ANALYTICAL
1320 pose other than analysis of traces of nitrogen dioxide. The method can be applied to fractionate a sample containing less 15 microns of nitrogen dioxide in order to raise the partial pressure into the less troublesome region above 15 microns. Time Required for Conditioning, Calibration, and Analysis of 20 Mixtures. The following estimate of instrument time applies to any mixture containing nitrogen dioxide plus lower molecular weight components. Calculating time requirements vary with the complexity of the mixtures.
than
Instrument Time NO2 conditioning 3 runs on NO2 to check conditioning, 2
plus background run additional NO2 calibration
Man-Hours 1.0 1.0
7.5
Anal. Chem. 1953.25:1320-1324. Downloaded from pubs.acs.org by KAOHSIUNG MEDICAL UNIV on 09/06/18. For personal use only.
5
on the particular mixture. If, however, collected at micron pressures all of the abovementioned difficulties are avoided and the sample can be truly representative. For a 3-liter expansion bulb in the mass spectrometer, a sample collected at 200 microns pressure in a 3-liter bottle is sufficient. The accuracy of analysis of nitrogen dioxide depends upon its concentration and on the other components in mixtures. The analyses of synthetic blends are indicative of the accuracy attainable. If a mass spectrometer can be devoted solely to analysis of nitrogen dioxide samples, this accuracy might be subject to improvement.
difficulties depending such samples
are
0.5 5.0
runs
20 samples
Man-Hours
Calculating Time calibrations plus background
0.5 ~5.0
20 samples
~5.5
Depending upon the type of mixture to be analyzed, frequent background runs may be required for accurate analysis. CONCLUSIONS
Reaction among the various components of the synthetic blends analyzed for this paper was avoided by mixing at micron pressures instead of millimeters. This procedure points the way toward successful sampling of mixtures in high temperature systems. If samples containing nitrogen dioxide, water, nitric oxide, oxygen, etc., are collected at or near atmospheric pressure, the cooling process will result in condensation of water, solution of nitrogen dioxide in the condensed water, reaction between nitric oxide and oxygen, formation of nitrogen trioxide, and other
CHEMISTRY
ACKNOWLEDGMENT
Gratitude is hereby expressed to the Consolidated Engineering Gorp. for financial support of this work under a cooperative agreement with the Bureau of Mines. The assistance of Joseph Malli, William Arnold, and Anthony Logar, Jr., is gratefully acknowledged. LITERATURE
CITED
Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., March 2 to 6, 1953. (2) Johnson, C. L., Anal. Chem., 24, 1572 (1952), (3) Mattraw, H., and Schacher, G. P., paper presented before Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., March 5 to 7, 1952. (4) Verhock, F. H., and Daniels, F.. J. Am. Chem. Soc., 53, 1250 (1) Frey, H. J., and Moore, G. E., paper presented before
(1931).
(5)
Whitnack, G. C., Holford, C. J., Gantz, E. St. C., and Smith, G. B. L., Anal. Chem., 23, 464 (1951).
Received for review January 8, 1953, Accepted May 1, 1953. Presented at the Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., March 2 to 6, 1953.
Determination of Phosphate by Differential
Spectrophotometry ALLEN GEE AND VICTOR R. DEITZ National Bureau of Standards, Washington 25, D. C. precise determination of phosphate in rocks, bones,
THE teeth, and in other materials containing calcium phosphate
a laborious process because the accepted gravimetric or volumetric methods require the complete separation of calcium and other interfering elements. The available colorimetric procedures are rapid and adequate for most routine purposes, but they are relatively low in precision. By using the techniques of differential spectrophotometry (with necessary refinements), a method has been developed in the present research which combines the essential simplicity of a spectrophotometric method with the precision of careful gravimetric analysis. The method is based upon the yellow complex formed by phosphate with vanadate and molybdate in acid solution. Molybdivanadophosphate (MVP) exists in a true solution and is more stable and reproducible than the molybdenum-blue complex of phosphate (5,17). The stability of this system, together with its relative freedom from interfering substances and its ease of color development, recommends its use for the quantitative determination of phosphate. The high precision which can be obtained with the procedure is based upon differential spectrophotometric techniques. The complex formation in unknown solutions of phosphate is adjusted to give an absorbance of about 2 in the near ultraviolet. The
has been
are determined not against water blank but against an accurately known standard solution of nearly the same absorbance. When the absorbance of the unknown solution is sufficiently close to that of the standard so that it can be determined to within 0.001, the precision is thus 1 part in 2000. Since the unknown solutions are developed almost simultaneously with the standard, the variations in the absorbance of uncombined molybdate and vanadate with time and temperature are automatically compensated.
