Esters are easily converted to alkyliodides (IO) but in some cases their solubility in aqueous hydriodic acid is very low. Suitable solvents were therefore sought. Among the recommended solvents (10, 11) phenol is the best, but in our procedure it interfered with the titration as a small part of it is extracted by benzene and titrated as an acid. Propionic acid and propionic anhydride are suitable solvents which do not interfere. Both solvents give quantitative results when used with 70% hydriodic acid . n a ratio of 1:l. The results of a series of esters analyzed by this procedure appear in Table I11 and show satisfactory recoveries. Addition of red phosphorus suppresses the formation of free iodine which appears when no solvent is used, and interferes with the deterinination. Unless extremely pure phosphorus is used, there is the possibility of fclrming phosphine, which is extracted and titrated as an acid. Various solvents such as pentane, hep-
tane, toluene, and benzene were tried for extraction of the alkyliodides from the reaction mixture. Benzene was the most suitable solvent,. The accuracy of the method was tested by analysis of highly pure n-hexylacetate, 1-hesadecanol, and l-octadecanol. Table IV shows that the described procedure gives results with a relative error of about 1%. The method was applied to a plasticizer, dioctyl phthalate (DOP, Table 111), and quantitative results were obtained with and without a solvent. It is therefore reasonable to assume that in the cases where the plasticizers are not sufficiently volatile to be determined by the method developed by Esposito ($), this procedure can be used. ACKNOWLEDGMENT
The authors thank Aaron J. Kalb for help in the preparation of the manuscript and Ephraim Katchalski for his kind interest during the course of this work.
LITERATURE CITED
(1) Anderson, D. M.W., Duncan, J. L., Talanta 8, 241 (1961). (2) Esposito, G. G., ANAL. CHEM.35,
1439 (1963). (3) Fritz, J. S., Lisicki, N. M., Ibid., 23, 589 (1951). (4) Ingold, C. K., “Structure and Mechanism in Organic Chemistry,” p. 413, G. Bell and Sons, Ltd., London, 1953. (5) Kirsten, W. J., Ehrlich-Rogozinski, S., Mzcrochim. iicta 4, 786 (1955). (6) Kretx, R., 2. Anal. Chem. 176, 421 (1960). (7) Patchornik, A., Ehrlich-Rogozinski, S., ANAL.CHEM.31, 985 (1959). (8)Patchornik, A., Ehrlich-Rogozinski, S.,Ibid., 33, 803 (1961). (9)Sporek, K. F., Danyi, M. D., Ibid., 34, 1527 (1962).
(10) Steyermark, A., “Quantitative Organic Microanalysis,” p. 422, Academic Press, New York, 1961. (11) Wilson, C. L., Wilson, 1).W., “Comprehensive Analytical Chemistry,” p. 669, Elsevier, New York, 1960. (12) Zeisel, S., Monatsch. 6, 989 (1885). RECEIVED for review Xovember 5, 1963. Accepted December 30, 1963. Work supported by research grant (AM-5098) from U. S.Sational Institutes of Health, Public Health Service.
Chemical Shifts of Methyl Protons in Polymethylnaphthalenes FOCH FU-HSIE YEW, ROBERT J. KURLAND, and BEVERIDGE J. MAlR Petroleum Research hboratory, Carnegie Institute of Technology, Pittsburgh, Pa. The proton chemical shifts of the methyl groups in 24 methyl-substituted naphthalenes have been determined. Those chemical shifts are correlated with empirically derived parameters related to the effects of methyl group substitution on the aromatic nucleus and to steric effects. The prediction of chemical shifts made possible by this correlation is valuable for the identification of the polymethylnaphthalenes in petroleum.
A
a Flrogram on the identification of dinuclear aromatic hydrocarbons from petroleum, the nuclear magnetic resonance spectra of 24 polymethylnaphthalenes were obtained. -4correlation of the data has provided an insight inxo the structural factors which affect the chemical shifts of the methyl protons. s PART of
cycles per second and a field strength of 14,560 gauss. The measurements were made on solutions containing 2 to 5 ’ of hydrocarbon in carbon mole % tetrachloride, with tetramethylsilane (ThfS) as an internal reference. The T M S signal was taken at 10 7 units ( 5 ) . The chemical shifts were measured by means of the audio frequency side band technique. The effect of concentration change, in the range 1 to 10 mole %, on the chemical shifts of the methyl protons of 1- and 2-methylnaphthalenes is given in Figure 1. The chemical shifts decrease 0.02 unit as the concentration is decreased from 5 to 2 mole %. According to Pople, Schneider, and Bernstein (IO), smaller shifts due to changes in concentration may be expected with increasing substitutions-i.e., greater molecular volume-of the aromatic ring. Therefore, the uncertainty due to concentration differences for the polymethylnaphthalenes is expected to be within d~0.01 p.p.m.
