resents the quantitative measurement of 0.015 y of copper. These results clearly show the suitability of neutron activation for this particular analysis. 'With the possible exception of zinc, the sensitivity is more than adequate for each element measured and exceeds t h a t obtainable by other methods with such small samples. PRECISION AND ACCURACY
The various sources of error to which neutron activation analysis is subject have been discussed in detail by Plumb and Lewis (6), and steps were taken to minimize these in this work. No treatm e n t at all was given the samples prior t o irradiation to prevent surface contamination. After irradiation, and before leaching, the samples were rinsed with water. It was not possible to apply any more stringent surface cleaning because of the e h e m e readiness of these very thin f i l m t o dissolve. Standard sampies of several of the elements of interest were irradiated simultaneously with the films t o eliminate a n y problems of fluctuations-in neutron level. Other factors affecting the accuracy of the results are leaching of contaminants from the backing, incomplete isotopic exchange during the chemical separa-
tions, and incomplete decontamination of the separated fractions from other activities. The first of these is probably not a serious problem as the dissolution of the film is accomplished in such a short time b u t could be evaluated more precisely by running a blank as discussed earlier. With regard to the second problem, the sample aliquots were heated in the presence of the carriers for some time i n - d e r t o assure isotopic exc h g e . Where possible, the results o b t s i d , & t h e separated fractions .were checked @nst those obtained by gamma spectrometry and found to be in good agreement. The possibility of incomplete decontamination was guarded against by checking the half life for each fraction showing a measurable activity. The factors affecting the precision of the measurements include the precision of the standard value, the precision of the chemical yield, the counting precision of the -le, and the precision of t h e sample aneight. I n this work the third factor mas h i t i n g in nearly all cases. N e precision for all the standards was &bout &0.5% (standard deviation). The chemical yields were measured to better than &l%,as discussed earlier. Except for copper, tungsten, and iron, where high counting rates were obtained, the counting precision for the
separated fractions was between 10 and 20%. This could have been improved somewhat by taking longer counting times; however, in this work the order of magnitude of the impurity concentrations was of most interest and a precision of AZO% (standard deviation) was quite satisfactory. I n the case of copper, tungsten, and iron, relatively high counting rates were obtained and the precision of these results is &2y0(standard deviation). However, as the precision of the sample weight is only *3%, the over-all precision of these nieasurements is about A4%. LITERATURE CITED
(1) Buschmann, E. C., Proc. Conf. Znd. A p p l . X-Ray Anal., 6th, Denver, 1957, 207. (2) Chu, W. L., Wolfe, J. E., Wagner, B. C., J . Chem. Phys., in press. (3) Heath, R. L., U. S. Atomic anergy Comm., 3Do-164084.4957). (4) Kleinberg, J., 'U. S. Atomic Energy Comm., LA4721 (Rev.) (1956). (5) Linder, M., U. S. Atomic Energy Comm., UCRL-4377 (1954). (6) Plumb, R. C., Lewis, J E., Nucleonics 13, No. 8, 42 (1955). (7) Thorn son, B. A., Strause, B. M., Leboeut M. B., ANAL. CHEM.30, 1023 (1958).
RECEIVED for review September 22, 1958. Accepted April 23, 1959.
Visual Titrations in Nonaqueous Solvents with Benzanhrone as mal Indicator *AM CHAND PAUL, JASWANT SINGH, and S c w USINGH SANDHU Department of Chemistry, Panjab hiversify, Hoshiarpur, lndia
A systematic study of the acid-base reactions in polar as well as nonpolar sdvents with benzanthrone as an internal indicator has been made. In phosphorus oxychloride, sulfuryl chloride, and thionyl chloride, solutions of pyridine, quinoline, stannic chloride, titanium tetrachloride have been used as titrants to indicate the reversible behavior of benzanthrone. In the case of arsenic trichloride, only the results given by basic titrants are subject to quantitative interpretation because the end point with acidic titrants cannot b e detected with precision. However, the behavior of benzanthrone in carbon tetrachloride and chlorobenzene is not reversible nor does it provide a means of quantitative assessment of acids and bases. Nevertheless, this difference in the mode of action of benzanthrone in polar and nonpolar solvents helps in understand-
ing the fundamental distinction between the two types of solvents.