absorbances of these solutions or a
FORMATION
OF THE COLOR COMPLEX
The absorption spectrum of the molybdivanadophosphate complex is shown in Figure 1. The blank, consisting of the reagents without phosphate, also had an appreciable absorption spectrum in the ultraviolet. Because the exact nature of the complex as a chemical entity is not known, it was necessary to examine the conditions under which the complex is formed in order to realize its full potentiality. The system was investigated with regard to variation in the concentration of vanadate, molybdate, and acid. With a series of solutions of 0.42 and 0.50 mM of phosphate per liter and with excess molybdate in 0.2 M perchloric acid, the absorbance increased upon addition of vanadate from that of the molybdiphosphate complex until the point was reached where the vanadate molarity equaled that of the phosphate (Figure 2). Since further addition of vanadate did not mcrease the absorbance, the result confirmed the fact that each
VOLUME
2 5,
NO. 9, SEPTEMBER
1953
Precision and rapidity are continuously sought in chemical analyses. The yellow complex which orthophosphate forms in the presence of a mixture of vanadio and molybdic acids was known to satisfy the requirement for rapidity. The precise method of differential spectrophotometry was investigated to attain the desired precision. With reagents more dilute than those used in previous procedures, the yellow complex obeys Beer’s law when measured differentially for absorbances compared with water as high as 3.8 at 390 mg. The Beckman spectrophotommust be eter with an incandescent light source provided with a suitable filter to minimize stray light. Many substances have little or no effect on the color development. Under optimum conditions phosphate concentrations can be compared spectrophotometrically with a precision of 1 part in 3000. of phosphate in calcium phosThe determination phates can now be made rapidly with a precision comparable to the best gravimetric procedures. The phosphate contents of two National Bureau of Standards phosphate rocks were determined with a relative error somewhat better than 1 part in 1000.
1321
(Figure 1). Since complex solutions containing phosphate much above 1 mM formed a precipitate upon standing, the amounts of excess molybdate and vanadate normally did not need to be more than about 50% of the stoichiometric quantities required to form the complex with 1 mM phosphate. The dependence of the absorbance of the blank upon acid concentration is illustrated in Figure 4. There was also a measurable change in absorbance over a period of a day. The difference in absorbance between a complex solution and a blank also varied slightly with time, but differential measurements among known solutions containing 0.4 to 0.8 mM of phosphate per liter when developed at the same time maintained remarkable constancy and showed no variation in 2 days. For high precision it rvas therefore essential to develop the solutions within about 5 minutes of each other. The perchloric acid concentration chosen was 0.21 M, less than that recommended in previous procedures (IS), but the formation of the yellow complex is more rapid when less acid is present (16). The difference between the acid concentrations of the unknown and standard solutions should not exceed 0.01 . The color-developing reactants should be present in the amounts shown in Table I when phosphate is present in concentrations up to 0.8 mM. A phosphate concentration of 0.8 mM yields an absorbance of about 3 at 390 mg for a 1-cm. light path, and the solution can stand for weeks after color development without precipitation. When the phosphate concentration is increased to 1 mM with the same amounts of reagents, the measured absorbance is 0.4% below that indicated by Beer’s law. For solutions more dilute in phosphate, where it may be
phosphate requires one vanadate to form the complex. This had previously been stated by Kitson and Mellon (12) and Misson (14). When the vanadate was in excess and the phosphate again held at 0.42 and 0.50 mM, the absorbance at 450 mg increased with molybdate concentration as shown in Figure 3. MeasureNo complex ments at 390 mg, also followed a similar curve. formation took place when the molybdate was present at a concentration below 0.5 mg.-atom of molybdenum per liter. If 0. 5 mg.-atom of molybdenum per liter is assumed to remain uncombined when phosphate is in excess, then the stoichiometric ratio of molybdenum to phosphorus in the complex is about 14 to 1. Misson (14) reported a ratio of 16 to 1 for the solid salt. While the results of these experiments are in general agreement with the composition of the complex stated by others (12, 14), they show in addition the strength of the complex formation. The sharpness of the bend in the curves of Figures 2 and 3 is a of this tendency. The low acidity employed made it measure unnecessary and undesirable to use large excesses of molybdate and vanadate which contributed to the absorbance of the blank
Figure 2. Dependence of Absorbance
on
Vanadate
Figure 3. Dependence of Absorbance
on
Molybdate
Concentration
Wove
length,
mg
Figure 1. Absorption Spectrum~4 of MolybdivanadophosPO, Compared with phate Complex Developed in 10 the Blank
Blank and molybdic acid solutions were compared with water; molybwas 17 mg.-atom per liter in 0.2 /V acid date concentration
Concentration
ANALYTICAL
1322
Table I.
