Citrate-Insoluble Residue from an Ammoniated Superphosphate X-RAY DIFFRACTION STUDY W. J. HECHT, JR., E. A. WORTHINGTON', E. D. CRITTENDEN T h e Solvay Process Division, Nitrogen Section, Allied Chemical & Dye Corp., Hopewell, Va.
M . A. NORTHRUP Central Research Laboratory, Allied Chemical & Dye Corp., Morristown, N . J .
A
MMONIATION of superphosphate has been practiced for many years, and a considerable body of empirical information regarding the factors limiting the amount of ammonia that can be used is available (6). However, detailed knowledge of the reactions occurring is lacking, although some progress has been made (7). The mixtures are so complex and the differences in composition of the various possible calcium phosphates so slight that a purely chemical approach seems hopeless. It would appear that the identification of the phosphatic components of ammoniated superphosphate would go a long way toward solving the problem. As a preliminary part of a program with that end in view, an x-ray diffraction study has been made of the ammonium citrate-insoluble residue, that fraction of the ammoniated superphosphate containing the phosphorus pentoxide rated unavailable to plants by the Association of Official Agricultural Chemists (1). It is recognized that some hydrolysis may have occurred during digestion of the sample in ammonium citrate solution. About 50 grams of the residue were prepared by treating the ammoniated superphosphate with ammonium citrate solution in the proportion (1 gram to 100 nil. of solution) specified in the A.O.A.C. procedure ( l ) ,using the technique developed by Jacob and his associates (6). The ammoniated base was made by reacting ordinary commercial, superphosphate with a solution of the composition 65.0% ammonium nitrate-21.7yo ammonia13.3% water a t the rate of 7 pounds of free ammonia per 20 pounds of phosphorus pentoxide. The product was stored in the presence of 10% water for 30 days a t 50" C., after which time about 10% of the phosphorus pentoxide was unavailable.
tervals of 0.02" 20 (0 = angle of deflection), setting the Geiger tube position manually and moving i t in the same direction to avoid gear backlash. This was found to give considerably better reproducibility than the alternative method of measuring intensities graphically from the automatically recorded spectrometer curves. Since the maximum point for each refilling and replacement of the sample holder varied slightly, the line in question was scanned over 0.1 O to 0.2' 28 for each determination The counting interval was 64 seconds, using the scale selector number &-Le., of the counts actually detected by the Geiger tube were recorded. The background intensity of the citrate-insoluble residue was high and erratic. From this it was apparent that a t least two of the major conditions for ideal x-ray analysis were lacking in these mixtures-namely, the existence of a t least one fairly sensitive line of each compound sought, in an angular position free of interference by lines of other constituents, and the presence of a nearly smooth and constant background free of the variation caused by overlapping of weak or broadened lines of the several constituents, fluorescence, and other disturbing factors These conditions made it difficult to obtain reproducible and unambiguous intensity measurements, and practically eliminated the use of any of the internal standard techniques. The somewhat approximate method eventually adopted was t o locate several low points on both sides of the line being measured by means of a spectrometer curve (0.5' 28 per minute scanning). These areas were explored with the Geiger count recorder and a straight line was drawn between minima on either side of the measured peak. The point of intersection of this line with a
PRELIMINARY
The chemical analysis and x-ray diffraction data for the citrate-insoluble residue and for a number of its possible components are given in Tables I, 11, and 111. All of the x-ray data were obtained with a North American Phillips Co. Geiger counter spectrometer, Type 12021, using a copper tube with nickel filter, and a Geiger tube with a mica window, Type 62019. Samples were carefully screened through 325 mesh, avoiding overgrinding, and placed in holders of polished aluminum or glass, following the technique described by McCreerv (8). , . Line intensities were determined from Geiger counts a t inI
Present address, Kaiser Aluminum and Chemical Corp., Permanente Cdif. 1
TABLEI. X-RAYPOWDER DIFFRACTION PATTERN OF CITRATE-INSOLUBLE RES~DUE d/n5, A. 8.8 6.7 4.66 4.22 3.88 3.75 3.45 3.36 3.25 3.15 3.06 2.88 2.80 2.71 2.62 2.53 2.50 2.46 2.28 2.24 2.19 2.13
*
Intensityb VW VW W hl
vw vw M vs vw VW VW Mhl
VW W VW VW
Compoundc AB AB A QD A QD AB BDQ
A
AC WD BD AB AB ADB ABD A
Una, A. 2 08 2.05 1 984 1.940 1.915 1 884 1.866
:::2":
1 808 1.794
1 783 1 769
Intensityb VW W W MVW VW VW
E?
