Spectrophlotometric Determination of Mercury(ll1) in Aqueous Potassium Iodide Media A. J. Pappas and H. B. Powell Department of Chemistry, University of Miami, Coral Gables, Fla. Spectrophotometric measurement in neutral 1.OM aqueous potassium iodide provides a convenient method for the determination of mercury(l1) in inorganic compounds. The measurements are insensitive to small changes in the iodide ion concentration or the pH of the solution. Although several cations interfere with tlhe analysis, cyanide, chromate, and dichromate are ‘the only anions which interfere.
NUMEROUS METHODS have been proposed for the quantitative determination of mercury(I1) (1). These methods generally suffer from interference of other ions or from the volatility of mercury(I1) compounds. In many cases the methods are severely limited by these problems. One of the better methods for the determinati.on of mercury(I1) in inorganic compounds is complexometric titration with thiocyanate ( I ) . A spectrophotometric method using thiocyanate has also been reported (2). Either method involving thiocyanate is limited by interference of halide and pseudohalide ions as well as metals such as silver (1). Because of the lack of a generally applicable method for the determination of inercury(II), especially in the presence of halide ions, a spectrophotometric determination in aqueous potassium iodide was investigated. Mercury(I1) is known to form the complex ions Hg13- and Hg14-2in aqueous solutions. The Hg!4-2 ion, which has an absorption maxima at 323 mp ( E = 2.34 X lo4) (3), is the principal mercury-containing species present in 1.OM aqueous potassium iodide (3). EXPERIMENTAL
Reagents. Reagent grade mercury(l1) iodide was purified by repeated sublimation in uacuo at 73”-80” C . The purified material [mp., = 254.7-255.2 uncorr.; Lit. 257 (4)] was stored in a desiccator over activated alumina. All other chemicals were A.. R. grade or equivalent and were used without further purification. Stock solutions of mercury(I1) were prepared by dissolving accurately weighed amounts of purified mercury(I1) iodide in 1 M aqueous potassium iodide. Although no effort was made to have identical concentrations of mercury(I1) in successive stock solutions, the stock solutions were approximately 2 X 10-.’M. Stock solutions of the anions used in this investigation were prepared by weight from the sodium or potassium salts. Stock solutions of the cations other than mercury(I1) were prepared by weight from the nitrate or chloride salts except for platinum(1V) which was prepared from sodium hexachloroplatinate(1V). Stock solutions of the Lewis Bases wlxe prepared by weight from the pure bases. All other solutions were prepared by appropriate dilutions (1) J. F. Coetzee, “Treatise on Analytical Chemistry,” Part 11, Vol. 3, I. M. Kolthoff and P. J. Elving, Eds., Interscience, New York, 1961, pp. :!31-315. (2) G. E. Markle alid D. F. Boltz, ANAL.CHEM., 26, 448 (1954). (3) C. Merrit, Jr., H. M. Hershenson, and L. B. Rogers, Ibid., 25, 572 (1953). (4) F. A. Cotton and G . Wilkenson, “Advanced Inorganic Chemistry,” 1st Ed., Interscience, New York, 1962, p. 485.
