How constant are Ritchie's constant selectivity relationships ? A general reactivity scale for n-, π-, and σ-nucleophiles

Article abstract of DOI:10.1021/ja021010y

The kinetics of 82 reactions of benzhydrylium ions (Ar₂CH⁺) with n-nucleophiles has been determined at 20 °C. Evaluation by the equation log k = s(N + E) delivered the reactivity parameters N and s for 15 n-nucleophiles (water, hydroxide, amines, etc.). All nucleophiles except water (s = 0.89) and ^-SCH₂CO₂^- (s = 0.43) have closely similar slope parameters (0.52 < s < 0.71), indicating that the reactions of most n-nucleophiles approximately follow Ritchie's constant selectivity relationship (s = constant). The different slope parameter for water is recognized as the main reason for the deviations from the Ritchie relationship reported in 1986. Correlation analysis of the rate constants for the reactions of benzhydrylium ions with the n-nucleophiles (except H2O) on the basis of Ritchie's equation log k = N₊ + log k₀ yields a statistically validated set of N₊ parameters for Ritchie-type nucleophiles and log k₀ parameters for benzhydrylium ions. The N and s parameters of the n-nucleophiles derived from their reactions with benzhydrylium ions were combined with literature data for the reactions of these nucleophiles with other carbocations to yield electrophilicity parameters E for tritylium, tropylium, and xanthylium ions. While the E parameters for tropylium and xanthylium ions appear to be generally applicable, it is demonstrated that the E parameters of tritylium ions can be used to predict reactivities toward n-nucleophiles as well as hydride transfer rate constants but not rates for the reactions of tritylium ions with π-nucleophiles. It is now possible to merge the large data sets determined by Ritchie and others with our kinetic data and present a nucleophilicity scale comprising n- (e.g., amines), π- (e.g., alkenes and arenes), and σ-nucleophiles (e.g., hydrides).

Full text of DOI:10.1021/ja021010y

Published on Web 12/10/2002
How Constant Are Ritchie’s “Constant Selectivity
Relationships”? A General Reactivity Scale for n-, π-, and
σ-Nucleophiles
Shinya Minegishi and Herbert Mayr*
Contribution from the Department Chemie der Ludwig-Maximilians-UniVersit a¨ t M u¨ nchen,
Butenandtstrasse 5-13 (Haus F), D-81377 M u¨ nchen, Germany
Received July 23, 2002; Revised Manuscript Received October 4, 2002; E-mail: Herbert.Mayr@cup.uni-muenchen.de
+
Abstract: The kinetics of 82 reactions of benzhydrylium ions (Ar
2
CH ) with n-nucleophiles has been
determined at 20 °C. Evaluation by the equation log k ) s(N + E) delivered the reactivity parameters N
and s for 15 n-nucleophiles (water, hydroxide, amines, etc.). All nucleophiles except water (s ) 0.89) and
-
-
SCH
2
CO
2
(s ) 0.43) have closely similar slope parameters (0.52 < s < 0.71), indicating that the reactions
of most n-nucleophiles approximately follow Ritchie’s constant selectivity relationship (s ) constant). The
different slope parameter for water is recognized as the main reason for the deviations from the Ritchie
relationship reported in 1986. Correlation analysis of the rate constants for the reactions of benzhydrylium
ions with the n-nucleophiles (except H
statistically validated set of N parameters for Ritchie-type nucleophiles and log k
2
O) on the basis of Ritchie’s equation log k ) N
+
+ log k
0
yields a
+
0
parameters for
benzhydrylium ions. The N and s parameters of the n-nucleophiles derived from their reactions with
benzhydrylium ions were combined with literature data for the reactions of these nucleophiles with other
carbocations to yield electrophilicity parameters E for tritylium, tropylium, and xanthylium ions. While the E
parameters for tropylium and xanthylium ions appear to be generally applicable, it is demonstrated that the
E parameters of tritylium ions can be used to predict reactivities toward n-nucleophiles as well as hydride
transfer rate constants but not rates for the reactions of tritylium ions with π-nucleophiles. It is now possible
to merge the large data sets determined by Ritchie and others with our kinetic data and present a
nucleophilicity scale comprising n- (e.g., amines), π- (e.g., alkenes and arenes), and σ-nucleophiles (e.g.,
hydrides).
2
Introduction
with increasing reactivity. The so-called “constant selectivity
relationship” (eq 1) found wide acceptance, but in 1986 Ritchie
Ritchie’s discovery that the rates of the reactions of stabilized
carbocations and diazonium ions with water, alcohols, and
replaced the electrophile-independent parameters N+ by several
sets of N+ values, which actually are relative reactivities toward
malachite green, tris(p-methoxy)tritylium, pyronin-Y, or the
p-(dimethylamino)phenyltropylium ion. Since then many authors
either have been using the “largest revised set of N+ parameters”,
which effectively are reactivities toward malachite green (or
several anions can be described by eq 1 marked a change of
paradigm in Physical Organic Chemistry.1
log(k/k ) ) N+
(1)
0
3
tris-p-methoxytritylium ion), or have been referring to Ritchie’s
log k ) N + log k0
(1a)
+
1
4
original parameters when analyzing new reactivity data.
log k ) electrophile-dependent parameter
0
(
2) Analyses of reactivity selectivity relationships: (a) Stock, L. M.; Brown,
H. C. AdV. Phys. Org. Chem. 1963, 1, 35-154. (b) Leffler, J. E.; Grunwald,
E. Rates and Equilibria of Chemical Reactions; Wiley: New York, 1963;
pp 162-168. (c) Johnson, C. D. Chem. ReV. 1975, 75, 755-765. (d) Giese,
B. Angew. Chem. 1977, 89, 162-173; Angew. Chem., Int. Ed. Engl. 1977,
N ) nucleophile-dependent parameter
+
1
6, 125-136. (e) Pross, A. AdV. Phys. Org. Chem. 1977, 14, 69-132. (f)
Equation 1 implies that the relative reactivities of two
nucleophiles are given by the differences of their N+ values,
which are independent of the electrophilicities of the reaction
partners. It thus contradicts the previously accepted reactivity-
selectivity principle, which postulates a decrease of selectivity
Young, P. R.; Jencks, W. P. J. Am. Chem. Soc. 1979, 101, 3288-3294.