VW
w-
VW VW VW. W VW W+ VW VW
Compound0 ABD AB
..
A D AD QD A AD A
..
d/nQ, A. 1.454 1 441 1.425 1.409 1.385 1.355 1.352 1.321 1,306 1,294 1,270 1.257 1.228 1.201 1.183 1.168 1.155 1.144 1.133 1,117 1.12
A 1.752 AB 1.722 B 1.692 1.673 M1.658 M A 1.640 AQBD 1.591 M WiVW ABD 1.510 VW MAD& 1.474 W d = interplanar spacing in Angstrom units and n = nth plane of atoms. W, weak; M, medium. s, stron : VW very weak A, apatite; B, @ - C a r ( h ) p ; C, 8aFz; b, CaHPOa.'2HzO or CsBO4.2HzO;
iQD
% !
..
1119
Intensitya W W
vw
VW
Wf
M - B VW
vw Vw w Vw w
$+ W vw W
W
vw VW vw
..
Q,quartz.
Com-
pound C AQ
..
A
..
Q Q
..
A
iB A'QC Q €2
2A 9
AQ AQ AC .4 Q
..
INDUSTRIAL AND ENGINEERING CHEMISTRY
1120
TABLE 11. CHEMICAL ANALYSIS OF CITRATE-INSOLUBLE RESIDUE Weight, % ' 29.53 1.17 0.79 21.46 2.43
Analysis Calcium oxide Iron oxide Aluminum oxide Phosphorus pentoxide Fluorine Sulfite Silicon dioxide LOBSon ignition a t 900' C. Oxygen equivalent of fiuorine Total
32.39 - 10.91 1.02 98.83
Vol. 44, No. 5
were determined by extrapolation to zero of the variation of intensity with quantity of compound added, ESTIMATION OF CALCIUM FLUORIDE
-
Since the amount of fluorine was far in excess of that required for apatite, even if all the phosphorus pentoxide were in that form (Table 11), it wae decided to examine the residue for calcium fluoride. No nonreinforced line of calcium fluoride was found. The variation of intensity with the addition of 0, I, 2, 4, and 8% C.P. calcium fluoride to the residue was determined for the strong calcium fluoride peak a t 28.2"20 ( d / n = 3.80) in spite of reinforcement, by a weak apatite line at 28.1 28 ( d / n = 3.85). An illustration of the spectrometer curves obtained is given in Figure 1. It was considered preferable to base the calculations on the ratio of intensities a t 28' and 29" (a nonreinforced apat)ite peak), in order to obviate differences between replicate determinations caused by variation in crystal orientation. The data in Table VI are subject to two corrections-since the apatit.e intensity a t 29", and presumably the apatite reinforcement of the calcium fluoride peak a t 28", decreased with addition of calcium fluoride, a correction must be made for this dilution effect; and the apatite reinforcement must be eubtracted from the intensity a t 28'.
