442
INDUSTRIAL AND ENGINEERING CHEMISTRY
99.9, 99.6, 100.4, 99.4, 99.3, 99.8, and 99.3% with gelatinglycerol solutions obtained from gelatin capsules. The arithmetic mean of all these determinations is 99.62% (2). The standard deviation is *0.43. SUMMARY
Val. 17, No. 9
consistent and duplicable. The arithmetic mean of representative determinations is 99.62%, while the standard deviation is *0.43. LITERATURE CITED (1) Assoc. Official Agr. Chem., Official and Tentative Methods of
Analysis, 5th ed., p. 480, 1940.
A method for the determination of glycerol in the presence of large quantities of gelatin has been developed to determine the glycerol content of gelatin capsules that also contain water and insignificant amounts of preservative and dye which do not affect the results of the analysis. After removal of the gelatin by precipitation as the tungstate. the glycerol content is determined by the official method of the .4.0..1.C. The results obtained are
(2) Crumpler, T.B.,and Yoe, J. H., “Chemical Computations and Errors”, pp. 125-34, New York, John Wiley & Sons, 1940.
(3) Kolhhoff, I. M.,and Sarver, L. A., J . Am. Chem. SOC.,52, 4179 (1930). (4) Kolthoff. I. hl., and Sarver, L. A., 2.Elektrochem., 36, 139 (1930). (5) Kolthoff, I. M., and Stenger, V. A., “Volumetric Analysis”, Vol. I, p. 133, New York, Interscience Publishers, 1942. PREBEKTED before the Division of Analytical and Micro Chemistry at the 108th Meeting of the AMERICAN CHEMICAL SOCIETY, New York, N . Y.
Determination of the Capacity of Shallow Jars EARLE R. CALEY Wallace Laboratories, Inc., N e w Brunswick,
N. J.
With jars having a smooth flat rim, this point may also be deH E accurate and rapid determination of the true capacity termined by following the decrease in the weight of the system of shallow jars of small volume commonly used as containers with time, rehtive constancy of weight being m,.ntained for a for certain cosmetic and pharmaceutical products is not so simple considerable interval after evaporation of the last portions of as may first appear. In contrast to volumetric flasks or specific water from the meniscus. Illustrative of this are the curves in gravity bottles, the orifice of such containers is great as compared Figure 1, which show the loss in weight during the first two to their volumetric capacity, and thus errors due to meniscus trials with jar I taken a t 5-minute intervals. Determination of effects are greatly exaggerated when standard calibrating liquids the end point from the appearance of the first minute bubble of such as water or mercury are used. Moreover, it is not easy to air is less laborious, however, and a t least equally accurate. It determine experimentally R ith accuracy the exact point a t which may, indeed, be slightly more accurate, since the volume of water such vessels are truly filled-the point a t which the level of the displaced by the air bubble is partly compensated by the volume calibrating liquid corresponds to that of an ideal plane deterof the remnants of the film of water that ordinarily still remains mined by the top surface of the rim. between the top of the rim and the glass plate a t the time the An obvious method of restricting the level when water is used bubble appears. At any rate, the results of determination of as a calibrating liquid is to fill the jar to overflowing, and to capacity with water shown by Table I are much more precise slide onto the top of the jar a square of plate glass in such a way if time is allowed for evaporation of water from the outside, and as to exclude air bubbles. Excess water may then be removed they are undoubtedly much more accurate. This is evident both from the outside of the jar and underside of the plate by careful from the above discussion and from a comparison with the rewiping, and the capacity determined immediately from the sults of the determination of the capacities of the same jars with difference in weight between the empty and filled system. H oWever, this simple and relatively quick procedure, which is often used in practice, gives noticeably high results, 3.510 in part because of the inclusion of the weight of water contained in a meniscus that forms between the outer edge of the rim and the glass plate. 3.500 More accurate results are obtained if the weight of the filled system is determined after standing long enough to allow the water in this outside meniscus 3.490 to evaporate. A positive error may still remain, however, because of inclusion of the weight of water that remains between the top and the rim and the glass 3.480 plate, If the rim is flat and wide on top, a common form, this error may be of appreciable magnitude. The results of trial determinations on two commer3.410 cial jars by this method are shown in Table I. The averages shown and the average deviations are based upon six successive trials. I was a glass jar with a 3.460 slightly rounded rim of ordinary width; I1 was a plastic jar with a flat rim of more than ordinary width. All the volumetric capacities shown in the table are uni3.450 formly based on 20” C. as the standard temperature. It will be seen that the initially determined capacities are appreciably higher than those found after allowing 3.440 the water in the outer meniscus to evaporate. I n these experiments the appearance of. the first minute bubble 0 PO 40 60 80 100 120 140 of air under the glass plate next to the rim was taken Time In minutes as the indication of complete evaporation of this Figure 1. Loss in Weight with Time and Final Temporary Constancy before water. Appearance of Bubble, B
T
ANALYTICAL EDITION
July, 1945 Table
Results of Experiments on Determination of Shallow Jars b y Different Methods
Average Capacity and Average Deviation Jar I Jar I1 Jar I11 Jar IV
31ethoda
M1. 4;917 0.013 4.895 0.0010 4.871 0.0003 4.80 0.020 4.82 0,034 4.89 0.015
A B
C D
E F a
I.