T
acid-base titrations in solvents other than water with the help of indicator dyes have been infrequently reported ( 2 , 3, 18, 19, 22). The first attempt to study acid-base titrations quantitatively in acid chlorides with crystal violet as the indicator for the detection of the end point is credited to Garber, Pease, and Luder (S), who employed thionyl chloride. a highly reactive liquid, as the medium. Recently, Lewis acids and bases have been titrated with crystal violet as reversible indicator in acetyl chloride ( I ? , 20) and benzoyl chloride (16). Benzanthrone gives a red color in concentrated sulfuric acid, which turns to yellow on dilution with water or addition of alkalies. The literature does not HE
provide any substantial data as to the behavior of this indicator in aqueous solutions. However, the titrations of acids and bases in acetyl chloride and benzoyl chloride have already provided useful results. I n this investigation, benzanthrone has been employed as a useful indicator for the detection of the neutralization point of Lewis acids such as stannic chloride and titanium tetrachloride, nith bases such as quinoline, pyridine, a-picoline, and dimethylaniline, all of which are tertiary bases and do not contain any reactive hydrogen. The solvents employed are phosphorus mychloride, sulfuryl chloride, thionyl chloride, and arsenic trichloride. I n these four solrents conductometric and other tj-pes of work have already been done and their polar behavior has been indicated (1, 3, 5-9, 1 2 , 12, 21). Apart from this, a n attempt has also been VOL. 31, NO. 9, SEPTEMBER 1959
1495
Table 1.
Reagents
Fraction Collected, BoilinDg Point, C.
Reagent Phosphorus oxychloride (commercial)
How Treated Fractionated in all-glass apparatus
Sulfuryl chloride (commercial)
Distilled
67-9
Thionyl chloride (commercial) Arsenic trichloride
Distilled
74-6
Chlorobenzene
Carbon tetrachloride Stannic chloride Titanium tetrachloride
Fractionated
104-6
129-31
Distillate Mixed with 10% dimethylaniline (by wt.) to remove acidic impurities. Refractionated to produce colorless distillate Distilled over mercury and colorless liquid obtained at take off ratio of 1 :10 Treated with 2% quinoline, refractionated Treated with 10% dimethylaniline, refractionated
Purified by washing with sulfuric acid. Material washed with potassium carbonate solution and water, respectively. Dried over phosphorus pentoxide and distilled Dried over phosphorus pentoxide and distilled Kept over pieces of tin. Fractionally distilled in all-glass apparatus in dry nitrogen Kept over copper turnings. Fractionally distilled in all-glass apparatus in atmosphere of dry nitrogen
Quinolinea Pyridinea a-Picolinea Dimethylanilinea Purified by usual methods and fractionally distilled.
Final Product Used Pressure, Boiling mm. Hg point, C. O
106
740
69
73 1
75.5
732
130
732
131
739
76
740
112 5-13
71 I
130-1
740
229-30 112-2.5 126-6.5 192-3
740 i37 735 735
Q
Table II.
Solvent POC1, SOYClZ SOClZ -4sCla
Color in Solvent Yellow Light yellow Yellow Orange
made to study such neutralization reactions in carbon tetrachloride and chlorobenzene. The reagents used and their preparations are described in Table I. TITRATIONS
I n the titrations of acids and bases solutions of both Lewis acids and bases have been used as titrants to illustrate the reversihility of the indicator. The substance to be titrated was taken in 20 ml. of the solvent and about 5 mg. of indicator was added. During the titration the reaction mixture was stirred with a magnetic stirrer and guarded from moisture by using dry calcium chloride tubes. The color of the indicator changed through various shades and usually a precipitate formed. Sometimes the indicator was temporarily adsorbed by the precipitate. Sufficient time was, therefore, allowed t o bring it back into solution. The precipitate generally obstructed the detection of the color change because it 1496
ANALYTICAL CHEMISTRY
Color Changes with Benzanthrone as Indicator
Titanium Tetrachloride Red Pinkish red Deep red Pink-violet
Stannic Chloride Orange-red Red Orange-red Red
Quinoline Light yellow Light yellow Yellow Light yellow
could not be judged so precisely. Frequently the stirring of the reaction mixture was stopped, the precipitate was allowed to settle, the color change near the calculated end point was noted, and the corresponding amount of the titrant solution was taken as the volume found. The color changes of the indicator in the presence of acids, bases, and pure solvents are recorded in Table I1 and the results of these titrations are presented in Table 111. Phosphorus oxychloride has been studied extensively as a solvent and ample evidence supports its polar character (9, 11). Acid-base titrations with benzanthrone as indicator give valuable results, Acidic as well as basic solutions in phosphorous oxychloride' have been used as titrants and the reversible nature of the indicator has been conclusively proved. Phosphorus oxychloride is a highly acidic solvent but the color of benzanthrone in phosphorus oxychloride is yellow and the color at the end point depends upon the titrant so1,ution
a-Picoline Light yellow Light yellow Yellow Light yellow
Diniethylani1in e Light yellow Light yellow Yellow Light yellow
(Table 11). I n acidic titrant solutions a t the end point, the indicator is a shade of orange, while in basic titrant solutions, it is yellow. The results of these titrations can help in comparing the solvent properties of phosphorus oxychloride with those of other polar and nonpolar solvents (Table 111). It is apparent from Table I11 that the results of the titration of pyridine against titanium tetrachloride in sulfuryl chloride, when the latter is used as the titrant, are inconsistent as compared with the values theoretically expected. B u t when pyridine solution is used as titrant there is a little improvement. This anomalous behavior of pyridine, where it seems to act as a weak base, is unwarranted. Stannic chloride gives encouraging results with pyridine as well as quinoline. The shade of color a t the end point indicates that the solutions are acidic even before the end point is reached. The obvious conclusion from this observation is that pyridine as a base is weak while stannic
Table 111.