EXPERIMENIAL
RESULTS
Nuclear magnetic resonance spectra were obtained for 24 methyl-substituted naphthalenes with a Varian V-4203 High Resolution N M R spectrometer, operating at a frequency of 60 mega-
The 48 chemical shift values measured for methyl groups from 24 methyl naphthalenes are given in Table I. The lettering and numbering system for the naphthalene nucleus is:
78w9,: 68
0
1
a
a
6
4
The methyl groups fall into three categories: (a) @-methyl groups, for which chemical shifts are between 7.50 and 7.70 T units; ( b ) a-methyl groups (except for dimethyl substitution in the 1,s- and 4,5-positions), for which chemical shifts are between 7.36 and 7.50 7 units; and (c) a-methyl groups, disubstituted in the 1,8- and 4,5positions, for which chemical shifts are in the range 7.14 to 7.26 T units. I n addition, two general trends may be observed: (a) The chemical shifts tend to increase as the number of substituents on the nucleus is increased; and ( b ) this increase in the chemical shifts, for a given number of substituents depends on the position of the other methyl groups. For example, the increase in the chemical shift of any methyl group with a n adjacent substitution is greater than that caused by a nonadjacent substitution in the same ring, which in turn is greater than that resulting from substitution in the VOL. 36, NO. 4, APRIL 1964
843
c -
2-me t hylnopht h a l e n r
-
cn
.-C &
¶
-
I - m e t h y Inaphthalenr
s 7.40 e
e
opposite ring. These two trends are also observable in the case of 1,8- and 4,5-dimethyl substitution. By assigning suitable values to the substitutional and steric factors suggested by the foregoing trends, a n empirical equation was developed to permit the calculation of the chemical shifts for all the methylnaphthalenes. The values of the constants were adjusted by the method of least squares;
where mole
percent
Figure 1. Chemical shifts of the methyl protons of the 1- and 2-methylnaphthalenes with respect to the concentration in carbon tetrachloride
Boa = 7 36 7 for 1-methylnaphthalene 6O8 = 7.52 I for 2-methylnaphthalene A61 = 0 10 p.p.m., contribution to the
=
Table I.
Measured and Calculated Values for Chemical Shifts of Methyl Protons in Polymethylnaphthalenes 6LYi 7 )
Type
tion
l-Methyl-c 2-Meth~l-~ 1,4-Dimethyl-e 1,5-Dimethyl-f lJ8-Dimethyl-u 2,3-llimethyl-e 2,6-Dimeth~l-~ 2,7-Dimethyl-f 1,2-Dimethyl-f 1,3-Dimethyl-e 1,7-Dimethyl-e 1,4,5-Trimethy1-e
a
13 2a 28 a, 8 ,
30r
2,3,6-Trimeth~l-~
2
1,2,4-Trimethyl-c
Posi- Measured Calculatedb tion ured Calculatedb
1
7 .36d
1,4
7.41 7.41 7.38 7.37 7.18 7.16
2
7. 52d
2,3 2;6 2.7 2' 3 7
7.65 7.54 7.54 7.60 7.55 7.52
$3 2
7.65 7.63 7.55 7.54 7.62 7.67
3
7.58 7.58
6
7.53 7.54
6
7.65 7.62 7.63 7.62
~
'J5
178 1 1 1 1
4 5
313
613( 7 )
Posi- Meas-
Compound"
7.48 7.41 7.38 7.43 7.18 7.14
7.46 7.41 7.37 7.42 7.21 7.17
a,213
lJ6,7-Trimethyl-*
1
7.49 7.43 7.42 7.42 7.42 7.42 7.41
4a
1,4,t5,8-Tetramethyl-e
413
2,3,6,7-TetramethyLh
174 53
7.26 7.22 7.26 7.22
1 5 1
7.46 7.52(7.46) 2' 7.40 7.39 3 7.44 7.56(7.43) 2 7 . 4 4 7.56 (7.43) 3 7.20 7.18 3 7.47 7.43 7.25 7.32 (7.26) 2 7.44 7.44 3 7.23 7.23 7.47 7.58 (7.45) 2,3 6,7
a,P
1
4
1,3,5-Trimethyl-e
1
5
2 a,@ 1,4,6-Trimethyl-e
1 4
2a, 213 1,2,3,5-Tetrameth~!-~ lJ2,3,4-Tetramethyl-'
4
1J3J5J8-Tetramethyl-e
3aJ13
3a, 213 1,2,3,5,8-Pentamethyl-a
13
5
1
5
8 2a, 413 1,2,3,4,6,7-He~amethyl-~ 1,4
7.51 7.46 7.42 7.38 7.42 7.42 7.38
7
7.64 7.64 7.69 7.60 7.63 7.63 7.60
7.62 7.53 7.53 7.62 7.57 7.53
7.64 7.64 7.73 (7.67) 7.68 (7.62) 7.77 (7.64) 7.77 (7.64) 7.63
7.72 7.74 (7.68) 7.62 7.69 (7.63) 7.62 7.79 (7.66) 7.67 7.68
Letters indicate source of compound. Values given in parentheses contain an additional correction for bteric effect (see text). API Research Project 58. A t a concentration of 3 mole % in CC14. 8 G. Dana Johnson, Kansas State University. Richard W. King, The Sun Oil Co. 0 Commercial sources. * From Petroleum by API Research Project 6 . a
b
0
~~
844
0
~
ANALYTICAL CHEMISTRY
=
A& =