33124
of the stock solutions using standard volumetric ware. Distilled water used in the preparation of 1.OM potassium iodide was deaerated by passing nitrogen gas rapidly through it for 2 to 3 hours before use. Potassium iodide solutions prepared with this deaerated water remained clear for several days when stoppered and stored in subdued light. SpectrophotometricMeasurements. A line-operated Beckman DU spectrophotometer equipped with a photomultiplier was used for all of the experiments in this investigation. The cells were 1.0-cm quartz cuvettes. Experiments with mercury(I1) in aqueous KI solution indicated the cells were matched to within +0.%005 absorbance unit at 323 mp. The absorbance of the mercury(I1) solutions was measured at 323 mp using 1.OM KI as a reference blank. No precautions other than normal stoppering were taken to exclude oxygen during the measurements. Although the absorbances of the solutions were usually determined the same day they were prepared, no appreciable change in absorbance was observed over a 48-hour period. RESULTS AND DISCUSSION
The absorbance of mercury(I1) in 1.OM aqueous potassium iodide was found to obey the Beer-Lambert law over the concentration range 0.5 to 4.0 X 10-5M. A least squares treatment of the data from 30 different solutions yielded a straight line which is described by the equation:
A
=
f0.0005
+ (2.050 X 104)C
where A is absorbance and C is the concentration of mercury (11) in moles per liter. The relative standard deviation of the ratio CIA for these 30 solutions was 1.8z. The intercept (0.0005) for the above line is equal to zero within the instru-
Table I. Determination of Mercury(I1) in the Presence of a Ten-fold Molar Excess of Various Anions Anion
Added
Founda
2.00 2.00 2.13 2.13 2.13 2.13 2.13 2.13 2.13 2.13 2.13 2.13 2.15 2.15 2.15 2.15 2.15 2.15
2.01 2.01 2.13 2.12 2.14 2.12 2.12 1.14 2.02 2.09 2.16 2.20 2.16 2.15 2.15 2.15 2.14 2.14
Error, +0.5 +O. 5 +0.2 -0.5 +O. 7 -0.5 -0.5 -46.0 -5.1 -2.0 $1.7 +3.3 +O. 5 0.0 0.0 +0.2 -0.2 -0.2
Each value is the average of 2 or more samples made from the same mercury(I1) iodide stock solution. * The ratio of anion to mercury is 2: 1. The ratio of anion to mercury is 1 :1.
VOL 39, NO. 6, M A Y 1967
579
ment readability (+0.001). The slope of the line (2.050 x lo4) is somewhat lower than the molar absorptivity reported for H&-* (3). This difference could be due to a difference in the path length of the cells used in this investigation and .ki the previous report. The absorbance of a series of 2.2 X 10-SM solutions remained unchanged when the iodide concentration was varied from 1.2 to 0.8M. In aqueous solution the Hg14-2ion is in equilibrium with the HgIa- ion; however, the equilibrium apparently lies far enough toward the Hg14-2 ion to make the concentration of Hg14-*relatively insensitive to changes in the iodide ion concentration under the conditions of this experiment. In order to assess the effect of pH on the absorbance of mercury(I1) in aqueous iodide media, several experiments were made in which either sodium hydroxide or perchloric acid was
Determination of Mercury0 in the Presence of
Table 11.
2 X 10-'M Concentrations of Other Cations Cation [Hg+f]x 105a Error,
None ~ 1 + 3
Sn+f
TI+' cu+z Pb+S
Fe+' Bi+a pt'
4
AU+~
1.93 1.92 1.91 1.91 1.97 1.97 1.96 1.99 2.02 2.02
Z
-0.5
-1.0 -1.0
+2.1 +2.1 +1.6
+2.9 +4.7 +4.7
2.07 +7.8 Pd+z Each value is the average of 2 or more samples made from the same mercury(I1) iodide stock solution.
Table 111. Determination of Mercury(I1) in the Presence of 2 x 10-5M Concentrations of Other Cations [Hg+*] X lo6
Added Found" Error, % 9.51 -0.8 Ag+ 9.65 9.71 +0.6 Zn+2 9.65 9.61 -0.4 cd+z 9.65 10.34 +7.1 TI+' 9.65 10.24 +6.0 FeA2 9.65 12.88 +33.6 Cu+' 9.65 27.9 +O. 4 AI+% 27.8 21.7 -0.4 Sn+' 27.8 Each value is the average of 2 or more samples prepared from the same stock solution. Cation
Determination of Mercury(I1) in the Presence of 10-'M Concentrations of Other Cations Cation [Hg+l] X 1Osa Error, 1.84 ... None 1.83 -0.5 Ag*' 1.83 -0.5 Cdf2 1.83 -0.5 Cri3 1.90 +3.3 Zn+l
Table IV,
2
x
cu+*
8.44
...
