Selectivity of stabilized benzhydrylium ions

Article abstract of DOI:10.1002/ejoc.200901327

Methyl carbonates (1-4-OCO₂Me) and phenyl carbonates (2-4-OCO₂Ph) were subjected to solvolysis in the series of aqueous ethanol, the product ratios were determined and the selectivities (k _E/k_w) of the correspon

Full text of DOI:10.1002/ejoc.200901327

SHORT COMMUNICATION
DOI: 10.1002/ejoc.200901327
Selectivity of Stabilized Benzhydrylium Ions
Bernard Denegri,[a] Mirela Matic´,[a] and Olga Kronja*[a]
Keywords: Kinetics / Solvolysis / Microsolvation / Stabilized benzhydrylium ions
Methyl carbonates (1–4–OCO2Me) and phenyl carbonates
(2–4–OCO2Ph) were subjected to solvolysis in the series of
aqueous ethanol, the product ratios were determined and the
selectivities (kE/kW) of the corresponding stabilized benz-
hydrylium ions 1–4 were calculated. The individual rate con-
stants for ethanolysis (kE) and hydrolysis (kW) have been cal-
culated. Ethanolysis proceeds via earlier more carbocation-
like transition state than hydrolysis. Due to more important
solvation, kE decreases slower than kW if the fraction of the
water in the binary solvent increases, which net effect is pro-
nounced dependence of the selectivity on the solvent po-
larity. The selectivities of less stable carbocations 5–8 are re-
duced and influenced by the leaving group because of pair-
ing and bending of the log k vs. E plot.
of water increases, while that of substrates that yield the
products from the solvent-separated ion pairs is S Ͻ 1.[3,4]
If more species exist as intermediates, the observed selectiv-
ity was interpreted as weighted average of the individual
selectivity values. The observations that the selectivity in-
creases with the ionizing power of the solvent and that bet-
ter leaving groups produce higher selectivities have been ra-
tionalized as a change in the relative population of the free
ions and the solvent-separated ions.[5–7]
Introduction
Carbocation intermediate formed by heterolysis reaction
in a binary solvent reacts with different rates with each
component of the mixture. The pseudo first-order rate con-
stant of such reaction represents the sum of the individual
pseudo first-order rate constants. Thus, if the carbocation
is generated in aqueous ethanol, the corresponding first-
order rate constant is presented with Equation (1).
Our investigations presented here rely on Mayr’s results
obtained with substituted benzhydrylium ions. First, it has
been shown that the stabilized benzhydrylium ions in aque-
ous ethanol can be treated as thermally equilibrated free
ions, since the reaction with the solvent is slow enough to
allow the counterion to diffuse away.[8] Therefore, the selec-
tivity values reflect the ratio of two rate constants only
(kE/kW). Secondly, Mayr et al. determined the nucleophilici-
ties of different solvents and solvent mixtures by measuring
the pseudo first-order rate constants for combination of the
substituted benzhydrylium ions with the series of binary
solvents, providing numerous experimental rate constants
(kЈ).[9] It has been shown that, similarly as was previously
demonstrated for the reaction of carbocations with variety
of nucleophiles,[10] the first-order rate constants for combi-
nation of the carbocation with the solvent follow the special
LFER relationship, see Equation (3).
kЈ = kЈE + kЈW = kE [EtOH] + kW [H2O]
(1)
in which kЈ is the pseudo first-order rate constants for reac-
tion with aqueous ethanol, kЈE and kЈW are the pseudo first-
order individual rate constants with ethanol and water,
resp., while kE and kW are the second-order individual rate
constants.
A selectivity is defined as a ratio of the individual sec-
ond-order rate constants (kE/kW) for the reaction of a given
carbocation with individual components of the binary sol-
vent.[1,2] Since the molar ratio of the alcohol and ethyl ether
([ROH]/[ROEt]) formed in reaction of carbocation R+ with
aqueous ethanol indicates the ratio of the individual first-
order rate constants (kЈE/kЈW), the selectivity can easily be
derived from Equation (2).
(2)
logk20 °C = s(E + N)
(3)
Generally, it has been accepted that in aqueous solvents
the selectivity of the substrates that solvolyze via formation
of the free carbocation is S Ͼ 1 and increases as the fraction
in which E represents the empirical electrophilicity param-
eter of the cation, N is an empirical nucleophilicity param-
eter of the solvent, and s is an nucleophile-specific param-
eter.
