Table 1. Determination of Copper after Separation from Various Interfering Substances
Interfering Ions
c u , hIg. Taken Found 24.87 24 99 24.98 25.00 Cd 16.68 16.66 16.69 16.66 E'?( 111) 16.62 8.29 8.33 8.33 ZIl 8.35 Xi 8.31 6.60 6.63 6.66 6.60 6.13 4.99 Go 5.00 5.04 Ni 5.02 4.80 3.33 3.12 3.30 3.33 1.64 1 66 1Lg 1.59 Fd 111) 1 64 1
100
Err or,
%
11g.)
-0 48 -0
12
+0 04 +o 12 +0 18 0 -0 2'4 -0 48 0 +O 24
-0 -0 -0 0 -0
24 91 48 91
-8 6
-0 20 +0 80 +0 40 -4 0 0 -3 3
-0 91 0 -1 2 -4 2 -1 2
c:itor. The decomposition was accompanied by a strong phenolic odor at all temperatures. When cuprous T P B was heated in an oven, a t 110' to 130" C., a mixture of white and reddish brown or sometimes gray powders was obtained, The white powder was
water-soluble and contained boron. The colored, insoluble portion contained all of the copper, and by its behavior when treated with various acids was found to be a mixture of cuprous and cupric oxides, as well as some copper metal. The stoichiometry was neither simple nor consistent, thereby eliminating the possible use of gravimetric procedures. Table I tabulates results obtained by the method outlined in the procedure section for a variety of copper solutions, each containing a metal ion of potential interference. The only dement found to intwfcre seriously \vas mercury, which caused lev- results. hIcrcury(I1) reacts with sodium T P B forming a precipitate. Apparently this precipitate coprecipitates with copper to some extent. A slight interference due to aluminum is indicated by Table I but this is believed to be due to the excess acidity caused by the aluminum chloride used as a source of ahiminum. Subsequent work with aluminum alloys indicates that accurate results can be obtained in the presence of aluminum. Silver, mercury(I), and lead form insoluble chlorides if hydrochloric acid is used t o acidify the copper solution, and should be removed by filtration before the addition of sodium TPB. Potassium precipitated upon the addition of sodium TPB and was filtered out before the addition of ascorbic acid. It was possible to substitute nitric or sulfuric acid for hydrochloric acid, without adversely affecting the results with pure copper solutions. If silver is present, however, it is ad-
vantageous to remove it as silver chloride, since silver TPB, which would form on the addition of sodium TPB, is difficult to remove by filtration. Two National Bureau of Standards alloys w r e analyzed as a further check on the accuracy of the method. For NUS aluminum alloy 85B the copper found was 3.98 and 4.00y0 as compared to the reported value of 3.99%. Zincbase alloy 94B contained 1.01, 1.00, and 1.00% copper, while the reported value was 1.01%. The average deviation of the method, as calculated from the data in Table I (excluding those solutions containing mercury or aluminum), \vas h0.04 mg. while the average error was -0.015 mg. The range of the method extends to about 1 mg. of copper in 50 nil. Bclow this amount the precipitation is not completc. For instance, only 95% of 0.6 mg. and 80% of 0.4 mg. was precipitated. LITERATURE CITED
(1) Barnard, A. J., Buechl, H., Chemist
-4naZyst 48, 44 (1959). (2) Flaschka, H. A., Mikrochemie oer Mzkrochzm. Acta 3 9 , 38 (1952); 40, 42
(1952).
(3) Flaschka, H. A., New Orleans, La.,
private communication, January 1960. (4) Gielmann, W. G., Gebauhr, W., 2. anal. Chem. 139, 101 (1953). (5) Neloche, V. M., Kalbus, L., h A L . CHEW28,1047 (1956). (6) Fesmeyanov, A. V., Sazonova, V. A., Liberman, G. S., Emel'yanova, L. I., Bull. ilcad. Sci. U.S.S.R., Dio. Chem. Sci. 1955, 41. RECEIVEDfor review March 30, 1960. Accepted June 29, 1960.