g) Ta-Shma, R.; Rappoport, Z. J. Am. Chem. Soc. 1983, 105, 6082-
095. (h) Richard, J. P.; Jencks, W. P. J. Am. Chem. Soc. 1984, 106, 1373-
1383. (i) Arnett, E. M.; Molter, K. E. Acc. Chem. Res. 1985, 18, 339-
(
6
3
46. (j) Jencks, W. P. Chem. ReV. 1985, 85, 511-527. (k) Argile, A.; Carey,
A. R. E.; Fukata, G.; Harcourt, M.; More O’Ferrall, R. A.; Murphy, M. G.
Isr. J. Chem. 1985, 26, 303-312. (l) Buncel, E.; Wilson, H. J. Chem. Educ.
1
1
987, 64, 475-480. (m) Formosinho, S. J. J. Chem. Soc., Perkin Trans. 2
988, 839-846. (n) Johnson, C. D.; Stratton, B. J. Chem. Soc., Perkin
*
To whom correspondence should be addressed. Fax: int.+49-89-2180-
Trans. 2 1988, 1903-1907. (o) Exner, O. Correlation Analysis of Chemical
Data; Plenum Press: New York, 1988. (p) Lee, I. Chem. Soc. ReV. 1990,
19, 317-333. (q) Exner, O. J. Chem. Soc., Perkin Trans. 2 1993, 973-
979.
7
7717.
(
1) (a) Ritchie, C. D. Acc. Chem. Res. 1972, 5, 348-354. (b) Ritchie, C. D. J.
Am. Chem. Soc. 1975, 97, 1170-1179.
286
9
J. AM. CHEM. SOC. 2003, 125, 286-295
10.1021/ja021010y CCC: $25.00 © 2003 American Chemical Society

A General Reactivity Scale for Nucleophiles
A R T I C L E S
Scheme 1. Abbreviations and Electrophilicity Parameters E of
Rearranging eq 1 leads to eq 1a, which emphasizes that log k0
in the Ritchie equation corresponds to a nucleophile-independent
electrophilicity parameter that was initially defined as the rate
constant for the reaction of an electrophile with water. During
the past 15 years, we have been studying reactions of carboca-
Benzhydrylium Ions
tions with π-nucleophiles,5-8 carbanions, and hydride donors
9
10-12
and demonstrated that the rates of these reactions can be
described by eq 2.
log k(20 °C)) s(N + E)
s ) nucleophile-specific parameter
N ) nucleophilicity parameter
E ) electrophilicity parameter
(2)
Equation 2 differs from eq 1/1a by the use of an additional
nucleophile-specific slope parameter, s. This parameter, which
was set to s ) 1 for 2-methyl-1-pentene,5 represents a correct-
ion term that improves the fit of data compared to eq 1. A theo-
retical interpretation of the physical meaning of s has recently
been published.13 The Ritchie equation (eq 1/1a) can be con-
sidered as a special case of eq 2 that holds for reactions with
groups of nucleophiles that have identical slope parameters s.
It was the goal of this work to investigate Ritchie-type
reactions (carbocations + n-nucleophiles) with our methodology
and to identify domains that can sufficiently be described by
Ritchie’s eq 1. In this way, it should become possible to search
for relationships between the reactivity parameters of eqs 1 and
,8
2
and to create a common reactivity scale for n-, π-, and
σ-nucleophiles.
Method
Recently, we have recommended 22 differently substituted
benzhydrylium ions as reference electrophiles for quantifying
the reactivities of various types of nucleophiles.7,8 (Scheme 1)
We have demonstrated that the same electrophilicity para-
meters E can be used for describing the reactions of benzhydryl
cations with alkenes, arenes, allylsilanes, allylstannanes, enol
(
3) Ritchie, C. D. Can. J. Chem. 1986, 64, 2239-2250.
(4) (a) Richard, J. P.; Amyes, T. L.; Vontor, T. J. Am. Chem. Soc. 1992, 114,
5
626-5636. (b) Richard, J. P. Tetrahedron 1995, 51, 1535-1573. (c)
Richard, J. P.; Toteva, M. M.; Crugeiras, J. J. Am. Chem. Soc. 2000, 122,
1
664-1674. (d) Okuyama, T.; Haga, N.; Takane, S.; Ueno, K.; Fueno, T.
Bull. Chem. Soc. Jpn. 1991, 64, 2751-2756. (e) Heo, C. K. M.; Bunting,
J. W. J. Chem. Soc., Perkin. Trans. 2 1994, 2279-2290. (f) Vigroux, A.;
Kresge, A. J.; Fishbein, J. C. J. Am. Chem. Soc. 1995, 117, 4433-4434.
a
From ref 8.
(
g) Dalby, K. N.; Jencks, W. P. J. Am. Chem. Soc. 1997, 119, 7271-7280.
ethers, ketene acetals, enamines, carbanions,9 and hydride
7
(
h) de Lucas, N. C.; Netto-Ferreira, J. C.; Andraos, J.; Scaiano, J. C. J.
1
0-12
Org. Chem. 2001, 66, 5016-5021.
donors.
We have now investigated the kinetics of the
(
5) Mayr, H.; Patz, M. Angew. Chem. 1994, 106, 990-1010; Angew. Chem.,
Int. Ed. Engl. 1994, 33, 938-957.
reactions of benzhydrylium ions with n-nucleophiles (Ritchie-
type nucleophiles) and analyzed the observed second-order rate
constants by eqs 1 and 2.