vertical line from the peak was taken as the corresponding background intensity. It was clear that the residue contained large quantities of quartz and material with the apat,ite configuration, but the great overlapping of spacings encountered in the suspected compounds rendered direct identification unlikely. Accordingly, an attempt, only partially mccessfuI, was made to effect a separation of the residue into two fractions (tricalcium and more basic phosphates and calcium fluoride in the heavier, and quartz and acidic phosphates in the lighter) by taking advantage of the differences in density. A sample was shaken with a bromoform-carbon tetrachloride mixture, having a density of 2.75 grams per ml., and the slurry was allowed to settle for 24 hours before TABLE 111. X-RAYPOWDER DIFFRACTION PATTERNS OF POSSIBLE COMPOXF;STS OB decanting the lighter fraction, which it CITRATE-Ih-SOLUBLE RESIDUE was hoped would contain most of the CaHPOa,2Hz0 &Cas(POa)eG 3Caa(POa)z.Ca(OI3zb 3Caa(POa)z,CaFzC CaFz quartz. Although occlusion of lighter d/n, Intend/n. Intend/n, Intend/n, Intend/n, Intenon heavier particles prevented a clean A. sityd A. sityd A. sityd -1. sityd A. aityd separation, the component's of the heavier 7.50 S7S 5.27 W 4.08 vw 4.05 w 3.15 s 4.22 8 4.07 W 3.43 iLI 3.43 s 1.935 VS fraction were concentmted sufficiently to 3.78 W 3.45 ILl 3.17 W+ 3.17 W 1.651 W+ 3.36 w 3 . 2 0 S I & + 3 . 0 3 W+ 1.369 W permit positive identification of trical3.04 >I+ 3.00 W 2.80 vs 1.254 W cium phosphate and to give evidence for 2.92 >I 2.87 VS 2.79 s 2 78 s1.118 W+ 2.85 VW 2.81 W 2.71 S 2.70 s iron oxide as shown by t,he x-ray dat'a in 2.62 M 2.76 W+ 2.62 >I2 63 W+ 2.43 W+ 2.71 W2.29 w 2 29 vwTables IV and V. 2.26 TV 2.60 S 2.26 T+ 2.26 W+ Since the density separation method 2.17 W+ 2.56 W 2.15 W 2.14 W 2 . 1 5 w 2 . 4 1 W + 2 . 0 6 VW 2.06 W did not prove as fruitful as hoped, it 2.10 w 2.26 W 1.945 ILI 1.038 M2 . 0 0 Tv 2 . 0 4 W + 1.890 W + 1.885 W was decided to continue the investiga1.87 W+ 1.935 M1.842 M 1.838 tion by less direct means-by intensity 1.86 VW 1.893 M1.798 W+ +;bl1.818 W+ 1.728 >I1,754 W measurements of convenient diffraction 1.797 w 1.553 W+ 1.722 W+ 1.725 W+ 1.710 w 1.468 W 1.476 W 1.474 mr lines of samples to which were added 1.561 W 1.126 W+ 1.455 W+ 1.454 W+ various amounts of suspected coma Furnished by W I, Hill U S Department of Agriculture B.P I Sample 2466-A. pounds, The original amounts present b Furnished by W: L: Hill: U: S: Department of Agriculture: B.P:I: Sample 2187.
!22::
::;;;
Prepared by heating at 900° C. for 5 days a mixture of CaFz and hydroxyapatite, which analyzed 41.10% Pi05 and 52.80% CaO. d W, weak; itl, medium; S, strong; VW, very weak.
TABLE IV. X-RAYPOWDER DIFFRACTION PATTERN OF FRACTION OF CITRATE-INSOLUBLE RESIDUEHEAVIERTHAN 2.75 GRAMSPER ML. d / ? & ,A.
Intensitye W W
Coni17d.b
d / n , A.
4.04 13.8 4.01 11.6 3.98 w 9.8 3.92 8.75 W 3.90 w 8.40 3.87 W AB 7.95 D 3.82 W 7.55 3.74 w 7.35 w 3.70 7.05 AB 3.66 W 6.75 3.59 W B 6.20 3.44 w 6.00 ., W 3.42 5.85 W B' 3.35 5.43 B 3.34 w 5.21 3.31 W 5.02 3.28 W 4 92 Q' 3.24 4 54 A w 3.19 4 52 W 3.17 B' W 4.39 3.13 V W 4 35 3.06 W 4 30 QD 3 .04 4 24 w+ .. 3.01 VW 4.18 A 2.97 VW 4.10 a W weak. M medium. S , stron b A,'apatitk; d, @.Ca3(P'04)2; C, 8aFz;
Intensity&
AB
A' D'
..
