April 15, 1931
INDUSTRIAL AND ENGINEEBlNG CHEMISTRY
be made with results on arsenical residue determinations. It should be remembered, however, in making such comparison that these results express approximately 80 per cent of the total residue, whereas arsenical residue results are always expressed in grains arsenic trioxide per pound, which in the case of lead arsenate is probably approximately 30 per cent of the total residue. One point to be remembered in following this procedure is that the volume of solution must be sufficiently large at all times to keep in solution all barium fluoride and barium fluosilicate that may be formed before the final precipitation of barium as barium sulfate. As shown in the table, these analyses were made on samples which had ,been subjected to considerably different treatments, including check samples which had received as many as five cover sprays, samples from the orchard immediately after spraying, and samples from commercial washing ex-
147
periments with a number of different solvents on apples which had received five cover sprays of barium fluosilicate. Naturally, under such conditions considerable variation in duplicates may be expected, owing to variations in sampling. Actually, however, the agreement between the duplicate analyses was very close in most cases. The difference between duplicates which does not appear to be a function of the amount of residue, has a mean value of 0.011 grain per pound under the conditions of these experiments. This corresponds to an average error in the mean of duplicate determinations of *7 per cent when the total residue approximates 0.08 grain per pound. This is probably within the limits of experimental error in sampling. Based on these results this method may be considered sufficiently accurate and reliable to give at least comparative results for the determination of barium fluosilicate as spray residue on apples.
Further Applications of the Centrifugal -Filtration Tube' Evald L. Skau and Louis F. Rowe JARVIS
CHEXICAL LABORATORY, TRINITY COLLEGE, HARTFORD, CONN,
Experimental data here reported show that the is pressed flat on the disk and centrifugal filtration tube may be used satisfactorily folded well down over the tion tube, a simple confor the approximate estimation of solubilities at differedges. The disk is then trivance for the rapid ent temperatures. The binary freezing-point diagram pushed down upon the shouls e p a r a t i o n of solids from of the system benzene-naphthalene, which has been der B so as to fit snugly to liquids at any desired temperaconstructed by means of this method, shows fair the walls of the tube. ture, has been described in agreement with the accurate diagram. The mixture of known coma previous publication (a) and In such cases, as well as in water solutions of solids, position is now weighed into experimental data were there the eutectic composition may be estimated by a single the tube H , tube G is weighed presented to show its usefuldetermination by means of this device. This is proved and fitted into place, and the ness in purification by reby experimental results for the above system and for sample is melted completely crystallization, especially in the systems water-sodium nitrate and water-potassium by heating. The whole tube the case of compounds which iodide. is then immersed for from 20 require low-temperature techThe centrifugal filtration tube should be applicable to 30 minutes in a thermonic for their separation.2 The in any plant-control work in which, for example, it stat set a t the desired temadditional applications of this is desired to carry a reaction to a stage where a certain perature, t, being protected device are based upon the fact consistency or freezing point is reached. Instead of from contact with the liquid that it can be used to detera rough determination of the freezing point, a deterof the bath. mine the approximate promination of the amount of solid separated at some Note-For temperatures above portions of solid and liquid in 0' C. a rubber sac such as may be arbitrary temperature could readily be made with a mixture at any definite temmade by cutting off the end of a t o y suitable accuracy. perature. It can therefore be balloon is suitable. For low temused in certain cases for the peratures where rubber cannot be rough construction of binary freezing-point diagrams and for used, the tube may be inserted within another glass tube. the rapid estimation of eutectic compositions and temperat#ures, It is then inverted quickly and immediately centrifuged for and has practical applications in plant control work. a few minutes. The liquid is thus thrown into chamber G, the bulk of the separation being accomplished within 20 secRough Construction of a Binary Freezing-Point Diagram onds of the time when the tube was removed from the bathInstead of determining the freezing point of a given compo- that,is, before the temperature has changed appreciably. The sition, the composition of the mixture having a given freezing solid phase is now removed with the disk and the weight point is found-that is, a mixture of known composition is of the liquid in G ascertained. The solid phase must be brought to equilibrium a t the desired temperature (below its identified. primary freezing point) and the proportion of solid and Since this liquid was at equilibrium with the solid phase liquid determined. The composition of the liquid can then a t the temperature of the thermostat, its freezing point is t. be calculated. If the original mixture consisted of a grams of A and b grams A circular piece of filter paper about 1.5 cm. larger in of B , and if 8 b grams of solid B have been separated out by diameter than the perforated porcelain disk C (Figure 1) centrifuging, the liquid having the freezing point t has the is folded along a diameter and then slit with a pair of scissors composition a b - Sg part way across the perpendicular diameter starting a t the of A and of B a f b - S b a + b - S b center. The wire E is passed through this slit and the paper The mother liquor and the crystals may now be recombined 1 Received October 23, 1930. and another determination may be made at other temperaThese tubes are now obtainable from Eimer and Amend, New York.
