Sept. 20, 1954
ORIENTATION I N
AROK~TIC NITRATION RRACTTONS
45Li
TABLE I ORGANIC PRODUCTS ISOLATED FROM Compound
1,it.a
THE
REACTION OF HYDRAZINE WITH BEXZYL SITRATE
M.p., "C. Found
Carbon, % Calcd. Found
Hydrogen. % Calcd. Found
Nitrogen, % Calcd Found
CsH,CHzSzH, , H S 0 3 ., 172.5-173.5 45.40 56.24b 5.99 6.46' 22.69 17.76' .. 61.07 6.23 15.26 (C~HjCHz)zKzHz.HSOj ( C ~ H ~ C H Z ) ~.HSOa XOH .. 188-188.5 69.02 69.42 6.34 6.49 11.50 11.04 (Ce"CHz)&z 139 139-140 85.6i 85.83 7.19 7.05 7.14 7.12 ( CsHjCHz)zX--l\j=CHCeH6e 87 83-84 83.96 84.39 6.71 6.96 9.33 8.5; a Ref. 17. This analysis indicated a mixture of mono- and higher benzylated hydrazine mononitrates. A hydrochloride was obtained from this product, m.p. 213-215' dec. 1,2-Dibenzylhydrazine monohydrochloride has m.p. 215-217' dec. It was not possible to improve either the m.p. or t h e ( A . F. Bickel and W. A. Waters, Rec. trau. chim., 69, 312 (1950)). analysis of this product. Hydrolysis with HC1 gave benzaldehyde an: a hydrochloride, m.p. 198.5-199.5'; N, 11.31 ; reported (ref. 17) for 1,l-dibenzylhydrazine monohydrochloride, m.p. 202 ; N (calcd.), 11.26. mixture after completion of a reaction, immediate extraction with ether, and distillation of the ethanol-ethyl nitrate azeotrope from the ether-soluble fraction. T h e a-naphthylurethan was prepared directly from the azeotropic mixture. Butanol was characterized by boiling point, refractive index, and by derivatization as the a-naphthylurethan. Attempts were made to detect aldehyde b y refluxing the ether-soluble fractions with dilute HCl (to hydrolyze any hydrazones or azines) and adding 2,4-dinitrophenylhydrazine (DNP). A D N P derivative was obtained from one experiment with isoamyl nitrate and was shown to be t h a t of isovaleraldehyde, m.p. and mixed m . p . 122-123'. S'o
evidence of aldehyde formation was obtained from any experiments using n-butyl or ethyl nitrate.
Acknowledgment.-The authors are indebted to Dr. C. A. VanderWerf and Dr. Howard W. Kruse for some of the starting materials; to Dr. S. K. Smith for the mass spectra; and to R. H. Pierson for some of the infrared spectra. Thanks are also due to Kathryne B. Clifton and Mario J. Falbo for assistance in various phases of this investigation. CHINALAKE,CALIF.
[CONTRIBUTION FROM THE DEPARTMENT OR CHEMISTRY A N D LABORATORY FOR NUCLEAR SCIENCEAND ENGINEERING, MASSACHUSETTS INSTITUTE O F TECHNOLOGY AND THE LABORATORY FOR ORGANIC CHEMISTRY, UNIVERSITY O F AMSTERDAM]
Orientation in Aromatic Nitration Reactions] BY
JOHN
F. L.J. S I X M A , ' H. CERFONTAIN' AND R.2 . 4 ~ ~ ;
D. ROBERTS,2~3 JANET K.
RECEIVEDMARCH8, 1954 T h e isotope dilution method has been employed t o determine isomer distributions in the nitration of chloro-, bromo- and iodobenzenes and the relative reactivity of iodobenzene with respect to benzene. For nitration a t 25', the following isomer distributions were found: chlorobenzene, ortho 29.6 1.7%, meta 0.9 5 O.O%, and para 69.5 f 1.7%; bromobenzene, ortho 36.5 f 1.7%, meta 1.2 O.l%, and para 62.4 f 2.8%; iodobenzene, ortho 38.3 f 2.0%, meta 1.8 f 0.3%, and para 59.7 f 2.0%. The reactivity of iodobenzene relative t o benzene a t 25' was 0.13 in acetic anhydride and 0.22 in nitromethane. T h e degree of correlation of the nitration reactivities of benzene derivatives with Hammett u-constants is discussed. Deviations from the Hammett relationship are interpreted in terms of important contributions of transition-state resonance.
