X-Ray Fluorescence Spectrometric Analysis of the Copper(II) a nd M ercury(II) Co m plexes of 6-ChIoro2-methoxy-9-t hiolac ridine KENNETH E. DAUGHERTY, REX J. ROBINSON, and JAMES I. MUELLER Department of Chemisfry and Ceramic Engineering Division, University of Washington, Seattle, Wash.
b The quantitative analysis of the copper(l1) and mercury(l1) complexes of 6-chloro-2-methoxy-9-thiolacridine employing x-ray fluorescence spectrometry is described. Normally the direct comparison method or the internal standard method is used for estimation; however, the method currently presented in this paper was found to be superior to either of these methods. Calibration curves were obtained by plotting the logarithm of the ratio of the fplfp, peak ,, peak area of mercury to the KK area of copper vs. the mole % of mercury and likewise by plotting the logarithm of the ratio of the copper peak area to the mercury peak area vs. the mole % o f copper. Almost a straight line was obtained in each case. This ratio method was faster and simpler than previous methods and had a relative error of f3% in the range of 0 to 6 mg. of copper and 0 to 20 mg. of mercury. It seems likely that the method may be generally applicable to the large group of organic agents.
I
N 1940, DASGUPTA@,4) reported the
synthesis of 6-chloro-2-methoxy-9thiolacridine (termed “thiolacridine” in this paper). He found that copper(I1) and mercury(I1) were complexed quantitatively with an excess of thiolacridine, copper forming a violetbrown precipitate and mercury an orange precipitate. The structures of these compounds are as follows:
1098
ANALYTICAL CHEMISTRY
sI
98 7 05
S
Hg
I
)3)3-0cH3
c1 mercury (11)-dithioacridine Hg(CidHgOSSC1)2 Thiolacridine was found by the authors to complex with a number of other metals. Silver(1) formed an orange-red precipitate, gold(II1) a deep red precipitate, platinum(IV), a purpleblack precipitate, and palladium(I1) a deep purple-red precipitate. All were amorphous in appearance. The metals of groups I1 and I11 of the periodic table coprecipitated in the determination of copper(I1) and mercury(I1) while those of groups IV and V had no effect. Only qualitative work was conducted with the metallic ions other than copper(I1) and mercury(I1). Thiolacridine or the acridine derivatives in general have not been exploited as yet in analytical chemistry. Using Das Gupta’s discovery that thiolacridine quantitatively complexed copper(I1) and mercury(I1) we undertook a project to simultaneously analyze these metal ions in a mixture. Spectrophotometric and fluorometric titrations of the copper and mercury complexes with thiolacridine were explored. But since these complexes break down in nonaqueous solvents and an excess of thiolacridine is needed to quantitatively precipitate copper(I1) and mercury(II), these methods of analysis could not be used. Experiments with x-ray fluorescence spectrometry were next tested since the two undesirable conditions mentioned previously are not involved in the x-ray analysis of the solids. Of the elements present in the samples only the metals being complexed were detected by the x-ray equipment used
for this investigation. The other elements, with atomic numbers less than 22, were not detected by this equipment and thus did not interfere. This is owing to their low emission energies being air attenuated before they reach the detector, and by their absorption by the diffracting crystal and by the counter window. EXPERIMENTAL
