VOLUME
23, NO. 4, A P R I L 1 9 5 1
613
proximately 2" C. above the saturation temperature. This last reading was made a t 27" C. Figures 3 :Lnd 4 were prepared from data taken in this manner. Detailed tables of data on potassium chloride and potassium bromate may be obtained from the author. APPLICATION AhD ACCURACY
Based on the accuracy ( *0.000035) with which refractive index readings can he made using the Bausch & Lomb dipping refractometer, this method of determining the saturation temperature of potassium bromate and potassium chloride solutions is estimated to give values t h a t are within *0.2' C. of the true saturation trmperature, as indicated by the solubility data used. This ckLgree of accuracy is attainable only if the refractometer reading is properly adjusted before use. The reading for dipping prism A v a s adjusted with distilled water according to the manufactuier's instruction. The reading for prism B was adjusted using a sodium chloride solution whose index could be read on both prism A and B. Frequent checks were made to assure proper adjustment of the refractometer. There are, however, certain salts to which the application of this method does not yield sufficiently accurate results to narrant its use. Salts having extremely small change - of solubility with tc,mperature give results which are only moderately accurate-
for example, preliminary work with sodium chloride yields data from n hich the saturation temperature can be read under the bmt conditions t o an accuracy of only * 1.0' C. The accuracy may vary in certain temperature regions for the same salt-for example, greater accuracy can be expected in the case of potassium bromate xhen working 1% ith solutions saturated in the temperature region from 45' t o 65' C. than can be attained in the region from 25" to 45" C. This results because the change in solubility x i t h temperature of potassium bromate is much greater in the higher temperature range; thus, there is a greater change in refractive index per degree change in saturation temperature. Preliminary work with sodium bromate arid ammonium dihydrogen phosphate using this method gives data which promise an accuracy approaching * 0 . l o C. LITERATURE CITED
(1) International Critical Tables, 5'01. 111, p. 106, Xew York, Mc-
Gral%-HillBook Co.. 1928. (2) Seidell, h.,"Solubility of Inorganic and Metal Organic Compounds," 3rd ed., Tol. I, pp. 687, 697, Kew York, D. Van Kostrand Co., 1940. (3) Urban, Frank, and Meloche, V.W., J Am. Chem. SOC.,50, 30039 (1928). (4) Washburn, E. R., and Oisen, A . L., Ibid., 54, 3212-18 (1932).
I
RECEIVED May 25, 1950.
Spectrophotometric Studies of Dithizone and Some Dithizonates Molecular Extinction Coeficient of Dithizone i n Carbon Tetrachloride STANCIL S. COOPER AND SISTER MARY LOUISE SULLIVAN'
S t . Louis Unicersity, S t . Louis, Mo. Uithizone may be prepared pure in solution, b u t not in t h e dry state. Solutions prepared from t h e dry material are of unknown composition. The dithizone content of these and other solutions can be found if its molecular extinction coefficient is known a t its wave length of niaximum absorption in t h e solvent employed. By complexing purified dithizone in carbon tetrachloride w-ith lead, zinc, silver, and mercury(II), coefficients for t h e primary and secondary absorption maxima (at 620 and 450 mp) of dithizone are found to be (34.60 * 0.88) X l o 3 and (20.30 * 0.82) X lo3, respectively. Coefficients for lead, zinc, silker, and mercury(I1) dithizonates are
Q
IT+4?,Y"I'TATIVE spectrochemical micromethods have been espnnded considerably in recent years through the use of dithizoiie (13, 15, 16). Although a large amount of work has heen done to develop new procedures in the application of dithiz o n ~aiid to improve existing ones, in general t,he development has 1)wn :ilong idnipirical lines. S t ~ c ~ r acontributions l treat the more theoretical aspects of clithizonp and its metal complexes (5-5, 8,11,1R,16). Asystematic applicntion of dithizoiie has been delayed by lack of estcnpive distribution studies and by inability to prepare solutions of the reagent in accurately known concentrations. Because quantitative distribution determinations will involve measurements of 1
I're-ent address, T h e B t . 3 I a r y College, Leavenworth, Kan.
also given. The value for dithizone a t 620 nip in carbon tetrachloride can be found by measuring t h e change in optical density of a carbon tetrachloride solution of excess dithizone when shaken with a n equal kolume of a water solution of h o w - n silver content. The coefficient is given as the ratio of the optical density change to the molar concentration of t h e silber i n the w7ater solution. The coefficient in chloroform a t 605 mp can be found in a similar manner. AIolecular extinction coefficients should be valuable in determining the concentration of dithizone and dithizonates. especially in distribution studies involving these compounds.
