to 1.002 cp, common basis at 20" C. The last column of the table gives the per cent deviation (in parentheses) between our measured and calculated data. The correlation represents the data with an average deviation of *0,17%, and with a maximum deviation of 0.49% at 40" C. It is interesting to note that Kampmeyer (12) also found the largest deviation from a similar correlation at approximately the same temperature. The fit was very satisfactory at higher temperatures where the literature data are most controversial. The excellent agreement of our data, up to 125" C, with the measurements of Hardy at the National Bureau of Standards is encouraging. The new correlation represents all tabulated reference data with an average deviation of k 0.0018 cp. Figure 3 shows a deviation plot of the data given in Table IV from the correlation. The correlation represented by the zero line is centered between the experimental data, while the correlation given by Kestin and Whitelaw, which is represented by the dotted line, remains consistently on the low side. Finally, Table VI lists the recommended water viscosity
values over the entire temperature range. Whenever experimental values were available, those were entered in the table, In reviewing the paper, R. E. Manning recommended the use of the following equation for interpolation: ?p40 log =
A(t
- 20)
+ B(t - 20)2
9 C+t where A = 1.37023, B = 0.000836, C = 109, t
=
"C.
This correlation represents the experimental data within f 0.05% average deviation.
ACKNOWLEDGMENT The authors thank A. S. Borsanyi for his help in the development of the apparatus. Viscosity measurements were made by T. R. Middleton.
RECEIVED for review June 8, 1967. Accepted September 18, 1967. Work sponsored by the Office of Saline Water, U. S. Department of Interior.
Study of Cerium(1V)-Thallium(1) Reaction and Analysis of Thallium(1) Mixtures by Kinetic Differences George H. Schenk and William E. Bazzelle Department of Chemistry, W a y n e State University, Detroit, Mich. 48202
The uncatalyzed cerium(lV)-thallium(1) reaction in 0.125M sulfuric acid exhibits an extremely slow but measurable rate in the dark, while in room light, the rate of disappearance of cerium(lV) increases, presumably because of thallium(1)-catalyzed photoreduction. Catalysis by manganese(ll1) was so effective that photometric titration of thallium(1) with cerium (IV) titrant was possible. Two types of thallium(1)metal ion mixtures can be analyzed on the basis of kinetic differences alone. In the first type, metal ions such as iron(l1) react rapidly with the cerium(lV) titrant in the absence of catalyst before thallium(1) reacts significantly. Then the catalyst is added and the thallium(1) i s titrated with cerium(1V). In the second type of mixture, the catalyst is added at the start, and thallium(1) reacts rapidly with the cerium(lV) titrant before the other metal ions or reducing agents react significantly. The reducing agents studied were mercu ry( I), a rsen ic( I I I), and ch romium( I II). Theoretical calculations indicate the reactions of these substances with cerium(lV) should be negligible, but weakcatalysis by manganese(ll1) apparently causes significant reaction above equivalent concentrations.
THAT OXIDATION-REDUCTION reactions involving the exchange of unequal numbers of electrons (and little structural change) tend to be slow has often been postulated (1, 2 ) . A survey of this type of reaction involving cerium(1V) in sulfuric acid (Table I) provides a useful test of this postulate. There is a fairly large spread in the magnitude of the rate constants, but only the cerium(1V)-uranium(1V) reaction appears to be rapid at high (0.1M) concentrations. For analytical (1) P. A. Shaffer, J . Am. Chem. SOC.,55, 2169 (1933) and J. Phys. Chem., 40, 1021 (1936). (2) J. Halpern, Can. J . Chem., 37, 148 (1959). 162
ANALYTICAL CHEMISTRY
purposes, all of the reactions in Table I are run at high temperatures, with catalysts, or in hydrochloric acid solvent. None of the reactions are as rapid as the cerium(1V)-iron(I1) reaction which has a rate constant of about 10GM-'set-1 in 0.5M sulfuric acid (3). An even slower reaction than those listed in Table I is the cerium(1V)-thallium(1) reaction. It has a complex rate law in 6.18M nitric acid ( 4 ) with rate constants of k l = 0.055 hr-1 and kz = 1.221V-~ hr-I at 54" C. No one has studied the uncatalyzed reaction in sulfuric acid although Shaffer (1) reported qualitatively that no reaction is observed. Photoreduction of cerium(1V) in sulfuric acid is slow compared to other acids, but Sworski ( 5 ) reported that thallium(1) enhances the rate of photoreduction. Because other potential interferences, such as those in Table I, react slowly with cerium(1V) in sulfuric acid, it is of interest to study the uncatalyzed reaction and to investigate catalysts for the purpose of determining thallium(1) by titration with cerium(1V) in sulfuric acid. Historically, of course, thallium(1) has been determined with numerous oxidizing agents. Koreman (6) has summarized the gravimetric and titrimetric methods (7-10) which do not involve cerium(1V).