(
6) (a) Mayr, H.; Kuhn, O.; Gotta, M. F.; Patz, M. J. Phys. Org. Chem. 1998,
1
1, 642-654. (b) Mayr, H.; Patz, M.; Gotta, M. F.; Ofial, A. R. Pure Appl.
Chem. 1998, 70, 1993-2000.
(
7) Mayr, H.; Kempf, B.; Ofial, A. R. Acc. Chem. Res. ASAP 8/30/2002
(DOI: 10.1021/ar020094c).
Experimental Section
(
8) Mayr, H.; Bug, T.; Gotta, M. F.; Hering, N.; Irrgang, B.; Janker, B.; Kempf,
B.; Loos, R.; Ofial, A. R.; Remennikov, G.; Schimmel, H. J. Am. Chem.
Soc. 2001, 123, 9500-9512.
Materials. Potassium hydroxide was purchased as an aqueous
standard solution (Merck). 2,2,2-Trifluoroethanol (TFE), 2,2,2-trifluo-
roethylamine, semicarbazide hydrochloride, hydroxylamine hydrochlo-
(
9) (a) Lucius, R.; Mayr, H. Angew. Chem. 2000, 112, 2086-2089; Angew.
Chem., Int. Ed. 2000, 39, 1995-1997. (b) Lucius, R.; Loos, R.; Mayr, H.
Angew. Chem. 2002, 114, 97-102; Angew. Chem., Int. Ed. 2002, 41, 91-
2 3 3
ride, sodium sulfite (Na SO ), sodium azide (NaN ), aqueous hydrogen
9
5.
(
10) (a) Mayr, H.; Basso, N.; Hagen, G. J. Am. Chem. Soc. 1992, 114, 3060-
peroxide, phenol, and p-nitrophenol were reagent-grade chemicals from
commercial sources and used without further purification. n-Propyl-
amine, morpholine, and piperidine were commercial samples (Acros)
and distilled over KOH before use. Aqueous hydrogen peroxide
3
066. (b) Mayr, H.; Basso, N. Angew. Chem. 1992, 104, 1103-1105;
Angew. Chem., Int. Ed. Engl. 1992, 31, 1046-1048.
(
(
11) Funke, M.-A.; Mayr, H. Chem. Eur. J. 1997, 3, 1214-1222.
12) Mayr, H.; Lang, G.; Ofial, A. R. J. Am. Chem. Soc. 2002, 124, 4076-
4
083.
solutions were titrated with potassium permanganate solution. H
2 2
NCH -
(
13) Schindele, C.; Houk, K. N.; Mayr, H. J. Am. Chem. Soc. 2002, 124, 11208-
1
1214.
CO Et was purchased as the hydrochloride salt, and the free base was
2
J. AM. CHEM. SOC.
9
VOL. 125, NO. 1, 2003 287

A R T I C L E S
Minegishi and Mayr
3 2 2 2
liberated with NH gas as described in ref 14. The free H NCH CO Et
1
was stored at -60 °C, and its purity was checked by H NMR before
use. HSCH CO Na was recrystallized from aqueous EtOH. Diazabicyclo-
2.2.2]octane (Dabco) was recrystallized from n-hexane. Sodium
tetrafluoroborate (NaBF ) was recrystallized from methanol.
Benzhydrylium tetrafluoroborates were prepared as previously
2
2
[
4
8
described.
Water was distilled and passed through a Milli-Q water purification
system. Dimethyl sulfoxide (DMSO, Fluka, puriss., stored over
molecular sieve, H O e 0.01%) was used without further purification.
2
Acetonitrile was distilled over diphenylketene.
Kinetics. The reactions of benzhydrylium ions with nucleophiles
were studied in aqueous solution or in DMSO. The benzhydrylium salts
used in this study are colored substances with absorption maxima in
the range of 585-634 nm, which differ only slightly from those reported
Figure 1. Determination of the second-order rate constant for the reaction
of (lil)2CH with OH in water (with 0.4% CH3CN, 20 °C).
8
+
-
in CH
2
Cl
2
(Supporting Information Table S6). All amines were used
-
-
-
-
as free bases. The anions SCH
2
CO
2
, HOO , PhO , and (p-
-
NO
2
)C
6
H
4
O
were generated in aqueous solution by treatment of the
sulfonic acids. Probably protonation at nitrogen and successive reaction
with water resulted in a decolorization of the solutions.
corresponding acids with KOH. Solutions of sulfite ion contained ca.
-
5
1
0
3
M hydroquinone (recrystallized from CH CN) to avoid decom-
Attempts to determine the reactivities of benzhydrylium ions with
1
5
position.
+
phenoxide ion in water failed, since none of the reactions of (lil)
2
CH ,
As the reactions of the colored benzhydrylium ions with n-
nucleophiles gave rise to colorless products, the reactions could be
+
+
+
(jul)
2 2 2
CH , (thq) CH , or (pyr) CH with an excess of phenoxide
showed an exponential decay of the carbocation concentration. The
1
6
followed by employing UV-vis spectroscopy. The rates of slow
reactions (τ1/2 > 10 s) were determined by using a J&M TIDAS diode
array spectrophotometer, which was controlled by Labcontrol Spectacle
software and connected to a Hellma 661.502-QX quartz Suprasil
immersion probe (5 mm light path) via fiber optic cables and standard
SMA connectors. The temperature of solutions during all kinetic studies
was kept constant (usually 20 ( 0.2 °C) by using a circulating bath
thermostat and monitored with a thermocouple probe that was inserted
into the reaction mixture.
+
+
2 2
reactions of p-nitrophenoxide with (ind) CH and (dma) CH were also
examined, but precise rate constants could not be obtained because of
the low reactivity of p-nitrophenoxide.