optical densities of both dithieone arid the dithizonates it is necessary to have a t hand a method of obtaining the concentration of dithizone in solution. Predetermined concentrations of dithizone are difficult, if not impossible, to prepare by dirwt n-eighing bccause the pure compound has not bren prepawd tate. Satisfactory mcthods of purification of the mxtwin1 in solution are in gcneral use; however, such purified solutions are not of known strength. The molecular estinrtion mefficient of dithizone at its wave length of maximum absorption will allow one to establish its concentration after purification, and to determine similar constants for the dithizonates. The limitrd data available for the molecular extinction coefficient of dithizone have been determined on samples obtained direct neighing and
ANALYTICAL CHEMISTRY
614 are not consistent. Values a t 450, 514, and 620 mp in carbon tetrachloride are given by Liebhafsky and Winslop ( l a ) as 19,000, 4740, and 30,400, and by Clifford ( 4 ) as 18,700 or 17,800, 4510 or 5400, and 31,100 or 29,700, respectively. The molecular extinction coefficient of dithizone in carbon tetrachloride a t 620 mp has been established in this work by complexing purified dithizone with known amounts of the metal ions: lead, zinc, silver, and mercury(I1). Of the four metal complexes, those of mercury(I1) and silver show no absorption, Mhile those of lead and zinc show considerable absorption a t 620 mp in carbon tetrachloride. Therefore, mixed color measurements on lead and zinc dithizonates, together with excess dithizone, cannot be employed, whereas such measurements can be made on mercury(I1) and silver dithizonates to establish optical densities of the excess dithizone a t this wave length. I n the case of mercury(I1) and silver, excess dithizone can be separated from the dithizonates without their decomposition, whereas with the lead and zinc complexes this ie not possible. With lead and zinc all dithizone present may be complexed in carbon tetrachloride if excess metal ion is employed. Optical density of the metal complex together with a determination of the amount of metal it contains allows its coefficient to be evaluated. Either dithizonate can be decomposed by shaking Tvith dilute acid; the metal passes to the aqueous solution n here it may be determined, and the dithizone is released in the organic layer where ita optical density may be measured. From the determined concentrations and measured optical densities, molar extinction coefficients of dithizone and the dithizonates can be evaluated. TF'ith mercury(I1) and silver a knon n concentration of metal ion, [ M + z ] moles per liter, in aqueous solution may be shaken with an equal volume of carbon tetrachloride containing excess purified dithixone of optical density Do D~ and concentration CO moles per liter. The dithixonate, NDz, [HgDz? (9, IO) or AgDz (IO)],is formed equivalent to the metal taken, and the excess dithizone of concentration C, remains in solution. For this mixed color solution we have D , x == D, ~z X -k Dmm. X where the optical densities of mixed color, of dithizone, and of
complex are D,, DeD ~ and , D,,,., respectively. At 620 mp U,,,. = 0, and because D = E 1C it f o l l o ~ sthat the change in optical density a t 620 mp, brought about by complexing with deficient metal ion, is
AD = D O D .- D*D$= ~ ( C -O C,) where ED^ is the molecular extinction coefficient of dithizone. 4 s the concentration of dithizone entering the complex is x [XI and is the same aE the change in dithizone concentration, it follows that
Co - Ce
x[M+"]
(1)
and, a t 620 mp
To obtain C D ~a t 620 m p or at other wave lengths the exceas dithizone may be extracted from the dithizonate-dithizone system, removed t o a volume of carbon tetrachloride equal to the original, and D, D ~ Xobtained. If DOD ~ Xis known it follows that (3) Solutions of dithizone prepared by direct weighing of a carefully purified dithizone (1) should give molecular extinction coefficients that approximate those determined through met'al complexes. REAGENTS AND APPARATUS
Borosilicate Glass-Distilled Water, prepared by redistilling distilled water from alkaline permanganate in an all-borosilicate glass system. This water was used in the preparation of all aqueous solutions. Ammonium Hydroxide. Concentrated C.P. reagent was distilled and ammonia absorbed in borosilicate glass-distilled water. The extracting ammonia solution was prepared by diluting this solution to an equivalent concentration of I part of concentrated ammonia to 100 parts of water (about 0.15 M ammonia). Ammonium Acetate Buffer. SOLUTIONA . A 0.1 M ammo-
Table I. Data for Determination of Molecular Extinction Coefficients in Carbon Tetrachloride of Lead and Zinc Dithizonates and of Dithizone through Lead and Zinc Complexes
Sample
Reacting Solutions Dithizone. [ll+ + I , Buffer mg./lite; mg./liter and p H (approx.) (1) (2) (3) 8 8 8 8 8 10
10
10 10
10
11
12 13 14
1.j
16 17 18 lY
20 21
10 10 10 in ~.