The above results prompted us to study the selectivities
of substrates that produce stable benzhydrylium ions, ex-
tract the individual rate constants for ethanolysis (kE) and
hydrolysis (kW) in a given aqueous ethanol, and interpret
[a] Faculty of Pharmacy and Biochemistry, University of Zagreb,
Ante Kovacˇic´a 1, 10000 Zagreb, Croatia
Fax: +385-1-48-56-201
E-mail: kronja@pharma.hr
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.200901327.
1440
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 1440–1444

Selectivity of Stabilized Benzhydrylium Ions
Results and Discussion
In order to measure the ether/alcohol product ratio
([X,Y-BhOEt]/[X,Y-BhOH]) and derive the corresponding
selectivity values, the series of methyl carbonates (1–
OCO2Me, 2–OCO2Me, 3–OCO2Me, and 4–OCO2Me) and
the series of phenyl carbonates (2–OCO2Ph, 3–OCO2Ph,
and 4–OCO2Ph) were solvolyzed for approximately 10 reac-
tion half-lives in aqueous ethanol mixtures (90–60% v/v)
at 20 °C with and without base added. The products were
analyzed by means of NMR spectroscopy. Product ratios
and the selectivity values are presented in Table 1.
As expected, the selectivities obtained for solvolysis of
both carbonates are kE/kWϾ1. The following trends can be
seen from Table 1. First, the S values increase considerably
without exception as the polarity of the solvent mixture in-
creases. Secondly, the S values obtained with the phenyl
carbonates are virtually equal to those obtained with the
methyl carbonates, supporting the formation of the free
ions. Furthermore, the selectivity values slightly but consist-
ently decrease as the stability of the carbocations decrease.
From Equations (1) and (2), taking the pseudo first-or-
der rate constants for reaction of the X,Y-substituted benz-
hydrylium ions with the series of aqueous ethanol[9] and the
selectivity values shown in Table 1, the second-order indi-
vidual rate constants for ethanolysis and hydrolysis can be
determined straightforwardly, see Equation (4).
selectivity values obtained on quantitative basis (Table 1).
The main criterion was to choose substrates that produce
carbocations that react with the series of aqueous ethanol
with the rates that are way bellow the diffusion controlled
processes. X,Y-Substituted benzhydrylium (Bh) ions 1–4
turned out to be suitable substrates for determination of the
selectivities. In order to avoid mixing problems, which may
cause inconsistent product ratios, it was necessary to use
substrates whose rates of the carbocation formation were
not too fast. This was achieved by adjusting the nucleofu-
gality of the leaving group. We found that methyl and
phenyl carbonates (MeOCO2 and PhOCO2) were suitable
leaving groups for these investigations.[11] The overall pro-
cess investigated here is presented in Scheme 1.
(4)
The calculated second-order individual rate constants kE
and kW along with the corresponding pseudo first-order
rate constants are presented in Table 2.
Scheme 1.
Table 1. Product ratios and selectivities (kE/kW) for solvolysis of benzhydryl carbonates in ethanol/water mixtures at 20 °C.
Cation (X,Y)
Solvent[a]
Methyl carbonate
[X, Y-BhOEt][b]
Phenyl carbonate
[X, Y-BhOEt][b]
S[c]
S[c]
[X, Y-BhOH]
[X, Y-BhOH]
1 (MeO,MeO)
90E
80E
70E
60E
90E
80E
70E
60E
90E
80E
70E
60E
90E
80E
70E
60E
12.2Ϯ0.2 (4)
7.2Ϯ0.4 (6)
5.2Ϯ0.2 (7)
4.1Ϯ0.2 (6)
12.1Ϯ0.2 (3)
6.8Ϯ0.1 (3)
4.9Ϯ0.1 (3)
3.9Ϯ0.1 (3)
10.8Ϯ0.2 (4)
6.3Ϯ0.2 (4)
4.6Ϯ0.1 (4)
3.7Ϯ0.1 (5)
11.5Ϯ0.2 (5)
6.3Ϯ0.1 (3)
4.5Ϯ0.1 (3)
3.1Ϯ0.1 (4)
4.4Ϯ0.1
5.8Ϯ0.3
7.2Ϯ0.2
8.8Ϯ0.2
4.4Ϯ0.1
5.5Ϯ0.1
6.8Ϯ0.1
8.5Ϯ0.2
3.9Ϯ0.1
5.1Ϯ0.1
6.5Ϯ0.2
8.0Ϯ0.2
4.1Ϯ0.1
5.1Ϯ0.1
6.2Ϯ0.1
6.8Ϯ0.1
2 (MeO,PhO)
3 (MeO,Me)
4 (MeO,H)
12.1Ϯ0.1 (3)
6.7Ϯ0.2 (4)
4.9Ϯ0. 1 (3)
4.2Ϯ0.1 (3)
12.3Ϯ0.2 (5)
6.8Ϯ0. 1 (4)
4.8Ϯ0.1 (4)
3.7Ϯ0. 1 (4)
11.5Ϯ0.2 (4)
6.2Ϯ0.4 (5)
4.5Ϯ0.1 (4)
3.4Ϯ0.1 (4)
4.4Ϯ0.1
5.4Ϯ0.2
6.8Ϯ0.1
9.1Ϯ0.2
4.4Ϯ0.1
5.5Ϯ0.1
6.7Ϯ0.1
8.0Ϯ0.1
4.1Ϯ0.1
5.0Ϯ0.1
6.2Ϯ0.1
7.4Ϯ0.2
[a] Aqueous ethanol mixtures are v/v at 25 °C; E = ethanol. [b] Data with standard deviation; entries in the parentheses present the
number of determinations; Bh = benzhydryl. [c] Data with standard deviation.