+
O- in Water. A solution
CN was
added to a mixture of trifluoroethanol (TFE, 2 mL) and aqueous KOH
0.491 M, 1.4 mL) in 500 mL of water. After stirring at room
temperature for 30 min, the organic layer was extracted with four 100-
mL portions of CH Cl . The combined organic layers were washed
with water and dried with MgSO . Then the solvent was evaporated to
give 117 mg of a 7:1 mixture of (dma) CHOCH CF and ((dma) CH)
according to H NMR) as a pale blue solid.
Product from (dma)
2
CH with CF
3
CH
2
+
-
of (dma)
2
CH BF
4
(200 mg, 0.588 mmol) in 10 mL of CH
3
(
2
2
Hi-Tech SF-61DX2 stopped-flow spectrophotometer systems (con-
trolled by Hi-Tech KinetAsyst2 software) were used for the investiga-
tion of rapid reactions of benzhydrylium ions with n-nucleophiles (τ1/2
4
2
2
3
2
2
O
1
(
<
10 s at 20 °C). The kinetic runs were initiated by mixing equal
volumes of solutions of the nucleophile and the benzhydrylium salt.
Nucleophile concentrations at least 10 times higher than the benzhydryl
cation concentrations were usually employed, resulting in pseudo-first-
order kinetics with an exponential decay of the benzhydryl cation
concentration. First-order rate constants k1Ψ (s ) were obtained by least-
squares fitting of the absorbance data (averaged from at least four kinetic
Results
The combination of a benzhydrylium salt with more than 10
equiv of a nucleophile usually resulted in an exponential decay
of the carbocation absorption, from which the pseudo-first-order
rate constant k1Ψ was derived. As shown for the reaction of
-1
+
-
runs at each nucleophile concentration) to the single-exponential A
t
)
(lil) CH with OH in Figure 1, k1Ψ increases linearly with
2
A
0
exp(-k1Ψt) + C.
the concentration of the nucleophile, and the slope of this
As shown for the reaction of (lil)
2
CH+ with OH- in Figure 1, k1Ψ
correlation corresponds to the second-order rate constant (k2).
All second-order rate constants reported in this paper have
analogously been derived from k1Ψ vs [nucleophile]0 plots, as
explicitly shown on pp S23-S77 in the Supporting Information.
In some cases, a bathochromic shift of the absorption
maximum up to 5 nm was observed in the final stages of the
reactions, when the carbocation concentrations became small
(>95% conversion). Since the reason for this shift is not known,
we have not evaluated the late stages of such reactions.
The rates of cation-anion combinations are known to depend
on ionic strength (I). However, Ritchie reported that for
aqueous solutions, changes of ionic strength are negligible when
I < 0.1 mol L . In accord with this report, the second-order
rate constant for the reaction of (lil)2CHBF4 with OH remained
almost unchanged when NaBF4 was added to realize a constant
ionic strength of I ) 0.005 or 0.01 mol L instead of I )
increases linearly with the concentration of the nucleophile, and the
slope of this correlation corresponds to the second-order rate constant
(k
2
). All second-order rate constants reported in this paper have
analogously been derived from k1Ψ vs [nucleophile] plots.
0
As a consequence of the poor solubility of the benzhydrylium
tetrafluoroborates, it was necessary to employ 0.4% (v/v) of a cosolvent
(TFE or CH
3
CN) for the kinetic investigations in water. Since aqueous
solutions of benzhydrylium salts with an electrophilicity parameter E
>
-7 (Scheme 1) are not stable, the rates of the reactions of these
-
electrophiles with OH or H
of benzhydrylium ions in CH
2
O were determined by mixing solutions
-
3
CN with equal volumes of water or OH /
17
water in the stopped-flow instrument. In some experiments with
+
-
(
dma)
2
CH BF
4
in water, small quantities of benzenesulfonic acid or
-1 18
p-toluenesulfonic acid were added to stabilize the aqueous solutions
-
of the benzhydrylium salts. We were not able, however, to stabilize
+
+
2 2
aqueous solutions of (mor) CH and (mfa) CH by the addition of
-
1
(
14) Kane-Maguire, L. A. P.; Kanitz, R.; Jones, P.; Williams, P. A. J. Organomet.
Chem. 1994, 464, 203-213.
(17) (a) Ritchie, C. D.; Skinner, G. A.; Badding, V. G. J. Am. Chem. Soc. 1967,
89, 2063-2071. (b) Bunton, C. A.; Huang, S. K. J. Am. Chem. Soc. 1972,
94, 3536-3544.
(18) Ritchie, C. D.; Minasz, R. J.; Kamego. A. A.; Sawada, M. J. Am. Chem.
Soc. 1977, 99, 3747-3753.
(
15) Ritchie, C. D.; Virtanen, P. O. I. J. Am. Chem. Soc. 1973, 95, 1882-1889.
16) (a) Mayr, H.; Schneider, R.; Schade, C.; Bartl, J.; Bederke, R. J. Am. Chem.
Soc. 1990, 112, 4446-4454. (b) Mayr, H.; Ofial, A. R. Einsichten-
Forschung an der LMU M u¨ nchen 2001, 20, 30-33.
(
288 J. AM. CHEM. SOC.
9
VOL. 125, NO. 1, 2003

A General Reactivity Scale for Nucleophiles
A R T I C L E S
solutions containing 0.4% TFE within a standard deviation of
3
% (Table 2). For that reason, the presence of 0.4% of
cosolvents in water will be neglected in the following discussion.
In the reactions of benzhydrylium ions with n-propylamine,
hydrogen peroxide anion, or 2,2,2-trifluoroethoxide in water,
competition of hydroxide with these nucleophiles has to be
+
considered. As explicitly described for the reaction of (lil)2CH
with n-propylamine in water (Table 3), the pseudo-first-order
rate constant reflects the reaction of the carbocation with
-
n-propylamine and with OH (eq 3).