Titrant Stannic chloride Titanium tetrachloride Quinoline Pyridine
Qiiinoline Pyridine a-Picoline Dimethylaniline
Stannic chloride Titanium tetrachloride Quinoline Pyridine
Stannic chloride Titanium tetrachloride Quinoline Pyridine
Substance Quinoline Pyridine Quinoline Pyridine Stannic chloride Titanium tetrachloride Stannic chloride Titanium tetrachloride Stannic chloride Titanium tetrachloride Stannic chloride Titanium tetrachloride Stannic chloride Titanium tetrachloride Stannic chloride Titanium tetrachloride Quinoline Pyridine Quinoline Pyridine Stannic chloride Titanium tetrachloride Stannic chloride Titanium tetrachloride Quinoline Pyridine Quinoline Pyridine Stannic chloride Titanium tetrachloride Stannic chloride Titanium tetrachloride
chloride is a comparatively strong acid. I n thionyl chloride, pyridine and quinoiine act as strong bases and give good results. Titanium tetrachloride acts as a weak acid because other factors influence the results when a quinoline solution is used as the titrant. The behavior of arsenic trichloride as a polar solvent has been recognized (I, 5 , 6). I n titrations of acids and bases in arsenic trichloride, benzanthrone does not appear to act as a reversible indicator. The addition of acidic solutions as titrants does not affect the color of the acidic solution of benzanthrone a t the neutralization point. I n the acidic solutions of titanium tetrachloride the color of benzanthrone remains red even in the presence of excessive quantities of the titrant solution. However, the results are tolerable when solutions of bases such as quinoline, pyridine, a-picoline, and dimethylaniline are used as titrants. These reagents have been tried in the acid-base titration in carbon tetrachloride and chlorobenzene. It has been observed that the precipitate adsorbs the indicator irreversibly and i t is difficult to assess the results quantita-
Titration with Benzanthrone
Titrant Concentra- Weieht tion, Gram of SGbEquivalent/ stance, Gram Equivalents Titrated Liter Gram Experimental Theoretical Solvent, Phosphorus Oxychloride 0.046 4.10 X lo-' 0.17832 3.57 x 10-4 0.17832 0.049 6 . 2 4 X 10-4 6.20 x 10-4 0.11980 0.061 5.52 X 4.73 x 10-4 0.10316 0.058 7.74 X 7.34 x 10-4 0.13907 0.049 3.87 X 3.76 x 10-4 0.19840 0.083 9.92 X 8.73 x 10-4 0.22820 0.088 5.93 X 6.77 x 10-4 0.22820 0.057 5 . 9 3 X 6.00 x 10-4 Solvent, Arsenic Chloride 0.18410 0.058 4.16 X 4.42 x 10-4 0.22654 0.090 11.89 X 10-4 9.48 x 10-4 0.34366 0.048 3.44 X 3.68 x 10-4 0.34366 0.045 5.16 X 4.74 x 10-4 0.33246 0.054 4.32 X 10-4 4.15 x 10-4 0.33246 0.058 6.65 X 6.11 X 0.26156 0.077 6 . 2 8 X 5 . 9 2 x 10-4 0.26156 0.038 5.23 X 4.03 x 10-4 Solvent, Thionyl Chloride 0.20910 0 074 5.54 X 5.73 x 10-4 0.09002 0 046 5.76 X lo-' 5.80 x 10-4 0.09144 0.042 3.12 X 3.25 x 10-4 0.09144 0.069 8 78 X 1 0 - 4 8.73 x 10-4 o 06017 o 059 i 57 x i o - 4 4.53 x 10-4 0 06017 0 053 5 72 X 5 , 5 8 x 10-4 0 15332 0 059 4 37 X 4.54 x 10-4 0 15332 0 025 2 61 X lo-' 2.63 x 10-4 Solvent, Sulfuryl Chloride 0 25864 0 086 6 68 X 0 25864 0 066 9 05 X 10-4 0 33420 0 061 2 34 X l o e 4 0 33420 0 038 2 50 X 0 17072 0 054 4 27 X 10-4 0 11298 0 036 3 84 X 0 30918 0 062 4 64 X 0 12660 0 062 6 52 X