*
*
Introduction Considerable disagreement exists as to the importance, in electrophilic aromatic substitution, of transition-state resonance interactions involving the orienting substituent groups and most easily discussed in terms of the customary pentadienate cation model (I) of the transition state.
Although Hammett's statement6 that relative reactivities in the nitration of substituted benzenes parallel the a-constants of the substituents would seem to exclude important resonance contributions of the following type for 0- and fJ-substitutions
I1
I ( 1 ) Supported in part by t h e joint program of the Office of Naval Research and t h e U. S. Atomic Energy Commission; also supported in part by a grant from t h e van't HoffFund of t h e Koninklijke h'ed. Akademie van Wetenschappen. Presented before t h e Division of Organic Chemistry a t the New York meeting of the American Chemical Society, September 13, 1954. ( 2 ) Massachusetts Institute of Technology. (3) Present address: Gates and Crellin Laboratories, California Institute of Technology, Pasadena 4, Calif. (4) U. S. Atomic Energy Commission Predoctoral Fellow, 19491951. ( 5 ) University of Amsterdam.
evidence for significant contributions of such resonance has long been accumulating. Kenner7 found substantial differences in the charge a t the p-position of a halobenzene when calculated from nitration rate data and from dipole moment data, respectively, attributed by him to transition state resonance during nitration. Transition state resonance involving the p-position also has been postu(6) L. P. Hammett, "Physical Organic Chemistry," McGraw-Hill Book Co., Inc., New York, N. Y., 1940, pp. 198-199. (7) G. W. Kenner, Proc. Roy. Soc., 8 1 8 6 , 119 (1946).
4320
J. D. ROBERTS, J. K. SANFORD, F. L. J. SIXMA, H. CERFONTAIN AND R. ZAGT
lated by Swain and Langsdorfs to account for the two curves correlating rates of bimolecular reactions of substituted benzyl bromides with trimethylamine. One curve accommodates unsubstituted and m-substituted benzyl bromides, the other accommodates p-substituted benzyl bromides. Prominent resonance contributions would be anticipated in 0- and p-substitution when the substituent can accommodate a positive charge as in the halobenzenes, cinnamic acid and certain related compounds. These compounds have been least satisfactorily described by quantum mechanicalYand other7,localculations. As not enough experimental data were available for an adequate test of the Hammett relation in this critical area, measurements of the sniall degree of m-isomer formation in the nitration of chloro-, bromo- and iodobenzene were made. Method of Analysis The isotope dilution procedure” was used for the determination of isomer distribution and the reactivity of iodobenzene relative to benzene for which only an approximate value has been reported.’l For the isotope dilution analysis of isomer distribution, halobenzenes containing radioactive Cl”, RrY2and were synthesized and nitrated. The percentages of each of the isomeric nitration products were determined, as shown schematically in Fig. 1, by (1) adding known Radioactive halobenzene
--f
Sitration iriixture
v I)ilulioti with inactive n i t r o l ~ a l o b e ~ i ~ e ~ i ~ s
the activities of the diluted forms with those of the starting materials. To eliminate highly active impurities, each diluted isomer was scavenged repeatedly with inactive forms of all suspected impurities (other nitrohalobenzene isomers, halide ion, halogen and halobenzene). Purification was deemed complete when the activities of the recovered scavenging species were less than that of the scavenged material or until the activity of the scavenged material remained substantially constant on successive treatments. The scavenged diluted isomers were recrystallized, distilled or chromatographed until samples of high chemical purity were obtained. These either were counted as such or converted to silver halides for counting. I n the determination of the (small) extent of formation of the m-isomers by isotope dilution, especial care was necessary to ensure complete removal of the relatively highly radioactive oand 9-nitrohalobenzenes. The most satisfactory procedure for this purification took advantage of the very large reactivity of the o,p-nitrohalobenzenes toward piperidine as compared to the corresponding nz-isomers. Generally, three or four piperidine treatments, using added inactive o,pnitrohalobenzenes as scavengers, reduced contamination by the radioactive o,p-isomers to negligible proportions. The general scheme for determination of the relative reactivities of iodobenzene and benzene, based on the radiochemical determination of 9nitroiodobenzene in the products from a competitive nitration, is shown in Fig. 2. From the amount Rciizene and radioactive iodobenzene
I
r.1 ~
Piirificatioti
~. .._.1 ~
~
4
l’urificat ion
.
