Acid-catalyzed hydrolysis of bridged bi- and tricyclic compounds. XXXVIII - Kinetics and mechanisms of 1- and 3-nortricyclanols

Article abstract of DOI:10.1002/poc.421

The disappearance of 1- and 3-nortricyclanols (1-OH and 2-OH) in aqueous perchloric acid was followed by capillary GC at different temperatures and acid concentrations. 1-OH is ca 1000 times more reactive than 2-OH. The activation parameters, solvent deut

Full text of DOI:10.1002/poc.421

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2001; 14: 854–858
DOI:10.1002/poc.421
Acid-catalyzed hydrolysis of bridged bi- and tricyclic
compounds. XXXVIIIÐKinetics and mechanisms of 1- and
3-nortricyclanols
Martti Lajunen* and Veli Lahti
Department of Chemistry, University of Turku, FIN-20014 Turku, Finland
Received 17 April 2001; accepted 6 June 2001
ABSTRACT: The disappearance of 1- and 3-nortricyclanols (1-OH and 2-OH) in aqueous perchloric acid was
epoc
followed by capillary GC at different temperatures and acid concentrations. 1-OH is ca 1000 times more reactive than
2-OH. The activation parameters, solvent deuterium isotope effects and parameters of excess acidity equations were
measured and the products were studied. Both isomeric nortricyclanols react according to the AdE2 mechanism, i.e.
the cyclopropane ring is protonated at the rate-determining stage of the reaction. The protonation causes, in the case of
1-OH, an isomerization called homoketonization with 2-norbornanone as the only product and, in the case of 2-OH,
hydration, i.e. the formation of hydroxyl-substituted norbornyl cations, the fast attack of which by water produces
several norbornanediols. Copyright  2001 John Wiley & Sons, Ltd.
Additional material for this paper is available from the epoc website at http://www.wiley.com/epoc
KEYWORDS: nortricyclanes; kinetics; acid catalysis; excess acidity; homoketonization; hydration; reaction
mechanisms
INTRODUCTION
This similarity is surprising, because 1-OAc is a
tertiary and 2-OAc a secondary acetate. Evidently, the
location of the acetoxyl group at a cyclopropane carbon
in the tertiary isomer makes the normal AAL1 ester
hydrolysis in this case less probable.2 1-Nortricyclanol
(1-OH) was considered to be a possible unstable
intermediate product of the hydrolysis for 1-OAc.1,4
Therefore, it was chosen as the subject of additional
studies on the comparison of 1- and 3-substituted nor-
tricyclanes, when accordingly 3-nortricyclanol (2-OH) is
the other isomer. Their disappearance rates were
measured in aqueous perchloric acid at different
temperatures and acid concentrations and in deuterioper-
chloric acid, and the products were analyzed in order to
elucidate the reaction mechanisms.
Recently, a mutual comparison of the kinetics and
mechanisms of acid-catalyzed hydrolysis (hydration)
of 1- and 3-X-substituted nortricyclanes or tri-
cyclo[2.2,1.02,6]heptanes (1-X and 2-X) was started at
our laboratory.1 In the first case, when X = OAc, the
disappearance rates and reaction mechanisms of the
isomers were very similar, although the reaction was
changed from AAC2 ester hydrolysis2 to the AdE2 or
A-SE2 hydration of the cyclopropane ring3 with in-
creasing acid concentration. Even the concentration
[c(HClO4) ꢀ 6 mol dmÀ3] where the portions of the two
reactions were equal was observed to be similar for both
isomers.1
EXPERIMENTAL
Materials. 1-Nortricyclanol (1-OH) was prepared by
LiAlH4 reduction of the corresponding acetate (1-OAc)
in dry diethyl ether.4 The substrate was unstable at room
temperature, so it was stored in a freezer as a concen-
trated ether solution of the crude product. 3-Nortricycla-
nol (2-OH) was obtained by alkaline hydrolysis of
2-OAc.5 The purities by GC were 85% for 1-OH
[contains 7% of 2-norbornanone (3) and in total 8% of
*Correspondence to: M. Lajunen, Department of Chemistry, Uni-
versity of Turku, FIN-20014 Turku, Finland.