10 10 10 10 10 10 10 10
-4-8.50 A-8.50 -4-8.50 A-8.50 -4-8.50 .4-8,47 -4-8.47 B-8.47 B-8.47
4
A-8 00 A-8 00 -4-8 00 -4-8 00 A-8 00 -4-8 00 A-8 00 A-8 00 A-8 00 A-8 00 A-8 00 -4-8 00
4 4 4 4 8 8 8 6
4
4 4
6 6 6 6
Optical Densities Dithisone, Dns Metal complex At At a t X* 620 m p 450 m p (4) l5) (6)
Concentrations, Moles per Liter x 108 Metal Dithicomplex zone (7) 18)
0.507 0.502 0.511 0.501 0.491 0.373 0.373 0,355 0.382
Lead (X* = 520 mp) 7.12 0,305 7.31 0.312 0.520 7.41 0.314 0 519 7.50 0 . 3 1 5 0.521 7.08 0.320 0.529 5.64 0 . 2 2 3 0.365 5.06 0.229 0.375 5.47 0.236 0.391 5.64 0.233 0.391
0.714 0.714 0.699 0.716 1.365 1.352 1.382 1.096 1.106 1.121 1,122 1.124
0 511 0.517 0 508 0 513 1 020 1.050 1.040 0.812 0,812 0.818 0,829 0.832
0,508
Zinc (X* = 535 m r ) 7.48 0.301 7.25 0.305 7.55 0.301 7.75 0.305 15.4 0.609 15.6 0.607 15,5 0.610 12.4 0.486 11.5 0.477 12.1 0.486 12.1 0,500 11.6 0.486
Molecular Extinction Coefficients x 10-8 Dithizone At At 620 mp 450 nip (10) (11)
Metal complex a t A* (9)
14.2 14.6 14.8 15.0 14.2 11.3 10 1 10.9 11.3
35.8 35.6 35.0 34.5 37.1 32.3 37.1 35.9 34.6 35.3 * 0 . 9 2.5
21.4 21.4 21.2 21.0 22.5 19.7 22.6 21.6 20.6 21.3 * 0 . 6 2.8
15 0 14.5 15.0 15.5 30.8 31.2 31.0 24.8 23.0 24.2 24.2 23.2
34.2 35.6 33.8 33.1 33.1 33.9 33.5 32.8 35.3 33.8 34 2 35.8 34 1 A 0 . 8 2.3 3 4 . 6 * 1.1 3.2
20.0
71.2 68.6 69.0 67.0 69.4 66.2 73.6 64.8 67.8 68.6 * 1 . 9 Av. 2.8 Av. der., 70 95.5 98.5 92.6 92.4 89.0 86.7 89.0 88.4 96.0 92.6 93.0 97.0 A 7.. 92 6 A 2 . 8 Av. dev., Yo 2.9 For Pb and Zn, av. Av. dev., 7%
21.0
20 0 19.7 19.7 19.5 19.7 19.6 20.7 20 0 20 6 21.0 20.1 * 0 . 5 2.5 20.6 * 0.8 3.7
615
V O L U M E 23, NO. 4, A P R I L 1 9 5 1
chloride as to give the approximate concentrations desired (cf. Tables I and 11). All dithizone solutions used for metal extractions w r e thus purified immediately before use. Glassware. Pyrex brand KO. 774 cleaned before use with a warm mixture of concentrated nitric and sulfuric acids, rinsed with borosilicate glass-distilled water and steamed for 15 minutes. Apparatus. Agitation of samples was by means of a mechanical shaker operated a t room temperature (24’ to 26” C.). p H measurements were made with a Leeds & Korthrup pH meter, Model 7661-hl. Transmittances, T , referred to solvents were determined with a Cenco-Sheard spectrophotelometer KO.12,317 employing an entrance slit of 1 mm. and an exit slit with nominal width of 10 mp, giving a spectral range of 14 mp, -4bsorption cells were “precision made,” 1-em. borosilicate glass cells, provided with Cores windows. Optical density, D = -log,,T.
i
I\
MOLECULAR EXTINCTIOh- COEFFICIENTS IN CARBON TETRACHLORIDE
8 0.4
Through Lead(I1) Complex. Specific data for each determination will be found in Table I.
Figure 1. Optical Densities Following Extraction of Carbon Tetrachloride Solutions w i t h E q u a l Volumes of Aqueous A m m o n i a (1 :100)
+
Solid curves. AgDz (2.32 X 10-6 &I excess dithizone) 1. At 620 m p (Amsx., primary for dithizone) 2. At 461 m y (Amax., for AgDE) 3. A t 450 m p (hmax.,secondary, for dithizone) Dotted curves. 4. At 620 5. At 490 6. At 450
HgDz2 (1.25 X 10-6 M my my ( X mar., for HgDz2) m y
+ excess dithizone)
nium acetate solution was adjusted to pII 8.5 with distilled ammonia. Metals were removed by shaking with successive portions of dilute dithizone in carbon tetrachloride until green persisted in the organic phase. Dithizone was removed from the buffer with isoamyl alcohol. The excess alcohol was extracted with carbon tetrachloride. SOLUTION B. Same as Solution A, except that dithizone was removed from the buffer by extracting with carbon tetrachloride. Ammonium Nitrate Solution, 0.1 ill C.P. reagent. Metals were removed by successive shaking with dithizone in carbon tetrachloride. Nitric and Hydrochloric Acids. Reagent grade concentrated acids were distilled; middle fraction was retained. Aqueous Solutions of Metal Ions. Reagent grade chemicals were used. Lead nitrate recrystallized from borosilicate glassdistilled water containing a little nitric acid, metallic zinc dissolved in hydrochloric acid (1-l), eilver nitrate dried at 110” and mercuric chloride recrystallized from dilute hydrochloric acid, were used to prepare stock solutions containing 1 gram of metallic ion per liter in suitable acid strength. Solutions of the concentrations indicated in Tables I and I1 were prepared by diluting these stock solutions with acidified borosilicate glassdistilled water. Carbon Tetrachloride and Chloroform, distilled over calcium oxide and middle fraction retained. Once used, the carbon tetrachloride ?as recovered by the method of Biddle ( 2 ) and used only once again. Dithizone. The Eastman product was purified according to the method of the Association of Official Agricultural Chemists ( 1 ) except for one modification. The final heating was omitted and the solvent was removed a t room temperature in a vacuum desiccator in the dark. Stock solutions of dithizone in redistilled carbon tetrachloride contained approximately 100 mg. of purified material per liter and were stored in the dark a t 10” C. Dilute Solutions of Dithizone. Suitable volumes of stock solution were “stripped” of dithizone just prior to use by means of 1 to 10 ammonia. After separation of phases the aqueous portion was acidified with hydrochloric acid and the precipitated dithizone was extracted with such a volume of fresh carbon tetra-