Recommended Concentrations
of Reagents
Maximum Concentration
Component Vanadate, milüatom V per liter Molybdate, milliatom Mo per liter Perchloric acid, molarity
_of
0.8 mM
Phosphate_
1.2 17
0.2-0.6
0.1 mM
0.4 6
0,2-04
CHEMISTRY
phorus pentoxide. Additional standard solutions containing increments of 0.1 to 0.2 mg. more of phosphorus pentoxide are required for the linear calibration. Use of more than two standards enables one to use the method of least squares in establishing the calibration. A single standard is allowable when the instrument has not been reset. Wait at least 5 minutes after adding the mixed reagent for complete color development. The standard and unknown solutions must be prepared concurrently.
The matching of absorption cells is not critical when the following method is used.
Fill one cell with the solution containing the least phosphate. Measure all the solutions (including standards) in a second cell against the solution in the first cell. Interpolate the phosphate content of the unknown solutions from the absorbances and phosphate contents of the standards. A linear extrapolation for measurements not bracketed by the standards can be made at reduced precision. INFLUENCE
OF FOREIGN SUBSTANCES DEVELOPMENT
ON COLOR
Few substances found with common phosphate materials caused any interference. Saits of the alkali and alkaline earth 4.5
1 1
PROCEDURE FOR COLOR DEVELOPMENT
To a 25-ml. volumetric flask containing 1.0 to 1.4 mg. of phosphorus pentoxide as orthophosphate and not more than 0.2 meq. of acid (unless compensated for by standard) in 10 to 20 ml. of water, add 2 ± 0.05 ml. of mixed reagent. Dilute to exactly 25 ml. and measure the absorbance in 1-cm. matched cells with respect to a standard which should contain from 0.0 to 0.3 mg. less of phos-
€
d(P04) O Filter 5970 • Filter 5113
C No filter 2.5
c
1
1
0.4
_1_
0.6 mM PO4
0.8
Figure 5. Apparent Decrease of Absorbance Index with Concentration
Substances Which Do Not Interfere with Color
Table II.
Development Substance
AgNO, AhfSOi),
Ca(CH,COO)2 CdSO.
, , HgCl; KBr LiOH MgSOi NaF
a
The mixed reagent is prepared by pouring 1 volume of Solution A into 1 volume of Solution B and stirring until any precipitate formed is redissolved. The mixed reagent can be diluted so that standard pipets can be used to measure the required amounts of material. To obtain the phosphate complex, the following procedure is used:
0 s
3.5
dA390
As suggested by Barton (1), the most convenient way to develop the color complex is to have just one reagent contain the requisite amounts of vanadate, molybdate, and acid. Such a solution was found to be rather unstable, particularly in warm weather when precipitation began in about a week. The reagents were therefore stored in two solutions and mixed before use.