-
k1Ψ ) k2,OH
-
[OH ] + k2,n-PrNH2[n-PrNH ]
(3)
2 eff
Since the concentrations of n-propylamine and hydroxide ion
21a
can be calculated from the known pKa values and the second-
-
order rate constant for the reaction with OH has independently
been determined in this work, one can easily derive the
contribution of k2,n-PrNH to the observed pseudo-first-order rate
2
Figure 2. Solvent effect on the second-order rate constant of the reaction
+
-
constant. Table 3 shows that the corrections due to the
of (lil)2CH with OH in acetonitrile-water mixtures (individual rate
-
constants are given on pp S23-S28 of the Supporting Information). Because
contribution of OH are marginal (<2%), and k2,n-PrNH can again
2
+
-
of the poor solubility of (lil)2CH BF4 , the rate constant in pure water
cannot be measured.
be derived from a plot of k1Ψ,n-PrNH2 vs [n-PrNH2]eff (see
Supporting Information, pp S49-S54).
Analogously, the observed pseudo-first-order rate constants
Table 1. Influence of Ionic Strength (I) on the Rate Constant of
+
-
-5
-1
-
the Reaction of (lil)2CH BF4 (1 × 10 mol L ) + OH in Water
for the reactions of benzhydrylium ions with trifluoroethoxide
(with 0.4% CH3CN)
3
(
)
pKa ) 12.4) and the anion of hydrogen peroxide (pKa
1
], mol L-1
I, mol L-1
, L mol-1 s-1
21b
[
KOH], mol L-
[NaBF
4
k
2
11.8) are only slightly affected by the competing reaction
0
0
0
0
.001-0.004
.001-0.004
.001-0.004
.002-0.008
0.001-0.004
0.005
0.01
2.16
2.24
1.93
1.18
with hydroxide. In accord with this interpretation, the reaction
0.004-0.001
0.009-0.006
0.048-0.042
+
-
of (dma)2CH BF4 with 99.6/0.4 (v/v) H2O/TFE in the presence
-
of OH gave preferentially the trifluoroethyl ether.
0.05
In several cases, the equilibrium constant for product forma-
tion was relatively small, and the carbocation absorbance
remained constant or only partially disappeared upon addition
1
0
.001-0.004 mol L- in the absence of the inert salt (Table 1).
-
1
+
Only when the ionic strength was set to I ) 0.05 mol L by
the addition of NaBF4 did the second-order rate constant
decrease to 50% of the value observed in the absence of an
inert salt (Table 1). Since all reactions studied in this investiga-
tion were carried out at I < 0.01 mol L , we did not enforce
constant ionic strength by adding inert salts.
To examine the influence of the cosolvents (trifluoroethanol
or acetonitrile) on the rate constants in water, we have studied
the reactivity of OH in water/acetonitrile mixtures of different
composition. As shown in Figure 2, the second-order rate
constant for the reaction of hydroxide with (lil)2CH decreases
from 2.16 to 1.03 L mol
water is increased from 0.4% to 50%. A dramatic enhancement
of reactivity is observed when the acetonitrile content exceeds
of the nucleophile. Thus, the reaction of (lil)2CH with
semicarbazide required a relatively large concentration of
-
3
-1
semicarbazide (1.4 × 10 mol L ) to achieve 90% of
conversion (Supporting Information, p S35).
-
1
When (lil)2CH or (jul)2CH were combined with N3- in
+
+
water,2 the carbocation absorbances decreased by less than 5%.
2
+
+
-
The reactions of (thq)2CH and (dma)2CH with N3 were also
incomplete in water, but they proceeded so fast that we were
not able to determine the rate constants. Only for the reaction
2
3
-
+
-
of (lil)2CH with N3 in DMSO, which proceeded with 65%
-
-5
-1
+
conversion at [N ] ) 2.6 × 10 mol L , the combination
3
0
-
1
-1
rate constant could be determined (Table 2).
s
when the acetonitrile content in
In previous work by McClelland and co-workers,2 it has
4a,b
been shown that the first-order decay of benzhydrylium ions in
-
1
-1
acetonitrile/water mixtures increases with [H O] at low water
7
5% (v/v) (11.2 L mol
analogous solvent dependence has been reported for the reaction
of malachite green with hydroxide ion19 and for the hydroxide-
s
for 15/85 (v/v) H2O/CH3CN). An
2
concentrations but remained almost constant for mixtures
containing more than 20% water. In agreement with this report,
we have found the pseudo-first-order rate constants for the
induced hydrolysis of p-nitrophenyl acetate in acetonitrile/water
+
2
0
consumption of (mfa) CH to increase by less than 16% when
mixtures. The small curvature of the graph in the water-rich
section on the left of Figure 2 suggests an insignificant difference
of reactivity in water and in 99.6/0.4 (v/v) H2O/CH3CN. In
accord with this interpretation, the rate constants measured for
the reactions of benzhydrylium ions with semicarbazide, sulfite,
hydroxylamine, or thiolatoacetate in aqueous solutions contain-
ing 0.4% of acetonitrile agreed with those determined in aqueous
2
the solvent mixture was changed from 91/9 to 50/50 (v/v) H2O/
(
21) (a) Evans, A. G.; Hamann, S. D. Trans. Faraday Soc. 1951, 47, 34-40.
(
b) Evans, M. G.; Uri, N. Trans. Faraday. Soc. 1949, 45, 224-230.
+
-5
-1
-
3 0
]
3
) (1-4) × 10 mol L-1.
-3
(22) (a) [(lil)
2
CH ]
2
0
) 1.5 × 10 mol L ; [N
+
-5
-1
-
-3
(
b) [(jul)
CH ]
0
) 2.2 × 10 mol L ; [N
]
0
) (0.4-1.4) × 10 mol
-1
L .
CH+]
CH ]
) 1.2 × 10 mol L-1; [N
-5
-
]
) (1-4) × 10 mol L-1.
-3
(
23) (a) [(thq)
2
0
3 0
+
-5
-1
-
-3
(
b) [(dma)
2
0
) 1.4 × 10 mol L ; [N
3
]
0
) (0.4-1.4) × 10 mol
-1
L
.