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 854–858

HYDROLYSIS OF BRIDGED BI- AND TRICYCLIC COMPOUNDS
855
Table 1. Rate constants for disappearance of 1- and
3-nortricyclanol '1-OH and 2-OH) in aqueous perchloric acid
at different temperatures and acid concentrations and in
deuterioperchloric acid
three impurities] and 99.5% for 2-OH. The substrates
were identified from their 13C NMR spectra [1-OH
(CDCl3, TMS) tentative assignments: C-1, 66.3; C-2 and
C-6, 18.0; C-3 and C-5, 34.2; C-4, 32.6; and C-7, 38.9;
2-OH;6 and 37] and from GC–FTIR and/or GC–mass
spectra.
c(HClO4)a
b
Substrate
T (K) (mol dmÀ3
)
X0
k (10À4 sÀ1 c
)
1-OH
288.2 0.249
298.2 0.248
298.2 0.250
298.2 0.249
298.2 1.014
308.2 0.247
318.2 0.246
298.2 1.013
318.2 1.005
328.2 1.000
338.2 0.994
348.2 0.988
288.2 5.459
298.2 5.427
298.2 5.427
308.2 5.388
318.2 1.510
318.2 2.002
318.2 2.361
318.2 2.953
318.2 3.488
318.2 3.939
318.2 4.383
318.2 4.381
318.2 4.904
318.2 5.358
1.130 Æ 0.005
3.69 Æ 0.05
Kinetic measurements. The disappearance of the sub-
strates in aqueous perchloric acid was followed by GC
(with a FFAP capillary column) using cyclohexanone
(for 1-OH) or 2-norbornanone (3, for 2-OH) as the
internal standard and dichloromethane as the extraction
solvent.1 The pseudo-first-order rate constants were
obtained from the slopes of the strictly linear
(r = 0.9990–0.99997) correlation of ln St vs t(S? = 0),
where St is the ratio of the GC integrals of the substrate
and the internal standard at time t. Each rate constant was
measured twice, the values being equal to within at least
4% (average 1.4%) for both 1-OH and 2-OH. The
impurities (8%, see above) did not affect the kinetic runs
on 1-OH by GC.
3.70 Æ 0.07d
1.360 Æ 0.015e
22.5 Æ 0.3
11.17 Æ 0.07
30.1 Æ 0.6
2-OH
0.0227 Æ 0.0003d
0.336 Æ 0.004
1.148 Æ 0.004
3.546 Æ 0.019
10.59 Æ 0.11
2.252 Æ 0.011
8.39 Æ 0.12
0.257
8.1 Æ 0.4d
26.9 Æ 0.4
0.362
0.467
0.549
0.699
0.856
1.007
0.723 Æ 0.004
1.384 Æ 0.004
2.184 Æ 0.008
4.267 Æ 0.025
7.975 Æ 0.026
13.64 Æ 0.08
14.15 Æ 0.07e
22.97 Æ 0.11
43.8 Æ 0.4
Product analyses. The only product of hydrolysis (and
isomerization) of 1-OH was identified by comparing its
retention time with that of authentic 2-norbornanone (3).
The formation of 3-d1 from 1-OH was also observed in
0.25 M DClO4 (D2O) (1.3 h at ca 321 K; ca 10t1/2) by
GC–MS. A 2H NMR spectrum of the isolated (by CH2Cl2
extraction) 3-d1 was recorded in CHCl3–CDCl3 (9:1)
with TMS-d1 (not enriched according to D) as the internal
reference and the spectrum was compared with those
obtained by Nickon et al.8
About 0.2 g of 2-OH was stirred magnetically with
50 cm3 of 3 M HClO4(aq.) for 0.5 h (ca 1t1/2) and 5 h
(ꢁ10t1/2) at ca 318 K in a tightly stoppered Erlenmeyer
flask. The solution was extracted several times with
CH2Cl2 and the combined organic solution was neutra-
lized and dried by allowing it to flow through anhydrous
K2CO3. The solvent was evaporated almost totally and
the residue was analyzed by GC, GC–FTIR and GC–MS,
the components being mainly identified by comparing
their spectra with those stored in the memories of the
spectrometers.
1.173
1.392
1.602
74.3 Æ 0.6
a
Temperature corrected.
b
c
d
e
Excess acidity, temperature corrected.10,11
Error limits are standard deviations.
Calculated from the activation parameters (Table 2).
Measured in DClO4 (D2O).
than 1-acetoxynortricyclane (1-OAc; 19 times).1 Hence it
is natural that 1-OH could not be detected by GC in the
hydrolysis of 1-OAc. The rate constants measured for
2-OH are in fair agreement with those measured earlier
under similar conditions.5 The activation parameters for
1-OH were measured in 0.25 M HClO4(aq.) and for
2-OH in 1.0 and 5.4 M HClO4(aq.). Their values are
shown in Table 2 together with the solvent deuterium
isotope effects. All values were calculated from the
second-order rate constants [ka = k /c(HClO4)].
