INDUSTRIAL AND ENGINEERING CHEMISTRY
1128
VOL. 32, NO. 8
The several products were TABLEXIII. EFFECTO F W E T PRESSROLLMANIPULATION ON THE SWELLING PROPERTIES characterized by sheet swellOF A LOW-VISCOSITY HARDWOOD SULFITE P U L P I N 18 P E R C E N T SODIUM HYDROXIDE SOLUTION ing in 18 per centcausticsoluAT
Sample No. A-1 A-2 A-3 B-1 B-2 B-3
Basis Weight 298 255 207 144 145 148
Density,
20”
c.
G./Co.
% WI
%S
% VI
0.451 0.462 0.500 0.256 0.335 0.483
448 445 446 594 537 460
145 150 184 66 95 171
117
% SH 6.5
121 141 45 72 132
86 .. 05 7.5 6.5 8.5
press rolls t h a t are ordinarily located before the dryer, a single set of experiments was made with a base fiber that was sheeted on a commercial Fourdrinier machine. The pulp was of hardwood sulfite origin and mas characterized by a somewhat lower cuprammonium solution viscosity than is common. Conditions of fabrication on the wire were held reasonably constant and changes were effected (a) by altering the basis weight and holding substantially uniform compactness, and ( b ) by holding a uniform basis weight and altering the compactness by means of the wet press roll pressures.
tion a t 20’ C. Results are given in Table XI11 and Figure 8. % SH % Voids Cc./G. S.0 10.5 5.5 I n general, a change in basis 0 weight while holding density 56 .. 72 63 .. 0 181 .. 92 6.0 1.4 6 5 fairly constant has only a 5.8 0.7 5.8 slight effect on swelling and 6.0 1.6 5.3 weight increase. On t h e other hand. the sheets that were compacted while wet and then dried behave very much like the dry compacted products described heretofore. I n other words, density of dry sheet does not materially influence the thickness-swelling volume. ThicknessSwelling Vol.,
Literature Cited (1) Jayme, G., and Steinmann, R.. Papier-Fabr., 35,337-60T (1937). (2) Richter, G. A., and Glidden, K. E., ISD. ENG.CHEM.,32, 480-6. (1940).
Composition and Properties of Superphosphate Estimation of Acid H. L. MARSHALL AND W. L. HILL
Requirements and Correlation of Experimental Results
Bureau of Agricultural Chemistry and Engineering, U. S. Department of Agriculture, Washington, D. C.
Suitable reference systems for the graphical comparison of the distributions of phosphorus among the principal forms in widely different superphosphates are discussed. The comparison is based primarily on the degree of acidulation, which is expressed quantitatively by a coefficient of acidulation or corrected acid-rock ratio, and takes into account the acid consumed by the aluminum and iron as well as the acid value of the fluorine in the superphosphate. A procedure is outlined for the interpolation of the acid requirements of phosphate rock, which will be useful to superphosphate manufacturers interested in the greatest possible acid economy. The range of the degree of acidulation used in commercial practice is estimated from plant data and from the analyses of commercially prepared superphosphates.
iX THE course of the bureau’s investigations on the com-
I
position of superphosphate (3-6) a large amount of data has been accumulated. An attempt to correlate the results disclosed the need for (a) a reference system and the definition of the variables required to determine the major changes that may be expected to occur in a particular superphosphate and (b) a careful consideration of scales, or criteria, for measuring the degree of acidulation as a basis for the comparison of the properties of superphosphates. Some of the more important relations obtained in the theoretical study are discussed in this paper; the experimental results will be presented in a later article. The conclusions are based on previous work, on results that will appear in a later paper, and on known phase relations. The terms “ingredient phosphate” and “ingredient acid” signify the phosphatic material (for example, phosphate rock or bone ash) and acid, respectively, used in making up the superphosphate and refer to these ingredients in the form in which they go into the mix. I n discussions of the composition of superphosphate the term “raw” phosphate is used in a somewhat restricted sense to refer to the constituent phosphate compounds-for example, fluorapatite-of the ingredient phosphate. This simple distinction seems preferable to the common practice in superphosphate parlance whereby one. term carries both connotations.