M1. 3.489 0.016 3.451 0.0018 3.457 0,0003 3.41 0.026 3.43 0.018 3.42 0,010
M1.
.. .... .... ,.
2.793 0.0003 2.72 0.022 2.72 0,038 2.80 0.011
Ml.
.. .. .. .. .. .. .. ......
....... I
4.308 0.0012 Taken as standard
.. . . . . . . . . .. . ..
.. . .. . .
"OVf,~~ Deviations
443 routine control purposes by weighing on an ordinary laboratory scale sensitive to 0.1 gram instead of on an analytical balance.
of Capacity Time Re-
szF$,!&-
M1. 0.015
mination Min. 15-20
0.0014
60
+
LEAD SHOT AS CALIBRATING MEDIUM mination Fair Fair
0.0005
15-20
0.023
5-10
Poor Good
0,030
5-10
Good
0.012
5-10
Good
A method that avoids the manipulative difficulties of working with mercury and at the same time makes possible weighing on an ordinary laboratory scale is based upon the use of very fine lead shot as a calibrating medium. At first thought this sort of method may not appear to possess much promise of either precision or accuracy, but actual experiments showed that useful results may be obtained.
The procedure is first to weigh the empty jar, place it in a level, shallow cardboard tray or other suitable receptacle, and pour into it fine lead shot from a container held just above it until the cavity in the jar is filled to overflowing. Then the shot is leveled off by passing a straightedge slowly across the rim while pressing it down firmly. Finally, the weight of shot contained in the cavity is determined by again weighing. From the apparent density of the shot, found by weighing the amount contained in a jar accurately calibrated with water or mercury, the volumetric capacity of a given jar is readily calculated. It is essential for successful results by this method that shot of very small diameter be used, such as that commonly sold for the filling of tare bottles in microanalysis. The average diameter of the shot should not exceed 1.0 mm., which was the average diameter of the shot used in the experiments here described Furthermore, a fixed procedure for filling and leveling must br rigidly followed, and the capacity of the jar being tested must he approximately that of the one taken as the standard.
A . Gravimetrically with water and glass plate. Initial wei ht used. B . Gravimetrically with water and glass plate. Final weigft taken. C. Gravimetrically with mercury. D. Gravimetrically with lead shot and scraper. E . Volumetrically with ordinary buret and no indicating device. F . Volumetrically with microburet and perforated glass plate as an indicating device.
Table Trial
II. Reproducibility of Fill with Fine Lead Shot Weight of Shot T h e n Filled Level
Corresponding Capacity
Grama
M1. 4.31 4.29 4.30 4 . 3 31 4.29 4.32 4.34 4.29 4.31
1 2 3 45 6 7 8 9 10
27.79 27.70 27.76 2 7 . 98 34 27.68 2 7 . 87 27.99 27.70 27.84 Av.