RESULTS AND DISCUSSION
The slightly or moderately negative entropies of
activation, À5 to À27 J KÀ1 molÀ1 at 298 K, and the
isotope effects, kH/kD = 1.62–2.72, greater than unity are
typical of the rate-determining protonation of the
cyclopropane ring (AdE2 mechanism; Schemes 1 and
2).6 In the case of 2-OH, the decrease in DS≠ from À5 to
À27 J KÀ1 molÀ1 with increasing acid concentration
(1.0–5.4 M) seems normal for the AdE2 mechanism.9
Thus, DS≠, À27 J KÀ1 molÀ1, for 1-OH in 0.25 M acid
is unexpectedly negative. The reason for this is not
known.
Rate constants, activation parameters and iso-
tope effects
The rate constants of disappearance for 1- and 3-
nortricyclanol (1-OH and 2-OH) were measured in
aqueous perchloric acid at different temperatures and
acid concentrations and in deuterioperchloric acid. The
pseudo-first-order rate constants (k ) are shown in Table
1. 1-OH is much more reactive than 2-OH (990 times in
1.0 M HClO4 at 298.2 K) and also clearly more reactive
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 854–858

856
M. LAJUNEN AND V. LAHTI
Table 2. Parameters of activation and solvent deuterium isotope effects at 298 K 'unless noted
otherwise) and parameters of the excess acidity equation [Eqn. '3)] at 318 K for the hydrolysis of
1- and 3-nortricyclanol '1-OH and 2-OH) in HClO4'aq.), with standard deviations
Substrate
c(LClO4)a (mol dmÀ3
)
DH≠ (kJ molÀ1
)
DS≠ (J molÀ1 KÀ1
)
kH/kD
1-OH
2-OH
0.25
1.01
5.43
4.38
81.2 Æ 0.7
103.8 Æ 0.4
86.9 Æ 1.8b
À27 Æ 3
À5.0 Æ 1.2
À27 Æ 6b
2.72 Æ 0.07
1.62 Æ 0.02c
‡
d
À1
m'
pKS'H
À2.04 Æ 0.26
m≠ m*
aA
log(k0/M sÀ1
)
2-OH
0.97 Æ 0.07
1.52 Æ 0.08
0.84 Æ 0.05
À4.85 Æ 0.03
a
L = H or D.
b
The Arrhenius plot is slightly curved; however, r = 0.9996 for the linear regression.
c
318.2 K.
aA = m≠ m*/1.80.10,13
d
The excess acidity method
if the hydroxyl substituent of the substrate S is also
protonated in a fast equilibrium producing S'H (the
‘impure’ AdE2 mechanism; Scheme 2).
‡
The excess acidity theory10,11 was applied to the rate
constants measured for the disappearance of 3-nortri-
cyclanol (2-OH; 1-OH being too reactive for the
corresponding measurements) at different acid concen-
trations at one temperature, 318.2 K (Table 1). The rate
constant should follow the equation10,11
In the above equations, k is the pseudo-first-order rate
constant measured in aqueous acid with concentration
cH
‡
[= c(HClO4)] and excess acidity X0 (cH and X0 are
‡
corrected according to temperature);10,11 m≠ is the slope
parameter indicative of the character of the transition
state and m* is the slope parameter dependent on the site
of proton attack (the cyclopropane ring in this case, when
m* = 1.80 Æ 0.10);11,13 k0 is the medium-independent
rate constant of the rate-determining step (r.d.s. in
log k À log c ˆ m mÃX0 ‡ log k0
ꢂ1†
ˆ
‡
H
if no protonation of the substrate other than the rate-
determining step on the cyclopropane ring occurs (the
‘pure’ AdE2 mechanism), and the equation1,12
Scheme 2); and cS and cS'H are the concentrations of
‡
the unprotonated substrate and of the substrate protonated
on the substituent, respectively.
0
‡
‡
In the case of substrate 2-OH, the hydroxyl group is
log k À log‰cS=ꢂcS ‡ c
†Š À log cH
S H
ˆ
ˆ m mÃX0 ‡ log k0
ꢂ2†
Scheme 1
Copyright  2001 John Wiley & Sons, Ltd.