AUGUST, 1940
INDUSTRIAL A N D ENGINEERING CHEMISTRY
Acidulation of Tricalcium Phosphate
REACTIONS. If tricalcium phosphate is agitated with aqueous phosphoric acid of the proper concentration, the ensuing reaction for 2 and 4 moles of acid may be represented by the following equations, in which the water of crystallization is disregarded :
+ +
+ 2CaHP01
Caa(POa)z 2H3P01 = Ca(HZP01)z Ca3(P04)2 4H3PO4= 3Ca(HzPO&
(1A) (1B)
If the ingredient acid is aqueous sulfuric acid, the initial reaction may be regarded as: Ca,(PO,),
+ 3H2S04= 3CaSOA + 2H3PO4
(2)
If the amount of tricalcium phosphate exceeds the proportion expressed in Equation 2 , the initial reaction will be followed by reaction between tricalcium phosphate and the liberated phosphoric acid. When the mole ratios of tricalcium phosphate to sulfuric acid are 1:1 and 1:2 , respectively, the corresponding equations, which are analogous to Equations 1A and l B , are:
+ +
+ 2CaHP04 + Ca(H2P04)z
Ca3(P0& H2SO4 = CaS04 Ca3(P04)z 2HZSO4= 2CaSO4
(3-4)
(3B)
I n either case the extent of the reaction a t equilibrium under atmospheric pressure depends upon the proportion of the ingredients and, for the range over which mono- and dicalcium phosphate are coexistent, upon the quantity of free water in the system and the temperature. Since the solubility of calcium sulfate is so small (6) in the presence of solid monocalcium phosphate that the sulfate content of the solution phase can be disregarded, the system obtained with the use of sulfuric acid for this range in composition can be treated as though it were the three-component system CaOP20sH20. SCALE OF ACIDULATIOS. I n the foregoing reactions water of constitution of the acid appears as water of constitution in the salts, and further consideration of this form of water is avoided by expressing the ingredient acid in terms of anhydrous Hap04 or H2SOd, instead of the corresponding oxides. -4s a consequence of the near equality between the molecular weights of and HzS04,the weight ratios H3P04/P205 and H2S04/P20sare, for purposes of discussion, regarded as identical and designated by h. Thus, in accordance with Equations 1-4, l B , 3A, and 3B, 2h and h (0.6905) are the factors for the conversion of P20sof the ingredient phosphate into H3POaand H2S04equivalents, respectively. Other quantities in parts per 100 parts of tricalcium phosphate to be considered are: S = Hap04 equivalent of total calcium as monocalcium phosphate, or more generally, of total metallic elements as monobasic phosphates s = HaPo4 equivalent of ingredient acid (plus that of any p
=
sulfate in ingredient phosphate) PzO6 in ingredient phosphate
Since the Hap04 equivalent of p is 2hp, S = 3 X 2hp = 6hp (4) A scale of acidulation, a, is given by the ratio of the H3P04 equivalent of the total free and combined acid in the system t o that of the total calcium as monocalcium phosphate; i. e., a=--s
+ 2hp s
s
--
+ 2hp 6hP
(5)
which in many respects is the simplest scale. -4closely related scale, u,which is adopted here, is given by the ratio of s to the acid required for complete acidulation-that is, to make a = 1-thus: u = s '4hp
(6)
1129
The two scales are compared in Figure 1. The numerical values of these ratios express the acidulation quantitatively, and they may be considered as coefficients of acidulation. When the ingredient phosphate is phosphate rock, u represents a corrected acid-rock ratio.
Distribution of Phosphorus in Equilibrium Systems SYSTEMS DEVOIDOF FREEW S T E R . For equilibrium systems derived from the reaction of tricalcium phosphate having the theoretical composition with c. P. sulfuric acid and the subsequent removal of free water, the fractions of the phosphorus present as monocalcium phosphate, dicalcium are determined phosphate, p2, and free phosphoric acid, by the degree of acidulation. I n view of Equations 1, 2 , and 3 , the distribution for certain values of u may be written a t once by inspection and with slightly greater difficulty also general expressions in terms of u (Table I) ; for example, in the lower range u = 0.5 to 1 pl = ( u - 0.5)/0.5 = 2u - 1 52 = 1 - 5 1 = 2(1 - u )
F,,
TABLE I. SYSTEMS
DISTRIBUTION O F PHOSPHORUS I N EQUILIBRIUM (DEVOID O F FREE WATER) OBTAINED FROM TRICALCIUM PHOSPH.4TE AND SULFURIC ACID XIonocalcium
U
Phosphate.
Dicalcium Phosphate,
Pl
Pl
Free Phosphoric Acid, p f
0 5 0 1 0 0 75 0.5 0 5 0 1 1 0 0 1 5 0 0 la 0 5tol 2u 1 2u0- U ) 0 1to15 3 - 2u 2(u 1) a The calcium In solution ia considered as calclum sulfate.
-
-
For systems derived from tricalcium phosphate and phosphoric acid, the calculations are complicated by the condition that the total phosphorus increases with u. If, however, the phosphorus of the ingredient tricalcium phosphate and that of the ingredient acid are regarded as separate entities throughout the discussion, this system can be treated in much the same way as the foregoing systems. The distinction is easily preserved with the aid of a dilution factor calculated from a balance of materials or, if the acid is substantially free from calcium, from the calcium contents of the ingredient phosphate and the product. At u = 0.5 the monocalcium phosphate in the system is equivalent to the ingredient acid, and the dicalcium phosphate is equivalent to the ingredient phosphate (Table 11); and if for higher values of u the dicalcium phosphate is regarded as coming from the tricalcium phosphate, the two types of systems have the same values of and gZ,respectively, on the interval u = 0.5 to 1. For still higher values of u the two types of systems diverge as a consequence of the fact that in this range monocalcium phosphate is consumed by sulfuric acid, whereas its amount is unaltered in similar systems obtained with phosphoric acid. On the other hand, if in the latter systems the free phosphoric acid (although it must be considered a part of the ingredient acid rather than a part of the ingredient phosphate) is expressed as a fraction of the ingredient phosphate, the values of 5,are numerically the same in both types of system. SYSTEL~Y CONTAISIXG FREEWATER. Monocalcium phosphate in contact with wvstter decomposes into dicalcium phosphate and free phosphoric acid and yields a solution having the composition of the liquid phase a t the equilibrium point, monocalcium phosphate-dicalcium phosphate-solution, which depends only upon the temperature. The equilibrium values of the ratio of free H3P04to free water, k , a t several temperatures are given in Table 111.
INDUSTRIAL AND ENGINEERING CHEMISTRY
1130
TABLE 11. DISTRIBUTION O F PHOSPHORUS IN EQCILIBRIUY SYSTEMS (DEVOID OF FREEWATER)OBTAINED FROM TRICALCIUM PHOSPHATE AND PHOSPHORIC ACID
HaPo,
u 0.50
0.75 1.00 1.50
Equivalent of Total Phosphorus 4hp 5hp 6hP 8hp
Fraction gredient Fraction of Ingredient phate Acid as: hlonohlonoD,icalcalcium calcium cium phosphosphosFree phat?, phate phate IIIPOI jjl 1 0 0 0 1 0 0 0.5 1 0 0 1 0.67 0 0.33 1
Fraction of Total Phosphorus as: hlonoDjcalcalcium cium phosphoaphate phate 0.5 0.5 0.8 0.2 1 0 0.75 0
of InPhosas: Dicalcium
phosphate. pn 1
:.5 0
TABLE 111. RATIOOF FREEH,PO, T O FREEWATERAT THE EQUILIBRIUM POIST,Ca(H,P04)z.H20-CaHP04-H20a
Temp.,
c.