.
4,
~
J I’urification
A
I
4
Kitratioti mixture
,_
I . . . __. .-. .. .
+ Chemical 31i:tly& for
o-Nitrohalol~eriz~~le ~ ~ ~ - S i t r o h : ~ l o I ~ cp-Nitroll:ilohct~zel~e t~zrt~e aroinatic nitro coinpouiids I __- _1
\1
Radioactivity aiialysis Fig. l.-Isoiner distribution analysis.
quantities of the inactive forms of each isomer to fractional parts of the rciaction mixture, (2) separating and purifying the diluted isomers to radiochemical and chemical purity, and ( 3 ) coniparing (8) C. G . Swaiu and UT.P. 1-angsdorf, Jr., THISJ O U R N A L , 73, 2813 (1951).
(9) (a) G. W. Wheland, ibid , 64, 900 (1942); (b) J. D. Roberts and A. Streitwieser, Jr., ibid , 74, 4723 ( l Y c 5 2 ) . (10) T h e high accuracy reported by T.R i and €1. Eyring, J . Cheuz. P h y s . , 8 , 433 (1940), for calculations of isomer distril,utic,ns from d i i i d e moment d a t a is probahly illusory since i t depends on assumption of a n arbitrary and physically unrealistic electrostatic model for the nitration transition state a s well us “semi-empirical” adjustment of 0 - p ratios. Vor t h e halobenzenes. the Iii and Eyring procedure leads t o proportions of m-isomers which are shown later in this paper t o be substantially incorrect I n our opinion, t h e quantitative significance of t h e Ri and Eyring calculations has been overestimated by L. h*.Ferguson, Chem. Revs , SO, 47 (1952), and A E . Rernick, “Electronic Interpretations of Organic Chemistry,” John IViIey and Sons, Inc., S e w York, N. Y., 2nd Ed , 1‘349, pp. 218-220. (11) F . C. Henriques, Jr. and C. hlargnetti, I n d . E n g . Chem , A n o i . E L , 18, 476 (1948). (12) hZ. L. Bird and C. K . Ingold, J . Chem. SOL.,918 (1938).
Vol. TO
_----1 $
T)ilution wit11 illactive p - r i i troiodobenzene
Radioactivity c- P-Sitroiodo- -t- I-’urifiration :tnJysis benzene Fig. 2.-Relative reactivity determination.