chemical shift of any methyl group by an adjacent methyl substitution 0.05 p.p.m., contribution to the chemical shift of any methyl group for each nonadjacent methyl group substituted on the same ring 0 01 p.p.m., contribution to the chemical shift of any methyl group for ea:h additional methyl group substituted on the opposite ring (excepting 1,8- and 4,5-substitution) -0.20p.p.m., contribution to the chemical shift of a methyl group at C-1 by the eubstitution of a methyl group at C-8 and vise versa
The chemical shifts calculated for each compound are listed in Table I. The differences between experimental and calculated values are less than 0.04 p.p.m., except for 1,2,3- and 1,2,3, 4-multisubstituted compounds. Those cases are discussed later. DISCUSSION
The constants in the empirical equation provide some insight into the factors responsible for the differences in the chemical shifts of the methyl groups. The positive contribution to the chemical shift of any methyl group has been noted for the methyl substituted benzenes (11) and for some dimethylnaphthalenes ( 6 ) . It has also been noted that the S M R signals of aromatic ring protons shift upfield as the number of methyl substitutions increases (6, 7 ) . There are several factors which may give rise to this upfield shift: ( a ) the perturbation of ring current by the substitutions; ( b ) the redistribution of charges caused by the resonance and inductive or field effect of methyl substitutions; ( c ) a possible magnetic anisotropy of the methyl group. Each of these effects may be both distance and direction dependent. Our attempt to interpret quantitatively the chemical shifts of the methyl protons in terms of one or the other of the effects individually was not SUCcessful; for esampie, Hammet con-
stants, derived from the expression given by Dewar (2’) for substituted naphthalenes and from the electronic atom-atom polarizabilities calculated by Coulson and Daudel ( I ) , did not correlate with the shifts nor did plots of the form (1 - 3 co!j2B)/R3, a formula representing anisotropy effects, correlate with the chemical sh[fts. We feel that any attempt to combine quantitatively the effects giving rise t o the shifts would not be meaningful in view of the large number of parameters involved; we therefore prefer to interpret these shifts on a qualitativv basis. The abnormally low chemical shifts for 1,s- and 4,5-methyl groups may be interpreted in terms of a steric effect. Reid ( 1 2 ) and MacLe:tn and Mackor (6) have observed that protons which are sterically hindered give NMR signals at lower fields than thcse which are unhindered. They also found that 4,saromatic protons in lI4-dimethylnaphthalene give lower chemical shifts than the corresponding protons in naphthalene. The downfield shift resulting from steric hindrance may be satisfactorily explained in terms of Pople’s treatment of nuclear magnetic deshielding in an electrostatic field (9), since distortion of the electron clouds in hindered groups is always expected. The relatively large negative contribution (A64 = -0.20) to the 1,s-or 4,5-
methyl shifts is probably due to the strong steric interaction between the methyl groups a t the 1 and 8, and a t the 4 and 5 positions. One may note that this interaction can be represented by Hirschfelder-Taylor molecular models. The strain is probably sufficient to distort the naphthalene nucleus into a nonplanar configuration. Intramolecular distortions of this kind have been revealed by both x-ray diffraction (3) and ultraviolet spectrometry ( 8 ) . Where three or four methyl groups are substituted on one ring, as in 1,2,3- and 1,2,3,4-methylnaphthalenes,large differences between the calculated and measured chemical shift values are observed. These also are probably due to steric interaction and distortion of the ring (IO). [X-ray measurements show, though not conclusively, that adjacent methyl groups in 1,2,4,5tetramethylbenzene cause a slight distortion of the ring (C).] If negative contributions to chemical shift ( - 0.06 p.p.m. for lI2,3-suhstitution and -0.13 p.p.m. for 1,2,3,4-substitution) are attributed to this increasing steric effect; the calculated chemical shift values are in satisfactory agreement with the experimental values (values in the parentheses in Table I). Methyl group shifts for additional compounds are needed to determine this effect with greater accuracy.