Eur. J. Org. Chem. 2010, 1440–1444
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1441

B. Denegri, M. Matic´, O. Kronja
SHORT COMMUNICATION
Table 2. The pseudo-first-order rate constants (kЈ) for reaction of the benzhydrylium ions with the series of aqueous ethanols at 20 °C
and the derived second and first-order individual rate constants for ethanolysis (kE) and hydrolysis (kW).
Substrate
Solvent[a]
kЈ/104 s–1 [b]
kE/104 s–1–1
kЈE/104 s–1
kW/104 s–1–1
kЈW/104 s–1
1 (MeO,MeO)
E
433[c,d]
253
151
71.3
9.44
868[c,d]
947
27.7
14.8
9.34
2.25
90E
80E
60E
W
234
132
57.3
3.37
1.60
0.255
19.1
18.4
14.0
3 (MeO,Me)
4 (MeO,H)
E
55.6
54.8
47.9
48.1
90E
80E
60E
W
80E
60E
W
866
680
516
14.1
9.39
6.03
1.4
14.5
11.1
80.2
108
139
788
655
78[d]
1220
1060
190[d]
74.1
74.9
1053
801
167
258
[a] Aqueous ethanols mixtures are v/v at 25 °C; E = ethanol, W = water. [b] Rate constants kЈ are taken from ref.[9] [c] 91% in acetonitrile.
[d] Rate constants are taken from ref.[13]
Effect of Microsolvation on Selectivity
Very pronounced changes in selectivities have been ob-
served if the ratio of the solvents has been changed. Thus,
the selectivities are about twice higher if the content of the
ethanol decreases from 90% to 60%, indicating that the in-
dividual rate constants change differently with changing po-
larity. We plotted the logarithms of kE and of kW against
the solvent ionizing power (YOTs).[12] The rate constants ob-
tained earlier with pure water and ethanol (91% in acetoni-
trile) have also been taken in account.[9,13] The plots ob-
tained for cations 1 and 3 are presented in Figure 1, while
those for 4 are shown in the Supporting Information.
The rates for combination of the cations 1–4 with both
solvents decrease as the content of water increases, because
of the lower nucleophilicity of the solvent. However, the
trends of decreasing rates for hydrolysis and ethanolysis are
not the same. Parts a and b of Figure 1 show essentially the
same phenomenon that the individual rates for ethanolysis
in the series of aqueous ethanol depend considerably less
on the solvent content than the individual rates for hydroly-
sis, hence different slopes of log kE vs. YOTS and logkW vs.
YOTS plots cause pronounced changes of the selectivities.
Beside the nucleophilicity of the solvent, the solvation of
a transition state also influences the reaction rate to some
constants for ethanolysis (black circle) and hydrolysis (black
Figure 1. Correlation plots of the logarithms of the individual rate
degree. Different slopes of logkE vs. YOTS and logkW vs.
YOTS lines can be rationalized if the different microsol-
vation of the two transition states is assumed. Since eth-
square) of cation 1 (a) and cation 3 (b) against the ionizing power
[12]
(YOTs
)
for the series of aqueous ethanols.
anolysis is more rapid than hydrolysis, the transition state that in the TS of hydrolysis of cation 1 in aqueous acetoni-
for combination of a given benzhydrylium ion with ethanol trile the formation of C–O bond is already 50–65% ad-
is more carbocation-like than the TS for the reaction of the vanced.[13]
same cation with water. The solvation of the earlier carbo-
In accord with the above considerations are the steeper
cation-like TS is important. On the other hand, because of slopes of the logk vs. YOTS plots for cation 1 in comparison
the more advanced hydrolysis in the TS, a transfer of the to the slopes obtained for cation 3 (Figure 1). While logkE
positive charge, which is delocalized additionally to oxygen, vs. YOTS for 1 has a slope of ca. –0.3, the slope of kE vs.