(3) G . Dulz and N. Sutin, Inorg. Chem., 2,917 (1963). (4) M. K. Dorfman and J. W. Gryder, Ibid.,1,799 (1962). (5) T. J. Sworski, J. Am. Chem. SOC.,79,3655 (1957). (6) I. M. Koreman, "Analytical Chemistry of Thallium," Daniel Davey and Co., Inc., New York, 1960, pp. 76-84. (7) A. J. Berry, Analyst, 51, 137 (1926). (8) G. S. Deshmukh, Anal. Chim. Acta, 12, 319 (1955). (9) E. Rother and G . Jander, Z . Aneu. Chem., 43,930 (1930). (10) C. D. Fresno and J. Valdes, Z . Anorg. Allgen. Chem., 183, 258 (1929).
~~~
~
~~
Table I. Reactions of Ce(1V) in Which Unequal Numbers of Electrons Are Exchanged (Uncatalyzed) k, 25" C, H2S04(M-1 sec-1) Rate law Reactants and ref. -d[CeIT] ~3 . 8 x 103 - 2 k[UIl][CeIV] 2 Ce(IV) U(1V) (11) dr (at constant H+) (2M HClO4)a 6.3 X 101 _-d[CeIY _ _ _ - 2 k[Sn1I][Ce1\] 2 Ce(IV) Sn(I1) (12) dr - d[CeIY] ~2.7 x 10-2 - k[HgZi2][CeIV] Ce(1V) Hg," (13) dt
+
+ +
3 Ce(1V)
+ Cr(II1) (14) kz = 1 . 6 X 10-3N-1sec-1
+
a
Complex; varies from 2nd order to 3rd order 2 Ce(1V) As(II1) (15) Extrapolated from rate constants given at temperatures other than 25" C.
Numerous methods of determining thallium(1) by titration with cerium(1V) have also been reported (16-18). All of these methods utilize catalysis by some species. The first was reported by Willard and Young (16). The reaction is carried out in 1.8M hydrochloric acid and at 50-60' C. The elimination of interference from chromium(II1) requires that the reaction be run at 35" C and in the presence of 3 . 6 M hydrochloric acid. A procedure involving iodine monochloride as catalyst was developed by Rao and Rao (17). This enabled the titration to be carried out at room temperature in 1 M hydrochloric acid. An amperometric method for thallium(1) was developed by Singh (28) in hydrochloric acid solution. All of the cerium(1V) methods apparently depend on the catalytic effect of chloride suggested by Duke and Borchers (19). The authors felt that the use of catalysis would permit the determination of thallium(1) by titration with cerium(1V) in sulfuric acid medium. In all preceding methods iron(II), arsenic(III), and/or chromium(II1) must be absent or present only in small quantities for meaningful results to be obtained. It was felt that a method could be devised, using selective catalysis, for the determination of thallium in the presence of the above mentioned ions. EXPERIMENTAL
Apparatus. A Beckman Model DB spectrophotometer was used to follow the reaction rates. A Bausch-Lomb Spectronic 20 was used for photometric titrations and was modified (20) for this purpose. Reagents. All reagents conform to ACS reagent grade specifications unless otherwise stated. CATALYST.Manganous sulfate (940 mg) was dissolved in 800 ml of water containing 12 ml of sulfuric acid. This ~
(11) F. B. Baker, T. W. Newton, and M. Kahn,J. Phys. Chem., 64, 109 (1960). (12) C. H. Brubaker and A. J. Court, J . Am. Chem. SOC.,78,5530 ( 1956).
(13) W. H. McCurdy and G. G. Guibalt, J . Phys. Chem., 64, 1825
(1960). (14) J. Y. Tong and E. L. King, J . Am. Chem. Soc., 82,3805(1960). (15) J. W. Moore and R. C. Anderson, Zbid.,66,1476 (1944). (16) J. J. Willard and P. Young, J . Am. Chem. SOC.,52, 36 (1930). (17) K. B. Rao and G. J. Rao, 2.Anal. Chem., 168,83 (1959). (18) D. Singh and V. S . Agarwala, J . Sci. Ind. Res., 21B, 212 (1962). (19) F. R. Duke and C. E. Borchers, J . Am. Chem. SOC.,75, 5186 ( 1955). (20) C. Rehm, J. I. Bodin, K. A. Connors, and T. Higuchi, ANAL. CHEM., 31,483 (1959).