ACKNOWLEDGMENT
Grateful acknowledgment is made to G. Dana Johnson for providing most of the samples used in this investigation. LITERATURE CITED
(1) Coulson, C. A,! Daudel, It., “Diction-
ary of Values of 1Iolecular Constants,” Vol. 11, p. 9, Centre de Chemie Theorique de France, Paris. (2) Dewar, M. J. S., Grisdale, P. J., J . rlm. Chem. SOC.84, 3548 (1962). (3) Donaldson, D. SI., Robertson, J. M., J . Chem. SOC.1953, p. 17.
(4) Harnik, E , Herbstein, F. H., Schmidt, G. AI. J., Ibzd., 1954, p,. 3288. ( 5 ) Jackman, L. ?*I., “Application of Nurlear Magnetic Resonance Spectroscopy in Organic Chemistry,” p. 47, Pergamon Press, Xew York, 1959. ( 6 ) MacLean, C., hlackor, E. L., J l o kcular Physics 3, 223 (1960). ( 7 ) Zbid., 4 , 241 (1961). (8) Mosby, IT. I,.>J . .4m. Chem. SOC.75, 3348 (1953). ( 9 ) Pople, J. .4.,M o l . Phys. 1, 199 (1958). (10) Pople, J. A., Srhneider, W. G., Bernstein, H. J., “High R,ysolution Xuclear Magnetic Resonance, p. 426, McGraw-Hill Xew York. 1959. (11) Ibzd., p. 263. (12) Reid, C., J . Mol. Spectr. 1, 18 (1957). RECEIVED for review Xovember 13, 1963. ilccepted ,January 2, 1964. This research was supported by a grant from the American Chemical Society Petroleum Research Fund.
Titrimetric Determination of Zinc with Tetracyanomo no - 1,I (I-phena nt hroI ine-f e rra te (II) ALFRED A. SCHILT and ANTHONY V. NOWAK’ Department o f Chemistry, Northern lllinois University, DeKalb, Ill.
b
Potassium tetracyano-mono-(1 ,I Ophenanthro1ine)-ferra te(ll) is found to be superior to potassium hexacyanoferrate(l1) as a reagent for the titrimetric determination (of zinc. An amperometric procedure is described that enables titration of 1Oe5M concentrations of zinc. Measurements of stoichiometry, solubility, sensitivity, interferences, and relative accuracy and precision are reported.
T
was undertaken to evaluate the eflectiveness of the divalent complex anion tetracyanomono - (1,lO - phenanthroline) - ferrate(11) as a precipitimetric reagent for the determination of zinc. The [Fe ~ h e n ( C N ) 4 ] -complex ~ ion merits study in this regard since i t is closely related to the hexacyanoferrate(I1) ion, [Fe(CN),j]-4, a well known and useful HE PRESENT WORH
precipitimetric reagent for zinc and certain other heavy metal ions. It was believed that several characteristics of the 1,lO-phenanthroline-substituted ferrocyanide derivative should prove advantageous in comparison to [Fe(CX)b]-4: the anion is intensely colored, thus small amounts may be more readily detected; its formula weight is greater, hence its metal saltq may exhibit lower solubility; and since i t has a divalent rather than a tetravalent charge, i t should exhibit a simpler, more analytically favorable stoichiometry in its precipitation with Zn+Zions. Various disadvantages associated with the classical zinc-ferrocyanide titration have been cited in the literature (1-3, 6, 6). These can be summarized as follows: Precipitate composition depends upon pH. The sohtion must be distinctly acidic to form KIZns [Fe(CN)6]2. Vari-
ations in p H can cause formation of the normal zinc ferrocyanide Zn2[Fe-
(CW, 1.
Attainment of equilibrium is slow unless a n elevated temperature (70” C.) is employed. The use of higher temperature is not only inconvenient but also decreases the sensitivity of the method by increasing the solubility of the precipitate. Use of internal oxidation-reduction indicators requires careful control of solution conditions, and false end points due to adsorption of excess of either reactant are troublesome. External indicators, such as uranyl nitrate, are neither sufficiently sensitive nor convenient for many purposes. Micro amounts of zinc are not satisfactorily determined using the ferrocyanide method. Present address, Department of Chemistry, University of Illinois, Urbana, Ill. VOL. 36, NO. 4, APRIL 1964
845