Concn. of Aqueous &PO4 a t Equilibrium Point,
I; = (fr-)b
REFEREXCE SYSTEMS. I n an equilibrium system a t 25” C. obtained by acidulating waterfree tricalcium phosphate under realizable conditions such that the monocalcium phosphate, but not the dicalcium phosphate or the calcium sulfate, is hydrated and that no water is lost from the system, the total water (exclusive of water of constitution) and also the free water (that is, the difference between the total water and the water of crystallization of the monocalcium phosphate) are expressible in terms of u and the concentration of the ingredient acid. Thus, the free water, in parts per 100 parts of tricalcium phosphate, put into the mixture is given by the equation:
Concn. of Aquesui H3P01
a t Equi-
where c
=
T:mp,
“/o
free n a t e r 0.211 0.241 0.305 0.381
25 17.4 70 0.553 35.6 30 19.4 80 0.661 39.5 YO 0.782 43.9 40 23.4 50 27.6 100 0.926 4s. 1 60 0.462 31.6 a T h e ratios were calculated from revised solubility data reported b y Elmore and Farr (a) b Free &PO4 and free water are defined (6) as H3P01 and Ha0 in excess of t h e Ca(HnP0a)n H20 equivalent of t h e calcium in solution.
-
e
=
wj = cu - e P i p water, in parts per 100 parts Ca3(P04)2, added with H2S04 at 100 per cent acidulation weight ratio of water of crystallization t o PZOj in Ca(HiP04)2.H20
By substituting this relation into Equations 7 and 8 and then solving for fi, and 51,respectively, we obtain ??, =
2(1-2)
If a quantity of water is introduced into a dry system (Tables I and 11)-for example, one in which u = 1-the quantity of free &Po4 a t equilibrium depends on the quantity of free water, w,,and is given by kt+, or more conveniently by kw,,where a quantity reckoned on the basis of the product of acidulation is indicated by a primed symbol. (If the amount of water of crystallization is not altered, the free water is equal to the added water.) Then, if the dilution of the ingredient phosphate arising from the addition of acid and the change in weight during mixing, etc., is designated by R , in general, 1 kwl p, = -(7) R 2hp and since the free Hap04 represents only half of the phosphorus of the decomposed monocalcium phosphate, -
p l = 2u - 1
- -kw’ L RhP
I n systems obtained with the use of sulfuric acid Rp = p ’ , and with this simplification the distribution of the phosphorus a t equilibrium for the lower range of acidulation (corresponding with the range 0.5 to 1.0 in Table I) is given by the set of equations consisting of 7 , 8, and 9: pz = 1
- ir, - P I
(9)
The extremities of the useful parts of the straight lines represented by these equations are marked by the acidulation a t which monocalcium phosphate disappears (with decreasing u)-that is, when F,. 6 2 = 1-and that a t which dicalcium phosphate disappears (with increasing u)-that is, when & 51 = 1. For the analogous systems obtained with the use of phosphoric acid, the distribution of the phosphorus from the ingredient phosphate lies between mono- and dicalcium phosphates and is given by Equations 8 and 10:
+
-
pz
+
=
1
- p1
(10)
The free H3P04required for equilibrium is, however, given by Equation 7.
VOL. 32, h’0. 8
[($-T)u+f] (11A)
For systems obtained with the use of 66.63 per cent (53” Be.) sulfuric acid and Ca3(P0J2,the constants in these equations are: c = 48.15, e = 0.1267, h = 0.6905, IC = 0.211 a t 25” C. (Table 111), and p = 45.79. With these numerical values of the constants, Equations 1 l A and 12-4 reduce to:
3,. pi
=
=
+
0 . 1 2 7 ~ 0.020 1 . 7 4 6 ~- 1.040
(11B) (12B)
The distribution of phosphorus among the three constituent compounds is shown graphically in Figure 1. G’H’ and A’E’ are segments of the curves represented by Equations 11B and 12B, respectively, whereas F’B‘ is the corresponding curve for dicalcium phosphate. A’ and F’ fall a t the acidulation a t which monocalcium phosphate appears. Dicalcium phosphate disappears a t the acidulation corresponding to B‘, H‘, and E’. An increase in u beyond the latter acidulation merely liberates phosphoric acid from monocalcium phosphate until a t some higher acidulation (represented schematically by C in Figure 1) this salt ceases t o exist as a solid phase. Under the assumed conditions, A’E’C, F’B’, and G‘H’D show the equilibrium distribution of phosphorus among monocalcium phosphate, dicalcium phosphate, and free phosphoric acid, respectively. For systems devoid of free water, the corresponding curves are AEC, F B , and B D , respectively. In the analogous series of Ca3(P04)2-H3P04systemsthat is, systems having the same values of u and w,’/Rp as the Ca3(P0,)2-H2S04systems, the curves for monocalcium phosphate and dicalcium phosphate are A’E”C” and F“B”, respectively. G’H“D“ is the curve for the free acid, which in this type of system is to be regarded as a part of the ingredient acid. With free water in the system the monocalcium phosphate fraction of the phosphorus attains (with increasing u) its highest value when u is greater than 1 (100 per cent), and this value of u depends upon the ratio w,‘/Rp. This particular
AUGUST, 1940
IIVDUSTRIAL AND ENGINEERING CHEMISTRY
value of u, designated by U,,,,marks the intersection of the two branches of the curves for monocalcium phosphate ( E ' , Figure 1) and free acid ( H ' ) , respectively. The properties of the product of acidulation should therefore differ, depending upon whether u is greater or less than Urn. For example, on the upper interval of acidulation p:/w; is greater than k , and the amount of monocalcium phosphate is independent of the free water content of the system, whereas this is not true on the lower interval.