..
Compd.*
VW VW
MXI -
-
>I W+ VW
w+ w+ w+
Q E
..
..
B
AD BD AQ
..
A'
w w+
-4B
w
B
W+ W+
w
A C A D
..
d/% A.
2.95 2.88 2.79 2.77 2.70 2.66 2.65 2.63 2.60 2.58 2.56 2.54 2.53 2.51 2.48 2.45 2.43 2.42 2.41 2.36 2.35 2.29 2.28 2.25 2.21
Intensityo
W
vs S 31 hI
+ w+
C0mpd.b D BD AB AB ABE , .
W
AD'
B B
..
SD A I? h
8 QB'
AD A
' '
2BDQ A
d / n , A. 2.20 2.18 2.16 2.14 2.13 2.10 2.07 2.06 2.05 2.04 2.03 2 02 2.01 1.978 1.970 1.956 1.938 1.919 1.896 1.888 1.881 1.873 1.845 1.839 1.818
D, CaHP04.2HzO or CaS04.2HaO; E, FenOs; Q ,quartz.
Intensitya
R.1 -
nf w w w
Compd.a E
B
ABD AD
Q
VW VW 15IT
BD B AB
wMw+ vw vw
aBD
M
AB c
M W
BD A AB D D ABE DQ
+
s7w
V
w
w VW w w4vw w
B
SD
d/n,A. 1.811 1.801 1.798 1 794 1 772 1,764 1.753 1.745 1.738 1.734 1.722 1.707 1,704 1.687 1.684 1.663 1.655 1.648 1.645 1.639 1.633 1.607 1,600 1.590 1,573
IntensityQ
VW M
M
M
Compd.b ABD D
-4
w
D AB
w
A
Wpi+
% ri;r
vw VW w vw Ww w
W VW W+
W+
W W W+
..
.. ..
AB B1) E B
8Q AC AB iiB
E A D
INDUSTRIAL AND ENGINEERING CHEMISTRY
May 1952
1121
+
Equation 1, R = (Ib Ic)/Ia. Thus, Ib/Ia must be subtracted from the ratios and the equation becomes
R
=
0.41
+ 0.273P
(2)
A t zero intensity (28"), R is zero and the percentage of calcium fluoride originally present can be calculated as 1.5%. Assuming fluorapatite or hydroxyapatite values for Ib/Ia instead o f the average would give L56% or 1.44%, respectively. ESTIMATION OF &TRICALCIUM PHOSPHATE
8% CaFz Added
Without CaFn
Figure 1. Spectrometer Curves for Estimation of Calcium Fluoride in Citrate-Insoluble Residue
CORRECTION FOR DILUTION.The ratio of intensities a t 28" and 29Ois E = -I b + I C
la
where l a is the apatite (solely) contribution to intensity at 29"; I b is the apatite contribution at 28'; and I C is the calcium fluoride contribution a t 28'. The value Ib/Ia was estimated from measurements of pure compounds to be 0.68 for fluorapatite and 0.76 for hydroxyapatite. Since at this stage it was not known whether a pure or mixed apatite predominated, an average value of 0.72 was selected. It would appear to be a reasonable assumption that the apatite contribution at 28' decreased with dilution in proportion to the change in intensity a t 29 '. Then, using the mean measured intensity at 29 Ib = 161 X 0.72 = 116 for the original residue;for samples with added calcium fluoride, Ib = 116 X Ia/161 The corrected ratio is O,
R =
(I6 f I C )
+ (116 - 116 X I ~ / 1 6 1 ) 161
Using corrected ratios, one can compute the equation for the straight line best fitting the data to be
R = 1.13
+ 0.273P
(1)
The original quantity of p-tricalcium phosphate was estimated by a variation of the technique applied to calcium fluoride. The sample of pure p-tricalcium phosphate used for this determination was supplied by the Bureau of Plant Industry, B P I 2466-a; its x-ray pattern checked existing published data. Beta-tricalcium phosphate is weakly reflective and its addition complicates the background in the mixture so that only its strongest peak a t 31.1" 28 ( d / n = 2.87) was sufficiently sensitive. Unfortunately, this line overlaps diffuse scatter from a very strong apatite line at 31.9" 28 ( d / n = 2.81), so that the uncertainty of peak height measurement was worse than for calcium fluoride. The wide shoulder a t 32" (Figure 2) is probably attributable to the very small crystal sizes and the presence of solid solutions in the citrate-insoluble residue. At any rate, it could not be duplicated by a pure apatite to determine its reinforcement of the 32"peak. It was necessary to solve graphically for the reinforcement by extrapolating the spectrometer curves a t 32O, as illustrated in Figure 3, taking the mean value for samples containing
TABLE VI. ESTIMATION OF CaF2IN CITRATE-INSOLUBLE RESIDUE
Run NO.