is more developed, so the importance of solvation is dimin- YOTS plot for 3 is almost parallel to the abscissa. Similarly,
ished, causing more pronounced decrease of the reaction the slope of the logkW vs. YOTS plot for 1 is steeper for ca.
rate, i.e., steeper logkW vs. YOTS plot. This conclusion is in 0.2 in comparison to the slope for 3. Since cation 3 is more
line with the results of Van Pham and McClelland, who on reactive than cation 1, the reactions with both solvents pro-
the basis of kinetic and equilibrium isotope effects showed ceed via earlier TS than for 1, the reactions with solvents
1442
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 1440–1444

Selectivity of Stabilized Benzhydrylium Ions
are less advanced and both transition states are more carbo-
cation-like than the transition states for cation 1. Because
of more important solvation of the early TS, the trends of
decreasing logkE vs. YOTS and logkW vs. YOTS plots for
cation 3 with decreasing nucleophilicity of the solvent (in-
creased water content) are diminished. It should be noted
that the relative differences of the slopes that correspond to
ethanolysis and hydrolysis for both cations 1 and 3 are al-
most same, causing similar changes of selectivity with sol-
vent polarity.
Selectivity of Less Stabilized Benzhydrylium Ions
At this point some selectivity values presented earlier in
the literature can be rationalized using the above approach.
We plotted logS obtained for cations 1–4 against the elec-
trophilicity of the given carbocations (E) and obtained
good linear correlation. This is not surprising, since, ac-
cording to definition [Equation (2)], logS equals the differ-
ence between two variables, logkE and logkW, which both
correlate linearly with E; see Equation (3). The logS vs. E
plot obtained in 80% aq. ethanol is presented in part b of
Figure 2 (all other plots see in the Supporting Information).
For the purpose of comparison, logarithms of some selec-
tivities obtained earlier with X,Y-substituted benz-
hydrylium chlorides[5] and p-nitrobenzoates[7] are also
added to Figure 2. Good logS vs. E correlation provides
reference selectivities expected for free cations if followed
the route presented on Scheme 1, whose values in turn can
be compared with the S values obtained experimentally for
cations 5–8.
There are some striking differences between the selectivi-
ties of the stable cations 1–4 and those of less stable (5–8)
determined earlier. First, the selectivities of cations 1–4 are
almost invariant with the structure, while those of 5–8 de-
crease considerably with decreasing stability of the cation
(Figure 2, b). Also the S values for 5–8 are lower than ex-
pected for free cation.[5,7] Secondly, an increase of the selec-
tivities of less stable ions with solvent polarity is less pro-
nounced (e.g. an increase of the water content from 90% to
60% in ethanol causes an increase in selectivities of 7–Cl
for only ∆S = 1.0 at 25 °C).[7] Furthermore, our results pre-
sented in Table 1 indicate that the selectivities of the benz-
hydrylium ions 1–4 are independent of the nucleofugality
of the leaving group, while those obtained with less stable
ions indicate that substrates with better nucleofuges pro-
duce higher selectivity. For example, 7–Br is more selective
than 7–Cl (SBr = 3.69 vs. SCl = 2.60 in 80% aq. ethanol at
25 °C).[7]
Figure 2. Comparison of the selectivities of stabilized (1–4) and less
stable (5–8) X,Y-substituted benzhydrylium ions in 80% aqueous
ethanol at 25 °C using (a) logkЈE vs. E and logkЈW vs. E plots and
(b) logS vs. E plot.
= 3.6) predictably undergo diffusion-controlled reaction
with 80% ethanol.[14] Numerous data show that logk vs. E
correlation plots are linear if the reaction rates are below
108, while for faster reactions the correlation lines bend and
asymptotically approach the diffusion limit region.[14,15]
Since the differences between kЈE and kЈW related to selec-
tivity [logkЈE – logkЈW = logS + log(H2O)/(EtOH), see
Equations (1) and (2)], it is clear (Figure 2, a) that in the
region of less stable carbocations (stronger electrophiles)
the selectivity is reduced.
Dependence of the selectivity on the leaving group can
be rationalized if the rates of the bond breaking and the
bond forming are compared. Stable cations, as are 1–4, re-
act slowly enough with the solvent to allow the nucleofuge
to depart leaving free ions, so the reaction rates with the
solvent are independent on the leaving group. On the other
hand, less stable ions react with the solvent with rates that
are approaching the diffusion limit, so the counterion can-
not diffuse away. In such cases pairing occurs and the selec-
tivity depends on the nucleofugality of the counterion in a
way that better leaving group, which diffuses away faster
and therefore increases the fraction of the free ions, pro-
duces higher selectivity. It should be emphasized that the
selectivities presented in the literature have been determined
for such less stable ions.