Ultraviolet Spectrophotometric Analysis of Solutions of BenzonitriIe and Benzamide MELVIN J. ASTLE and JAMES 6. PIERCE' Department of Chemistry and Chemical Engineering, Case lnsfitute o f Technology, Cleveland, Ohio
,A spectrophotometric method has been devised for analyzing solutions containing benzonitrile and benzamide in the presence of water and alcohols. In this method the number of moles of absorbing species i s held constant, and mole fraction is used as a unit of concentration. The use of this unit of concentration permits a simplification in the calculations and in the experimental techniques since a two-component system may be analyzed by observing the transmittance or absorbance of the unknown sample a t only one wave length. The use of mole fraction has also resulted in a simplified rela-
1322
0
ANALYTICAL CHEMISTRY
tionship between the first-order rate constant and the absorbance of the solution.
T
HE ANALYSIS of a two-component mixture by ultraviolet spectrophotometry has been accomplished b y different methods. Bastian, Weberling, and Palilla ( 1 ) discuss the use of a highly absorbing blank. This method, if not properly executed, can lead to errors that might be minimized by using a proper procedure (6). Beroza ( 2 ) uses a method of differential analysis in which the highly absorbing blank
approximates the unknown sample in composition. He reports an improvement in precision. De Vries and Gantz (3) analyzed mixtures of nitroaminoguanidine and nitroguanidine by employing two different wave lengths, at one of which the nitroguanidine was transparent and the p-nitrobenzal derivative of nitroaminoguanidine absorbs strongly. Levy et al. ( 8 ) use a base
1 Present address, Department of Chemistry, Lowell Technological Inst.itute, Lowell, Mass.
hne method in the determination of penicillin. Stearns (9) discusses techniques for interpreting data, among which is the solution of simultaneous equations to obtain individual concentrations. Fry, Nusbaum, and Randall (4) use calibration curves to determine the individual absorptivities in multicomponent mixtures. These values are then used in the usual equations derived from Beer’s law, I n an unconventional approach Vaughn and Stearn (10) analyze a threecomponent mixture b y determining the absorbance of the three possible pairs of compounds a t all concentrations and b y plotting the difference in absorbance found at two other wave lengths. The morking plot turns out t o be a slightly diitortcd triangular graph. The distortion result% from deviation from Beer’>lam. Hirt, King, and Schrnitt (5) present an interestinq technique that may be applied to the equations derived from Beer’s law. These authors search for a n isoabqorptive point R here two components have the same absorptivity. This wave length is used n i t h one other in analyzing a two-component mixture. Then by forming a ratio of the two absorption equations a simplified solution results. EXPERIMENTAL
In this work. solutions containing benzonitrile and benzamide were analyzed in the presence of n a t e r and alcohols. Materials. T h e methanol used as solvent was C.P. grade redistilled from hlg(OCHs)? and had a maximum boiling range of 0.5’ C. T h e benzonitrile n as Eastman’s chemical 487 a n d shon ed only t h e slightest trace of impurity on t h e mass spectrograph. T h e benzamide ?I as Eastman’s chemical 278, n hich a as recrystallized from 1,Zdichloroethane before use. All standard solutions were dilutions of two primary standards which were 0.100M in benzonitrile and benzamide, respectively. Both samples were l O M with respect to water and were diluted to volume with methanol. Apparatus. A Beckman ratio recording spectrophotometer, hlodel DK-2, seiial number 118227, was used. All volumetric flasks were of borosilicate glass and had a class A tolerance. T h c 0.1-ml. pipet uspd was of borosiiicate glass graduated t o read t o 0.01 ml., and had a tolerance of &0.005 1111. Procedure. Samples Tvere prepared for a n a l y w by measuring a knonm q u a n t i t j of standard solution into a 10-in].volumetric flask a n d diluting t o t h e mark Rith methanol. T h e same 0.1-ml. piaet n-as used for all measurements. T h e pipet was cleaned between samples by rinsing with methanol and dried by attaching t h e pipet
Table 1.
Absorbance
Concentration, hfoles/Liter, X 10 BenzoBenznitrile amide 1.0 2.0
Mole Fraction ___
s,
1.0 2.0 3.0 4.0 5.0 6.0
1 .o
0.143 0,286 0.429 0.572 0.715 0.858 1.000 0.858 0.715 0.572 0.429 0.286 0 143
Wave Length, M 274.8 Abeorbance, A.