(over limited concentration range)
solution was cooled to approximately 10" C, and then 200 mg of potassium permanganate was added. This solution was maintained at 10" C for 2 hours with periodic stirring. The mixture was then diluted to 1 liter, covered with aluminum foil to protect from light, and allowed to come to room temperature. This gave a colloidal solution which contained some precipitation of manganese dioxide. This solution is stable for about 1 week; then the manganese(II1) begins to disproportionate. A 0.01N solution was prepared by dissolving THALLIUM(~). thallous nitrate (99.9z, K and K Laboratory) in 0.1M sulfuric acid. In some cases the solution was prepared by dissolving thallium metal in 9 M sulfuric acid and adding 5 to 10 drops of nitric acid per 2 liters. The thallium(1) solutions were standardized by precipitation of thallous chromate. CERIUM(IV). A 0.037M stock solution in 0.21M sulfuric acid was prepared from cerium ammonium sulfate (G. F. Smith, specially prepared for protein bond iodine experiment) and standardized against a standard arsenic(II1) solution using osmium tetraoxide as catalyst. CHROMIUM(III) SULFATE. A 0.01 6 N aqueous solution was prepared from KCr(S04)2* 12 H20. IRON(II). A 0.04M stock solution in 0.21M sulfuric acid was prepared from ferrous ammonium sulfate. CERIUM(III).A 0.02M stock solution in 0.21M sulfuric acid was prepared from Cez(SO& 8 HzO. ARSENIC(III). A 0.01N solution was prepared by dissolving arsenic(II1) oxide in base and neutralizing with sulfuric acid, Reaction Rate Procedure. The rates of the uncatalyzed and silver(1) catalyzed cerium(1V)-thallium(1) reactions were studied at a constant temperature of 25.0" C. Flasks were wrapped completely in aluminum foil or left uncovered in subdued room light away from sunlight. Aliquots were withdrawn and the unreacted cerium(1V) was measured at wavelengths from 430 to 470 mp without quenching because of the slow reaction rate. Thallium(1) and thallium(II1) have been found to have no effect on the light absorption of cerium(1V) solutions (3, so that a calibration curve could be made from cerium(1V) alone. Photometric Titration Procedure. Pipet an aliquot of a solution containing 0.15 to 0.35 meq of thallium(1) into a 125-ml titration flask modified to be used for photometric titration (20). Stir the manganese catalyst solution and add about 15 ml to the titration flask. Stir the resulting solution for 2 minutes. The total volume of solution should be from 60 to 90 ml, and the acid concentration of this solution should be no lower than 0.1M in sulfuric acid. A sulfuric acid concentration of 0.10 to 0.125M is recommended. Below O.lOM, precipitation of cerium(1V) may occur, and above 0.125M the reaction rate begins to decrease. Record the initial absorbance reading at 425 mp. VOL. 40, NO. 1, JANUARY 1968
163
I
Table 11. Rates of Uncatalyzed and Silver(1)-Catalyzed Reaction at 25.0 i. 0.1" C 0.005M Ce (IV), 0.005M TI (I), 0.125M H2S04 Molarity of Ce(1V) after given time at conditions below Subdued Darkness 1 X Time, hr. lighta Darkness lO-3M Ag(1) 0 5 . 0 0 x 10-3 5.00 x 10-3 5.00 x 10-3 70 4.73 x 10-3 4.80 x 10-3 4.55 x 10-3 240 4.42 x 10-3 4.61 x 10-3 4.32 x 10-3 409 4.04 x 10-3 4.42 x 10-3 3.88 x 10-3 1177 2.40 x 10-3 3.60 x 10-3 3.08 x 10-3 a By "subdued light" is meant room light for ca. 14 hours a day. The solutions were not exposed to direct sunlight.
I
I
I
I
I
I
I
+
Add the cerium(1V) titrant in units of 0.25 to 0.5 ml, and allow the solution to stir for 1 minute before reading the absorbance. Continue this until the change in absorbance is 0.04 to 0.06 for 0.25-ml increments or 0.09 to 0.11 for 0.5-1111 increments of 0.037N cerium(IV), then take 3-4 readings further. Plot the absorbance, corrected for dilution, us. milliliters and extrapolate to the end point.
1
I
E n
N
0
0.1
tI
0.0
1 I
I
1.0
I
4.0 Ce(lV)
2.0 3.0 mi
I
I
5.0
I
I
6.0
7.0
I
THE CERIUM-THALLIUM REACTION
Figure 1. Photometric titration of 0.222 meq of thallium(1) with 0.037N cerium(1V)
The Uncatalyzed Reaction. The rate of uncatalyzed cerium(IVbthallium(1) reaction was studied in 0.125M sulfuric acid in subdued light (room light) and in the dark (Table 11). The rate of the reaction in subdued light is variable and somewhat greater than the rate in the dark. This is consistent with Sworski's (5) finding that thallium(1) catalyzes photoreduction of cerium(1V) in 0.4M sulfuric acid, [We found that in the absence of thallium(I), 0.005M cerium (IV) solutions in 0.125M sulfuric acid are stable in subdued light for over 160 hours.] The data in Table I1 for the reaction in the dark appeared to fit the integrated second order rate law
great as the uncatalyzed reaction. Silver appears to be a somewhat better catalyst for this reaction than for the cel ~cerium(1V) rium(1V)-mercury(1) reaction. At 1 0 - 3 M l e ~ eof and mercury(I), the initial rate at 50' of 1.7 x 10-5M-1min-1 of the uncatalyzed reaction (8) was increased to 4.3 X 10-5M-1min-1 by 2.4 X 10-3Msilver(I) catalyst (21). No mechanism is suggested for the siIver(1) catalysis ; in 1.5714 perchloric acid the following mechanism has been suggested (22) for silver(1) catalysis of this same reaction:
(a
- 2 b)kt
=
b(a - 2 x) In a(b - X )