In order to examine the trends of the individual pro-
cesses, we plotted the logarithms of the individual first-or-
der rate constants (kЈE and kЈW) obtained in 80% ethanol
against E and obtained linear plots; see Equation (3). Ex-
trapolation of those plots toward higher E indicates that
the individual rates for ethanolysis are reaching the dif-
fusion limit. This is in line with Mayr’s results that carbo-
Experimental Section
General: All substrates were prepared as described previously.[11]
cations which are less stable than di-p-tolylmetyl cation (E Typically, 40–50 mg of substrate was dissolved in 0.10–0.15 mL of
Eur. J. Org. Chem. 2010, 1440–1444
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1443

B. Denegri, M. Matic´, O. Kronja
SHORT COMMUNICATION
dichloromethane and injected into 50 mL of solvent (2–3ϫ10–3 ).
Solvolyses were carried out under the thermostatted conditions at
20Ϯ0.1 °C. After 10 half-lives the solvent was evaporated in vacuo.
The residue was dissolved in deuterated chloroform, and NMR
spectrum was recorded. The individual S values were calculated
according to the Equation (2) from the observed peak area ratios
of α-H of ether (5.3–5.4 ppm) and alcohol (5.7–5.8 ppm). Solvo-
lyses were also conducted in the presence of pyridine (5–
11ϫ10–3 ). The buffered solutions gave individual S values within
the experimental error in comparison with the unbuffered solu-
tions. The S values presented are average values obtained from at
least three measurements, which included both the buffered and the
unbuffered solvolyses.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
J. M. Harris, D. C. Clark, A. Becker, J. F. Fagan, J. Am. Chem.
Soc. 1974, 96, 4478–4483.
Y. Karton, A. Pross, J. Chem. Soc. Perkin Trans. 2 1978, 595–
598.
J. M. Harris, A. Becker, J. F. Fagan, F. A. Walden, J. Am.
Chem. Soc. 1974, 96, 4484–4489.
Y. Karton, A. Pross, J. Chem. Soc. Perkin Trans. 2 1977, 1860–
1863.
P. R. Luton, M. C. Whiting, J. Chem. Soc. Perkin Trans. 2
1979, 646–647.
D. J. McLennan, P. L. Martin, J. Chem. Soc. Perkin Trans. 2
1982, 1099–1105.
H. Mayr, A. R. Ofial, Pure Appl. Chem. 2009, 81, 667–683.
[9] S. Minegishi, S. Kobayashi, H. Mayr, J. Am. Chem. Soc. 2004,
Supporting Information (see also the footnote on the first page of
this article): logS vs. E plots for X,Y-substituted benzhydryl phenyl
and methyl carbonates in 60%, 70%, 80% and 90% aq. ethanol
with logS for X,Y-substituted benzhydryl chlorides p-nitro-
benzoates.
126, 5174–5181.
[10] a) H. Mayr, T. Bug, M. F. Gotta, N. Hering, B. Irrgang, B.
Janker, B. Kempf, R. Loos, A. R. Ofial, G. Remennikov, H.
Schimmel, J. Am. Chem. Soc. 2001, 123, 9500–9512; b) H.
Mayr, B. Kempf, A. R. Ofial, Acc. Chem. Res. 2003, 36, 66–77.
[11] B. Denegri, O. Kronja, J. Org. Chem. 2007, 72, 8427–8433.
[12] T. W. Bentley, G. Llewellyn, Prog. Phys. Org. Chem. 1990, 17,
121–159.
Acknowledgments
[13] T. Van Pham, R. A. McClelland, Can. J. Chem. 2001, 79, 1887–
1897.
We gratefully acknowledge the financial support of this research
by the Ministry of Science, Education and Sport of the Republic
of Croatia (Grant No. 006-0982933-2963).
[14] S. Minegishi, R. Loos, S. Kobayashi, H. Mayr, J. Am. Chem.
Soc. 2005, 127, 2641–2649.
[15] H. Mayr, A. R. Ofial, Angew. Chem. Int. Ed. 2006, 45, 1844–
1854.
[1] R. Ta-Shma, Z. Rappoport, Adv. Phys. Org. Chem. 1992, 27,
239–291, and references cited therein.
Received: November 18, 2009
Published Online: February 3, 2010
1444
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 1440–1444