1131
fate in the bone ash. The acidulation may be stated in two ways :
sc -- so so - net acidulation on basis of Ca,(PO,),
u=--
2 , = - = S C
gross acidulation
(15) (16)
Division of numerator and denominator of Equation 15 by S,, substitution from Equation 16, and insertion of 8, for S,/S, give an equation connecting u and v,
Acidulation of Bone Ash Complete acidulation of bone ash, in which magnesium and the alkalies are to be considered in addition to calcium and phosphorus, requires that the total acid in the system shall be the equivalent of the metallic elements as the monobasic phosphates, S . Since the metals and phosphorus in bone ash are not present in tricalcium phosphate proportions (that is, S is greater than 6hp) Equation 4 is replaced by
S
= 6hp
+ So
(13)
where S, = acid equivalent o f excess of metallic elements over tricalcium phosphate proportion. Furthermore, neglecting for the moment the sulfate content of the bone ash, the acid equivalent, S,, demanded for complete acidulation is given by
S, = 4 h p
+ So
(14)
Then the ingredient acid required for complete acidulation is the difference between S,and the acid equivalent of the sul-
The two scales of acidulation are identical for tricalcium phosphate, whereas for bone ash the u scale represents a stretching of the u scale from 1, or 100 per cent, where u = v. The divergence of the two quantities is shown by a rearrangement of Equation 17: v - u =
(1 - u)v,
(18)
Superphosphates made from bone ash are properly compared with one another and with the tricalcium phosphate reference system on the basis of u. Since vo is zero for tricalcium phosphate or any material in which the metallic elements and phosphorus occur in the tricalcium phosphate proportion, and increases as the amount of metallic elements relative to the phosphorus increases, the numerical value of this quantity for a particular case gives the extent of departure of the phosphatic material from tricalcium phosphate. In other words, vo is the fraction of the ingredient acid needed for complete acidulation (disregarding the aciduphosphatic material) that is lation value of the sulfate in the . . required by the excess of metallic elements over tricalcium phosphate proportions.
Acidulation of Phosphate Rock
aZ'O.667
lOO,U=
1 50
I I
!
60
I
I.&
l.000
70
1
80
30
100
110
/ZO
I30
I40
I
150'
A cidula?;on
FIGURE 1. EQUILIBRIUM DISTRIBUTION OF THE PHOSPHORUS OF THE INGREDIENT PHOSPHATE IN REFERESCE SYSTEMS AT 25" C. OBTAINEDB Y MIXIKG TRICALCIUM PHOSPHATE WITH SULFURIC OR PHOSPHORIC ACID System containing free water in a definite proportion t o the acidulation is represented by solid lines. system devoid of free water with the exception of monocalcium phosphate curves E'C a n d E"C" a n d the free phosphbric acid curve H ' D . which represent both types of system in the corresponding range of acidulation, is represented by broken lines -- -. A'E'C, F'B', and G'H'D show t h e fractions of the phosphorus as monocalcium phosphate dicalcium phosphate and free phosphoric acid respectively, in systems obtained with tril calcium phosphate a d d 66.63 per cent sulfuric acid. A'E"C" F"B" and GtH;DD"tt h e fractions a8 the same components. respectively, in ans'logous sys'tems dbtained with trrcalcium phosphate a n d phosphoric acid.
I n order to use the curves in Figure 1 as a basis for the comparison of complex superphosphate systems containing fluorine, aluminum, and iron, it is necessary to know the acid equivalent of these elements as they exist in the superphosphate. Although detailed knowledge of the reactions would be highly desirable, such information is not essential. It is sufficient to know the compounds formed by these elements in superphosphate. Whereas the natures of the fluorine and R103 compounds of superphosphate are not well established, available data make possible a tentative choice of equivalence factors that lend a measure of consistency t o the results for a wide variety of superphosphates. The problem is made difficult by the condition that superphosphate, particularly fresh material, is subject to continual change and departs to a greater or less extent from equilibrium relations, a matter that requires home elaboration a t this point. If hydroxyapatite, with the appropriate amount of acid to satisfy the acidulation requirement of the apatite in accordance with Equations 14 and 17, had been used instead of tricalcium phosphate in the reference systems already discussed, the character of the resulting mixtures would have been the same. Thus, mixtures of hydroxyapatite and phosphoric acid prepared and brought to equilibrium at 25' C. in a closed system (to avoid loss of water) would, depending upon t h e
INDUSTRIAL AND ENGINEERING CHEMISTRY
1132 loo
t
Conversion of Raw Phosphate
-
I
~ / o oMonocalcium 3 Phosphate
!