where a is the initial cerium(1V) concentration, b is the thallium(1) initial concentration, and 2 x and x are the respective amounts of each that reacted after time t . In 0.125M sulfuric acid solvent, k was found graphically to be 0.03M-l hr-l. By comparing the initial rate of reaction of 0.005M thallium(I)-O.O02M cerium(1V) with the initial rate of reaction of 0.005M thallium(I)-0.001M cerium(IV), it was verified that the order in cerium(1V) was approximately one. This is in accord with the rate law in Equation 1. The kinetics of the reaction are under further study. The rate of the reaction thus varies enormously with the acid solvent. In warm 3M hydrochloric acid, the reaction is fast enough to be used for direct titration (16). In 6M nitric acid, the half life is 10 hours at 54" C. In 0.125M sulfuric acid, the half life is calculated from k to be 2700 hours at 25" C under the conditions of Table 11. The Silver(1)-Catalyzed Reaction. A 1 X 10-3M level of silver(1) catalyst does increase the rate of the reaction (Table 11), but not as much as needed for titration purposes. Estimation of the initial rates of the uncatalyzed and catalyzed reactions by determining graphically the slope of concentration us. time curves at zero time indicated that the initial rate of the silver (1)-catalyzed reaction was over four times as 164
ANALYTICAL CHEMISTRY
Ce(1V)
+ Ag(1) e Ce(II1) + Ag(I1)
Ag(I1)
+ T U ) -,&(I) + TKII)
(2) (3)
Thallium(I1) is oxidized in a fast step by cerium(1V) to thallium(II1). The Manganese(II1)-Catalyzed Reaction. Manganese(II1) was found experimentally to be an excellent catalyst for the cerium(1V)-thallium(1) reaction. Krishna and Sinha (23) reported that in 0.49N sulfuric acid, the reaction was almost complete in 5 minutes in the presence of 1.75 X 10-4N manganese(II1). At 35" C, they found the reaction was second order in cerium(1V) with a rate constant of 3.4 X 10-3M-lmin-1 at 3.5 x 10-5Nmanganese(III). With this type of catalytic efficiency, it was only necessary to devise suitable titration conditions. By using a 1.5 X 10-3M concentration of manganese(II1) catalyst in the titration vessel, a fairly rapid photometric titration of thallium(1) could be achieved. The end point was detected by the appearance of excess yellow cerium(1V). A typical photometric titration curve is shown in Figure 1. The initial absorbance is the result of colored manganese(II1) (21) G. G. Guibault and W. H. McCurdy, J . Phys. Chem., 70, 656 (1966). (22) W. C. E. Higginson, D. R. Rosseinsky, J. B. Stead, and A. G. Sykes, Discussions Faraday SOC.,53,49 (1960). (23) B. Krishna and B. P. Sinha, 2. Physik. Chem., 212, 149 (1959).
catalyst which remains constant throughout the reaction. No decrease in the intensity of color of the catalyst is apparent after mixing the catalyst with thallium(1). Krishna and Sinha (24) declined to comment on the mechanism of the catalysis by manganese(II1) but did postulate that manganese(1V) acted in the following manner :
+ Mn(1V) -(rapid)+ Mn(I1) + Ce(1V) -(slow)+ Mn(II1) + Ce(1V) -(slow)-. Tl(1)
+ Mn(I1) Mn(II1) + Ce(II1) Mn(1V) + Ce(II1) Tl(II1)
(4)
Table 111. Titration of Thallium(1) with 0.03687N Cerium(1V) Meq. No. of Meq. Error, Abs. std. trials taken dev. found
z
0.1665 0.2220 0.2775 0.3330
3 11 6
0.1675 0.2221 0,2752
0.60 0.05 0.84
3
0.3329
0.03
0.001 0.001 0.003 0.002
(5) ANALYSIS OF TI@)MIXTURES
(6)
However, Rechnitz, Rao, and Rao (25)found that cerium(1V)manganese(I1) reaction (Equation 5 ) is really reversible, not irreversible, in sulfuric acid. They also found that the rate of this reaction increases with decreasing sulfuric acid concentration. The apparent second-order rate constant of 0.435M-'sec-' reported (25) for the cerium(1V)-manganese(11) reaction in 3M sulfuric acid does not appear to be large enough for effective catalysis, although this rate constant should be much larger in the 0.125M sulfuric acid solvent used in the photometric titration. Without further data, we can not exclude the possible participation of manganese(I1) in the catalyst cycle of the photometric titration. Photometric Titration of Thallium(1). The photometric titration illustrated in Figure 1 was performed at various concentrations of thalliuni(I), and the results are summarized in Table 111. The results appear to be satisfactory for this end point method. In developing the titration procedure it was found that the rate of the titration reaction is dependent upon the hydrogen ion concentration of the media. The greater the hydrogen ion concentration, the slower the reaction rate (23). The sulfuric acid concentration of the titrant has to be maintained at 0.2M to prevent hydrolysis of the cerium(1V). In the solution to be titrated, the sulfuric acid concentration must be greater than or equal to 0.1M. Under these conditions the precipitation of cerium(1V) hydroxide is slow relative to the reaction of cerium(1V) with thallium(1). The recommended acid concentration was arrived at by taking a compromise between the reaction rate at higher concentrations and the precipitation of cerium(1V) hydroxide at lower acidity. Initially the reaction is rapid-i.e., occurs with 20 to 30 seconds. But as the hydrogen ion concentration is increased by adding more titrant, the reaction begins to slow down. This is shown in Figure 1 where the absorbance begins to increase before the end point is reached. It was also found that cerium(II1) has no visible effect on the reaction rate or where the curvature of the titration graph began. At high ratios of thallium(1) to cerium(IV), the rate of reaction is independent of the thallium(1) concentration also (23). However, in titrating 0.15 to 0.35 meq of thallium(I), the end point curvature always began at a point about 2 ml from the end point. If the curvature were independent of the thallium ion concentration entirely, it should always start at the same per cent reaction in the titration regardless of the concentration of thallium(1). This indicates that at the low concentrations of thalliurn(1) found at the end point, the order of the reaction with respect to thallium is increasing. (24) B. Krishna and B. P. Sinha, 2. Pkysik. Chem., 213, 177 (1959). (25) G. A. Rechnitz, G. N. Rao, and G. P. Rao, ANAL.CHEM., 38,1900 (1966).