27.81
Av. deviation 0 . 0 8 4
4.31 0.013
mercury, shown next in the table. Though accurate and precise results may thus be obtained gravimetrically by the use of water as a calibrating medium, this method is much too slow for routine work. Even the approximate method based upon the weight of the water initially contained in the jar is a little too slow and cumbersome for routine control work where the capacity of a large number of jars must be determined rapidly. Wien mercury is similarly used as the calibrating medium, no error occurs from the formation of an outside meniscus and generally none of the calibrating medium is trapped between the top of the rim and the glass plate. The actual procedure is to pour mercury into a weighed jar until it is full to overflowing and then to place a very clean glass plate on top and press down firmly to squeeze out the excess of mercury. After mercury globules are brushed away from the outside of the jar, the plate is lifted off and the weight of the mercury found. Manipulation is difficult if the plate is left in place for weighing, since one of convenient size and weight floats on the mercury surface, and even a larger and heavier one has a marked tendency to slide off. Determinations of the capacities of the same two jars used in the experiments with water as a calibrating medium are shown in Table I under Method C. Results of experiments on two other jars are also shown. Jar I11 was of exceptionally small capacity and IV a n experimental jar prepared from a commercial jar by grinding down the rim so as to make it very even. With water the capacity of jar IV was found to be 4.310 ml. Determinations with mercury as the calibrating medium are very Precise and are probably very accurate. Unfortunately, determination of the capacity of such jars by this method is not free from manipulative difficulty and is not very well suited to rapid routine work. Unlike determinations with water as a calibrating medium, those with mercury may, however, be made with ample accuracy for
In Table I1 is shown the result of an experiment to test the precision of this method. Jar IV was used and the capacitiee shown are based upon the average capacity as equal to that found by water and mercury calibration. On a single jar the precision is adequate for practical testing purposes. In Table I appear the results of actual determinations on commercial jars, using the average result obtained on jar IV as the standard. Ten trials were made on each jar. The capacities found are in fairly satisfactory agreement with those found by the use of mercury, except for jar I11 where there is a difference of over 0.1 ml. between the average results. This illustrates the necessity of using as a standard a jar having a capacity close to that of the .sample jar. I n practice a calibration curve constructed from the results on a series of jars of known capacities is convenient for calculating the results of capacity determinations by this method. The capacities of these small shallow jars may be rapidly estimated volumetrically by means of a buret, the jars being placed on a leveling table for filling. When an ordinary buret is used and the coincidence of the level of the liquid and the rim is judged by sighting across the top of the jar when a t the level of the eye, the results are neither very precise nor very accurate, as are indicated by the data in Table I, Method E These were obtained by a single observer from ten trials on each jar. Another observer working similarly found an average capacity of 3.35 ml. with an average deviation of 0.028 ml. for jar 11, and an average capacity of 2.72 ml. with an average deviation of 0 020 ml. for jar 111. I n order to eliminate these marked personal errors of observation, some mechanical means of indicating the point of proper fill must be used. The use of a straightedge or a taut wire placed across the rim is not satisfactory, since the calibrating liquid is usually drawn Up by capillary action to form a continuous surface across the bottom Of the straightedge or wire before the jar 1s completely filled, .thus giving a premature indication of the end ,point F~~example, one observer found by the use of a straightedge an average capacity of only 4.62 ml. with an average deviation
444
I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY
Vol. 17, No. 7
Of 0.052 ml. for jar 1, and an average capacity Of 2.56 ml. with an average deviation of 0.028 ml. for jar 111. The most satisfactory device was a square of plate glass with a central hole about 5 mm. in diameter. In use this perforated plate is placed on a jar with the tip of the buret a t the top of the hole and water is run in until i t just reaches the bottom of the hole: Care must be taken that no air bubbles are trapped under the plate, and it is sometimes necessary to tip the assembly slightly in order to bring trapped air bubbles to the hole. Observation is much facilitated if the water is colored with a suitable dye, and accuracy is increased if the rim is greased slightly with stopcock grease before the perforated plate is placed on the jar.
Again ten trials on each jar were made. By comparing these 6 t h those under Method E, their superiority in both accuracy and precision will be evident. For routine work, this volumetric method is sufficiently rapid, and the results arp PUToses. Of the methods amp1e in accuracy for here described it is probably to be preferred for routine control purposes.