E
i /’\
~
I
, 0
75
I 1 8 8 ’ 1 85 95 105 115 Resultant Acidulation, ,u + , u ~ I . per cent
/25
FIGCRE 2. DISTRIBUTION OF PHOSPHORUS IN ORDINARY SUPERPHOSPHATE PREPARED IN THE
LABORATORY
OlSuperphosphate (90 days old) prepared by the authors from Florida land pebble (No. 912) and 66.6 per cent sulfuric acid; lightly-drawn curves are from Figure 1. acidulation and the concentration of the ingredient acid, consist of monocalcium phosphate monohydrate and solution, or of monocalcium phosphate monohydrate, anhydrous dicalcium phosphate, and solution. The use of sulfuric acid introduces calcium sulfate as a solid phase, but as already pointed out, i t has a negligible effect on the composition of the liquid phase in the composition range involved in superphosphate manufacture. In all these cases the chemical reaction between the tricalcium phosphate or the apatite and phosphoric acid, either introduced directly or liberated with sulfuric acid, involves two distinct processes. Accordingly, dissolution of the apatite is the primary process, which without mechanical hindrance will go to completion, and with a sufficient amount of liquid phase no other process will occur. Precipitation of one or the other of the products is a secondary process, although it may occur simultaneously with the primary process. Thus, it is not permissible to look upon Equations 1A and 1B or their apatite analogs as representing reversible chemical reactions. I n mixtures simulating the superphosphate type, in which the volume of liquid phase relative to that of the solid phases is small, some apatite may persist for long periods as a result of surface coatings of mono- or dicalcium phosphate. The presence of this slowly disappearing, residual raw phosphate, however, need not prevent a close approach to equilibrium conditions as regards the precipitation of salt from the liquid phase. I n other words, a practical state of equilibrium as regards the secondary process does not necessitate a state of completion of the primary process but merely requires a sufficient difference between the rate a t which the two processes occur. The existence of the liquid phase in a thin layer in intimate contact with the precipitated salt favors a rapid attainment of equilibrium between them, whereas the dissolution of raw phosphate is hindered by surface coatings, which may’ sometimes remove it entirely from contact with the liquid phase. Thus, after a preparation has stood for a time under constant external conditions, the departure of the mixture from ultimate equilibrium will be determined largely by the
VOL. 32. NO. 8
completeness of the dissolution of the apatite or, in other words, by the proportion of residual raw phosphate. If water is allowed to escape from the system it may, and in practice often does, happen t h a t a mixture formulated to yield a final product containing mono- and dicalcium phosphates and prepared in a n open vessel turns out t o consist of monocalcium phosphate, apatite, and a vanishingly small amount of liquid phase. This type of product results from the removal of water from the sphere of reaction a t such a rate that the composition of the liquid phase-that is, its free acidwater ratio-never falls below the critical value (Table 111) for the precipitation of dicalcium phosphate. Thus, monocalcium phosphate continues to separate from the liquid until the amount of the latter becomes too small to support further reaction at an appreciable rate. Such a material, which is obviously not in a state of equilibrium, would fall in the dicalcium phosphate field of Figure 1 without actually containing this salt. As a convenient means of approach to a discussion of superphosphate mixtures, fluorapatite and hence phosphate rock will be assumed to behave with respect t o phosphoric acid or sulfuric acid in the same manner as hydroxyapatite and tricalcium phosphate. The points of difference arising chiefly from the additional components, fluorine and R203, may then be derived from the experimental data. The principal differences between the calculations for phosphate rock and the materials already considered have to do with the allowances that should be made for the presence of elements, aluminum and iron. Since a fluorine and the R203 part of the fluorine of the rock is volatilized during the manufacture of superphosphate and this element is also often a constituent of phosphoric acid, a proper allowance for its effect on the acid requirements presupposes knowledge of the fluorine content of the superphosphate. Accordingly, it is advantageous to neglect the fluorine entirely and calculate the imposed acidulation, u or v, which is the proper basis for studying factors that affect the volatilization of water and fluorine during the mixing operations, and later to apply the proper correction, uF’or vF’, for the fluorine in the superphosup‘ or v phate to obtain the resultant acidulation, u up*,of which u up’is the correct scale for studying the changes that occur after the mixing operation. Resultant acidulation does not take into account the completeness of the reactions.
+
+
TABLEIV.
EQUIVALENCE FACTORSFOR CALCULATION OF H3PO4REQUIREMESTS FOR ACIDULATION
Constituent Expressed as: C30 ZIgO
HaPOi/Constituent Weight Mole factor factor 2 2
2 KZO 2 A1110 2 hl2Os 2 a Arerage apparent factor superphosphates (Table V ) . Xa30
+
3.496 4.861 3.162 2 081 2.763 1.923
Constituent Expressed as: FenOa
PZOS
HsPOI/Constituent Mole Weight factor factor 2 1,227 2 1.381
Fz
SOa HzSOI
2 2
2 448 1 998
determined b y a balance of constituents in 7
in acidulation with sulfuric IMPOSED ~ ~ C I D U L A T I O ~Since, . acid to prepare ordinary superphosphate, the final stage of the reaction is substantially the same as in acidulation with phosphoric acid, and since the sulfate unquestionably combines with calcium, i t is logical to calculate the phosphoric acid requirements for the desired acidulation value and then obtain the sulfuric acid requirement with the aid of the appropriate factor. Equivalence factors for the elements considered in this investigation are given in Table IV. Evidence will be presented in a later paper to show that the iron and
AUGUST, 1940
INDUSTRIAL AND ENGINEERING CHEMISTRY
1133
TABLEV. BALAXCEOF CONSTITUENTS IS SUPERPHOSPHATE
NO. ES7 ESG J
4 3 2 1 a c
d e