2.21 +20.1 1.W ... 3. 9gb ... Fe+2 a Each value is the average of 2 or more samples made from the same mercury(I1) iodide stock solution. Precipitation occurred in some of these samples and results obtained on the filtrates were highly erratic. Sn+P Pb+l
added to the solutions. The absorbance of these solutions was found to be insensitive t3 the pH of the solution between pH 5 and pH 9. At a pH of 4 the oxidation of iodide ion to triiodide ion became appreciable; however, with care, it was possible to perform the measurements with errors of less than 1.0%. Measurements of the absorbance of mercury(l1) solutions at a pH of 10 were within 1.0% of the expected value but were consistently low. Low results might be expected ai higher pH values because of the formation of species of the type H d O H h (5). Effect of Extraneous Anions. Several anions (Table I) were investigated for possible interference with the determination of mercury(I1) in aqueous iodide. Each of the values listed in Table I and the following tables is an average of two or more samples made from the same Hg12 stock solution. The relative deviation of individual samples from the average for a particular set was 1.0% to 1 . 5 z . Interference by an anion, as well as for the other species in this investigation, was defined as results which were either consistently low or high by 1.3 or more. The only anions, of the 16 investigated, which interfered when present at a mole ratio of anion to mercury(I1j of 10 :1 or less were cyanide, chromate, and dichromate (Table I). The interference of both the chromate and dichromate was due to the absorbance of these ions at 323 mp rather than reaction with either the iodide or mercury(I1). The interference of cyanide ion, which is a nonlinear function of the cyanide ion concentration, is probably due to the formation of complex species of the type Hg(CN)(,I,-,, (6). Effect of Extraneous Cations. Various concentrations of the cations listed in Tables 11, 111, and IV were added to samples of the mercury(I1) iodide solutions prior to the determination of mercury(I1). As shown in the tables, most of these cations produced a positive interference over at least part of the concentration range used. Silver(I), cadmium(II), and chromium(II1) which produced no interference when present at concentrations of 2 X lO-'M or less, were exceptions. Because copper(II), platinum(IV), and gold(II1) are known to oxidize iodide ion to triiodide under conditions similar t c $he conditions in these experiments, the intense yellow to red colors produced when these ions were dissolved in 1.OM KI were assumed to be due in part to triiodide. Consequently, samples of these solutions were treated with acidified thiosulfate prior to the analysis. The thiosulfate reduced the intensity of the color in all cases but completely removed the interference only in the samples containing gold(II1). Separate experiments indicated that acidified thiosulfate would quantitatively remove triioaide from the 1 .OM KI solutions. Iron(II), lead(I1). bismuth(III), and thallium(1) produced visible precipitates in 1.OM KI solution when their concentrations were 1 X 10-4M or greater. Although no precipitate was visible at low concenrrations, interference by smai! amounts of finely divided precipitates containing these ions cannot be precluded. The slight interference of iron(II! wher: present at the lower concentrations could also be due to tne presence of iron(II1) in these szrnples, as iron(I1) is rapidi?. oxidized to iron(1lIj in neutra; aqileous solution. Effect of Neutral Donor Ligands. Several neutral donor ligands were added to the rnercury(I1) iodide solutions prior to the determination of mercury(Xl) (Table V). These i n , ______ .. .
~ _ .
(5) A. B. Garret and A. E. EIirshIer,
ANALYTICAL CHEMISTRY
Am. Chem. Soc.,
60,294
(1938).
( 6 ) R. A, Pennernan and L. H.
19 (1961). 586 *
4 . .
JQiia,
2. Inorg. Muci'. Chert:.. 2&
ciuded ligands whose donor atoms were oxygen, nitrogen, sulfur, phosphorus, and arsenic. Triphenylphosphine, trip henyiarsine, and tri(pto1yI)thiophosphate were insoluble in water and formed precipitates when their alcoholic solutions were added to the a.queous Hg12/KIsolutions. Determination of the mercury(II1 in the filtrates from the above solutions yielded erratic results. The erratic results were possibly due to failure to remove consistently the finely divided precipitate; however, the low results (Table V) found for triphenylphosphine are probably due to the formation of an insoluble complex with mercury(I1) iodide (7). Ligands containing nitrogen or oxygen donor atoms did not interfere with .the determination of mercury(I1) (Table V) when present at concentations of 2 X 10-4M or less. Thioacetamide, which is a sulfur donor ligand, did not interfere at 1 y IO-'M or less but caused a slight positive interference at 2 X IO-". Increased interference by sulfur donor ligands would be expected since mercury(I1) has a greater tendency to coordinate with sulfur than with nitrogen or oxygen (8). The large errors in the determination of mercury(I1) when N,N-dimethyldithiooxamidewas present were due, in part, to the yellow color of the ligand but could also reflect the greater stability of complexes involving chelate ligands. In summation, spectrophotometric measurement in 1.OM aqueous potassium iodide provides a relatively simple and iapld means of determining mercury(I1) in inorganic compounds. The method can provide analyses with a probable error of 1.3z.The insensitivity of the determination to changes in the pH 3f the solution and to changes in the iodide (7) R. C . Man, H. S . Peiser, and D. Purdie, J. Chem. SOC.,1940, iX)
i 209. F. 4.