Scheme 2
J. Phys. Org. Chem. 2001; 14: 854–858

HYDROLYSIS OF BRIDGED BI- AND TRICYCLIC COMPOUNDS
857
‡
partly protonated in the moderately concentrated acids
[c(HClO4) = 1.0–5.4 M] used in the present work. Thus,
of proton transfer from a hydroxonium ion (H3O ) to the
substrate at the transition state (0 ꢃ m≠ ꢃ1; m≠ = aA, i.e.
the excess acidity Bro¨nsted a). Thus aA = m≠ m*/1.80 =
0.84 Æ 0.05, which is a typical value for the protonation
of nortricyclanes in HClO4(aq.) and in excellent agree-
ment with the isotopic Bro¨nsted a (a1 = 0.85 Æ 0.02)
measured for 2-OH in 0.1 M LClO4 (L2O) (L = H or D) at
348.2 K.18
the plot log ka vs X0 (ka = k /cH ) obeying Eqn. (1) is not
‡
linear (Fig. 1). In order to estimate the protonation
correction term, Àlog[cS/(cS ‡ cS'H)], of Eqn. (2), the
equation is changed to a non-linear form:1,12,14
ˆ
log k À log c ˆ m mÃX0 À log‰1
‡
H
m0X0
Now the correction term, Àlog[cS /(cS ‡ cS'H
Eqn. (2) can be calculated by Eqn. (4) from the values of
m' and pKS'H in Table 2. If the left-hand side of Eqn. (2)
vs X0 is drawn, a strictly linear (r = 0.9992) plot is
obtained (Fig. 1). The difference between the two plots in
Fig. 1 [according to Eqns (1) and (2)] shows that the
effect of protonation of the hydroxyl oxygen is marked on
the hydration rate of the cyclopropane ring of 2-OH.
‡)], of
0
‡
‡
‡ ꢂcH =KS H †10 Š ‡ log k0 ꢂ3†
‡
with the excess acidity equation:10,11
logꢂcS H =cS† À log cH ˆ m0X0 ‡ pKS H
ꢂ4†
0
0
‡
‡
‡
derived for the protonation equilibrium (Scheme 2). In
these equations, m' is the slope parameter indicative of
the protonation of the hydroxyl group and KS'H the
‡
thermodynamic dissociation constant of the substrate
protonated on the substituent. Non-linear least-squares
minimization can be used for the evaluation of the four
Hydrolysis products
According to the product analyses (see Experimental),
the only product of hydrolysis for 1-nortricyclanol (1-
OH) is 2-norbornanone (3), which was also slowly
formed from 1-OH during storage. The process (Scheme
1), called homoketonization,4,8,19 is comparable with the
keto–enol tautomerism,20 both isomerizations being
catalyzed by acids and bases. The mass spectrum of 3
formed in 0.25 M DClO4 (D2O) shows that it contains one
deuterium atom per molecule and the 2H NMR spectrum
indicates that ca. 90% of deuterium is situated at the
endo-6 position (ꢀD = 1.31) and ca. 8% at the exo-6
position (ꢀD = 1.60; besides 2% found at 1.84). The
exo/endo ratio of 8:92 is in excellent agreement with that
observed by Nickon et al.8 for the hydrolysis of
1-acetoxynortricyclane (1-OAc) in D2SO4–DOAc–D2O
([DOAc]:[D2O] = 1:1 or 2:1) and shows that the rate-
determining deuteration (protonation) occurs dominantly
at C-2 or C-6 by retention in aqueous acids (Scheme 1).
The formation of a thermodynamically stable oxo-
parameters of Eqn. (3), i.e. m≠ m*, KS'H
‡
(or pKS'H
and k0 (or log k0), by iteration from the experimental
values of k , cH and X0 (Table 1) and from the estimated
‡
), m'
‡
approximate values of the parameters. The iterated best
values of the parameters are given in Table 2.
The excess acidity parameters for the AdE2 hydration
of 3-nortricyclanol (2-OH) seem reasonable. The par-
ameters dependent on the protonation of the hydroxyl
group (m' = 0.97 and pKS'H
‡
= À2.0 at 318 K) are very
close to those measured earlier for other alcohols.9,12,15
The combined parameter m≠ m* consists of two terms,
the latter being dependent on the protonation site (in this
case a cyclopropane ring) and the former on the character
of the transition state of the AdE2 mechanism, i.e.
according to Kresge and co-workers,16,17 on the progress
‡
carbenium ion (SH ) is the probable reason for the
relatively high protonation rate of 1-OH [k (1ÀOH)/
k (2ÀOH) ꢀ 1000; see Table 1).