(In moles per 10,000 grams of superphosphate) 4 5 6 7 8 9 Positive Radicals hIgO Imposed Citrate- Total Minus CaO KzO Superphosphate A c i d u l a t i o n , Insol. Free-Acid mmus NarO h I a d e from: V P~OP pno6 SOa Rz01 .\In0 Total % % Moles Moles Moles Moles Moles 10.10 1.46 0.34 8.30 7.43 0.10 105.5 Tenn. brown rock No. 908 11.19 0.79 0.30 10.10 8.20 0.00 106.3 Fla. land pebble No. 912 11.38 0 . 7 7 0 . 3 1 10.30 9.08 0.04 106.9 Fla. land pebble No. 912 11.87 0.80 0.32 10.76 9.96 0.03 101.5 Fla. land pebble S o . 912 15.23 0 . 8 6 0 . 3 2 14.05 12.50 0.00 96.2 Fla. land pebble No. 912 0.33 17.92 16.60 0.99 14.67 0.33 90.9 Fla. land pebble No. 912 16.89 1 . 0 0 0 . 3 3 l 5 . 5 6 e 13,946 1 . 5 5 85.6 Fla. land pebble No. 912
2
1
3
+++
Per cent of superphosphate. b Reault in column 9 minus t h a t in column 5. Result in column 10 divided by t h a t in column 11. Caloulated from analytical data obtained on superphosphate a t 30 days of age. Result corrected for citrate-insoluble portion o i constituent; thus PiOs 1.09, CnO 4.74, and
aluminum more or less leisurely go into compounds of the type CaO.RnO3,2P2O5.zH20. From this condition it is tentat,irely concluded that the t,heoretical acid requirement for these two elements is given by the ratio 2H3P04:R205. RE~ULTAST ACIDCLATIOS.The fluorine content of superphosphate can hardly be assigned to a single compound. Some of the possible compounds are apatite (residual raw phosphate) , CaF2, CaSiFG, and Ca(P03F)2. I n comparison with the Pz05in monobasic phosphates, the acidulation effect or neutralizing value per mole of fluorine is 1 in CaF2, l / 3 in CaSiF6, and 0 in Ca(P03Fj2,so that the observable average or apparent Pz05equivalent, of the fluorine in a superphosphate containing two or more of these compounds would be less than 1. With the use of the probable P205-R203equivalence indicated in the preceding section, the PzOs equivalent of the fluorine can in some superphosphates be estimated by a balance of the bound constituents shown by careful analyses. I n order to avoid troublesome corrections for residual raw phosphate, which cannot be applied a t all unless the relative proportions of calcium, phosphorus, and fluorine in the citrate-insoluble fraction of the sample are substantially the same as in the ingredient rock, it is necessary to choose superphosphates that carry negligible amounts of raw phosphate as indicated by a very low figure for citrate-insoluble phosphorus. A balance of constituents for seven superphosphates prepared in the laboratory is given in Table V. Other basic constituents not considered in the calculation amounted to less than 0.1 per cent of the superphosphate, and silica was neglected. Complete d a t a are shown for the materials a t 90 days of age. The difference between the total positive radicals (column 9) and the bound PzOj(column 5) is shown in column 10. This difference divided by the bound fluorine (column 11) gives a ratio (column 12) that expresses the equivalence of fluorine to PZOSin the material a t 90 days of age. For comparison, the ratio calculated in the same way from analyses a t 30 days is shown in column 13. The averages of the ratios for the seven superphosphates a t the two ages are 0.58 and 0.64, respectively. Accordingly, the average apparent Pz05 equivalent of the fluorine in these materials is given implicitly by the relation FP= 0.6 Pz05from which the factors given for fluorine in Table IV were calculated. The equivalence ratios show greater variations with the superphosphate than with age, which may mean that t’heacidulation efTect of the fluorine is markedly affected by the conditions of manufacture. On the ot,her hand, the ratios estimated for three commercially prepared superphosphates, which for lack of determined alkali figures were omitted from Table V, fall within the limits shown. Thus, it would appear that the equivalence factor derived from the d a t a in Table V can be tentatively adopted as a working basis
F2 0.42
10
PtOd Moles 2.67 2.99 2.30 1.91 2.73 3.25 2.95
12
11
Total E~~~~~ hiinus over Free-Acid
F2
13
Equivalence Ratio for superphosphate a t : 90daysc 30daysd
Moles 4.00 4.39 4.16 4.71 5.02 5.08 5.14e Average
0.67 0.68 0.55 0.41 0.54 0.61 0.57 0.58
0.71 0.74 0.58 0.55 0.58
0.74 0.62 0.64
mole, respectively.
for calculating the acid value of the fluorine content of superphosphate. Accordingly, the correction to be added to the imposed acidulation on account of the fluorine content of the superphosphate is given by the equation,
I n order to illustrate the method of comparing superphosphate data with the aid of the reference system (Figure 1j and to show the measure of agreement with the reference system that may be expected, the distribution of phosphorus in a series of ordinary superphosphates is given in Figure 2. The difference between the water-soluble and free-acid phosphorus was taken as the measure of the monocalcium phosphate, whereas the difference between the total and citrate-insoluble phosphorus was taken as the measure of the conversion of raw phosphate to soluble forms. The free acid was determined by ether extraction (4); the other phosphorus determinations were made by official methods ( I ) . These superphosphates (Figure 2 ) were prepared in open vessels and lost sufficient water to keep the free acid-water ratio above the critical value (Table 111)for the separation of dicalcium phosphate. Accordingly, the dicalcium phosphate curve was omitted from Figure 2 , and its position is merely indicated by short marginal lines. Dotted curve M shows the position the monocalcium phosphate curve would occupy if the factor for fluorine corresponded with the relationship, F = H3P04; and Q shows its position if the fluorine were neglected.
Interpolation of Acid Requirements From Equations 13 to 17 it is obvious that the acid requirement for complete acidulation (neglecting the fluorine) of a phosphate rock is defined in terms of p , of either S, So, or u, (calculated from the analysis of the rock; the latter is best suited to plotting), and of the sulfate content of the rock. If v, is known, S , is obtainable from the equation, Sc
4hp/(l - v,)
(20)
The acidulation value of the sulfate of the rock is conveni-. ently expressed as the ratio of its H3P04equivalent to S , and designated by vjo,. Thus, the values of p , u, and which are for convenience called “acidulation constants”, for a rock give information that is essential to its economical use in the manufacture of superphosphate and, indeed, in-. formation that is customarily obtained by studying trial batches. Rocks differing in grade by more than 3.5 per cent of Pz04 often have nearly equal v, values. Thus, to cite a single exT ample, the v , values of two Florida land pebbles t h a t con-.