Solution A, ammonium molybdate (80 grams per liter). Dissolve 40 grams of ammonium molybdate tetrahydrate, analytical reagent (ACS specification), in 400 ml. of hot water. Cool and dilute to 500 ml. This solution is stable for several months. Solution B, ammonium metavanadate (4 grams per liter in 5 M perchloric acid). Carefully pour 250 ml. of 70% perchloric acid into 200 ml. of water, add 2.0 grams of ammonium metavanadate, and warm until dissolved. Cool and dilute to 500 ml. This solution is stable. Standard Phosphate Solution, 0.005 M. Weigh accurately 0.67 gram of potassium dihydrogen phosphate (Baker’s primary pH standard dried overnight at 105° C.), dissolve in water, and add 5 ml. of 70% perchloric acid and sufficient water to make 1 liter. The perchloric acid is added to compensate for the acid used in dissolving basic calcium phosphates. The use of properly recrystallized potassium dihydrogen phosphate as a phosphate standard was not studied, but it appeared to be good to 0.1%. NBS standard phosphate rocks may be more suitable.
1
8
Figure 4. Dependence of Absorbance of Blank on Acid Concentration after 1 Hour (Solid Lines) and after 1 Day advantageous to use 10-cm. absorption cells, the quantities in the last column of Table I are recommended.
1
b e
d
Gram/ Liter
Substance
2. 1 1.7 0. 2.9 1.0 1.4
NaCl NaiPaO:
3.8 0.6 1.1
CuSOa C0CI2
Gram/ 5.4° .
(NHOaSaOe
SrCU ZnCla CeHsCOOH
1.8 2.3 1,4 1.8C
0.7c 0.7 c
MnCh »
0.6«*
b
.
1.2
0.01
Slight increase in absorbance (0.01) at this concentration. Some hydrolysis to orthophosphate after 1 day. Correction for color of cation from 0.004 to 0.007 applied. Small increase of color with time, possible attack on glass.
Table III.
Substances Requiring Excess Molybdate
Substance
Change in Absorbance, % After ^After 1 day
Grams/Liter 1.6
Lithium oxalate Ammonium citrate
Ammonium
2.5 1.4
tartrate
Arsenic trioxide
:
Qualitative)
0 0
0
small
-2 —
3
Small
VOLUME
2 5,
NO.
SEPTEMBER
9,
1953
elements in general had no effect in quantities up to a few grams per liter. Substances which had negligible effects on the color development are listed in Table II together with the concentrations tested. The phosphate concentration was 0.6 mM (0.043 The absorbgram phosphorus pentoxide per liter) in all cases. ance was measured at 390 mg (and at 420 mg in most instances) with respect to an unadulterated solution after 1 hour and after 1 day. Except as noted, none of the substances in Table II influenced the change of color with time. The slight color of the Cu + +, Co++, and Mn + + ions was the same as in water. Except for a small number of discrepancies (notably chloride and fluoride), this list conforms with the results for the many substances tested by Kitson and Mellon {12), who permitted much larger tolerances. Four substances which appear to form colorless complexes with To get full color one of the reactants are given in Table III. development, higher concentration of molybdate was required (6), and even then the color faded slightly after one day. Since the complex is reduced rather easily, complete absence was required of certain ions such as ferrous, stannous, hydrazonium, and iodide. Some idea of the reactivity of the complex can he gaged from the fact that it can slowly oxidize metallic mercury. Lead acetate, tetramethylammonium bromide, and tetraethylammonium bromide formed precipitates when present to the extent of 0.1 gram per liter. Materials causing moderate interference when present at the time of color development are listed in Table IV. These are not materials which one ordinarily encounters in phosphate analyses. When nitrate is present originally with the phosphate, care must be taken that no decomposition to nitrite occurs.
Table IV. Substance
KSCN Na2WOl XaNOa
Causing Moderate Interference
Gram/Liter
Change in Absorbancy, %a 390 µ 420 µ After 1 day at 390 µ -49 -7 +3
1.5 1.7 1
1
Sucrose
1.1 250
Based
0,6 mM phosphate.
XazSOs
a
Materials
on
-4
+ -
7 1
0
Precipitate
-3 -
1
0
Reduction to blue Reduction to blue
1323
Table V.
