0 18.1

(c. P.)

a Acid ooncentration, 80% HsPO4; acid temperature, 100'

F.

Effect of Chemical Composition of Rock A series of runs was made to determine the effect on the composition of the rock, and the results are summarized in Table V. The fluid stage decreased with decrease of PZOs content of the rock sample, or conversely with increase in impurities which were principally silica and alumina. In the course of this investigation it was found that finely divided phosphate rock formed a very stable suspension or dispersion in concentrated phosphoric acid when the amount of rock thus dispersed was less than one-fifth that required to react with the acid. It was found that the fluid stage was very much shortened by adding a small portion of the rock

IN

TABLEVI. EFFECT OF PHOSPHATE ROCKDISPERSED IN ACIDQ Rock Dispersed

Fluid Rock Fluid Stage Dispersed Stage % SeC. % See. 0 11.5 5.8 10 5 8.2 1.9 15 Average composition of rock 28.10/, PnOs 39.07 CaO. average particle diameter, 53 microns; acid honcentration: 80% fLP01:' acid temperature 100' F. Q

OF PARTICLE SIZEOF ROCKDISPERSED IX TABLEVII. EFFECT ACIDQ

Av. A v. Particle Particle Diam. of Diam. of Fluid Fluid Dispersed Dispersed Stage Stage Rock Rock Microns SeC. Microns Sec. 180 12.6 49 4.6 125 11.2 23b 3.0 104 7.8 19b 2.0 62 5.0 Sb 2.0 a 10% of rock dispersed; average composition of rock 28.1% PaOa, 39.0%. CaO: acid concentfation, SO7 HaPO4; acid temperature, 100' F. b Determined microscopicayly.

FEBRUARY, 1937

INDUSTRIAL AND ENGINEERING CHEMISTRY

acid dispersions and the great effect of fine materials further strengthened this conclusion.

Conclusions 1. The length of the fluid stage, in the reaction between concentrated phosphoric acid and Tennessee brown phosphate rock, decreases with increase in initial acid temperature in the range 50-250O F. (10-121' C.). It also decreases with increase in acid concentration in the range between 60 and 80 per cent H3PO4. Between 80 and 85 per cent ?&PO4, a minimum fluid stage is observed. 2. The length of the fluid stage is proportional to the average particle diameter in closely sized rock samples, and the time required for the mixture to become dry is approximately proportional to the square of the average particle diameter. I n rock samples with two predominant ranges of particle sizes-i. e. mixtures of fine and coarse materialsthe fine material causes a shortening of the fluid stage slightly more than its proportion would indicate, and also causes a much more pronounced shortening of the plastic stage.

179

3. The fluid stage is longer for rock samples of high P,Os content than for lower grade materials. 4. The length of the fluid stage, while affected by the rate of the chemical reaction between the phosphoric acid and the constituents of the phosphate rock, depends primarily upon physical phenomena, such as the absorption of the liquid by the solid phase.

Acknowledgment The authors wish to acknowledge the valuable aid of other members of the T. V. A. chemical engineering staff in obtaining the data presented in this paper. Literature Cited (1) Copson, Newton, and Lindsay, IND. Ewo. CHEM., 28, 923 (1936). (2) Curtis, Chem. & Met. Eng., 42, 488 (1935). 28,1182 (1936). (3) N e w t o n and Copson, IND.E N G .CHEM., (4) Willard and Winter, IND.E N G .CHEM., Anal. E d . , 5, 7 (1933). RECEIVED September 23, 1936. Presented before the Division of Fertiliaer Chemistry at the 92nd Meeting of the American Chemical Society, Pittp. burgh, Pa., September 7 to 11. 1936.

VAPOR PRESSURE OF

Commercial High-Boiling Vapor pressure curves have been determined for eleven samples of commercial highboiling solvents, covering the region20' to 150" C. Thedetermined values are in agreement with data given in the chemical literature, in the few cases where such data are available.

Organic Solvents GEORGE S. GARDNER AND J. EDWARD BREWER Brewer & Gerdner, Consulting Chemists, Philadelphia, Pa.

'I'

HE chemical literature contains little vapor pressure data on organic solvents. Therefore the purpose of this investigation was to obtain sufficient data to construct vapor pressure-temperature curves in the region 20" to 150" C. for eleven commercial high-boiling organic solvents. The samples of solvents used were not specially purified compounds but were technical products as actually used in the chemical industries. Materials The samples of solvents upon which the vapor pressure measurements were made were obtained from the following sources : Carbitol diethylene gl col. monqethyl etier), Butyl d r b i t o l (diethylene glycol monobutyl ether) Dimethyl phthalate, dibutyl phthalate Terpineol, terpenyl acetate, benzyl alcohol benzyl acetate Hexalin 6oyclohkxanol), Tetralip' (tetrahyronaphthdene), Decalin (decahydronaphthalene)

Carbide and Carbon Chemioals Corporation Commercial Solvents Corporation

E. I. du Pont de Nemours & Co.

Apparatus and Methods The procedure followed in determination of the vapor pressure was that originally described by Ramsey and Young (6) and more recently by Young (9), in which the boiling point is obtained at various pressures. It is particularly important that a large insulated bottle be inserted in the system in order to obtain more constant pressure readings. Pressures below 50 mm. were read with a mercury vacuum TABLEI. PHYSICAL COXSTANTS OF SOLVENTS -Mean B. P.Temp. Pressure -Density-a C. Mm. d Carbitol 196.0 763 1.023 Butyl Carbitol 232.1 763 0.957 Dimethyl phthalate 283.8 763 1.188 Dibutyl phthalate 340.7 763 1.047 Terpineol 221.1 763 0.936 Terpenyl acetate 220-231" 759 0.961 206.9 763 1.045 217.0 763 1.065 162.3 763 0.945 210.5 763 0.978 Decalin 193.8 763 0.884 0 Slight decomposition.

On each sample of solvent a sufficient number of physical it::$ constants was obtained to establish its identity definitely. ; : ; ; : $ I n all cases the mean boiling point a t atmospheric pressure, the density, and the refractive index were determined. These constants are given in Table I.

Substance

A:",",':

C. 20/20 20/20 25/25 20/20 20/4 20/4 20/4 20/4 20/4 20/4 20/4

Refractive Index nD c. 1.4242 26 1.4258 27 1.6138 20 1.4900 20 1 4 88 25 1:4!50 20 1 . 6 20 20 1.5200 20 1.4602 20.7 1.5439 20 l.ai(68 20