INDUSTRIAL AND ENGINEERING CHEMISTRY
1134
VOL. 32, NO. 8
the brown rocks, though greater, corresponds to less than 1.5 per cent error in S,. The acidulation effect of the sulfate in these phosphate rocks is shown graphically in Figure 4. Thus it appears that the acid requirements for any desired imposed acidulation of a rock belonging to these two types can be determined with an accuracy of about 2 per cent by analyzing i t for P205, CaO, AZO3,and Fe203,calculating the value of the ratio of the &Po4 equivalent of the calcium and RZ03 to the H3P04equivalent of the Pz05,and interpolating vo and vso, with the aid of Figures 3 and 4. For example, suppose it is desired to find the acid requirement for complete acidulation of a Florida land pebble, the analysis of which is: P20633.7, CaO 48.0, A12030.8, Fez03 1.6, and F 3.90 per cent. With the use of the appropriate factors (Table IV) the &Po4 equivalent of (CaO Rz03)is 167.7 1.6 1.9 or 171.2, and that (2hp) of the PzOj is 46.5, which gives 3.68 for the ratio [H3P04 equivalent of (CaO R203)]/2hp. According to the curve in Figure 3, the value of u, corresponding to this ratio is 27.0 per cent. Substitution of the numerical values of u, and 2hp in Equation 20 gives S, = (2 X 46.5)/(1 - 0.270) = 127; that is, if the sulfate and fluorine in the rock were disregarded, the acid needed for complete acidulation would be 127 parts of &PO4 per 100 parts of rock. The correction to S, on account of the sulfate in the rock is obtained by noting that the PzOscontent of the rock is 33.7 per cent and reading the corresponding value of vso, from Figure 4. Thus, use, is found to be 1.5 per cent, which merely means that the sulfate in the rock is equivaor 1.9 parts lent to 1.5 per cent of 127 (8,) of H3P04 per 100 parts of rock. Thus, the acid requirement, neglecting only the fluorine, would be 127 - 1.9 or 125.1 parts of HJPOI
+
+
+
+
I
s?
I
I
I
Tennessee Brown Rock LY
2 A
A
.,
A
4
6 0 5 , p e r cent of moisture-free rock
FIGURE3 (above). COMPARISON OF v, FRACTIONS OF ACID REQUIRED FOR COMPLETEACIDULATION (NEGLECTING SULFATECONTENTOF ROCK) OF FLORIDA LAND-PEBBLE AND TENNESSEE BROWN-ROCK PHOSPHATES FIGCRE4 ( b e l o w ) . ACIDULATION VALUE( V S O ~ ) OF SULFATECONTENTOF FLORIDA LAND-PEBBLE AND TENNESSEE BROWN-ROCK PHOSPHATES Florida land pebble: o Mulberry district, 0 other deposits, o' 0 unpublished analyses Tennessee brown rock: A Wales district, A Mountpleasant district, A- A- unpublished analyses
tained 31.8 and 35.4 per cent of Pz05 were 26.3 and 26.5 per cent, respectively. It is therefore obvious that a satisfactory interpolation curve for vo cannot be obtained by plotting the latter against the phosphorus content of the rock. The v, values of twentyfive Florida land pebbles, including four flotation concentrates from the Mulberry district, and fourteen Tennessee brown rocks (pyrite-free) ranging from 30.17 to 35.55 per cent of PpOjare compared in Figure 3 by plotting oo against the ratio (H3P04equivalent of CaO and R203)/2hp. For comparison, vo of binary mixtures of fluorapatite and calcium carbonate is also shown for the same range of this ratio. The analyses of nearly half of these rocks have been published (7, 8). The rocks fall along a curved line that passes through the point (not shown) corresponding to fluorapatite or fluorapatite plus silica. The largest deviation of the land pebbles from this average curve is 0.5, which amounts to less than 1 per cent error in the value of S, calculated with the aid of Equation 20. The deviation of
TABLEVI. M
~
facturer
No. -
IMPOSED ACIDUL-4TION USED I N COMMERCIAL PRACTICE
Acidulation ~ ~ - Constants of Rock k
P
Formula of Mixa
HzSOd
%
%
%
ratio Parts acid/ 100 parts
%("Be.)
TOCk
~
Imposed Acidnlation By ingredient
H&&*
v*
U
%
%
R
98-90 99-92 111-103 98-92 96-90 95-89 106-100 94-89
100-92 101-94 112-105 100-94 98-92 97-91 108-101
100-89 101-91 116-107 100-91 97-87 96-67 110-102 94-87
99-98 100-99 97-95
101-98 102-99 99-96
Florida Land Pebble 3 2 1
4 9 5 6 10
7 10
8
32.0 32.8 33.0 33.3 33.3 33.7 33.8 35.0
26.2-31.6 25.3-30.3 25.0-30.0 24.7-29.5 24.7-29.5 24.2-26.9 24.1-28.7 22.7-26.8
3 3 . 5 26.0-28.2 3 3 . 5 26.0-28.2 3 4 . 3 25.4-27.5
2.0 1.8 1.7 1.6 1.6 1.5 1.4 1. O
1.7 1.7 1.6
69.95(55.2) 69.65(55.0) 70.87(55.8) 71.17(56.0) 71.17 (56.0) 70.41(55.5) 69 6 5 ( ? 5 0) 6 9 . 6 5 (oa.0) Tennessee Brown 68.69(54.5) 70.11 (55.d) 68.13(54.0)
83.6 86.0 95.0 84.4
82 7 83.2 93.8 84.8
Rock 90.2 89.1 90.8
95-90
102-98 102-98 99-95
0 D a t a obtained in June, 1938, b y I (. D. Jacob of this bureau in correspondence with a number of large producers of superphosphate. b T h e range was interpolated from a graph (not shown) obtained b y plotting the 1'0 values (Figure 3) against the PzO? content of the rock. c Read from the curves In Figure 4 . d Reported in Baume degrees a t 15.56' C . (60' F,),corrected t o re resent true acidity. e Percentage form of the ratio of the H i p 0 1 equivalent of the inyrecfient sulfuric acid to 8,; the latter was obtained from the result in column 3 with the aid of Equation 2 0 . * Results in column 7 plus the figure i n column 4.