tively. I n the case of chlorobenzene also, the results cannot be interpreted quantitatively and have some qualitative value only so far as color changes of the indicator in acidic and basic solutions of the solvents are concerned ( I S ) . Noreover, the reaction seems to be slow, which indicates the absence of ions in these nonpolar solvents. On the wholr, these results of acidbase titrations in different solvents lead to a remarkable distinction between the two types of the solvents; polar and nonpolar. I n the case of phosphorus oxychloride, sulfuryl, and thionyl chlorides, and arsenic trichloride, the various color changes of the indicator, benzanthrone, in the presence of Lewis acids or bases can be explained on the basis of the existence of ionic species, which are the analogs of hydrogen and hydroxyl ions in water. The existence of ionic species in these solvents has already been reported. Acid-base neutralization reactions in acetyl chloride (4, 1 4 ) and benzoTl chloride (15) have been explained on the basis of the formation of solvo acids and solvo bases involving the reaction of the Leivis acids and bases n-ith thc
6.66 X 8.36 x 4.73 x 4.81 x 4.15 X 3.81 x 4.76 x 6.53
x
10-4 10-4 10-4 10-4 10-4 10-4 10-4
Color of Indicator a t End Point
%
Error -14 8 - 0 64 -16 7
Orange-yellow Orange-yellow Deep yellow Deep yellow Yellow Light yellow Light yellow Orange-yellow
- 5 45
- 2 92 -13 5
12 4 1 17
5.88 -25 4 6 52 - 8 86 - 3 94 - 8 84 - 6 08 -29 7
Light yellow Red Yellow Red Yellow Deep red Yellow Red
3 31 1 03 4 00 - 0 57
Orange-red Orange Red Red Light yellow orange Light yellow Yellow
- 0 88 - 2 51
3 74 0 76
- 0 30 - 8 37
50 5 48 0 - 2 89 - 0 79 2 10
0 15
Orange-yellow Orange-yellow Orange-yellow Orange-yellow Yellow Orange-yellow Light yellow Orange-yellow
solvent and by their mutual interaction to neutralize each other. Reactions in these polar solvents can also be explained similarly. For example, in arsenic trichloride solutions (IO), the Lewis acids, stannic chloride and titanium tetrachloride, which are dibasic acids produce the solvo acids after undergoing solvation. 2 AsC13 SnC14 ( . 4 ~ C l ~ ) ~ S nF?C l ~ 2 AsClz+ SnCIGz2 AsC13 $- Tic14 F! (A~C12)~TiC16 e TiClez2 AsC12f
+
+ +
The organic tertiary bases, quinoline and pyridine, are munoacid bases and undergo solvation to produce solvo bases.
6:
+ A4sC1,e
?I$ +
A
i,l
AsC13 e
ASCl:,
o:,: ~
VOL. 31, NO. 9 , SEPTEMBER 1959
4 IJ+AsC4 + c1.ksC13 :
1497
I n arsenic trichloride AsCla+ and C1are the ions characteristic of the solvent and correspond to H30+ and OH- in aqueous solutions. Thus the neutralization of the Lewis acids, stannic chloride and titanium tetrachloride, in arsenic trichloride by a monoacid base pyridine can be represented as :
+
[AsC12]2+[SnC16]2- 2
2 C1-
[
0
+AsC12
+
e O A s C l 2 ] SnCle + 2 hsC13 2
G [0 +
[AsCI~]2+[TiCl~]2-2
2 C1-
e
AsC12
AsC12] Tic16
+
+ 2 AsC13
2
From this account i t appears that in these titrations a fundamental difference exists in the two types of solvents, which has been recognized since early times. LITERATURE CITED
(1) Anderson, H. L., Lindqvist, I., Acta Chem. Scand. 9, 79 (1955). (2) Davis, M. M., Schuhmann, P. J., Lovelace, M. E., J . Research Natl. Bur. Standards 41, 27'( 1948). (3) Garber, E. B., Pease, L. E. D., Luder, W. F., ANAL.CHEM. 25, 581 (1953). (4) Goyal, K., Paul, R. C., Sandhu, S. S., J . Chem. SOC.1959, 322. (5) Gutmann, V., Monatsh. 83, 159 (1952). ( 6 j Zbid., 84, 1191 (1953). (7) Zbid., 85, 393 (1954). ( 8 ) Zbid., p. 404. (9) Zbid., p. 1077. . (10) Gutman, V., 2. anorg. u. allgem. Chem. 266, 331 (1951). (11) Zbid., 270, 179 (1952).