Some Materials
Causing Serious Interferences
at 390 mg
Material
f
Quantity Present, \Vt. Ratio to P2Os)
Fe(XO;i)3 Xa2HAsO,
1
0.
JGPOa
XaPOj
a
Expressed
so.oond column.
as
AS2O5
as
0.1
P2O*
as P2O5
). 04
as P2O5
APPLICATION
1
0.05
Xa2SiOs
equivalent
to
Fe2Os
0.3
0.07
K2Cr20: KCrfS04)2
as
1
as O2O3 as CrsOs as S1O2
Interference' 0.016 0.16 0.01 0.16 0.3 0.02
0.8
part by weight of the oxide given in
OF DIFFERENTIAL
SPECTROPHOTOMETRY
High precision in the absorbance measurement of a solution is attainable when it is compared with an accurately known standard (3, 9). Successful applications of this technique have been made to several nearly ideal systems (2, 3, 8). In addition to greater precision, this comparative method applied to the complex provides an automatic compensation for the blank and for the apparent variability7 in the absorbance of the uncombined vanadate and mol>7bdate with time and temperature (1, 4, 16). This can be realized when the standard solutions are developed at the same time as the unknown. The optimum range of wave lengths for measuring the absorbancy of the complex is between 380 and 450 mg. Figure 1 shows this to be well removed from the absorption peak of the phosphate complex. The upper limit of this range is restricted by low absorption, and the lower limit is restricted by7 the rising absorbance of the blank which is preferably kept below 0.1. The shorter wave lengths within this range arc favored for differential absorbance measurements. When complex solutions of less than 10% transmittance (relative to water) at any7 wave length between 380 and 450 mg are measured in the Beckman DU spectrophotometer, apparent There are some obvious readeviations from Beer’s law occur. sons for this behavior. For a given slit opening of the instrument and with the use of a tungsten lamp, the blue-sensitive photocell has 10 times the response at 530 mg as it does at 390 mg. Since wave lengths can pass almost unattenuated through the complex solution, the effect of any stray light is exaggerated over that of a system measured at an absorption peak in the middle of the visible spectrum. The effect of stray7 light has been discussed by7 many authorities (7, 10, IB). Nonuniformity of absorption over the wide spectral band (10 mg) causes a similar error. Stray light can be almost entirely eliminated by7 the insertion of an appropriate glass filter anywhere in the optical path of the inThe ultraviolet filter furnished with the Beckman strument. instrument is suitable for measurements at 390 mg. Standard 2X2 inch glass filters which cannot be placed in the filter holder can be conveniently accommodated by a 0.25-inch spacer inserted before the photocell compartment. Two standard filters have been tested: the blue (Corning 5113, half-thickness) with maximum transmittance at 405 mg and 1 % at 470 mg, and the ultraviolet (Corning 5970) with maximum at 370 mg and 1% at 405
light of longer
,
Table V lists some substances, several of which are commonly found in phosphate rocks, that may cause serious interference. These substances formed a yellow color with the color-developing reagent. For measurements made with 0.60 mM phosphate at 390 mg the amount of interference was moderate for iron where 1 part ferric oxide was equivalent in absorbance to 0.016 part phosphorus pentoxide. Interference by silica was serious where 1 part silica was equivalent to 0.8 part phosphorus pentoxide. For silica the interference at 450 mg decreased to about half of that at 390 mg. In some cases, the exact magnitude of the interferences depended on the acidity and to a slight extent on the phosphate concentration. The interference of arsenate could be markedlyreduced by raising the final acidity from 0.21 to 0.30 M. The interference of metaphosphate increased about 40% after the developed solution stood for 20 hours, probably because of hydrolysis to orthophosphate. On the other hand, the interferences of ferric and dichromate ions decreased slightly (less than 10%) after the same period of time For precise phosphate determination the interfering substances in Table V should be kept low or, particularly in the case of silica, completely removed. While only the interference of ferric iron has been found to conform with Beer’s law, the values given in the last column of Table V for the other substances can be used to estimate the correction required if present in small amounts. Interferences which may arise because of simultaneous presence of several foreign substances and/or other changes in the conditions for color development have not been fully investigated.
mg.