INDUSTRIAL AND ENGINEERING CHEMISTRY
AUGUST, 1940
1135
years prior to the manufacturers' data, it appears that these three manufacturers haye not changed their acid-rock ratios to any extent during this Superphosphate V _ _ _ _ From period. From
COEFFICIENTS OBT.4INED FROM ACID-ROCK RATIO TABLE VII. ACIDULATIOK AXD FROM THE ASALYSIS OF SUPERPHOSPHATE Description of hlanufacturer No. 9 9
6
10 a Table
1-0. Alade from: 1719 Fla. land pebble 1423 Fla. land pebble 1315 Fla. land pebble
D a t e of manufacture
June, 1938 Feb., 1935 Spring, 1933
From
From analysis
acid-rock ratio5
analysis
n /O
70
%
%
846
92-98 92-98 101-108
746 91 100
87-97 87-97 102-110
94 100
acid-rock ratlo"
Nomenclature Quantities reckoned on superphosphate are indicated by primes.
1316 Tenn. brown rock Spring, 1933 101 99-102 101 98-102 VI. b T h e manufacturer is unable t o account for this unintended low acidulation.
per 100 parts of rock in comparison with the figure 126.5 calculated directly from the complete analysis of the rock, and with 122.6 calculated directly from the analytical results for P205,CaO, A1203, Fe303, and SO3 only. The fluorine can be taken into account and the proper correction to the acid requirement made therefor, provided the amount that will be volatilized during acidulation can be estimated. Suppose, for example, that the plant experience in the acidulation of similar rock shows an average loss of 20 per cent of the fluorine. Accordingly, the probable amount of fluorine that will remain in the superphosphate may be taken as 80 per cent of 3.90 or 3.12 per cent of the rock. The acid value of this residual fluorine is given by the product of this figure and the apparent equivalence factor for fluorine (Table IV); thus 3.09 X 3.12 = 9.6 parts of HaP04per 100 parts of rock. Deduction of the latter result from 125.1 gives i n round numbers 116 parts of H3POrper 100 parts of rock for the proportion of ~ ~ that p should 0 ~be added to the mix. The equivalent amount of &So4 is 58.1 parts per 100 parts of rock.
Commercial Acidulation The range of acidulations used by several of the larger superphosphate producers is given in Table VI. Since the only known characteristic of the rock is its phosphorus content, only the limits of vo, and hence those of the acid requirement for complete acidulation and of the coefficients of acidulation, can be stated. The data serve a double purpose in that they show both the range of acidulations practiced in commercial superphosphate manufacture and the variation that may be expected in t'he acid requirements Of rock Of a given phosphorus content. With one exception noted, the coefficients of acidulation calculated from the analysis of superphosphates (Table VII) agree well lvith the range found from the manufacturers' data (Table VI). Since three of the superphosphates (Table VII) mere produced more than 3
ratio total equivalent of acid radicals t o total equivalent of metallic elements c = water added with ingredient acid a t 100% acidulation e = weight ratio of water of crystallization t o PzOi in Ca(HzP04)z.Hz0 F' = fluorine content of superphosphate h = weight ratio H8PO4/P2Oa or HzSO4/PZO5 k = weight ratio, free H8P04/freeHzO in solution phase at equilibrium point, Ca(H2PO&.HzO-CaHPO4solution P = phosphorus content (PZO,) of ingredient phosphate is,, pz, F, = fractions of p that exist in superphosphate as monocalcium phosphate, dicalcium phosphate, and free phosphoric acid, respectively R = dilution factor for ingredient phosphate S = H3P04equivalent of metallic elements on basis of weight factors in Table I V SC = difference between S and H3P04 equivalent of p = S - 6hp (Equation 13) = H3P04 equivalent of ingredient acid plus that of S sulfate in ingredient phosphate u = net imposed acidulation (disregarding presence of fluorine) on basis of tricalcium uhosuhate UP' = correction'for fluorine (Equation is) u $. UF' = resultant acidulation on basis of tricalcium phosphate Vm = (u U P ! ) a t which PI reaches (with increasing u) its highest value in equilibrium system v = gross imposed acidulation (disregarding presence of fluorine) VF' = correction for fluorine on basis of gross aridulation oo = So/#, (see discussion following Equation 18) V Q O ~ = acidulation value of sulfate in phosphatic material = Hap04 equivalent of sulfate in ingredient phosphate expressed as fraction of S, W J = free water in system a
=
p
S
O
+
Literature Cited (1) Assoc. Official Agr. Chem., Methods of Analysis, 4 t h
ed., pp. 21--3
(1935). (2) (3) (4) (5) (6) (7)
Elmore and Farr, IXD. EKQ.CHEM.,32, 580-6 ( 1 9 4 0 ) . Hill and Beeson, J . Assoc. Oficial A g r . Chem., 18, 244-60 (1935). I b i d . , 19,328-38 ( 1 9 3 6 ) . Hill and Hendricks, IKD.EKG.CHEBI., 28, 440-7 ( 1 9 3 6 ) . Hill and Jacob, J . Assoc. Oficial A@-. Chem., 17, 487-505 (1934). Hoffman and Lundell, J . Research XutZ. B U T .Standards, 20,
607-26 (1938). (8) Jacob and others, U. S. Dept. -kgr., Tech. BUZZ. 364 (1933).