2 8 O 28, I b 4- IC
29' 28, la
Corrected Ratio of Intensities 28"/29Oa, ~ ~ ~ ~ ( I~b +i I C$ ) ' (116 , ~ 118 X
28'/29', (Ib Ic)/Ia
+
+ --
Iu/161) 0.72 161
Citrate-Insoluble Residue
1 2 3 4 5 6 7 8
176 172 185
1.11 1.09 1.08
1 2 3
220 227 227
159 154 158
1 2 3 4 5
257 260 243 240 264
154 155 146 154 169
1 2 3
335 338 322
4% CaFl Added 153 2.19 146 2.32 139 2.32
475 549 474 510 524
158 160 128 125 153
1%
CaFz Added 1.38 1.48 1.44
0 66 0.72 0 70
2 % CaFf Added
where P is the percentage of calcium fluoride added. CORRECTION FOR ilP.4TITE REINFORCEMENT A T 28". In
TABLEV. X-RAYPOWDER DIFFRACTION PATTERN OF HANDSEPARATEDDARK MATERIALFROM FRACTION OF CITRATEINSOLUBLE RESIDUE
a
Intensity=
Compd.
vw vw M S
vw W+ vw vw vw vw W vw vw M
Fez08
... ...
... I . .
... ...
FkiOs
...
d/nb, A.
6.06 4.13 3.66 3.31 2.69 2.52 2.28 2.198 1.835 1.687 1.598 1.484 1.444 1.364 1.305 1.256
W W+ M+ M+
F&O'
1 2 3 4 5
F&O* Fea01
1
W W+
Fe;O* FezOs Fez01 FeOa FezOa Fez01 FezOr Fee08 FetOa
Intensity0
vw W
vvw M
vw vw W vw vw vw
W weak. M medium: 6,strong' VW very weak. 45 minutea a t 300b t o 460' C.
b Rdsidue h e a d d
0 91 0 92 0.86 0 80 0 89 1 40 1.45 1.38
8% CaFI Added
(Residue heavier than 2.75 g./ml.)
d/n. A. 7.08 6,lO 4.16 3.32 2.67 2.41 2.27 2.105 1.918 1.804 1.710 1.528 1.446 1.364
1.67 1.68 1.67 1.56 1.56
Compd.
...
3.00 3.44 3.70 4.08 3.43
2.24 2.70 2.68 2.61 2.67
Bureau Plant Industry Sample 2187 Hydroxyapatite
2 3 4
203 199 207 197
274 260 264 261
0.74 0.76 0.78 0.76
0.74 0.76 0.78 0.76
Fluorapatite 332 491 0.68 0.68 1 337 506 0.67 0.67 2 348 513 0.67 0.67 3 338 480 0.72 4 0.72 a Mean rmtio for fluorapatite and hydroxyapatite, 0.72: mean l a for citrate-insoluble residue = 161. 161 X 0.72 = 116.