Cotton and G. Wilkenson, "Advanced Inorganic Chem-
'stry," 1st Ed., Interscience, New York, 1962, p. 488.
Table V. Determination of Mercury(II) in the Presence of Neutral Donor Ligands (Ligand Concentratration = 2 X lo-* M ) [Hg+7 X 10' Ligand Added Founds Error, 9.60 9.56 -0.4 Urea 9.60 9.56 -0.4 Ethylenediamine 9.60 9.56 -0.4 Triethylenetetraamine 9.60 9.54 -0.4 Acetamide Thioacetamide 9.60 9.89 $3.0 Pyridine-N-oxide 9.60 9.56 -0.4 Ethylenediaminetetraacetic acid 9.60 9.52 -0.8 N,N-dimethyldithiooxamide 9.60 22.3 ... Triphenylphosphine 20.9 1 .o Triphen ylarsine 20.9 23.0 .". 20.9 43.9 ... Tri-ptolylthiophosphate a Each value is the average of two or more samples made from the same mercury(I1) stock solution. .
I
.
ion concentration gieatly simplifiesthe preparation of reagents. This method is suitable for the determination of mercury(I1) in the presence of a variety of extraneous species. However, the method should be most useful for the determination of rnercury(I1) in the presence of halide and pseudohalide (except cyanide) ions which severely interfered with the previously reported methods. ACKNOWLEDGMENT
We are indebted to Marshall Moorman and Douglas Gegen for the preparation of many of the solutions used in this investigation. RECEIVED for review January 1, 1967. Accepted March 13, 1967.
Spectrophotometric Determination of Primary Aromatic Amines with Thiotrithiazyl Chloride Application to Determination of Toluene-2,4-diisocyanate in Air Vaughn Levin, 8 ,W. Niopoldt, and R. E. Rebertus ','miral
Research Laboruiorks, Minnesota Mining and Manujacturing Co., St. Paul, Minn.
Yighiy coiored intermediates form during the reactions thiotrithiazyi chloriae with some primary aromatic amines. Tne molar absorptivities range from about X C O O !iters molt!-1 cm-1 for the red product from m,)henylenediamirie to about 300 liters mole-' cm-' or the green product from aniline. Rapid color d~veiopmenti s xhieved at room temperature either 3 v direct combination of solid thiotrithiazyl chloride with a solution cjf the amine in methanol-chloroform, or by percolating the amine solution over a bed of thiotrithiazyl chloride mixed with dry sand. Although the colorea species generally decompose within a few mindes, iinear calibration curves are readily obtained by antroiling reaction conditions. These observations w v e been applied to the determination of airborne taluene-2,4-diisocyanate d t concentrations as IQW as 9.01 ppm by volume. df
THEREACTIONof undiluted organic amines with thiotrithiazyl chloride (S4N3+CI-)to give highly colored products has been reported by Cohen, Kent, MacDiarrnid, and Marcantonio ( I ) . Although MacDiarmid (2) has qualitatively tested for thiotrithiazyl ion with aniline, there are no reported examples of the determination of amines with thiotrithiazyl chloride. In this paper we describe a new spectrophotometric method for determining small quantities of some primary aromatic amines with thiotrithiazyl chloride. The sens1tiv:ty of the (1) M. E. Cohen,
R. A. Kent, A. G. MacDiarmid, and N. H. ,Marcantonio, U. S. Depf. Comm., Ogfce Tech. Serc. P B Repf., 161,883(1960).
(2) A. G.
MacQiarmid, J . Am. Chem. Soc., 78, 3871 (1956). VQL 39, NO. 6, MAY 1967
581