Reagents. PREPARATION OF 6CHLORO - 2 - METHOXY - 9 - THIOLACRIDINE REAGENT.This reagent was prepared from 6,Q-dichloro-2-methoxyacridine (Sldrich Chemical Co.) by a procedure developed by Kitani (5). A solution of thiolacridine was prepared by dissolving 0.10 gram in 100 ml. of 95% ethyl alcohol mixed with 3.5 ml. of 0.1N aqueous sodium hydroxide solution. This solution was filtered and used the same day as prepared. METAL IONSOLUTIONS.Standard solutions (0.1M) of copper and mercury were prepared from Mallinckrodt analytical reagent grade cupric acetate and mercuric acetate in distilled water. The solutions were then standardized gravimetrically with thiolacridine by the procedure used by Das Gupta (4). It was ascertained that freeing the precipitate of thiolacridine by washing with ethyl alcohol was indicated by the disappearance of fluorescence in the filtrate in ultraviolet light. Values of 6.55 mg. of Cu and 20.10 mg. of Hg per ml. of solution were obtained through analysis as the thioacridines. A value of 6.57 mg. of Cn per ml. was obtained through analysis by the electrolytic deposition method. Tenth molar solutions of AgNO,, AuCls.HCl .311z0, HzPtC16, and PdClz were prepared but not standardized since only qualitative work was conducted with them. PREPARATION OF COPPER(I~) AND MERCURY(II) COMPLEXES.A series of solutions (Table I) was prepared rangicg
RESULTS AND
'040
20
-
40 60 MOLE PER CENT COPPER
EO
I
Figure 1. Cu/Hg peak ratio vs. mole copper
70 of
from 0 mole yocopper(I1) and 100 mole % mercury(I1) to 100 mole % copper and 0 mole % mercury. Using a calibrated 1-ml. syringe graduated in 0.01ml. divisions, calcul3ted quantities of the standard copper (11) and mercury(11) solutions were measured into 250-ml. beakers, and diluted with 100ml. portions of disti led water. These solutions were made, just alkaline by the addition of solid sodium carbonate and then just acidic with hydrochloric acid. Thiolacridine ijolution was added to each solution in sufficient quantity to complex quantitahively the copper and mercury. The solutions were stirred for several minutes and allowed to stand overnight. The supernatant liquid was siphoned off and the precipibate was washed three times with 9501, ethyl alcohol, or until there was no fluorescence of the supernatant liquid under ultraviolet light. The precipitates were then dried in a vacuum desiccator. Apparatus. The x-ray analyses were made with a Philips x-ray fluorescence spectrometer with a tungsten target tube operated a t 40 kv. and 40 ma., a lithiuin fluoride analyzing crystal, and a proportional counter detector. Method of Analysis. The precipitate samples were ground to 200 mesh and transferred to cellophane envelopes. This need not be done quantitatively. Thl3 capacity of the regular plastic samp e holder normally used with the x-ray equipment was too large as the sample weight for each run was only about 70 mg. Cellophane envelopes proved convenient both for containing the samples for analysis and for permanently storing the samples. The eivelopes were attached to the back of the plastic sample holder with a spring clip. Other sample containers were tried but did not prove as successful as cellophane.
DISCUSSION
x1.0
X-ray fluorescence patterns of each of the complexes of thiolacridine with Cu(II), Hg(II), A m , Au(III), Pt(IV), and Pd(I1) were obtained separately. A pattern was obtained also of a sample containing all the complexes mixed together in equal portions. The peaks of all six metals were quite distinctive and are listed in Table 11. The peak of tungsten always appeared in the patterns as an impurity from the tungsten target tube. X-ray patterns were obtained for the 11 samples of the various mixtures of the complexes of copper(I1) and mercury(I1) (Table I). Each sample was scanned from 20 = 47' to 20 = 30" a t a scanning speed of 0.25" per minute. ~ , of The Ka,Ka, and the K B ~ Kpeaks copper were not resolved. The L , peak of mercury appeared as did the L,,L, peak of mercury; the latter was not resolved. Additional lines due to tungsten, copper, and mercury were present but did not interfere with the analysis. The peaks of Ag(I), Au(III), Pt(IV), and Pd(I1) did not overlap the copper and mercury peaks when these elements were present in the mixture (Table 11). The K,K, peak of copper and the Lp,L, peak of mercury were selected for use in the quantitative analysis of these elements. The L, peak of mercury is a stronger peak than the LB,L& peak, but was not selected for use because of interference of the tungsten peak. The proper settings of the
Table 1.
Mole 70metal cu Hg
Mg. of Metal
Hg
0.00
20.10 18.09 16.08 14 07 12.06 10.05 8.04 6.03 4.02 2.01
Table II.