(
19) Bunton, C. A.; Huang, S. K.; Paik, C. H. Tetrahedron Lett. 1976, 18, 1445-
448.
20) Um, I.; Lee, G. J.; Yoon, H. W.; Kwon, D. S. Tetrahedron Lett. 1992, 33,
023-2026.
(24) (a) McClelland, R. A.; Kanagasabapathy, V. M.; Banait, N. S.; Steenken,
S. J. Am. Chem. Soc. 1989, 111, 3966-3972. (b) Pham, T. V.; McClelland,
R. A. Can. J. Chem. 2001, 79, 1887-1897. (c) Minegishi, S.; Mayr, H.
Unpublished results.
1
(
2
J. AM. CHEM. SOC.
9
VOL. 125, NO. 1, 2003 289

A R T I C L E S
Minegishi and Mayr
Table 2. Kinetics of the Reactions of n-Nucleophiles with Benzhydrylium Tetrafluoroborates at 20 °C
1
solventa
, L mol-1 s-1
solventa
nucleophile
H2O
electrophile
k
2
, L mol- s-1
nucleophile
n-PrNH2
electrophile
k
2
(thq)2CH+
2.20 × 10
5.57 × 10
-3 b
-3 b
99.6/0.4 W/AN
99.6/0.4 W/AN
W
(lil)2CH+
7.89 × 10
3.07 × 10
1.23 × 10
3.12 × 10
2.44 × 10
1.87 × 10
7.78 × 10
2.05 × 10
3.99 × 10
1.33 × 10
8.43 × 10
9.43 × 10
4.22 × 10
4.31 × 10
3.93 × 10
1.12 × 10
2.06 × 10
6.61 × 10
7.72 × 10
7.50 × 10
1.20 × 10
3.83 × 10
7.06 × 10
1.50 × 10
4.62 × 10
1.17 × 10
3.23 × 10
7.94 × 10
1.13 × 10
3.19 × 10
6.67 × 10
2.51 × 10
2.88 × 10
3.09 × 10
3.87 × 10
9.67 × 10
1.61 × 10
1.68 × 10
1
2
3
3
4 d
5 d
2
3
3
4
4
2
3
4
3
4
4
4
3
3
4
4
4
5
4
5
5
5
5
5
5
6
5
5
5
5
6
6
99.6/0.4 W/AN
99.6/0.4 W/AN
99.6/0.4 W/AN
99.6/0.4 W/AN
91/9 W/AN
91/9 W/AN
DMSO
DMSO
DMSO
DMSO
DMSO
99.6/0.4 W/AN
99.6/0.4 W/AN
99.6/0.4 W/AN
DMSO
DMSO
DMSO
+
+
(
(
(
(
(
pyr)2CH
(ind)2CH
+
-2 b,c
-1 b
+
dma)2CH
mor)2CH
mfa)2CH
2.6 × 10
(pyr)2CH
+
+
3.31 × 10
50/50 W/AN
50/50 W/AN
67/33 W/AN
67/33 W/AN
67/33 W/AN
67/33 W/AN
99.6/0.4 W/AN
95/5 W/AN
(dma)2CH
(mor)2CH
+
b
+
3.78
+
5 b,c
+
ani)2CH
1.3 × 10
9.1 × 10
2.1 × 10
3.2 × 10
2.16
(pfa)2CH
(lil)2CH
+
+
5 b,c
6 b,c
7 b,c
+
ani(tol)CH
H2NCH2CO2Et
+
ani(Ph)CH
(jul)2CH
(ind)2CH
(thq)2CH
+
+
+
(
tol)2CH
(lil)2CH
OH-
+
+
1
1
1
1
2
1
.90
.65
.05
.03
.81
(dma)2CH
HOO-
(lil)2CH
+
91/9 W/AN
+
80/20 W/AN
50/50 W/AN
25/75 W/AN
15/85 W/AN
99.6/0.4 W/AN
99.6/0.4 W/AN
50/50 W/AN
99.6/0.4 W/AN
99.6/0.4 W/AN
50/50 W/AN
50/50 W/AN
50/50 W/AN
99.6/0.4 W/TFE
99.6/0.4 W/TFE
99.6/0.4 W/TFE
99.6/0.4 W/TFE
99.6/0.4 W/AN
99.6/0.4 W/TFE
99.6/0.4 W/AN
99.6/0.4 W/TFE
99.6/0.4 W/TFE
99.6/0.4 W/TFE
99.6/0.4 W/TFE
DMSO
(ind)2CH
+
(dma)2CH
+
n-PrNH2
(lil)2CH
(jul)2CH
1
+
.12 × 10
+
(
(
jul)2CH+
3.44
(ind)2CH
+
1
+
ind)2CH
1.08 × 10
(thq)2CH
DMSO
-
+
8.56
SO32
(lil)2CH
99.6/0.4 W/TFE
99.6/0.4 W/AN
99.6/0.4 W/TFE
99.6/0.4 W/TFE
99.6/0.4 W/TFE
99.6/0.4 W/TFE
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
99.6/0.4 W/TFE
99.6/0.4 W/AN
99.6/0.4 W/TFE
99.6/0.4 W/TFE
99.6/0.4 W/TFE
DMSO
1
2
1
3
3
(
(
pyr)2CH+
4.85 × 10
+
(jul)2CH+
dma)2CH
1.31 × 10
+
9
.83 × 10
(ind)2CH
+
+
(
(
mor)2CH
1.06 × 10
6.67 × 10
3.32
(thq)2CH
+
+
mfa)2CH
(pyr)2CH
+
+
H2NNHCONH2
HONH2
(lil)2CH
morpholine
piperidine
(lil)2CH
(jul)2CH
+
+
1
1
2
2
+
(
(
(
thq)2CH
pyr)2CH
2.86 × 10
5.56 × 10
1.20 × 10
+
(ind)2CH
+
+
dma)2CH
(thq)2CH
+
1
.20 × 10
(lil)2CH
(jul)2CH
(lil)2CH+
6.59
.37
+
+
6
(ind)2CH
+
(
(
(
(
jul)2CH+
9.58
(thq)2CH
+
1
2
2
1
2
3
4
1
1
2
3
-SCH2CO2-
+
ind)2CH
pyr)2CH
2.94 × 10
1.24 × 10
2.52 × 10
2.26 × 10
1.44 × 10
3.09 × 10
1.65 × 10
3.79 × 10
7.06 × 10
4.35 × 10
2.14 × 10
(lil)2CH
+
+
(jul)2CH+
dma)2CH
+
+
CF3CH2NH2
CF3CH2O-
(lil)2CH
(ind)2CH
+
+
(
(
(
ind)2CH
DMSO
DMSO
DMSO
99.6/0.4 W/TFE
99.6/0.4 W/TFE
99.6/0.4 W/TFE
99.6/0.4 W/TFE
(thq)2CH
+
N3-
+
dma)2CH
mor)2CH
(lil)2CH
(jul)2CH
+
+
6
>2 × 10
DMSO
+
(lil)2CH
jul)2CH
+
(
(
(
+
thq)2CH
+
dma)2CH
a
Mixtures of solvents are given as ratios (v/v). Solvents: W ) water, AN ) acetonitrile, TFE ) trifluoroethanol. Unit is s-1. c Counterion is
b
2
4a d
4
-cyanophenolate or acetate.