c.,
Ten milliliters of acetate buffer (Solution A or €3) were mixed with enough stock lead solution to give the concentration indicated in column 1, dilut,ed to 25 ml. with water, and shaken for 15 minutes with 25 ml. of freshly purified dithizone in carbon tetrachloride (column 3). The two phases were separated, and the optical density (column 4) of the organic layer, containing only lead dithizonate (PbDzZ), was measured a t 520 mp, the wive length, h*, of maximum absorption of the dithizonate. .In aliquot of the dithizonate complex was destroyed by shaking with dilute nitric acid ( 1 drop of concentrated acid in 25 ml.). The dithizone equivalent to the lead in the complex remained in the organic phase while the metal ions passed into the aqueous phase. Optical densities, D D ~of, the reformed dithizone in the organic layer were determined a t 620 mp (column 5 ) and a t 450 mp (column 6), the wave lengths of the two absorption maxima for dithizone in carbon tetrachloride. The lead in the acid extract v-as determined b i ~standard mixed color procedure with measurement taken at 520 nip. These values, corrected for the aliquot taken, gave the concentration of lead existing as dithizonate (column 7 ) . The concentration of the dithizone in the organic layer (column 8), after decomposition of the dithizonate, according to the formula PbDzt ( 4 , 6, 7 , IO), is tw.ice that of the lead as dithizonate. The molecular extinction coefficients were calculated by the equation,
EX
=
a, where Dx is the optical density at wave
DX
length h, C is the concentration in moles per liter, and 1 is the thickness of the absorption cell in centimeters (1 cni. in all determinations). Columns 9, 10, and 11 list the coefficients in carbon tetrachloride of lead dithizonate a t 520 mp and of dithizone a t 450 and 620 mp, respectively. Through Zinc Complex. The procedure for experiments employing zinc was the same as for lead, but a shaking time of 10 minutes was employed. Data for the zinc determinations are recorded in Table I. Through Silver Complex. TESTO F STABILITY OF COMPLEX TO Amros~.iEXTRACTIOXS. Twenty-five milliliters of a mixture of 2.32 X M silver dithizonate with excess dithizone were successively extracted with 25 ml. of aqueous ammonia (1 to 100) and optical densities of the carbon tetrachloride solution were determined after each extraction at 450 and 620 nip (hma,. values of dithizone) and a t 462 mp (Amsx. for silver dithizonate). Data are plot,ted as curves 1, 2, and 3 of Figure 1. DETERMIXATIOSS INVOLVING SILVER(listed as samples 22 to 63 in Table 11). Twenty-five milliliters of aqueous solutions of silver ion were employed with concentrations indicated in column 1 and p H indicated in column 2. Samples 44 to 53 and 59 to 63 contained 10 i d . of 0.1 .If ammonium nitrate (Solution C). Each sample was extracted for 10 minutes with 25 ml. of a carbon tetrachloride solution cont,aining excess dithizone. Before the extracting dithizone solution was added, its optical density, Do, was measured at 620 me (colun~n3) and a t 450 m p (column 4). After equilibrium was attained, the layers were separated, and the organic phase was stripped of excess dit,hizone by twice shaking with aqueous ammonia (1 to 100). Optical densities of the stripped silver dithizonate, a t 462 mp3 are given in column 5. The alkaline extract TTas acidified, the dithizone redissolved in 25 ml. of fresh carbon tetrachloride, and the optical density of this recovered excess, D,, was measured at 620 mp (column 6 ) and at 450 mp (column 7 ) . Erceptions to the above procedure are:
ANALYTICAL CHEMISTRY
616
Table 11. Data for Determination of Molecular Extinction Coefficients in Carbon Tetrachloride of Silver and RIercurj-(II) Dithizonates and of Dithizone through Silver and JIercury(l1) Complexes [SI+"I
=
hIeta1 ~ XIoles/Liter x 10: (1)
Sample
~
~
and pH 12)
~ ~
Optical Densities Originala Excess f Dithizone, ~f ~ D ~o ~ hIetal l ~ Dithizone, ~ De , At At complex, At At 620 rnp 450 mp a t A* 620 m p 450 mp (3) (4) f5) (6) (7)
Optical Density Difference, AD = Do - De At At 620 inu 450 nip (8) 9)
~
hIolecrilar Extinction Coeffir,ients X 10-3 Dithieone hIetal complex .it At a t X* 810 m u 450 inp (10) (11) (12)
Silver, X * = 462 mp 22-24 25-28 29-33 34-38 35-43 44-48 49-53 .54-58 59-63
4 64 2 32 2.32 2.32 2 32 2 32 2 32 a 32 0.92i
3 3 3 3 3 2 3 3
60b 6Ob 60b 60b 606 73d 764
64 65-68
0,300 0,500
3.60b 3.60b
60b 4 00d
1 0 0 0 0 0 0 0
78 947 958 967 944 592
951 960 0 666
0.602 0 602
0: iG2 0.568 0.570 0.566 0 584 0.568
...