1.00 (r
e
W
b
5z
1 0.10
I
0.01
I
20
MOLE
Figure 2. mercury
I
I
3
40 60 SO PER CENT MERCURY
Hg/Cu ratio vs. mole
%
electronic circuit panel were made to keep the peaks on the chart paper when scanned a t 0.25" per minute. The areas under the copper and mercury peaks were evaluated by cutting out the peaks and weighing the paper on a Mettler balance. The paper weights of the peaks along with the copper/mercury and the mercury/copper peak ratios are recorded in Table I. The ratios were then plotted in Figures 1 and 2 on semilog paper us. the mole % copper and
Experimental Data for Copper(l1) and Mercury(l1) X-Ray Fluorescence Analysis Areas of peaks,
cu 0.66 1.31 1.97 2.62 3.28 3.93 4.59 5.24 5.90 6.55
la0
0.00
0.0
10.3 20.5 30.6 40.7 50.7 60.7
70.6 80.5 90.2 100.2
~~
100.0 89.7 79.5 69.4 59.3 49.3 39.3 29.4 19.5 9.8 0.0
mg. of paper cu Hg
Ratios of peak aread Cu/Hg Hg/Cu
0.00
60.18 .. _ _
n on
m
2.98 23.70 53.05 76.94 86.15 181.91
49.58 137.21 184.15 103.85 65.10
0.060 0.173 0.288 0.741 1.323
16.64 5.79 3.47 1.35 0.756 0- 292 0.077 0.053 0.034 0.000
m.i2
191.95 181.90 190.20
53 0.5 _. ._
14.05 10.24 6.15 0.00
3- . 4211 --_
13.03 18.75 29.58 m
X-Ray Emission Peaks in 20 Degrees for Metals Complexing with Thiolacridine
At.
Element Co per Pafadium Silver Tungsten Platinum Gold
Mercury
No. 29 46 47 74 78 79 80
Ka1 44.96 16.70 15.95 5.95 5.26 5.12
Ka, 45.08 16.85 16.10 6.06 5.41 5.26
K81
K8a
40.43 14.86 14.18 5.24 4.64 4.52
40.46 14.86 14.20 5.26 4.67 4.55
La,
Lb1
LO2
38.05
32.29 31.19 30.19
31.76 30.81 29.93
35.92
VOL. 36, NO. 6, MAY 1964
1099
the mole % mercury. Essentially straight lines were obtained. From the calibration curves (Figures 1 and 2), the mole fractions of mercury or copper in a sample can be obtained. In addition, if the copper and mercury are the only metal ions complexed, their exact quantities can be calculated. The product of the mole fractions of copper and mercury and the weight of the precipitated metal complex mixture originally formed establishes the weights. This analytical procedure was tested with copper in the range of 0 to 6 mg. and mercury in the range of 0 to 20 mg. The results shown in Table I indicate a relative error of i=t3%. This method for the estimation of metals in metal-organic compounds by relating concentration to the ratio of peak areas has certain advantages over other methods (1, 8, 6) of estimation. Secondary fluorescence (enhancement) is corrected in the binary metal curve. Variations in sample preparation and in x-ray conditions do not affect the ratio of the two elements. Furthermore this
method permits rapid analysis of samples when a calibration curve h w once been prepared. There are several reasons for complexing the metals with an organic reagent before x-ray analysis. First, certain interfering metal ions may be eliminated and thus, the interpretation of the patterns is simplified. Second, the bulk of the sample is increased greatly which facilitates handling. Finally, a single matrix is established throughout the broad range of peak ratios which lessens changes in absorption effects of the x-rays. This method of analysis has been applied to the determination of copper(I1) and mercury(I1) with thiolacridine as this reagent was of special interest to the authors, but the method is not limited to applications of this single compound or solely to determinations of these two metal ions. Rather it seems probable that the method is much more general and can be extended to other organic reagents. Analysis of the same metal thioacridines by x-ray diffractometry was
found to be feasible but x-ray fluorescence spectrometry proved to be superior, chiefly because of its simplicity. LITERATURE CITED
(1) Birks, L. S “X-ra Spectrochemical Analysis,” Ch. 5, %terscience, New York, 1959. (2) Cullity, B. D., “Elements of X-ray Diffraction,” pp. 415-17, AddisonWesley, Reading, Mass., 1956. (3) Das Gupta, S. J., J . Indian Chem. SOC. 17, 244 (1940). (4) Das Gupta, S. J., Zbid., 18, 43 (1941). (5) Kitani, K., J. Chem. SOC. Japan 75, 396-8 (1954). (6) Liebhafsky, H. A., el al., “X-ray
Absorption and Emission in Analytical Chemistry,” Ch. 7, Wiley, New York, London, 1960.