(12) Henry, M. C., Hazel, J. F., McNabb, W. M.,Anal. Chim. Acta 15, 187 (1956). (13) Luder, W. F., McGuire, W. S., Zuffanti,. S.,. J . Chem. Educ. 20, 344 (1943). (14) blanhas, B. S., Paul, R. C., Sandhu, S. S., J . Chem. SOC.1959, 325. (15) Paul, R. C., Chander, K., Singh, G., J . Indian Chem. SOC.35,869 (1958). (16) Paul, R. C., Sandhu, S. S., Singh, J., Eingh, G., Zbid., 35, 877 (1958). (17) Paul, R. C., Singh, J., Sandhu, S. S., Chem. & Znd. (London) 1958, 622. (18) Rice, R. V., Zuffanti, Saveno, Luder, W. F., ANAL. CHEY. 24, 1022 (1952). (19) Seaman, William, Allen, Eugene, Zbid., 23,592 (1951). (20) Singh, J., Paul, R. C., Sandhu, S. S., J . Chem. SOC.1959,845. (21) Spander, H., Brunneck, E., 2. anorg. u. allgem. Chem. 270, 201 (1952). (22) Tomirek, O., Sbornik Celosthtni Pracovni Konf. Anal. Chemikai 1 , 246 (1952).
RECEIVED for review September 5, 1958. Accepted May 4, 1959.
Thermometric Titration of Acids in the Presence of Hydrolyzable Cations F. J. MILLER
and P. F. THOMASON
Analytical Chemistry Division, Oak Ridge National laborafory, Oak Ridge, lenn.
,A
thermometric titration method is presented for determining acid in aqueous solutions of zirconyl ion in hydrofluoric acid, uranyl ion in sulfuric acid, uranyl and copper(l1) ions in sulfuric acid, thorium(1V) ion in nitric acid, and uranyl ion in nitric acid. This "free acid" is determined by titrating the test solution with a standard solution of a base and recording the change in temperature during the titration. Temperature, as detected by a thermistor, is plotted automatically by a recording potentiometer vs. volume of titrant added. The end point of the titration is obtained by extrapolating the straight-line portions of the curve to their point of intersection. Milliequivalent quantities of free acid can be determined. For 0.1 meq., the relative standard error is f1 to &2% a t the 95% confidence level; for 0.01 rneq., it increases to about
&5%.
S
methods are being used for the determination of "free acid"i.e., acid in the presence of hydrolyzable cations. Most of these are discussed by Booman et al. (1). I n general, these methods involve the addition of a complexing agent followed by one of several different types of titrations. EVERAL
1498
ANALYTICAL CHEMISTRY
Thermometric methods of titrations have been known to analytical chemists. Jordan and Alleman ( 3 ) have applied this method most recently. They have shown that by prior consideration of thermodynamic data, the feasibility of a thermometric titration can often be determined without experimentation. EXPERIMENTAL DETAILS
Apparatus. STIRRING MOTOR, A. S. LaPine a n d Co., Catalog No. 38286. T h e slow-speed shaft that operates at 320 r.p.m. was used. The stirring motor should have a cooling coil of I/*inch diameter copper tubing wound around it. A small flow of water is passed through the copper tubing t o prevent excessive temperature rise in the motor during operation. GLASSSTIRRER. A flat piece of glass, inch wide, twisted into a spiral 11/* t o 2 inches long, was attached t o a glass rod inch in diameter for use as a stirrer. THERMISTOR. A thermistor that is suitable for measuring temperature changes is Type 32-a-1, manufactured by the Victory Engineering Corp., Springfield Road, Union, N. J. It has the following characteristics: Sensitivity in the region of 25' C., 0.04 ohm/ohm/" C. Resistance at 25' C., -2,000 ohms Thermal time lag,