...
... .
..
0 .i 7 5 0,583 0,584
0.65lC 0.685 0.672 0 $49
0,200 0.172 0.174 0.171 0.176 0.148 0.133 0.153' 0.339f
34.0 * 0.1 ,.. 3 3 . 1 * 0 . 4 19 2 1 0 . 6 3 3 . 8 10 . 7 1 9 . 8 * 0 , 4 31.3 0 . 4 19 1 1 0 . 2 3 3 . 1 1 0 . 3 19 8 1 0 . 3 36 4 10 . 3 2 1 . 3 1 0 1 35 3 1 0 . 2 20 5 1 0 . 1 34.8 1 0 . 1 ... 35.3 * 0 . 3 ... 31.A1 * 0 . 8 8, 2 0 ox * 0 . 7 3.5 2.5
o:iis
0.109 0 124 0.111 0.089 0 083
l I e r c u r y ( I I ) , X* = 490 nip 0 i50 0.252 ... 0.350 . .. 0.250f 0.357 , For Ag and Hg(I1). a v . Av. dev., 70
...
70.0
..
...
35 0 35.7 A O . 2 ... 34.64 * 0.88 20.03 = 0.74 2.5 3.5 .
.
I
Approximate concentrations, ing. per liter, of original purified dithizone solutions: samples 22-24, 14; samples 25-58, 7.5; saniples 59-68, 1.0 Water employed instead of amnioniuni nitrate solution. Strinned once with dilute ammonia solution. '~ - nnlr d 10 ml. of O.Y-3 ammonium n i t i a t e ~ a d d e d . ~ e Excess dithizone not removed by stripping. hence measurement not inade. f Sleasurenient a t 620 m p taken with both complex and excess dithizone in solution a b C
~~
~
~~~
~~
samples 39 to 43 in which the excess dithizone remaining in the organic phase of the reaction mixture was stripped with ammonia only once; and samples 59 t o 63 in which the optical densit,y of the excess dithizone was determined (without stripping) a t 620 mp in the presence of the dithizonate. The differences in optical densities, AD, of the original extracting dithizone solut,ion and the recovered excess dithizone are given in columns 8 and 9, respectively. The Concentration of the metal complex is the same as the metal ion concentration in the original aqueous phase (column 1). From silver dithizonat'e, .4gDz ( I O ) , the concentration of dithizone in the comples is the same as that of the complex (Equation 1). The remaining portion of Table I1 is similar to that of Table I. The extinction coefficients of the dithizonate a t 462 mp are shown in column 10. The dithizone coefficients for 620 and 450 nip (columns 11 and 12) were obtained by Equation 2 or 3 by dividing values of ADD, in columns 8 and 9, respectivrlg, by thr corresponding value in column 1. Through Mercury(I1) Complex. Successive extraction of a carbon tetrachloride solution of mercury(I1) dithizonate and dithizone with aqueous ammonia proved that the mercury(I1) complex was not destroyed, and the dithizonate of mrrcury(II), like that of silver, does not absorb a t 620 mp (see curves 4, 5 , and 6 of Figure 1 and curve 4 of Figure 2). The procedure followed in determinations with mercurj (11), except for sample 64, was the same as that for samples 54 to 58 with silver. With sample 64 the excess dithizone was stripped away, and the optical density taken a t 490 mp, the wave length of maximum absorption of HgDz2. Transmittancy curves were determined for various solutions involved in sample 64 and are shown in Figure 2. Specific data (for samples 64 t o 68) are found in the lower part of Table 11. The concentration of dithizone in the mercury(I1) complex, HgDz2 (IO),is twice the concentration of the complex. Values of t,he coefficient for dithizone a t 620 mp (column 11) were obtained by Equation 2 by dividing values of ADD.. (column 8) by two times the concentration of complex (column 1). For sample 64, the coefficient of HgDzz was obtained by dividing the optical density in column 5 by the concentration in column 1. By Direct Weighing. Ten milligrams of purified dithizone were dissolved in 100 ml. of carbon tetrachloride. Five separate samples, with dithizone concentrations between 2 and 8 mg. per
liter, were prepared from this stock and optical densities determined a t 620 and 450 and at 515 inp (the absorption minimum) in carbon tetrachloride. l\lolrcular extinction coefficients froin these data are: (33.8 * 0.2) X lo3, (19.9 * 0.3) X lo3, and (4.8 * 0.2) X lo3 a t the respective xwve lengths. -2 second sample of purified dithizone gave for six separate determinations
2ol Ol
I
400
I 600 WAVE LENGTH, MILLIMICRONS I
I
500
I
I
700
Figure 2. Per Cent Transmittance as a Function of Ware Length 1. 2. 3.
