633
Crampton and Robotham
affected by steric factors rather than the relative basicities of
the amines. It is likely that the proton to be transferred from 4
will be hydrogen bonded to a DMSO molecule in the solvent
(13, 27, 28) and this, together with steric hindrance to the
approach of the reagents, will reduce values of rate constants
below the diffusion limit.
2,4-dinitronaphthyl ether, 11, with that for phenyl
2,4,6-trinitrophenyl ether, 8b, shows decreases in K1k2, K1kAn,
and K1kDabco by factors of 250, 24, and 7, respectively. These
changes may be attributed (2, 3) to two main factors: the de-
crease in ring activation in 11 compared to 8b, and the reduc-
tion in steric crowding at the reaction centre in the zwitterionic
It is interesting to speculate on values of rate constants for
the proton-transfer step. It has previously been estimated (15,
16) that the presence of the trinitrocyclohexadienate ring in
zwitterions such as 4 will acidify the adjacent ammonium pro-
tons by a factor of ca. 500, as indicated in eq. [11]:
[11] Ka(4)/Ka(AnH+) = 500
Use of eq. [1] allows the comparison of the relative acidities
of 4 and protonated Dabco as shown in eq. [12]:
500
5.8 × 10−6
[12] Ka (4) ⁄ Ka (DabcoH+) =
= 8.6 × 107
intermediate 14 compared to that in the corresponding inter-
mediate, 9b. The first factor will result in a decrease in the
value of K1 and hence in the value of K1k2. That the decreases
observed for K1kAn and K1kDabco are smaller than that observed
for K1k2 may be attributed to the second factor. A reduction in
steric crowding in 14 compared to 9b may allow the easier
approach of a base molecule to accept a proton from the zwit-
terionic intermediate. Hence values of kAn and kDabco may be
larger in the 2,4-dinitronaphthyl system. The kAn/kDabco ratio is
somewhat decreased in the dinitronaphthyl system and this
may indicate that the situation is being approached in which
the reduction in the electron-withdrawing ability of the ring
system renders the proton transfer from 14 to aniline thermo-
dynamically unfavourable.
+
The ratio obtained corresponds to kDabco / kDabcoH and since
kDabcoH has been found experimentally to be 0.2 dm mol–1 s–1
3
+
we obtain a value for kDabco of 1.7 × 107 dm3 mol–1 s–1. This
allows the calculation of values for kAn of 8 × 106 dm3 mol–1
s–1 and for K1 of 7 × 10–8 dm3 mol-1. Since the ratio given in
eq. [11] is not known precisely, the values calculated using it
should be regarded only as estimates.
The data in Table 3 indicate that values of rate constants for
proton transfer are larger in the formation of 1 from TNB, than
in the formation of 5. Here reaction occurs at an unsubstituted
ring position so that there is less steric congestion at the reac-
tion centre. A calculation analogous to that given above leads
here to a value for kDabco in excess of 109 dm3 mol–1 s–1.
The formation of 6 from 5 involves general acid-catalysed
expulsion of the ethoxy group. Values of the rate constants for
reaction with the anilinium ion, 860 dm3 mol–1 s–1, and Dab-
coH+, 0.06 dm3 mol–1 s–1, may be used in conjunction with
eq. [1] to calculate a Brønsted α value of 0.80.
Phenoxide is known to be a considerably better leaving
group than ethoxide (17, 23), and we did not observe interme-
diates in the substitution reactions of the phenyl ethers 8a–c.
We interpret the increases in the rate constant shown in Figs. 2
and 3 as base catalysis. Previous work (19, 21), at lower base
concentrations, did not find evidence for base catalysis. The
effects we have observed are not particularly large. However,
values calculated for K1kAn and K1kDabco for reactions of 8a–c
are similar to those found for reaction of 3, where there is
genuine base catalysis. These similarities are to be expected if
proton transfer from zwitterion, 9 or 4, to base is rate limiting.
The presence of a phenoxy group rather than an ethoxy group
at the 1-position is not expected to drastically affect the value
of K1 for formation of the zwitterion (29). Nor will this change
be expected to greatly alter the values of kAn or kDabco for the
proton-transfer step. A major difference between the phenyl
ethers, 8, and the ethyl ether, 4, is that in the former case much
of the reaction flux involves the direct uncatalysed decompo-
sition of the zwitterions, 9, by the k2 step. This step is likely to
involve intramolecular proton transfer from nitrogen to oxy-
gen coupled with carbon–oxygen bond cleavage. Leaving-
group expulsion is part of the rate-limiting step here, so that
reaction of the 4-nitrophenoxy derivative, 8c, is six times faster
than that of the phenoxy derivative, 8b (Table 3).
Experimental
Ethyl 2,4,6-trinitrophenyl ether (16), 3, phenyl 2,4,6-trini-
trophenyl ether (17), 8b, and phenyl 2,4-dinitronaphthyl ether
(17), 11, were available from previous work. The 4-substituted
phenyl 2,4,6-trinitrophenyl ethers 8a and 8c were prepared by
reaction at 45°C for 3 h of picryl chloride (1 equiv.) with
sodium hydroxide (1 equiv.) in an excess of the appropriate
4-substituted phenol containing a little water. On completion,
water was added to remove any picric acid produced and re-
crystallization from ethanol yielded 8a, mp 100°C (lit. (30) mp
103°C) and 8c, mp 163°C (lit. (30) mp 157°C). 2,4,6-Trini-
trodiphenylamine, 6, and N-phenyl-2,4-dinitronaphthylamine
were available from previous work (22). Solvent, amines, and
amine salts were prepared and (or) purified as described pre-
viously (22).
1H NMR spectra were recorded using Varian-200 XL or
VXR-400 spectrometers. UV–visible spectra and kinetic
measurements were made with Beckman Lambda 2 or Applied
Photophysics SX.17 MV stopped-flow spectrophotometers at
25°C. Reported rate constants are the means of several deter-
minations and are precise to ±5%.
Acknowledgements
We thank the Engineering and Physical Sciences Research
Council (EPSRC) for a studentship and for a research grant to
purchase the stopped-flow spectrophotometers. Mrs. J. Say is
thanked for assistance with the NMR measurements.
Comparison of data for reaction of phenyl
© 1998 NRC Canada