INDUSTRIAL AND ENGINEERING CHEMISTRY
1122
Vol. 44, No. S
such as fluorosilicates. warrants the conclusion that no in01 r C~TRATE-~NSOLUBLE than verv small amounts are uresent. one can calculate from the chemical analysis in Table I'I the quantity of fluorine (antl Backgronnd Geiger Calcd,,Counts lean Corrected phosphorus pentoxide) in the form of apatite, The residue plus BackReinforceBack- Background plus 8-Caz(POa)z contained 2.43y0 fluorine; this minus the fluorine for 1.5Yo Run ment+ Total ground Reinforcement Intensitye h-o. grounda calcium fluoride gives 1.7% fluorine as apatite, requiring x Citrate-Insoluble Residue minimum of 19.1% phosphorus pentoxide (&% of the total)foi. 1.31 1 pure fluorapatite. There are not sufficient data to state whetzher 1.23 2 .. the apatite is in pure or mixed form, such as 3Ca3(PO4)2.CaF(OlI), ? .. but it should be noted that the phosphorus pentoxide requirement 43.5 ,. 6 would be higher for mixed apatites. Thus, fluorapatite probably 1% ,9-Caa(POa)z Added predominates in this sample. 1.83 533 284 1.18 .. 1 1.98 526 294 The 2.1y0 6-tricalcium phosphate found corresponds to 1.OC;. 1.27 2 512 293 4i.i 3 phosphorus pentoxide. This, together with the requirements for 2 % p-Caa(P0a)z -4dded calcium fluoride, calcium sulfate, and apatite, accounts for 934% 1,95 540 296 1.21 .. 1 of the total phosphorus pentoxide and 95% of the total calcium 2.00 536 301 1.31 .. 2 .. 565 298 oxide. The remaining phosphorus pentoxide is insufficient to 3 *. .. 587 287 445 4 .. assume that all of the iron and aluminum is present as phosphate, 4 % p-Caa(POdz Added but there is good evidence that some of the iron is in the form of 1.28 1.94 577 296 ,. 1 oxides (Table V). 1.84 596 287 , . 1.21 2 619 295 A question which may be raised is that of the possible presence 2.01 1.20 120 3 of hydrated tricalcium phosphate, the existence of which has been 6 % B -Caa(POa)n Added a subject of controversy for some time. Its preparation has been LQ5 614 285 1.28 .. 1 610 286 .. 1.26 1.90 2 reported by Hendricks and associat,es ( 4 ) and by MacIntire, .. 760 291 . . .. 3 Palmer, and Marshall ( 9 ) , but in a review article, Eisenberger, .. 724 292 , . .. 4 697 289 .. 5 Lehrman, and Turner ( 3 )have claimed t,hat the preponderance of 1 :66 634 288 6 1:27 .. 2.00 649 284 1.25 7c evidence favors the hypothesis that no such compound exiats ah a 2.02 660 289 .. 1.27 7L distinct chemical entity and that in aqueous media there is a con663 287 1.84 447 1.25 8 tinuous series of solid solutions between dicalciurxi phospliatc antl 1.93 1.25 426 Man hydroxyapatite. The principal evidence for i t 6 independent esa From autoinatic spectrometer chart, arbitrary units. 6 Obtained graphically, arbitrary units. istence is from x-ray data-prolonged heating at 900" C. of a c Intensity contribution of @-Cas(POa)e. sample corresponding in composition t,o :3Caf).P~C5.XI120 causes a change in diffraction pattern from that of an apatitr to that of 0-tricalcium phosphate; no such chauge occurs when 0, 1, 2, 4,and 6% added tricalcium phosphate. The data are in hydroxyapatite is heated. However, the same results are obTable VII. The mean value thus obtained is 426 counts for tained with a dry mixture of hydroxyapatite and dicalcium phosbackground plus reinforcement. BY Plot,ting the intensity a t phat,eof the salne average composition. 31 a minus 426 against percentage of ~-tricalcium phosphate The foregoing results indicate t,hat fluorapatite is apparen:ly added, and extrapolatirlg to zero intensity, one obtains 2.1 as the the chief form of the revert,ed phosphorus pentoxide in superphoeapproximate percentage present in the original residue phntes ammoniated, stored, and subjected to analysis as reprc-
TABLE VII. ESTIMATIOK OF p-Ca&?O:). RZSIDUE
t
.