These experiments have been performed during late stages of this investigation. For that reason, they are not considered
in Tables S1, S2, and S3.
Table 3. Competing Reaction of n-Propylamine and OH- with (lil)2CH+ (7.39 × 10-6 mol L-1) in Water at 20 °C
, mol L-
1
k
1Ψ, s-1
[OH-]eff,a mol L-1
[n-PrNH
]
eff, mol L
-1
k
,
OH-[OH-]eff, s-1
2
k2,n-PrNH2[n-PrNH ]
eff, s-1
[
n-PrNH
2
]
0
2
2
-
-
-
-
-
3
3
3
3
3
-2
-1
-1
-1
-1
-4
-4
-3
-3
-3
-3
-3
-3
-3
-3
-3
-2
-1
-1
-1
-1
1
2
3
4
5
.29 × 10
5.63 × 10
1.05 × 10
1.59 × 10
2.39 × 10
3.02 × 10
5.96 × 10
6.93 × 10
1.32 × 10
2.00 × 10
3.05 × 10
3.77 × 10
1.30 × 10
1.78 × 10
2.19 × 10
2.70 × 10
3.01 × 10
5.50 × 10
1.03 × 10
1.57 × 10
2.36 × 10
2.99 × 10
-4
.15 × 10
.01 × 10
.30 × 10
.16 × 10
8.24 × 10
-3
1.01 × 10
-3
1.25 × 10
-3
1.39 × 10
a
+
Calculated from pKa(n-PrNH3 ) ) 10.7; ref 21a.
CH3CN.24c For that reason, the first-order rate constants for the
reactions with water in Table 2 referring to different water/
acetonitrile mixtures can directly be compared with each other.
previously published E parameters (Scheme 1) by linear
regressions on the basis of eq 2 (Table 4).
The small value of the standard deviation (factor 1.14)
between experimental rate constants and those calculated by
Discussion
8
eq 2 from E (Scheme 1) and the N and s values given in Table
Figure 3 shows that the rate constants of the reactions of
n-nucleophiles with benzhydrylium ions correlate linearly with
the electrophilicity parameters E, which have been derived from
the reactions of these benzhydrylium ions with a set of
π-nucleophiles.8
4 corroborates the suitability of the previously published
electrophilicity parameters E for describing reactions with
n-nucleophiles. Comparisons between calculated and experi-
mental rate constants can be found in Table S1 of the Supporting
Information.
8
It was, therefore, possible to derive N and s parameters for
n-nucleophiles from the rate constants given in Table 2 and the
Table 4 shows that, with the exception of water (s ) 0.89)
and 2-thiolatoacetate in water (s ) 0.43), all nucleophiles
290 J. AM. CHEM. SOC.
9
VOL. 125, NO. 1, 2003

A General Reactivity Scale for Nucleophiles
A R T I C L E S
Figure 3. Correlation of the rate constants (20 °C) for the reactions of
n-nucleophiles with benzhydrylium ions (Ar2CH ) toward the electrophi-
+
Figure 4. Analysis of the rate constants for the reactions of benzhydrylium
ions with n-nucleophiles (20 °C) in water (W) or dimethyl sulfoxide (D)
according to the Ritchie formalism (eq 1/1a) compared with crystal violet
licity parameters E. Solvents: W ) water, D ) dimethyl sulfoxide, 50AN
)
50/50 (v/v) H2O/CH3CN. Data from Table 2.
(
4-NMe2)3T. a Data for (4-NMe2)3T from ref 3 and Table S3; all other data
from Table 2.