Analysis Based on Kinetic Differences. The enormous difference between the rates of the catalyzed and uncatalyzed reactions permits the analysis of two types of thallium(1)metal ion mixtures on the basis of kinetic differences alone, In the first type, addition of the catalyst is omitted; the cerium(1V) is used to titrate only metal ions which react rapidly with cerium(1V) before thallium(1) reacts. Then the manganese(II1) catalyst is added, and the cerium(1V) reacts with thallium(1). A good example is the analysis of a mixture of iron(I1) and thallium(1) discussed below. In the second type of analysis, the catalyst is added at the beginning of the titration. The cerium(1V) is used to titrate only thallium(1) which reacts rapidly before other metal ions present can react. Good examples of substances which react slowly with cerium(1V) in sulfuric acid at room temperature are those found in Table I: mercury(I), arsenic(III), and chrorniurn(I11). All are discussed below. If rate constants are available, the time for quantitative reaction for a metal ion such as iron(I1) or the time for 0.1 reaction of a metal ion such as chromium(II1) can be calculated as a guide for experimental work. Such calculations are not definitive, however, because they ignore unexpected catalysis, as in the case of arsenic(II1). Iron(I1)-Thallium(1) Mixtures. The reaction of cerium(1V) and iron(I1) is known to be fast; in 0.5M sulfuric acid the rate constant is 1.3 X 1OGM-*sec-lat 25" C (3). Just how fast this reaction will be during the photometric titration in 0.125M sulfuric acid can be estimated using the integrated second order rate law where the initial concentrations of iron(II), a, and cerium(IV), b, are not equal: (a
- X) - b)kt = In b(a ___ n(b - X)
(7)
For a typical case, a might be 0.006M and enough cerium(1V) titrant might be added to bring b to a hypothetical 0.001M before the absorbance is read. [Actually, cerium(1V) reacts continuously as it is added so that the 0.001M level is not reached.] After 99.9 reaction (b - x ) will be 10-6M and (a - x) will be about 0.005M. The time for 99.9 reaction of the ceriurn(1V) is then related to the rate constant as follows
z
z
1.34 t99.9%
=
x
103
k
(8)
If one accepts that the rate constant in 0.125M sulfuric acid is the same order of magnitude as in 0.5M, then the time for
z
99.9 reaction of the first increment of cerium(1V) titrant will be of the order of 10-3 second after mixing. In contrast the data in Table I1 indicate that the uncatalyzed cerium(1V)thallium(1) reaction is extremely slow in subdued light or in the dark. No significant amount of thalliurn(1) ought to react with cerium(1V) while it is reacting with iron(I1). (Using the rate constant of 0.03 hr-*M-1 and Equation 1, it can VOL 40, NO. 1, JANUARY 1968
165
s
1 I
0.6
0.5
-
0.3 u)
n
a 0.2
t O.I
t
0
1'
i 4.0
8.0
ml
12.0 Ce(lV)
16.0 20.0
0.1 1
0.0 24.0
Figure 2. Simultaneous titration of 0.55 meq of iron(II1) and 0.2102 meq of thallium(1) with 0.03412N cerium(1V) a.
Catalyst added at this point
be estimated that 0.1% reaction of 0.01M thallium(1) will require 3 hours.) Theoretically it should be routine to titrate iron(I1) before any thallium(I), and this was found to be true in practice. A typical titration curve is shown in Figure 2. The iron(I1)
Table IV. Titration of 0.2102 meq. of Thallium(1) in Presence of Iron(I1) with 0.03412N Cerium(1V) Meq. Meq. Tl(1) Found Abs. dev. Error, Fe(1I)a 0,2102 O.oo00 0.00 0.00 0.2094 -0.0008 0.38 0.20 0.2110 $0.0008 0.38 0.39 0,2080 -0.0012 0.51 0.59 0.2112 t0.0010 0.48 0.55 a The meq of iron(I1) found at the first end point was always within &0.5% or less of the meq of iron(I1) present.