In Table I under Method F appears a series of results obtained by this means, a microburet being used to increase accuracy.
The technical assistance of Florence Pines in the preparation of this paper is hereby gratefully acknowledged.
ACKNOWLEDGMENT
Polarographic Determination of Iron and Zinc in Phosphate Coatings JACOB KNANISHU AND THOMAS RICE Rock Island Arsenal, Rock Island, 111. A polarographic method for the simultaneous determinations of iron and zinc in commercial phosphzte coatings has been developed. The analyses are conducted in a supporting electrolyte containing 0.3 molar ammonium oxalate. The calibration charts for iron and zinc are presented in order that in similar analyses the iron and zinc content can be calculated from them. Complete analysis for iron and zinc can be completed in less than one hour.
T
HE need for a method for determining the composition of phosphate coatings on ferrous surfaces has long been recognized. Present methods of appraisal of phosphate coatings are based on accelerated breakdown tests or microscopic examination. Salt spray tests are being used to compare the relative protection of phosphate coatings, but are not indicative of the composition of the coating, and are not necessarily a reliable index of the thickness of the coating. Microscopic examinations are difficult, time-consuming, and limited in application by many factors.
by which a simultaneous determination of iron and zinc could be obtained on a sample of a few milligrams. The use of a polarographic method appeared feasible. An examination of a table of half-wave potentials of inorganic substances ( 1 ) revealed that zinc has a half-wave potential in 0.3 molar ammonium oxalate at -1.30 volts, and iron (both ferrous and ferric) has a half-wave potential a t -0.24 volt in 1 molar potassium oxalate. It was decided, therefore, to investigate the possibility of polarizing phosphate coatings containing iron and zinc in 0.3 molar ammonium oxalate solutions. Some previous work with iron and zinc in 0.3 molar ammonium oxalate solution in this laboratory had shown that sharp curves should be obtained for both iron and zinc without interference due to the presence of both metals. The half-wave potential of iron in this electrolyte was found to be -0.305 volt, and of zinc -1.406 volts. The iron remained in solution in the slightly ammoniacal ammonium oxalate solutions used in the experiments.
Table Curve No,
I.
Zn in 50 MI.
Analysis of Standard Zinc Solutions (Shunt, none) Galvanometer Deflection Microampere
M8.
2 3 4 5 6
Figure I. Typical Zinc Curve
Although the determination of the composition of phosphate coatings will not give the thickness, density, or relative protection of the films, such analyses will give the per cent by weight of iron and zinc phosphates in the coating. By removing the coatc ing from a unit area, the relative amounts of the iron and zinc phosphates in different phosphate coatings can be compared by the method described in this paper. In order to obtain a sample for the analysis of the phosphate coating on a metal article such as a machine gun link, it is necessary to scrape a small amount of the unoiled phosphate crystals from the surface. As only a few milligrams of coating will be available for analysis in many instances, a method was desired
7 8 9 10 11 12
0.04 0.08 0.12 0.I6 0.20 0.24 0.28 0.32 0.36 0.40 0.50
Table Curve No.
4 7 10.7 14.4 17.8 21.3 24.5 28.1 31.5 35 43.7
16
17 18 19 20 21 22 23
-1.36 -1.38 -1.375 -1.41 -1.4 -1,425 -1.42 -1.4 -1.45 -1.45 -1.4
0.08 0.14 0.21 0.288 0.356 0.426 0.49 0.562 0.63 0.70 0.874
II. Analysis of Standard Iron Solutions
Fe in 60 M1.
(Shunt, none) Galvanometer Deflection Microampere
Mo. 14 15
Half-Wave Volts
0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
4.2 6.7 8.6 12.0 14.0 16.1 18.7 21.3 23.3 26.0
0.084 0.134 0.172 0.240 0.280 0.322 0.374 0,426 0.466 0.520
Half-Wave Volt
-0.300 -0,285 -0,285 -0,300 -0.315 -0,310 -0.300 -0,330 -0.300 -0,320