The corrective effects of the two filters are evident in Figure 5 where the apparent differential absorbance index at 390 mg is plotted against phosphate concentration. These values were obtained by7 dividing the observed difference in absorbance by the difference in phosphate concentrations. Without a filter, the index decreased rather badly. With the blue filter which essentially cuts out all the stray light above 450 mg, the situation was considerably improved. With the wave length setting at 390 mg about 0.1% of the photocurrent was produced by background light of 450 to 650 mg. This percentage of stray light With the ultraviolet filter was little changed by7 slit width.
ANALYTICAL
1324
Results for Analyses of PzOe in Phosphate Rocks, NBS Samples
Table VI.
Material0
Analysis
_PiOt,
390 m/n 33.16 ± 0.02c 32.92 ± 0.02 35.20 ± 0.006 35.19 =fc 0.006 35.17 ± 0.01
%!»_
420
µ
33.03 ± 0.04c NBS 56a 32.91 ± 0.04 2 NBS 56a 35.23 3 NBS 120 NBS 120 4 5 NBS 120 ° The NBS recommended value for sample 56a is 32,90% and for sample 120,35.20%. 6 Values given are the averages for measurements on 3 aliquots together with the standard deviation of the mean. c Value high probably because of incomplete digestion. 1
(No. 5970), the absorbance index was constant to absorbances exceeding 3.0. The difference between the results for the two filters gives some indication of the effect of the further elimination of stray light. Deviations arising from nonuniformity of absorption over the wide band employed were also reduced. A practical way to select a filter to compensate for a system like the moiybdivanadophosphate complex (which has low absorbance where the spectrophotometer response is greatest) is to determine the wave length at which the photocurrent is highest at a constant slit opening with the filter in the optical path of the spectrophotometer. The band 10 to 15 µ on each side of this point is the approximate useful range of the filter tested. Thus, the range is 375 to 400 µ for Coming filter No. 5970, and 420 to 450m/i for Corning filter No. 5113. These measurements made with an air path appear to be adequate, although insertion of an actual solution may give some improvement. DETERMINATION OF PjOs IN STANDARD PHOSPHATE ROCK
This simplified method was tested on two NBS standard phosphate rocks, using the following procedure.
After drying 1 hour at 105° C., a weighed sample of phosphate rock (0.35 to 0.38 gram) was digested with a mixture of 5 ml. of perchloric acid and 3 ml. of nitric acid in a 50-ml. Erlenmeyer flask on a hot plate. Because of the possible explosive nature of perchloric acid mixtures, care must be taken that nitric acid is initially present in the perchloric acid. Because of the interference of nitrogen oxides, all of the nitric acid should be driven off. Heat was applied for about 0.5 hour until all the nitric acid had boiled off, at which time the fumes receded from the neck of the flask. After cooling, the flask was half filled with water. The resulting solution was filtered and the precipitate washed with water. The filtrate was collected with the wash water in a 200-ml. volumetric flask. After diluting to the mark, three 1.8- to 2.1-ml. aliquots (actually weighed in this work to minimize volumetric errors) were placed in 25-ml. volumetric flasks. The color was developed as previously described. To get an estimate of the ultimate precision of the method, three standard solutions using 3.0, 3.5, and 4.0 mi. of standard phosphate solutions (again weighed) were employed for calibration. All absorbances were compared with the most dilute standard at 390 µ with the Coming 5970 filter in the Beckman DU spectrophotometer with a 1-mm. slit opening. The apparent amounts of phosphate in the three aliquots of the unknown phosphate rock solution were interpolated from the best straight line found for the standards. A correction for iron was made as shown in Table V, using the value for iron certified by the National Bureau of Standards. The results for five analyses on two standard phosphate rocks are presented in Table VI. The first analysis is somewhat high, possibly because of incomplete expulsion of nitrogen oxides or insufficient dehydration of silica during digestion of the sample. The standard deviation computed from measurements of different aliquots of the same solution is 0.011% of phosphorus pentoxide (coefficient of variation is 1 part in 3000). For the entire procedure and with omission of the first analysis, the standard deviation is 0.02% of phosphate. This precision compares very favorably with gravimetric work of Hoffman and Lundell _