Titration of 0.2220 Meq of Thallium(1) in Presence of Arsenic(II1) and Mercury(1) Meq Meq TI(1) As(II1) found Abs. dev. Error, 0.000 0.2223 $0.0003 0.14 0.103 0,2235 +0.0015 0.68 0.206 0,2209 -0.0011 0.50 0.206 0.2190 -0.0013 0.59 ... ... ... 0.2580 0.25P ... ... 0.20b 0.238 ... 0.66 ... 1.16 0.206 0.2198
Table V.
0
b
No end point could be found graphically. Meq of mercury(1).
166
ANALYTICAL CHEMISTRY
I
I
I
I
I
2.0
4.0 rnl
6.0
8.0
10.0
Ce(lV)
Figure 3. Effect of cerium(II1) on interference of chromium(111) in titration of 0.2102 meq of thallium(1) (1) TI(1) only (2) TI(I), 0.64 meq Cr(III), no Ce(II1). Smaller volume than solu-
tion (1) (3) TI(I), 0.64 meq Cr(III), 0.70 meq Ce(II1). Smaller volume than solution (2)
is titrated photometrically before adding the catalyst. After the absorbance rises to 0.15 to 0.20 unit, the catalyst is added and the thallium(1) is titrated in the usual way. [The initial rise in absorbance is the result of the formation of yellow iron(III).] The conditions in Figure 2 are approximately those treated theoretically above : 0.006M iron(I1) and increments of roughly 0.07 mmole cerium(1V) added to ca. 70 ml solution (0.001Mcerium). This procedure is recommended in general if other metal ions which react rapidly with cerium(1V) are present. A remotely similar approach has been used in potentiometric titrations of iron(I1)-vanadium(1V) mixtures (26). A statistical evaluation of the determination of both iron(I1) and thallium(1) in mixtures is given in Table IV. The per cent error rises slightly as the meq ratio of iron to thallium is increased from 1:1 to 3 :l. Thallium(1)-Arsenic(II1) Mixtures. The uncatalyzed reaction of cerium(1V) and arsenic(II1) is known to be slow in sulfuric acid (IO). As indicated in Table I, the rate law varies in 0.5M sulfuric acid and the rate constant of 1.6 X 10-3N-kec-1 applies only over a limited range. After about 30% reaction, the reaction appears to exhibit third order kinetics ( I O ) . To estimate an upper limit of the interference of arsenic(II1) after cerium(1V) has reacted completely with thallium(I), second order kinetics will be assumed. A typical titration might involve the addition of three to six 0.5-ml increments of 0.037M cerium(1V) in excess of thallium(1). After the addition of the first increment, the cerium(1V) concentration will be roughly 2 X 10-4M. At a typical arsenic (111) concentration of 4 X 10-3N, Equation 7 can be used to estimate that the upper limit of the time required for just 0.1 % reaction of cerium(1V) with arsenic(II1) is almost 170 seconds. (26) K. B. Rao, P. 1. Joseph, and Devavaj, Chemist-Analyst, 55, 103 (1966).
As an absorbance reading is taken 60 seconds after the addition of each increment of cerium(IV), theoretically the reaction of the first increment of excess cerium(1V) with arsenic(II1) should be negligible. The reaction of succeeding increments of excess cerium(1V) should not be significantly greater than the first. In practice, it was found that there was little interference from arsenic(II1) if the ratio arsenic(II1) to thallium(1) is less than one to one (Table V). The slope of the absorbance curve after the end point decreases with increasing concentrations of arsenic(II1). [The fact that equal amounts of arsenic(II1) can be tolerated before this decrease in slope affects the end point indicates that the simultaneous determination of thallium and arsenic would be difficult.] The possible existence of traces of iodide in the media was investigated as traces of this ion have a tremendous catalytic effect on the rate of the cerium(1V)-arsenic(II1) reaction. The removal of this interference from the water used in the titration and preparation of reagents proved ineffective in the elimination of the interference. The cerium(1V) reagent is a commercial grade especially purified to remove traces of iodide. The rate of the uncatalyzed cerium(1V)-arsenic(II1) reaction under the conditions of the photometric titration was checked by adding cerium(1V) to a solution of arsenic(II1). The absorbance readings were the same as when cerium(1V) was added to a sulfuric acid solution. This was further evidence that the uncatalyzed reaction was not responsible for a decrease in slope of the absorbance curve after the end point. Apparently the manganese(II1) catalyst also catalyzes the cerium(1V)-arsenic(II1) reaction. It has been reported (10) that 10-4M manganese(I1) increases the rate of this reaction sevenfold in sulfuric acid. The unexpected interference of large amounts of arsenic(II1) exemplifies one of the dangers of analysis based on kinetic differences. Any effective catalyst for the analytical reaction may also catalyze side reactions. Thallium(1)-Mercury(1) Mixtures. The rate constant (Table I) for the uncatalyzed cerium(IV)-mercury(I) reaction indicates that this reaction should interfere with the determination of thallium more than the cerium(1V)-arsenic(II1) reaction. The rate constant was determined in 2 M perchloric acid containing about 0.01M sulfuric acid. The initial rate of the reaction in 2M perchloric acid was 3.6 times greater than in 1 M sulfuric acid and 4.3 times greater than in 0.5M perchloric acid (8). This indicates that interference from mercury(1) should be somewhat less in 0.125M sulfuric acid. I n a calculation similar to that performed for the cerium(1V)arsenic(II1) reaction, Equation 1 can be used to estimate that 2 X 10-4M cerium(1V) will require about 20 seconds for 0.1 % reaction with 1.2 X lO-3Mmercury(I). The determination of thallium(1) in the presence of mercury(1) was not investigated extensively; the two results in Table V indicate a rough estimation of thallium(1) is possible at one-to-one ratios of thallium(1) to mercury(1). The concentration of mercury(1) is about 1 X 1OP3M. Thallium(1)-Chromium(II1) Mixtures. It can be deduced from Table I that chromium(II1) ought to interfere less than mercury(1) but more than arsenic(III), with the determination of thallium(1). To calculate the time for 0.1% reaction of chromium(II1) at the end point of the cerium(1V) titration of thallium(I), the rate law in Table I must be simplified. At the conditions of Table VI after addition of 0.5 ml of excess cerium(IV), typical concentrations might be: 2 X 10-4M ceriurn(IV), 3 X lO-3M cerium(III), and 2 X 10-aM chromium(II1). The rate law then simplifies to a zero order law:
Table VI. Titration of 0.2220 Meq of Thallium(1) in Presence of Chromium(II1)
Meq Cr(II1) 0.00
0.20 0.25 0.30 0.40
Meq
Tl(1) found
Abs. dev.