of p-nitroiodobenzene formed and the isomer tlistribut.ion determined above, the amount of iodobenzene nitrated was calculated. Chemical analysis for the total amount of aromatic nitration gave, by difference, the amount of benzene nitrated. Preparation of Radioactive Halobenzenes Radioactive chloro- and iodobenxenes were prepared by decomposition of benzenediazoniunl compounds in the presence of the respective radioactive halide ions. The iodobenzene synthesis could be achieved readily by standard procedures,I3 but the published procedures for the synthesis of chlorobenzenes from the diazonium salts required large excesses of chloride ion (four14to sixtyi5moles) (13) L. G a t t t r m a n n and H. Wieland, “Laboratory Methods of Organic Chemistry,” translated from t h e 24th German Edition by W. IrlcCartney, T h e hlacmillan Co., S e w York, h-. Y., 1938, p. 81; H. J. Lucas and E . R . Kennedy, OYR.Syntheses, 19, 55 (1939). (14) A. Angeli, Gnzs. chim. ; i d , 21, 11, 258 (1891). (15) L. H. Welsh, THISJ o u x ~ a 63, ~ , 3276 (1911).
Sept. 20, 1954
ORIENTATION IN
AROMATIC NITRATION REACTIONS
4525:
352s
ISOMER DISTRIBUTION IN Sui>stance nitrated
Conditions, 25'
TABLE rI THE NITRATIOX OF HALOBENZENES
Method of analysis
Distribution, %
para
mcla
OVlilO
Investigators
CsH6C1 AcKO3 in CH,SO? Isotope diln. 29.6 0.9 69.5 This research ( I ) h N 0 3 in CH3SOZ CsHjCl Cryoscopic 30.2 0 .0 69.8 Ref. 12 CsH5CI AciYOa in C H z C S Cryoscopic 30.2 0.0 69.8 Ref. 12 CsHjC1 H N 0 3 (heterogeneous,'" Cryoscopic 33.1 0 .0 66.9 Ref. b CdljBr Isotope diln. 36.5 1.2 62.3 H N O I in CH3NO? This research ( I 1 and 111 1 Cryoscopic 42.0 0 ,0 58.0 CoHjBr A c S 0 3 in Ac2O Ref. 12 CsHiBr H N 0 3 (heterogeneous)" Cryoscopic m.4 I).0 59.6 Ref. c COHjI AcKOs in CH3SO2 Isotope diln. 38.3 1.8 .59,7 This research (IWL.111 i caI-151 HNOI in CHISO:! Isotope diln. 41.8 1.5 56.7 This research ( I X ) "0s (heterogeneous)" CcHjI Cryoscopic 43.2 0.0 56.8 Ref. d Extrapolated from experiments a t 0" and -30". A. F. Holleman and E . R . de Bruyn, Rec. t r m . chim., 19, 188 ( l W ( L , .I.F. Holleman and E. R . de Bruyn, ibid., 19, 364 (1900). A . F. Holleman, ibid.,32, 134 (1913). '$
and, hence, were not well suited to the preparation of radioactive chlorobenzene. After a number of trials, radioactive chlorobenzene was found to be prepared best by decomposition of anhydrous benzenediazonium chloride in the presence of radioactive hydrochloric acid in anhydrous dioxane with copper powder as a catalyst. The yield was T-4-170based on total chloride ion. The reaction is similar to those carried out by Hantzsch and Semple,I6 except for the use of the copper catalyst which permitted a more reasonable reaction rate. Radioactive iodobenzene also was prepared by oxidative, direct iodination. l7 Since, in this synthesis, nitric acid is used as oxidant, the purity of the iodobenzene was checked for small aniounts of radioactive nitroiodobenzenes which might have been formed during the reaction. Radioactive bromobenzene was prepared in Ci4-GTc,', yields (based on total bromine) by bromination of benzene with a mixture of potassium bromate, radioactive bromine and sulfuric acid a t ;30",I. Experimental Results 'The radioactive halobenzenes were nitrated by the method of Bird and Ingold.12 Data obtained in a single determination of the isomer distribution in the nitration of chlorobenzene, two determinations for bromobenzene and five determinations for iodobenzene are