270.0
.vs
0.286 0.429 0.572 0.715 0.858 1,000
4.0 5.0 6.0 7.0 1.0
~
0.143
3.0
2.0 3.0 4.0 5.0 6.0 7.0 6.0 5.0 4.0 3.0 2.0
of Samples a t Several W a v e Lengths
0.143 0.286 0.429 0.572 0.715 0.858
t o a vacuum line. T h e volumrtric flasks were always rinsed four times with methanol before a new sample was prepared. The spectrophotometer was used with the follo~vingsettings: scanning speed, 2 minutes; scale expansion, 1 x; scale reading, 0 to 100% transmittance; sensitivity, 50; time constant, 0.1; multiplier phototube, 1 X ; light source, hydrogen discharge tube. The 1007, line on the plotting paper was set a t 9570 transmittance a t 340 mw. At this wave length the nitrile and amide are both transparent. A single pair of matched silica glass cells was used for a11 determinations, The sample cell had a path length of 1.000 cm., and the reference cell had a path length of 0.996 cni. Cells were always rinsed four times with new sample before use, and the esterior optical surfaces of the cells were always polished with clean lens paper and inspected visually before introducing the cell into the instrument. The reference sample in all instances was methanol which was permitted to reside in the instrument for a t least one-half hour before any measurements were tahen. Samples were introduced immediately before measurements were made. For a given series of samples the instrumental adjustments were not changed in any way. DISCUSSION
Tn the present method the equations in concentration and absorbance are used, but a considerable simplification occurs when the mole fraction of absorbing species is substituted for the concentration terms normally used in spectrophotometry. Because of the dependence of one mole fraction upon another in a given mixture, a two-
0 0864 0 161 0 255
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 340
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
338
528 624 0719
12T 193 2550 301 365 425 455 480 513 531 560
0
277.2
__
0.0861 0.158 0.253 0.332 0.42; 0.515 0.607 0.0569 0.0899 0.137
0531 0899 134 176 228 270 311 0645 104 158 201 241 297 340 328 330 338 322 320
0.179
0.212 0.255 0.297 0.344 0,380 0.428 0.470
0 0 0 0 31s
5!13
p ___
0.515 0 . 560
component iyatem may be analyzed by determining the abeorbaiice at only one wave length; or if the analyst wishe.. a n equation may be developed in n hich measurements a t t n o n a v e lengths are utilized. This method presupposes the validity of Beer’s lam, and standard curves have been plotted to show thnt the system under investigation does conform to this requirement. When a t n o-component system i i considered and the concentrations are expressed in terms of mole fraction (A7) (79,Beer’s law takes the form: A. = b(aiA-i
+
~2x2)
(1)
and bince N z = l-AT1the following derivation results: A, = b(aiXi
andif b
=
+ uz -
- U Z ) ~+ I UZ]
A, = b[(ai
(2)
1 cm.
A. = (a, N1
=
- U2)‘Vl
I:-[
+ a2
A, - a ai
- a.
(3) (4)
Equation 3 may be w e d to determine the mole fraction of either component at a single wave lergth, providing the absorptivities are known and the total number of moles of absorbing species is known and remains constant. Absorbances may Le determined a t t a o wave lengths, and from these values two equations such as Equation 3 may be written. If one of these is divided b y the other, the following expression results:
where the primed values refer to the VOL. 32, NO. 10, SEPTEMBER 1960
1323
second wave length at which nieasurements are made. Equation 5 expresses a n absorbance ratio and may be regarded as an extension of the work of Hirt et d. (6). Solving for the mole fraction gives: a2 S I
=
A,
- -4,
I
Q2
(a:- a;) -
(6) (a,
- a,)
which has the same limitation. as Equation 9. I n Equations 4 and 6 the A , values are determined experimentally from the samples of unknown composition. The absorptivities are constant values characteristic of each individual component a t a single wave length. From Beer’s law, A = a b c
it is evident that the absorptivity is numerically equal to the absorbance at unit concentration and unit cell length. Thus, the use of mole fraction as a unit
erty of these equations is that errors resulting from variations in the adjustment of the spectrophotometer from day to day may be minimized by observing the absorbance of a qtandnrd solution of pure component 2 a t unit concentration. This is poqsible since the absorptivities of component 2 appear in the numerator?, nhile only differences of absorptivitici appear in the denominators. The latter will remain constant for a given combination of instrumental aettings, ivhile the former may vary slightly because of the error in adjusting the instrument to the predetermined settings. If, hon-ever, the combination of settings is altered, the differences in absorbances may be expected to change also. To test Equations 8 and 9, a series of standard solutions was prepared, and the absorption spectra were determined between 250 and 300 mp. The data collected from theqe spectra are presented in Table I, and the calculated mole fractions are compared with the experimental values in Table 11.