PbDzz (5.1 X 10-8 H) in CCli Dithizone (approximately 17 X 10-6 M) in CCI4 HgDzz (5.0 X 10-6 M ) dithizone (approximatel) 7 X 103 .If) in
4.
HgDm (5.0 X IO-' M ) i n CCh DithiLone (approximately 15 X 1 0 3 .VJ) in CHCla
+
CCl4
5.
V O L U M E 23, NO. 4, A P R I L 1 9 5 1 the respective values (33.6 and (5.0 * 0.2) x 103.
*
0.5) X lo3, (20.5
617
*
0.2) X lo$,
RIOLECULAR EXTINCTION COEFFICIENTS IN CHLOROFORM
For purposes of comparison a stock solution of purified dithizone n a~ prepared in chloroform, from xvhich six separate deterniinationb of the molecular extinction coefficients were made a t 605 and 440 mp (the primary and secondary absorption maxima) and at 505 m r (the absorption minimum) in chloroform. Values a t thr respective wave lengths are (41.5 * 0.5) X lo3, (16.0 =t 0.2) X 103, and (6.0 * 0.1) X IO3. DISCUSSIOS
Reference to Figure 2 shows that lead dithizonatr (curve 1) in a concentration of 5.1 X N absorbs to the extent of 5% (1-cm. crll) at 620 nip, the wave length of maximum absorption (curve 2) for dithizone in carbon tetrachloride, and that mercury(I1) dithizonate (5.0 X lo-‘ MI curve 4)does not absorb a t this wave length. -4t 620 mp zinc dithizonate absorbs more greatly than does Itlad dithizonate a t the same concentration. All mercury (11) dithizonate solutions employed were of the same concentration as is represented by curve 4 and no significant absorption by silver dithizonate a t 620 mp was observed for any solution used. Therefore measurements on mixtures of mercury(I1) dithizonate or silver dithizonate’and dithizone a t 620 mp were representative of dithizone alone, whereas this was not true of mixtures of lead dithizonate or zinc dithizonate with dit,hizone a t this wave length. I n the reactions of excess lead and zinc Yith purified dithizone (Table I ) the mole ratio of metal to dithizone rvas approximately 2.5 and 4.1 for lead and 4.9, 7.8, and 9.8 for zinc. These rat,ios in each case were sufficiently large a t the pH employed to bring about quantitative complexing of all the dithiaone. The resulting solutions were shaken for sufficient time, 15 minut,es for lead solutions and 10 minutes for zinc solutions, for extraction to be complete. The optical densities of the purified dithizone employed (column 3, Table I ) were the same, within the limits of measurement as those on t,he recovered dithizone from the metal complex (columns 5 :rnd 6). Mixed color analyses of both lead and zinc showed quantitative recovery-that is, the metal reverted from the metal coniplex (column 7 ) t,ogether with that left when the buffer phase n.as rxtracted by dithizone in carbon tetrachloride was the same as that taken (column 1). I n calculating the concentration of dithiaone (column 8) the formulas of lead and zinc dithizonates were taken as PbDz? and ZnDzt. This has been shown for lead 1,s Fischer ( 6 ) , Clifford (41,Irving and coworkers (IO), and Hibbits ( 7 ) employing the method of continuous variation. ZnDz,?v a s assumed here by analogy. The fact that €620 and elj0 for dithizone from the zinc complex are the same as from PbDz2 (and also from HgDz? and AgDz) is confirmation that the formula of t,he zinc complex as assumed is correct. The curves of Figure 1 show that neither mercury(I1) nor silver dithizonate is decomposed by extraction with as many as five separate portions of 1 to 100 ammonia and that any excess dithizone in a mixture with mercury(I1) or silver dithizonates is removed by t\To such extractions. Thus these dithizonates can be freed of excess dithizone in order to determine the optical density of either a t whatever wave length is desired. In order to obtain the optical density of dithizone a t 450 mp, a separation from either mercury(I1) or silver dithizonates must be made (compare curves 1 and 4, Figure 2). Reference t,o the curves of Figures 1 and 2 show that a 620 mp absorption is due wholly to dithizone. Therefore if one is interested in ED%620 only, measurements of the optical density decrease on complexing mercury(I1) or silver with excess dithizone will be a direct measure of the dithizone entering the complex. One needs only to know the formula of the complex to evaluate ED^ 020 from AD620 and the concentration of metal employed. Rearrangement of Equation 2 ’ ] plot . of ADS20 against [.Ag+] gives AD620 = C D ~~ ~ n ~ [ ~ f + h
should be linear with a slope which represents ED^ 620. Such a plot for the data of samples 22 to 63 (columns 1 and 8, Table 11)gives a straight line with a slope of 34.3 X lob for enz 6 2 0 . Dat’a for samples 64 t o 68, employing mercury(II), fall on the silver curve if x is taken as 2. This is confirmation of the formula HgDz? for mercury(I1) dithizonate if that of the silver complex is AgDz. Irving and coworkers (10) arrive a t the same conclusion by their “reversion” procedure, and have prepared and analyzed thr mercury(I1) complex (9) and show it to be HgDzl. Therefore Equation 2 gives CD, 6 2 by ~ either procedure of determining ADB2o. Equation 3 is valid for calculating (50, as the curves in Figure 1 show that only the excess dithizone is extracted with ammonia. I n complex formation with silver the mole ratio of dithizone to silver was approximately 1.1, 1.2, and 2.2, giving an excess of dithizone over metal of 10, 20, and 120% (assuming AgDz), sufficient to give complet,e complex formation a t t,he pH values employed. For mercury(I1) the ratio was 3.5, giving an excess of dithizone (assuming HgDz2) of about 150%. Samples that were treated with ammonium nitrate (samples 44 to 53 and 59 to 63, Table 11) give consistently greater values for €620 and e4)0 for dithizone than did samples without this material. This was probably caused by incomplete demetaling of this solution. I t has been shown (1.4) for the purest dithizone obtninGd in this laboratory that the most probable value for the ratio of opticaul DDZ620 densities in carbon tetrachloride at 620 to 450 mp is -= 1.68.