CHEMISTRY
(11) who used two precipitations. No detectable change occurred for solutions in which the complex had been developed on the previous day. These results should be compared with values ranging from 34.8 to 35.8% of phosphate obtained for NBS standard sample 120 by direct absorbance measurements relative to water or a blank, using a slightly different chemical procedure (1). For routine analyses of phosphate rocks, it would be more convenient to take larger aliquots containing 4.2 to 5.2 mg. of phosphate from a more dilute phosphate rock solution and to develop the color in 100-ml. volumetric flasks. The amount of colordeveloping reagent as well as the quantity of phosphate should be increased in proportion to maintain the concentrations used in this procedure (Table I). When a large number of samples are to be analyzed at one time, it may be necessary to add the colordeveloping reagent from a rapid-flow pipet so that development of all the solutions is begun within a 5-minute interval. When the measurements were made at 420 µ with the Corning No. 5113 filter, the results were still very satisfactory (Table VI). Since the absorbance was only half as great as it was at 390 µ, the standard deviations were twice as large. CONCLUSION
This spectrophotometric analytical procedure is capable of comparing phosphate concentrations with a precision of 1 part in 3000. In addition, the method is simpler and more rapid than others of comparable precision for analysis of calcium phosphates. Since interfering substances are few, the procedure has a wide application. The simplified method of bringing phosphate materials into solution with a mixture of nitric and perchloric acid appears generally applicable but may require further investigation, particularly in regard to removal of different forms of silica. The data indicate possible modifications of the procedure to meet different conditions. The procedure described has been successfully employed to analyze synthetic basic calcium phosphates and tooth enamel after a simple solution in a one-fold excess of hydrochloric acid where it was necessary to detect very small changes in the ratio of calcium to phosphate. ACKNOWLEDGMENT
This investigation was sponsored as a joint research project undertaken by the Sugar Research Foundation, Inc., and the National Bureau of Standards. LITERATURE CITED (1) Barton, C. J., Anal. Chem., 20, 1068-73 (1948). (2) Bastían, Robert, Ibid., 23, 580-6 (1951). (3) Bastían, Robert, Weberling, Richard, and Pallila, Frank,
22, 160-6 (1950).
(4) Bridger, G. L., Boylan, D. R., and Markey, J. W.,
336-8 (1953).
Ibid.,
Ibid., 25,
(5) Clausen, D. F., and Shroyer, J. H., Ibid., 20, 925-6 (1948). (6) Epps, E. A., Ibid., 22, 1062-3 (1950). (7) Goldring, L. S., Hawes, R. G., Hare, G. H., Beckmann, A. O., and Stickney, . E., Ibid., 25, 869-78 (1953). (8) Hiskey, C. F., and Firestone, David, Ibid., 24, 342-7 (1952). (9) Hiskey, C. F„ Rabinowitz, Jacob, and Young, I. G., Ibid., 22,
1464-73 (1950). (10) Hiskey, C. F., and Young, I. G., Ibid., 23, 1196-1201 (1951). (11) Hoffman, J. I., and Lundell, G. E. F„ J. Research Natl. Bur. Standards, 19, 59-64 (1937). (12) Kitson, R. E., and Mellon, M. G., Ind. Eng. Chem., Anal. Ed., 16, 379 (1944). (13) Koenig, R. A., and Johnson, C. R., Ibid., 14, 155 (1942). (14) Misson, G., Chem. Ztg., 32, 633 (1908). (15) National Bureau of Standards, Circ. 484 (1949). (16) Willard, . H., and Center, . J., Ind. Eng. Chem., Anal. Ed., 13, 81 (1941). (17) Woods, J. T., and Mellon, M. G., Ibid., 13, 760 (1941).
Received April 14, 1953. Accepted July 6, 1953. Presented before the 122nd Meeting of the American Chemical Society, Atlantic City, N. J., September 1952. The senior author is a research associate at the National Bureau of Standards, representing the Sugar Research Foundation, Inc.