I
IK
.
,
I
ESTIiMATION OF 3C~a(P04)z.Ca(OH)z
Since no method of direct identification or estimation of hydroxyapatite was found because of the similarity in x-ray diagrams of fluorapatite and hydroxyapatite and the broadness of lines experienced in the residue, a chemical method for estimation was investigated. This method consists of analyzing for the free calcium oxide formed on heating hydroxyapatite with calcium fluoride for a prolonged period a t 900" C. The calcium fluoride replaces or substitutes the calcium group, Ca(OH)z,or water of hydration, .XR,O, or attaches to the tricalcium phosphate radical, XCa3(P04)?: thus calcium oxide is formed only when a calcium radical containing oxygen is replaced or substituted. Laboratory data indicated that this was true for the phosphates, suspected in the citrateinsoluble residue. No calcium oxide was found on heating the citrate-insoluble residue with calcium fluoride, thus indicating little or no hydrosyapatite. It is recognized that this method is limited by the improbability of complete mixingi.e., complete reaction of chemicals in the dry state.
.
DISCUSSION
3
4
-
29
30
31
32
33
34
35
DEGREES 20
Assuming that the lack of evidence for fluorine compounds other than calcium fluoride,
Figure 2.
Spectrometer Curve for Estimation of j3-Tricalciuin Phosphate in Citrate-Insoluble Residue
36
May 1952
INDUSTRIAL A N D ENGINEERING CHEMISTRY
L-
I 3
l
-tz
ACKNOWLEDGMENT
l
I 1
I
0
2.1a
Figure 3.
1123
2
3
I
I
I
I
4
.
The authors are indebted to W. L. Hill of the U. S. Department of Agriculture, Beltsville, Md., for kindly furnishing samples of fluorapatite, hydroxyapatite, beta-tricalcium phosphate, and hydrated tricalcium phosphate; also, to W. H. MacIntire of the University of Tennessee, Knoxville, Tenn., for furnishing samples of hydrated tricalcium phosphate. The chemical analysis of the citrate-insoluble residue was made by Mrs. J. C. Redding of The Solvay Process Division, and the x-ray data were obtained by Ruth Grimm of Central Research Laboratory.
7, B E T & T I I I C A L C I U L I P W O S P U A I I A D D T D
Graphical Estimation of P-Tricalcium Phosphate in Citrate-Insoluble Residue
LITERATURE CITED
(1) Association of Official Agricultural Chemists, "Methods of Analysis," 7th ed., Washington, D.
sented by this sample. This, however, is not in agreement with previously published data (5, 7 ) , which state that a large proportion of the reverted material is hydroxyapatite. Considering these conclusions correct, reversion could be reduced markedly by reducing the formati6n of fluorapatite-Le., by volatilizing larger amounts of fluorine in the manufacture of superphosphate to decrease the fluoride content, and maintaining a low moisture content and temperature throughout the storage of the fertilizer materials to decreaw the mobility of the fluorides. This conclusion would appear in agreement with the data of Datin, Worthington, and Poudrier (R), which indicates that the phosphdrus pentoxide reversion of a group of ammoniated superphosphates is a linear function of the fluorine content in the range of 0.47 to 1.2%.
c.., -.._ i ~ n .