DD, 450
This ratio for the dithizone solutions employed for determinations involving lead (samples 1 to 9) was 1.66 * 0.01, zinc (samples 10 to 21) was 1.69 * 0.01, and silver (samples 22 to 58) was 1.69 =t 0.01. The recovered excess dithizone in the determinations with silver showed a ratio of 1.59 * 0.07. For two separate samples of dithizone from which e620 and ~ 4 5 0were obtained by direct weighing, the optical densit,y ratios were 1.64 * 0.02 and 1.69 * 0.04, respectively. The existence of a lon value for the optical dmsity ratio indicates the presence of an oxidation product of dithizonc, (diphenylthiocarbadiazone) or, in the case of silver determinations, some silver dithizonate, either of which absorbs a t 450 m,u but not 620 mp. Tahles I and 11 list the data for a total of 68 separate detwminations of B 2 in ~ carbon tetrachloride based on metal complvx formation. The average of these 68 values is (34.64 * 0.94) x lo3. The necessity of determining lead and zinc by a mixed color procedure makes these values less reliable than the more direct determinations through mercury(11) and silver, even though the purity of the dithizone employed in all cases is comparable. The average of all values from silver and mercury(I1) dithizonates is (34.64 * 0.88) X lo3, giving a best value of (34.60 =t 0.84) X 103 with an average deviation of 2.4%. In samples 22 to 53 thr recovered excess dithizone was less pure than the parent dithizonr 450). Samples 44 to 53 and 59 to 63, t o (compare DD, WO:DD, which ammonium nitratcs was added, are subject to possible error by metals which may wire from this source. Therefore, of thc, silver determinations, those subject to least error are samples 54 to 58, which give an average value of ED^ e20 = (34.8 =t 0.1) X 103. Because this agrees so closely with all values through metal co111plex formation (34.64 * 0.94) X lo3 and with all values through silver and mercury(I1) (34.60 * 0.84) X lo3, the latter value is considered as more probably correct. In establishing ED^ F20 for one’s own use the procedure folloxved for samples 54 to 58 is recommended-complexing a purified dithizone with a known but drficient quantity of silver and determining the optical density change a t 620 mp on the mixed color solution in carbon tetrachloride. The value of (33.7 * 0.4) X lo3 for ED^ 620 from eleven direct aeighings of dithizone is considered as approximate only. The fact that values from direct weighing agree reasonably well with those based on assumed formulas of metal complexes indicates that these assumed formulas are correct. Tables I and I1 also show 50 separate determinations of ED. 4jg
ANALYTICAL CHEMISTRY
618 in carbon tetrachloride based on metal complex formation, the average for which is (20.29 * 0.82) X 103. Eleven determinations by direct 1Teighing of purified dithizone gave (20.2 =t0 2) X l o 3 for this value. The values for ED%460 and of E for the metal dithizonates a t their wave length of maximum absorption (Tables I and 11) are to be considered as less accurately known than ED. 620, because the thiocarbadiazone, from oxidation of dithizone, is nonreactive to metal and absorbs in this region of wave length. For ED^ 450 the presence of traces of metal dithizonates would produce abnormally large absorption a t 450 mp. The values of the molecular extinction coefficients of dithizone in chloroform are reported with no claim as to accuracy, as they were determined on weighed “purified” dithizone. Because dithizone is actually purified in solution ( 1 ) and its molar extinction coefficient a t 620 mp in carbon tetrachloride is knon.n, or can be ascertained, it becomes unnecessary t o obtain the solid material in a highly purified condition in order to prepare dithizone solutions of known concentration. A stock solution of dithizone in carbon tetrachloride or chloroform can be purified by stripping with dilute aqueous ammonia, phase separation made, and the dithizone reverted t o fresh carbon tetrachloride. The optical density a t 620 mp for this purified dithizone can be measured and its concentation ascertained by use of 34.6 X 108 for its molecular extinction coefficient. Molecular extinction coefficients for the metal dithizonates can now be determined from the known value for dithizone. SUMMARY
The preparation of dithizone solutions of known concentration from “purified” solid dithizone is not possible because the dry product is not pure. Dithizone can be prepared pure in solution and its concentration can be determined from its optical density and molar extinction coefficient. By complexing purified dithizone in carbon tetrachloride with lead, zinc, silver, and mercury(I1) the molar extinction coefficients of dithizone a t its primary and secondary absorption maxima,