Error, %
0.2223 0.2220 0.2235 0.2242 0.2368
$0.0003 O.oo00 +0.0015 +o, 0022 +0.0146
0.14 0.00
0.68 0.99 6.68
Table VII. Effect of Cerium(II1) on Interference of Chromium(II1) in Titration of 0.2102 Meq of Thallium(1)
Meq Cr(1II) present
Meq Ce(II1) added
0.00
0.00 0.00 0.70 0.70 0.80 1.00
0.64 0.64 0.64 0.80 1.04
Meq TU) found 0.2110 0.1977 0,2098 0,2090
0.2105 0.2085
Error, 0.38 5.95
0.19 0.57 0.14 0.81
d(Crv‘) _ _ -- 1.6 X 10-10Msec-1 dt
(9)
It is easily calculated that the time required for 0.1 % reaction [the appearance of 2 X 10-6M chromium(IV)] will be of the order of l o 4 seconds. It is clear that the production of cerium(II1) from the cerium(1V)-thallium(1) reaction has increased the time for 0.1 % reaction. In practice, it was found that the reaction between cerium(1V) and chromium(II1) is slow enough for an equivalent amount of chromium to be tolerated. A 2 X lO-3M concentration of chromium(II1) caused a large error [see 0.4 meq Cr(II1) entry in Table VI]. There appears to be a weak catalytic effect of manganese(II1) operating. The rate law (Table I) indicates that the rate of the reaction between chromium(II1) and cerium(1V) can be decreased by adding cerium(II1) to the reaction media. The reaction of cerium(1V) and thallium(1) is independent of the cerium(II1) concentration so that increasing the concentration of cerium(111) will have no effect on the rate of this reaction. The results of the experimental check of this hypothesis are given in Table VII. A graphical representation of the first three determinations is given in Figure 3. In the presence of a threefold excess of chromium(II1) a large negative error due to the decrease in the slope of the cerium(1V) absorbance curve was obtained. In the presence of cerium(II1) the slope is almost restored to normal. The decrease in slope is due to the reaction between chromium(II1) and cerium(1V) which is competing with the thallium(1)-cerium(1V) reaction and becomes important near the end point. Comparison with Other Methods. The manganese(II1)catalyzed photometric titration has a number of advantages over the methods of Willard and Young (16) and Rao and Rao (17). Che latter methods will determine only the total of thallium(1) and metal ions, like iron(II), which react rapidly with cerium(1V) in hydrochloric acid. The photometric titration will “see” and measure metal ions like iron(I1) independently if the catalyst is omitted. Arsenic(II1) will interfere in methods (16, 17) using hydroVOL 40, NO. 1, JANUARY 1968
* 167
chloric acid as solvent, and mercury(1) will interfere by precipitating in this solvent. Rao and Rao (17) did not study the interference of chromium(II1) in their method, but willard and Young (16) reported that chromium trichloride did not interfere at a ratio of 6 meq of chromium(II1) to 1 meq of thallium(1) at 35” C. This is slightly better than the photometric titration in the presence of added cerium(III), as seen in Table VII.
ACKNOWLEDGMENT The authors wish to acknowledge their appreciation to G . A. Rechnitz for valuable discussions and for use of facilities ~ l ~ . at the State University of N~~ York at ~ ~ f f The able comments of D. B. Rorabacher are also acknowledged. RECEIVED for review July 31, 1967. Accepted September 28, 1967.