This method of reasoning may be extended to the determination of reaction rate constants by the follon ing sequence of operations. If the firstorder rate equation is expressed in term- of the mole fraction of a single component, ~ h i c hcould be the ratedetermining reactant, it has the form: 111
s s,* = - Ict
(12)
Hon-ever, in a system in which reactant 1 disappears to form product 2 the initial mole fraction of reactant designated by ATl* is equal to unity, which gives Equation 12 the simplified form:
By taking the logarithm of Equation 8 and substituting with Equation 13, the rate constant may be expressed as a function of time and absorbance.
or k Table II.
Comparison of Experimental and Calculated Mole Fractions
Calcd., Wave Length, Mp Exptl. Eq. 13, 277.2 m p Eq. 14, 270.0; 277.2 M p Dev. AN’ Dev. ?A N‘ (AN’) 7 XIi (An“’) N x 100
1V
0.858 0.715 0.572 0.429 0.286 0.143
0,841 0.728 0.569 0.431 0.286
0.017 0.010 0.003 0.002 0.000 0 003
0.140
2.0 1.4 0.52 0.47 0.00 2.1
of concentration causea the absorptivity of a single component a t a given wave length to be numerically equal to the absorbance of the pure material a t some arbitrary unit concentration a t which it is convenient t o carry out the analysis. Thus, we may write for a two-component system : al = AT and
a2
= A:
(7)
Using this notation Equation 4 becomes
and Equation 6 becomes A2
s,=
A’(Ar’
- A2 A:‘
- A*’) -
(A?;- A : )
0
ANALYTICAL CHEMISTRY
0.013 0.020 0.033 0.024 0,014 0.016
1.5 2.8
5.8 5.6 4.9 11
The b e d precision was obtained when measurements were made a t 277.2 mp and low concentrations of benzamide when Equation 8 was used. The reverse was true when Equation 9 was applied, because of the lolver precision obtained a t a second wave length. Therefore, the choice between these two equations depends upon the experimental conditions. I n the system discussed in this uork both components absorb a t all 1%-ave lengths that may conveniently be used for analysis. If a wave length may be found a t which only one component of a two-component mixture absorbs, Equation 8 becomes:
(9)
A: When i t has been established that a system obeys Beer’s law, Equations 8 and 9 may be applied if the absorbances of the pure components have been determined. Another convenient prop-
1324
0,845 0.735 0.605 0.405 0,272 0.159
and the secoiid component may be determined b y subtracting N1 from unity. AT2
1
- x1
(11)
=
- 2.303 - log t
T1ii.j equation has all of the conveniences and limitations of Equation 8. Holyever, in kinetic studies Equation 15 iq very convenient, since i t eliminates the necessity for determining concentrations when only the rate constant is desired. This technique should also be applicable to related systems in Tvhich reactant goes to form product on a n equimolar basis with the formation of no by-products. LITERATURE CITED
(1) Bastian, Robert, Weberling, Richard, Palilla, Frank, AXAL. CHEM.22, 160 (1950). 12) Beroza. Rforton. Ibid.. 25. 112 (1953). (3) De Vries, J. E., Gantz,’ E. St. C:, Ibid., 2 5 , 1020 (1953). (4) Fry, D. L., Susbaum, R. E., Randall, H. M., J . B p p l . Phgs. 17,I50 (1946). (. 5.) Hirt, R. C., King, F. T., Schmitt, R. G . , ’ A N A L . C H E M . 2 6 . 1271l(1954). ’ ( 6 ) Hiskey, C. F., Youne. -, I. G.. Ibid.., 23., 1196 11951). (7) Hughes, H . K., Ibid., 24, 1349 (1952). (8) Levy, G. B., Shaw, Denham, Parkinson, E. S., Fergus, David, Zbid., 2 0 , 1159 (1948). (9) Steams, E. I., “Analytical Absorption Spectroscopy,” M. G. Mellon, Ed., Wiley, New York, 1950. (10) Vaughn, R. T., Stearn, A. E., AXAL. CHEM. 21,1361 (1949). RECEIVED for review December 17, 1958. Resubmitted April 22, 1960. Accepted May 6, 1960. Taken from the Ph.D. thesis of James B. Pierce to Case Institute of Technology, June 1958. Work made possible by financial assistance from the Standard Oil Co. of X e w Jersey.