HE centrifugal filtra-
T
'
148
ANALYTICAL EDITION
H
E
F
B
VOl. 3,
KO,
2
tures on the same sample if desired. It is recommended, however, that a fresh sample be used for each determination. I n this way the whole freezing-point diagram of the system may be constructed, the freezing points of the pure constituents being obtained by one of the usual methods.
mination is comparatively simple, since the temperature should remain constant during freezing until the sample is practically completely solidified. With systems showing simple compound formation there would be two eutectics, and these could be determined by using the same procedure on two samples, one rich in A and the other in B. The formation of solid solutions would cause deviations from the true diagram.
Eutectic Composition Found in One Determination
Table I shows the agreement in the benzene-naphthalene system between the results obtained by the use of centrifugal filtration tubes and those determined by the more accurate freezing-point method ( I ) . The benzene used was purified by repeated recrystallization and the final product had a freezing point of 5.5" C. The naphthalene was purified by sublimation and melted a t 80.0" C.
If the eutectic temperature were known the eutectic composition could easily be found by the above method. If, however, the eutectic temperature is unknown but lies below room temperature, the eutectic composition can be found by a single determination, provided one of the pure constituents is a solid a t room temperature, by cooling a known mcxture having an excess Figure I-Centrifugal FiltraOf that constituent to well below tion Tube (Inverted) the eutectic temperature, thus solidifying the mixture completely in H, and then centrifuging it in the inverted tube while it is warming up so that any liquid is whirled off through the disk as soon as it is formed. Under these conditions the liquid collected in G will have the eutectic composition, and from its weight its composition can be calculated as above. Let us consider, for example, the simp!e eutectic system benzene-naphthalene. If an equimolecular mixture (having a primary freezing point of 58.8' C.) be cooled to -20" or -40' C. and then allowed to come slowly to room temperature, liquid having the eutectic composition would start to
G
Results
Table I-Roueh
CeHs
Construction of Benrene-Naphthalene System MOLFRACTION C6HBIN
MOL TEMP. cloHe FRAC-BEFORE MOTHER$R : TION CENTRI- LIQUOR LEFT FUOING CBH~
L1'r'oR
Expt.
I
Grams Grams ' C. Grams 12.8 7.40 5.29 2 1 , l 5 0 . 2 9 1 9.155 7 . 2 7 2 7.868 0 . 6 0 3 -105 2 5 . 5 12.285 7.272 7.868 0.603 4.645 3 . 9 8 5 7 . 8 8 5 0 . 4 5 3 -14a 5 0 . 5 10.285 3.985 7.885 0.453
DIFF.
t
CioHs CioHs CIOHB CioHs CloHs
a Determinations of eutectic composition. -3.50 c .
p:,";"' 1 I
Eutectic temperature,
Two cryohydric points were also determined and the results are shown in Table 11. Merck's "Highest Purity" sodium nitrate and Mallinckrodt's C. P. potassium iodide were used t o make up these solutions. It was found in preliminary experiments that the best results are obtainable when there are approximately equal amounts of the solid phase and of the mother liquor to be
Table 11-Estimation of Composition a t Cryohydric Points (Temperature before centrifuging, -70"
c
W T .OF SOLUTE
KI NaKOs 4
Grams 20.0 15.85
WT. OF HzO
Grams 11.02 12.05
CONCN. OF SOLVT~
wt.70 64.5 56.8
EUTECTIC
TEMP
c. -23.2a -17.5b
MOTHER LIQUOR
Grams 21.231 14.5g5
PHASE
KI
Y a SOa
1 1
SOLUTE I N MOTHER 1,IOUOR
Expt.
JVt. 9% 48.1
1
Theoret
88 5
F:z 38.4b
1
DiR
1 I
wt.% -3.8 1
.+o.
Kremann and Kershbaum, 2. anoyg. Chem., 66, 218 (1907). J. A M . Chem. Soc , 40, 1204 (1918).
t, Rodebush,
appear at the eutectic temperature (-3.5' C.) and would continue to form at the expense of the two solids, the temperature remaining constant, until all of the solid benzene was used up. The excess of naphthalene would then start to dissolve in the eutectic liquid and the temperature would start to rise again until at room temperature, t o C., the mixture would consist of solid naphthalene suspended in a liquid whose freezing point is t o C . If this warming-up process took place during centrifuging as described above, however, the liquid formed at the eutectic temperature would be removed at that temperature before it had dissolved any more naphthalene, and the excess of naphthalene present at that temperature would be left on the disk, since its melting point is 80" C. Thus the amount of liquid and solid obtained would be exactly the same as if the mixture had been completely melted and then brought to equilibrium in a thermostat set a t the eutectic temperature. The eutectic temperature can be found by determining the freezing point of the liquid collected in G. This deter-
separated. Thus the large deviation in the case of the last determination in Table I is probably due to the fact that the solid fraction consisted of only 1.5 grams-that is, only 10 per cent of the whole sample. I n general the method is accurate to within 3 or 4 per cent. It will be shown in forthcoming papers that this compares quite favorably with the accuracy often attained by the ordinary Beckmann freezingpoint method when applied to organic binary systems. Acknowledgment
The senior au'thor wishes to express his thanks for grants from the Gyrus M. Warren Fund of the American Academy of Arts and Sciences and from Curtis H. Veeder, which have made this and other investigations possible. Literature Cited (1) International Critical Tables, Vol. IV, p. 177. (2) Skau, J . Phys. Chem., 33, 951 (1929).
4