Acidic Denitrosation and lodometric Determination of N -Nitroso Compounds Joseph Gal, E. R. Stedronsky, and Sidney I. Miller Department of Chemistry, Illinois Institute of Technology, Chicago, 111. 60616 An iodometric method of analysis is described for eight N-nitroso compounds: N-methyl-N-nitrosoaniline, Nnitrosopiperidine, N methyl N nitroso p- toluenesulfonamide, N-nitrosocarbazole, N-methyl-“-nitroN-nitrosoguanidine, N-nitrosophenylbenzylamine, dimethylnitrosamine, and N-nitroso-N-methylacetamide. The compounds are denitrosated in an inert atmosphere with hydriodic acid to produce iodine, which can then be titrated with standard thiosulfate. The group displays a wide range of reactivity, which is highly dependent on the medium. Methanol and dimethylformamide, or their aqueous solutions, are suitable as solvents for denitrosation of the more reactive nitroso compounds. However, quantitative denitrosation of compounds that are either unreactive or enter side reactions, e.g. in hydroxylic solvents, can be accomplished with dry hydrogen chloride and potassium iodide in acetonitrile and glacial acetic acid. Acetone is also a fast solvent, but reacts with the iodine. I t would appear that the present determination of N-nitroso compounds is as convenient as other titrimetric analyses and is probably subject to less interferences. Moreover, the iodometric determinations of C-nitroso compounds involve sufficiently different reactions and problems which distinguish them from N-nitroso compounds.
-
- -
-
As THEY ARE normally prepared, certain aryl N-nitrosoamides, ArzCHN(NO)OCAr, may contain the starting amide, ArzCHNHOCAr, and the decomposition product, ArzCHOOCAr (1). When dissolved, these nitroso compounds decompose to give up nitrogen and ester in ca. 30 minutes at 39” (1, 2). Faced with the problem of assaying the impure and labile compounds, we developed a volumetric iodometric determination for N-nitroso compounds. Methods for the determination of the nitroso group include reduction with Mo(III), Cr(II), Ti(III), Cu, Cd, etc., as well as acid-base and decomposition methods (3-5). Because our aryl groups could contain the nitro group as a substituent, (1) E. R. Stedronsky, J. Gal, R. A. More O’Ferrall, and S. I. Miller, J . Am. Clzem. SOC.,in press. (2) P. A. S. Smith, “The Chemistry of Open-Chain Organic Nitrogen Compounds,’’ Vol. 11, W. A. Benjamin, Inc., New York, N. Y., 1966, pp. 470f, Chap. 13. (3) M. R. F. Ashworth, “Organic Analysis,” Interscience, New York, Vol. 15, Part 1 (1964), Part 2 (1965). (4) W. W. Becker and W. E. Shaefer, “Organic Analysis,” Vol. 2, Interscience, New York, 1954, pp. 93f. ( 5 ) D. Tiwari and J. P. Sharma, Ind. J . Chem., 2, 173 (1964). 168
0
ANALYTICAL CHEMISTRY
and because our compounds were unstable, we felt that these methods were unsuitable. On the other hand, it was known that acid could denitrosate N-nitroso compounds (2, 6). By including iodide ion in the solution, we would have only to do an iodine titration to obtain the quantity of N-nitroso compound, provided that the denitrosation step was faster than other decomposition routes: RZN-NO NO+
+ I-
+ H’ +
‘/z
$ RzNH
12
+ NO’
(1)
+ NO
This procedure was anticipated for one compound, N-nitrosodiphenylamine (4). [Iodometric determinations of C-nitroso compounds involved several other reactions and will be touched on later (7-9).] In this paper, we examine eight stable N-nitroso compounds of sufficiently different structure to establish the generality of the analytical method. EXPERIMENTAL Materials. Reagent grade potassium iodide, sodium acetate, hydrochloric acid, and solvents were used. It should be noted that some N-nitroso compounds may be carcinogenic ( I O ) , some cause persistent eye inflammation ( I I ) , and some may disrupt the olfactory sense for long periods (12). Available N-nitroso compounds were purified by distillation or recrystallization, as required. N-nitroso-N-methylacetamide was prepared from the amide and nitrogen dioxide (13). The physical properties included in Table I, as well as IR and NMR spectra, were used to check the purities of these compounds. Procedure. The standard analysis is carried out in an electrolytic 250-ml beaker. As shown in Figure 1, the beaker is fitted with a seven-hole rubber stopper through which (6) B. A. Porai-Koshits and V. Shaburov, Zhur. Org. Chem., 2, 1666 (1966), and previous papers. ( 7 ) M. M. Lobunets and E. N. Gortins’ka, Uniu. Etat. Kiev, Bull. Sci. Rec. Chim. No. 4 , 37 (1939); Chem. Abstr., 35, 1356 (1941). (8) R. Aldrovandi and F. de Lorenzi, Ann. Chim. (Rome), 42, 298 (1952). (9) R. D. Tiwari and J. P. Sharma, Proc. Natl. Acad. Sci., India, Sec. A., 33, Pt. 3, 379 (1963). (10) J. H. Weisburger and E. K. Weisburger, Cliem. Eng. News, 44 (6), Feb. 7, 124 (1966). (11) L. P. Kuhn, G. G. Kleinspehn, and A. C. Duckworth, J. Am. Chem. SOC.,89, 3859 (1967). (12) R. H. Huisgen and H. Reimlinger, A ~ I ~599, z . , 161 (1956). (13) E. H. White, J . Am. Chem. SOC.,77,6008 (1955).