620 and 450 mp, are found to be (34.60 =t 0.84) x l o 3 and (20.30 * 0.82) X 103, respectively. Coefficients for dithizonates a t the wave length of their maximum absorption in carbon tetrachloride are found to be: (68.6 * 1.9) X lo3 a t 520 mp for PbDz2; (92.6 * 2.8) X l o 3a t 535 mp for ZnDz,; (27.2 * 1.8) X lo3 a t 462 my for AgDz; and 70 X lo3 a t 490 my for HgDz,. The coefficient for dithizone a t 620 mp is most easily determined by measuring the change in optical density of a purified dithizone in carbon tetrachloride when shaken with an equal volume of aqueous Ag+ whose concentration is known, [Ag+],but which is in deficient quantity to complex all the dithizone. The coefficient is given by
e620 =
AD020 __
IAg’I‘
LITERATURE CITED
(1) Assoc. Offic. Agr. Chemists, “Official and Tentative hlethods of
Analysis,” 1945. (2) Biddle, D. -4., ISD.ENG.CHEM.,ANAL.ED.,8, 99 (1936). (3) Biefeld, L. P., and Patrick, T . M., Zbid., 14, 275 (1942). (4) Clifford, P. A., J . Assoc. Oj%. Agr. Chemists, 26, 26 (1943). (5) Clifford, P. A., and Wichmann, H. J., Ibid., 19, 130 (1936). (6) Fischer, H., 2.angew. Chem., 47, 685 (1934). (7) Hibbits, J., thesis, St. Louis University, 1950 (8) Greenleaf, C. A., J. Assoc. Ofic. AQT.Chemists, 24, 337 (1941). (9) Irving, H., Andrew, G., and Risdon, E. J., J . Chem. soc., 1949, 545.
(10) Irving, H., Risdon, E. J., and Andrew, G., Ibid., 1949, 537. (11) Kolthoff, I. hI., and Sandell, E. B., J . Am. Chem. SOC., 63, 1906 (1941).
(12) Liebhafsky, H. A., and Winslow, E. H., Ibid., 59, 1966 (1937). (13) Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Chap. IV, See. I, New York, Interscience Publishers, 1944. (14) Tipton, G. hI., S.J., unpublished thesis, St. Louis University, 1949 (to be published later). (15) Welcher, F. J., “Organic Analytical Reagents,” Vol. 111, New York, D. Van Nostrand Co., 1947. (16) Wichmann, H. J., IND. EXG.CHEM.,ANAL.ED., 11, 66 (1939). RECEIVED July 21, 1950. Abstracted from the thesis presented by Sister Mary Louise Sullivan, S.C.L., to the faculty of the Graduate School of St. Louis University in partial fulfillment of the requirements for the degree of doctor of philosophy.
Densities and Refractive Indexes for Propylene Glycol-Water Solutions GORDON
M A C B E T H AND A. RALPH THOMPSON University of Pennsylvania, Philadelphia, P a .
A
SALYTICAL data which may be used for determining the compositions of aqueous propylene glycol ( lJ2-propanediol) solutions are presented in this paper. These were obtained by making careful measurements on a number of solutions for density a t 35 O C. and for refractive index a t 25 O C. Values for the pure organic material have been reported in the literature (1- 5 , 5 ) , but only a graphical representation of the specific gravity for solutions of propylene glycol in water could be found (1), PREPARATION OF SOLUTIONS
Commercial grade, 99%+, propylene glycol was fractionated a t approximately IO-mm. absolute pressure in an 18-inch glass column, 0.5 inch (1.25 cm.) in diameter,, packed with 3/16-in~h glass Fenske rings, using a high reflux ratio (about 30 to 1). To protect this hygroscopic material, the distillation was carried out in an entirely closed system and the distillate receivers were vented through drying tubes containing anhydrous calcium sulfate. Only the middle third of the constant boiling point distillate \vas used, the balance being discarded.
The refractive index, ~ Z D ,of the glycol at9 received was 1.4314 a t 25 O C. and that of the purified material was found to be 1.4316, which agrees exactly with that determined by Schierholtz and Staples (6). The specific gravity, compared t o water in a separate determination a t 20” C., was found to be 1.0381. This is in precise agreement with two previously reported values ( 1 , 8 ) and very close to others (5, 5 ) . The water content of the purified propylene glycol was determined by means of the Karl Fischer reagent t o be 0,04y0 by weight. Boiled distilled water from a laboratory Barnstead still waa used in making up the solutions. Nine solutions of varying glycol concentration from 10 to 90 weight % were prepared by pipetting first the water and then the glycol into 50-ml. ground-glassstoppered Erlenmeyer flasks. The actual amounts of each constituent added were determined by weighing t o 0.1 mg. on an analytical balance. DENSITY AND REFRACTIVE INDEX DETERMINATIONS
Density measurements were made using 10-ml. Weld specific gravity bottles and a constant temperature bath maintained a t