3-Nortricyclanol (2-OH) in HClO4(aq.) probably
produces five or six isomeric norbornanediols (4–9;
Scheme 2), the proportions of which do not vary
markedly within 1–10 half-lives of the substrate. Attemps
were made to identify them by GC (retention times), by
GC–FTIR (the isomers could not be separated, but two
ꢁOH absorptions were detected: 3653 cmÀ1 being typical
of the free and 3601 cmÀ1 of the hydrogen-bonded
hydroxyl group; cf. ꢁOH = 3657 cmÀ1 for 2-OH) and by
‡
Á
GC–MS [the molecular ion (M = 128) was sometimes
hard to detect owing to the easy elimination of one or two
water molecules]. The retention times and spectra were
also compared with those recorded for six methoxynor-
borneols formed in the hydration of 3-methoxynortri-
cyclane (2-OMe).21 The approximate amounts of the
Figure 1. Excess acidity plots [Eqns '1) and '2)] for the AdE2
hydration of 3-nortricyclanol '2-OH) in HClO4'aq.) at 318.2
K: '&) Eqn. '1); '*) Eqn. '2). The correction term,
Àlog[cS/'cS ‡ cS'H‡)], was calculated with Eqn. '4) from the
m' and pKS'H values in Table 2
‡
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 854–858

858
M. LAJUNEN AND V. LAHTI
5. Lajunen M, Hirvonen P. Finn. Chem. Lett. 1974; 245–249.
6. Lajunen M. J. Chem. Soc., Perkin Trans. 2 1986; 1551–1555.
7. Lippmaa E, Pehk T, Paasivirta J, Belikova N, Plate A. Org. Magn.
Reson. 1970; 2: 581–604.
8. Nickon A, Frank JJ, Covey DF, Lin Y-I. J. Am. Chem. Soc. 1974;
96: 7574–7576.
considered isomeric norbornanediols are shown in
Scheme 2. Accordingly, the C-1—C-6 edge of the
cyclopropane ring opposite to the 3-substituent is cleaved
most easily, producing mainly 2,7-norbornanediols (4
and 5), in agreement with earlier observations.6,21–23
9. Lajunen M, Latva-Nirva E. J. Chem. Soc., Perkin Trans. 2 2000;
719–723.
10. Cox RA. Acc. Chem. Res. 1987; 20: 27–31.
11. Cox RA. Adv. Phys. Org. Chem. 2000; 35: 1–66.
12. Lajunen M, Virta M, Kyllia¨inen O. Acta Chem. Scand. 1994; 48:
122–126.
13. Cox RA, Yates K. Can. J. Chem. 1981; 59: 2116–2124.
14. Kresge AJ, Yin Y. J. Phys. Org. Chem. 1989; 2: 43–50.
15. Perdoncin G, Scorrano G. J. Am. Chem. Soc. 1977; 99: 6983–6986.
16. Kresge AJ, More O’Ferrall RA, Hakka LE, Vitullo VP. J. Chem.
Soc., Chem. Commun. 1965: 46–48.
Acknowledgements
We are grateful to Dr Robin A. Cox for a copy of Ref. 11
and to Miss Kirsti Wiinama¨ki, Dr Martti Dahlqvist, Mr
Petri Ta¨htinen and Mr Jaakko Hellman for recording the
2
GC–mass, GC–FTIR, H NMR and 13C NMR spectra,
respectively.
17. Kresge AJ, Mylonakis SG, Sato Y, Vitullo VP. J. Am. Chem. Soc.
1971; 93: 6181–6188.
18. Lajunen M. Acta Chem. Scand. 1991; 45: 377–382.
19. Nickon A, Frank JJ. Tetrahedron Lett. 1975; 16: 4335–4338.
20. Lamaty G. In Isotopes in Organic Chemistry, Vol. 2, Isotopes in
Hydrogen Transfer Processes, Buncel E, Lee CC (eds). Elsevier:
Amsterdam, 1976; chapt. 2.
21. Lajunen M, Jantunen J, Koiranen P. Acta Chem. Scand. 1998; 52:
728–732.
22. Kantolahti E. PhD Thesis, University of Jyva¨skyla¨, 1975.
23. Werstiuk NH, Timmins G, Cappelli FP. Can. J. Chem. 1980; 58:
1709–1724.
REFERENCES
1. Lajunen M, Yli-Mannila R, Salminen O. J. Phys. Org. Chem.
2000; 13: 133–138.
2. Euranto EK. In The Chemistry of Carboxylic Acids and Esters,
Patai S (ed). Wiley: Chichester, 1969; chapt. 11.
3. DePuy CH. Top. Curr. Chem. 1973; 40: 73–101.
4. Nickon A, Lambert JL, Williams RO, Werstiuk NH. J. Am. Chem.
Soc. 1966; 88: 3354–3358.
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 854–858