summarized in Table I which also includes the reaction times and conditions. In Table 11, the average results are compared with those from earlier investigations where the metaisomer content was assumed negligible. Aside from the meta-isomer contents, agreement is good. Iieactivity calculations were made as described 1 ) ~ . Ingold, Lapworth, Rothstein and b7ard1g except in those experiments where 0-p and m-p ratios were determined for separate fractions of the reaction mixture. In Table 111 are listed the isomer distributions aiid relative reactivity data for the halobenzenes and some other benzene derivativesz0 studied by (16) A. Hantzsch and W. Semple, B e y . , 33, 2533 (1900). (17) F.B. Dains and R . Q . Brewster, Ovg. Synlheses, 9 , 46 (182gI). (18) I?. Kraftt, Ber , 8 , I044 (1875) ( 1 9 ) C K Ingold, A . Lapvi.orth, E. Ruthstein and D. Ward, J . C'heiiz. Suc., lY.39 (1931). ( 2 0 ) T h e relative reactivity d a t a for p i w l i c d o r positions in 'lable I I I h e r e cr,mpiited froin the ox-er~allnitration reactivities in aretic ailh \ i l n d c and I m i n F r disti-ihutiiin~ nhtaincrl in a c r l i c anhyilriilc or nitro
other investigators along with the u-constants of the substituents. A semi-logarithmic plot of relative reactivities against u-constants is shown in Fig. 3. TABLE I11 REACTIVITIESI N
hTITRATION O F hfONOSUBSTITUTEU
ZESES IN
Isomer distribution, Cb
CHaCsJ
mela
-CO?22Hsei/
iizela
H
para ..
I"'
iiieta
pava
paFa Cih,'
mela
Reactivity.a log k / @
*
4.3 0.2 3 7 . 4 Yk 0 1 72 0
0.491 1.731 -2.098 -3.061 (0.000)
4.0
....,, . ......... 85.41
...... -0.073 -2.975 -0,832 -2.967 -0.950 -2.154 -0.33'3
*
0.9 0.0 para 69.5 k 1 . 7 mefa 1.2 0.1 para 62.3 i 2 . 9 ineta 1.8 i0.3 p o v a 59 7 2.0
+
+,i I'
BEN-
ACETICANHYDRIDE AT 25"
*
0
Y 1'
-0.060 - ,170 ,3340 .4028 000
0 0-15
.. . .
...
+ +
+ +
+ + + + +
,062 ,373 ,227 ,391 ,232 ,352
,276
080 ,014
. 0:18 0 13 065 053 ,054 .0a I ,043
. OZfi . 067
a Ratio of rate of nitration a t the single l i t - or />-position to rate a t a single benzene position. Median deviation of u , see ref. 6. Calculated from data at 0" and Ref. 19. 30'. e C. I(. Ingold and Il.S. Smith, J . Chenz. Soc., 905 (1938). Data for 18". J , D . Roberts and Ll'. T. Moreland, J r . , THISJOURNAL, 75, 2267 (195X). Ref. 12. Orientation in acetic anhydride or nitromethane ha5 not been determined; listed values for nitration in nitric acid taken from A. F. Holleman, Chew. Rws., 1 , 187 (1924). See ref. 12 for a comparison of orientation in nitric x i t i and acetic anhydride solution with chloro- and I,roniobeii zenes. 3 See Tables I and 11.
'
@
Discussion Although the Hammett relation has been sirigularly useful in correlating equilibrium and rate data for many reactions of p a - and p-substituted benzene derivatives where the seat of reaction is not in the benzene ring, i t appears from the scatter of many of the points in Fig. 3 that i t applies rather less accurately to aromatic nitration. From theoretical considerations, one would expect that the Hammett relation should hold for electrophilic substitution reactions in the benzene ring only in those cases where there are no i n portant differences in resonance interaction involving the substituent and aromatic ring between methane. All available evidence indicates t h a t , althou-h ( I \ er a l l nitration reactivities are difierrnt in nitrorncthane and u e t l c a n h y dride, t h e isomer distributions are insensitive t o this denree , I f -