(2) Datin, R.C., Worthington, E. A., and Poudrier, G. L., IND.ENG. CHEM.,44, 903 (1952). (3)Eisenberger, S., Lehrman, A,, and Turner, W. D., Chem. Revs., 26, 257 (1940). (4) Hendricks. S. B..Hill. W. L.. Jacob. K. D.. and Jefferson. M. E.. IND.ENG.CHEM.,23, 1413 (1931). (5) Jacob, K. D.,Hill, W. L., Ross, W. H., and Rader, L. F., Jr., Ibid.,22, 1385 (1930). (6)Jones, R. M., and Rohner, L. V., J . Assoc. Oflc. Agr. Chemists, 25, 195 (1942). (7) Keenen, F. G.,IND.ENG.CHEM.,22, 1378 (1930). ( 8 ) McCreery, G. L., J . Am. Cernm. SOC.,32, No. 4, 141 (1949). (9) MacIntire, W.H.,Palmer, G . , and Marshall, H. L., IND.ENG. CHEM.,37, 164 (1945). RECEIVED for review August 23, 1951. ACCEPTED November 28, 1951. Preaented before the Division of Fertilizer Chemistry at the 120th Meeting O f the AMZRrcAN CHEMICAL SOCIETY, NEWYork, N. Y.
Stability of Carotene Concentrates J
H. L. MITCHELL, W. G. SCHRENK, AND RALPH E. SILKER Kansas Agricultural Experiment Station, Kansas State College, Manhattan, Kan.
c
ONSIDERABLE study has been devoted to the preparation of stable carotene concentrates for use in supplementing vitamin A-deficient rations. Such concentrates were made by dissolving extracted plant lipides or crystalline carotene in vegetable or mineral oils. The stability of the carotene of such concentrates has been modified by incorporation of various antioxidants into the concentrates ( I ) . A more convenient form of concentrate can be prepared by mixing the extracted plant lipides with finely ground solids, such as various feed ingredients (8, 5 ) , resulting in a free-flowing mixture. One advantage of such concentrates over the oil type is the greater ease of incorporation into the rations of animals. Furthermore, certain carriers exert some stabilizing influence on the carotene ( 2 ) . Additional studies on free-flowing concentrates have been made with three carriers to study the effect of the addition of various stabilizing substances. All three of the carriers have some inherent ability to improve carotene retention when mixed with either extracted plant lipides ( 2 )or with alfalfa meal (3). EXPERIMENTAL. The lipide fraction of dehydrated alfalfa meal was extracted and freed of chlorophyll and xanthophylls as described elsewhere ( 2 ) . The extractives were dissolved in Skellysolve B and the concentration of carotene was determined with a Beckman spectrophotometer at a wave length of 4360 A. The final solution contained 2700 micrograms of carotene per ml. The stabilizers constituted 0.2% of the final concentrate, except in the case of the rice bran extract. The latter consisted of the
Skellysolve B extract from 50 grams of unconverted rice bran (supplied by the American Rice Growers Cooperative Association, Houston, Tex.). Each stabilizer (0.15 gram) or rice bran extract was dissolved in Skellysolve B in a 600-ml. beaker. About 30 ml. of the carotene solution were added and the volume was reduced on a steam plate to 10 to 15 ml. Seventy-five grams of the desired carrier were added to the beaker and the contents mixed thoroughly. The mixture was transferred t o a sheet of paper and blended with a spatula. I t was spread out in a thin layer and kept in a dark cabinet for several hours to permit evaporation of the solvent. Each concentrate waa placed in a 4-ounce screw-cap bottle, analyzed for carotene (4), and stored a t 25" C. in a darkened constant-temperature room. The initial carotene content of each concentrate was about 1000 micrograms per gram. RESULTS
From Table I it will be seen that in the absence of added stabilizers unconverted rice bran contributed the greatest stability to the carotene. It is apparent also that the relative synergistic abilities of the stabilizers varied from one carrier to another, For example, the addition of diphenylamine or Caromax to cottonseed meal and rice bran concentrates did not reduce carotene destruction, while some reduction did occur when these were added to soybean meal concentrates. Dimethylaniline had a