RECEIVED for review January 3, 1964. Accepted February 4, 1964. Paper presented at the 18th annual Northwest Regional Meeting of the ACS at Western Waahington State College, Bellingham, Wash., June 17, 1963. This work was supported in part through fellowships by the Du Pont de Nemours and Co. and the Shell Oil Co. to Kenneth E. Daugherty.
Statistical Methods and Beer’s Law An Ultraviolet Absorption Study of Binary Solutions Containing a Nitroparaffin R. L. FOLEY, W.-M. LEE, and BORIS MUSULIN’ Deparfmenf of Chemisfry, Southern Illinois University, Carbondale, 111.
b Solutions of nitroparaffins in CC14 or HzO at 25’ C. obey Beer’s law in concentration ranges from 0.01 0 to 0.050M and from 0.12 to 0.30M. Criteria are established whereby statistical methods are used to determine whether Beer’s law is obeyed by solutions.
S
6, i’,l8) have indicated the interaction
EVERAL ULTRAVIOLET STUDIES (1,
of nitromethane molecules in binary solutions. The original purpose of this paper was to compare the behavior of the next three homologous members of the nitroparaffin series in binary solutions with CCl, or HzO as the solvent. However, during the course of study, another purpose, that of rigorously defining how close these binary systems obey Beer’s law, assumed equal importance. This rigorous definition is 1 Present address, University of Wkconsin, Theoretical Chemistry InRtitute, Madison, Wis. 53706.
1100
ANALYTICAL CHEMISTRY
made with use of statistical tooh which have not been previously applied to Beer’s law problems. EXPERIMENTAL
Part of the measurements were taken on a Beckman Model DU spectrophotometer equipped with a Beckman Model DU power supply. Several checks were made with a Beckman Model D U spectrophotometer not equipped with a Beckman hfodel D U power supply t o ensure that no effects were being recorded due to the power supply. Calibrated, 1.0-cm. Beckman silica cells were used for all measurements. The remaining measurements were taken on a Recording Cary spectrophotometer, Model 11, using a set of calibrated, 2.0-cm. cylindrical silica cells. Recordings were made on No. 1100 chart paper from the Applied Physics Corp. The data from the two instruments were treated independently and the results are tabulated separately. Reagents. High purity nitroparaffins (CHgX02, 99.8%; CzHsNOz, 99.4%; 1-C3H7NO2, 95.4%; and Z CaH7N02, 99.4%) were used Rithout Apparatus.
further purification. Fresh samples were used in measurements separated by large time intervals to avoid any decomposition difficulties. Fisher technical grade carbon tetrachloride, Eastman Spectrograde carbon tetrachloride, and distilled water were used as solvents. The technical grade carbon tetrachloride was purified by drying with phosphorus pentoxide and fractionally distilling through a Todd fractionating column a t a reflux ratio of 10 to 1. The fraction boiling between 76.0’ and 77.0’ C. (uncorrected) was retained for experimental use. This fraction had a refractive index of 1.4542 which compared to 1.4550 for the Eastman Spectrograde chemical a t the same temperature. Both solvents were used in the investigation and showed negligible differences spectrophotometrically in the region from 260 to 360 mp. The distilled water was run through a Bantam Demineralizer once before use. Procedure. Pnrt of the solutions were prepared by quantitative dilution of a stock solution which had been prepared by weighing out a specific amount of nitroparaffin on an analytical balance and adding to it dis-