Table 4. Nucleophilicity (N) and Slope (s) Parameters for 15
Ritchie-type Nucleophile Solvent Systems
Table 5. Comparison of N+ Values for n-Nucleophiles with
Respect to Different Series of Electrophiles
nucleophile (solvent)
N
s
nucleophile (solvent)
Ar
2
CH+ a
Ar
3
C+ b
Pyronin-Yc
Ar-Tropd
5.11a
0.89a
H2O (water)
-
b
-
OH (50AN)
10.19
10.47
11.05
11.41
12.15
12.66
0.62
0.61
0.52
0.55
0.65
0.59
0.56c
0.67
0.55
0.64
0.56
0.67
0.71
0.43
OH (water)
)4.75 (5)
4.60 (4)
4.83 (4)
5.16 (5)
5.92 (4)
5.99 (4)
6.22 (4)
7.35 (3)
7.43 (5)
8.11 (4)
8.26 (5)
9.20 (4)
9.61 (4)
9.72 (4)
)4.75
)4.75
)4.75
-
OH (50AN)e
-
OH (water)
H2NNHCONH2 (water)
HONH2 (water)
H NNHCONH (water)
3.73
5.05
4.86
5.06
3.42
3.82
2
2
HONH (water)
2
CF3CH2NH2 (DMSO)
CF CH NH (DMSO)
4.70
7.33
3
2
2
-
-
CF3CH2O (water)
CF CH O (water)
5.66
7.20
3
2
c
n-PrNH2 (water)
13.33
n-PrNH (water)
2
-
H2NCH2CO2Et (DMSO)
14.30
15.40
15.70
16.83
16.96
17.19
22.62
HOO (water)
8.52
6.54
7.88
8.01
-
HOO (water)
H NCH CO Et (DMSO)
2
2
2
n-PrNH2 (DMSO)
n-PrNH (DMSO)
8.40
7.91
9.17
9.32
2
2
-
2-
SO3 (water)
SO3 (water)
7.50
morpholine (DMSO)
piperidine (DMSO)
morpholine (DMSO)
piperidine (DMSO)
-
-
-
-
SCH2CO2 (water)
SCH CO2 (water)
9.09
2
a
From first-order rate constants; correlation not shown in Figure 3. b 50/
0 (v/v) H2O/CH3CN. c These numbers have been corrected after performing
the correlation analysis. For the calculations in Tables S1 and S3, N )
a
This work; least-squares fit; number of reactions given in parentheses.
b
From ref 3, relative reactivities toward malachite green. c From ref 3. From
ref 3, relative reactivities toward the p-(dimethylamino)phenyltropylium ion.
Solvent mixture, 50/50 (v/v) H O/CH CN.
d
5
e
1
3.57 and s ) 0.53 were used, which have been derived from the rate
2
3
+
constants in Table 2 without the rate constants for n-PrNH2 + (mor)2CH
+
and (mfa)2CH .
fixed E parameters from ref 8, while the log k0 parameters in
eq 1 are fully optimized for the data of this investigation. Table
S2 in the Supporting Information gives an explicit comparison
of the experimental rate constants with those calculated from
log k0 (Figure 4) and N+ (Table 5) by the Ritchie eq 1/1a.
investigated in this work have slope parameters of 0.52 < s <
0.71, indicating that most carbocation nucleophile combinations
follow almost constant selectivity relationships: For constant
values of s, eq 2 transforms into eq 1/1a. The exceptionally
high value of s for water given in Table 4 is consistent with
Ritchie’s report that in reactions of tritylium ions the ratio kOH-/
As expected, the N+ values thus derived from the reactions
of the n-nucleophiles with benzhydrylium ions differ only
slightly from those reported by Ritchie (Table 5). Since the
kH O decreases as the electrophilicities of the tritylium ions
2
3
increase.3
different N values given in Ritchie’s 1986 paper for reactions
+
with tritylium ions, tropylium ions, and pyronines are not based
The narrow range that embraces most s parameters in Table
on rate constants for series of reactions but refer to only a single
electrophile of each class, we will not attempt an interpretation
of the differences of N+ from different sources.
4
suggests evaluating the rate constants of Table 2 by the Ritchie
equation (eq 1/1a, Figure 4). When water with the different s
parameter is excluded, the standard deviation between calculated
(
1
eq 1) and experimental rate constants is somewhat larger (factor
.20) than that obtained by employing eq 2 (factor 1.14), despite
the fact that the linear regression according to eq 2 uses the
Evaluation of the kinetic data in Table 2 by the Ritchie
equation eq 1/1a also provides log k0 values for benzhydrylium
ions (Figure 4), which allow us to compare the benzhydrylium
J. AM. CHEM. SOC.
9
VOL. 125, NO. 1, 2003 291

A R T I C L E S
Minegishi and Mayr
Scheme 2. Electrophilicity Parameters E (according to Eq 2) for Ritchie’s Electrophiles from Reactions with n-Nucleophiles (at I < 0.1)
Compared with Electrophilicity Parameters for Benzhydrylium Ionsb
a
For further tropylium ions, see Table S3 (Supporting Information). b The E parameters for tritylium ions given in this scheme must not be used for
predicting reactivities toward π-systems (see the text).
ions studied in this work with those electrophiles previously
investigated by Ritchie (Table S2).
However, because of the much wider applicability of eq 2
compared to eq 1 (in contrast to eq 1, eq 2 also holds for
reactions with π-systems), it is more advantageous not to employ
log k0 (Figure 4) but to use the electrophilicity parameters E as
defined by eq 2 for comparing electrophiles of different type.
Therefore, the previously reported rate constants for the reactions
of tritylium, tropylium, and xanthylium ions with the 15
nucleophile systems characterized in this work (Table 4) were
292 J. AM. CHEM. SOC.
9
VOL. 125, NO. 1, 2003

A General Reactivity Scale for Nucleophiles
A R T I C L E S
Table 6. Comparison between Calculated and Experimental Rate Constants for the Reactions of Tritylium Ions with Hydride Donors
electrophile (E)a
nucleophile (N, s)b
k
calc (20 °C), L mol s-1
c
-1
k
exp, L mol-1 s-1
-
-
-
-
4
3
2
1
-3 d
-3 d
-2 d
-2 d
(
(
(
(
4-MeO)3T (-4.35)
4-MeO)2T (-3.04)
4-MeO)T (-1.87)
4-Me)3T (-1.21)
cycloheptatriene (0.52, 0.97)
cycloheptatriene (0.52, 0.97)
cycloheptatriene (0.52, 0.97)
cycloheptatriene (0.52, 0.97)
1,4-cyclohexadiene (0.09, 0.98)
1.9 × 10
3.6 × 10
4.9 × 10
2.1 × 10
3.9
1.6 × 10
1.1 × 10
2.0 × 10
2.7 × 10
3.2 × 10
8.2 × 10
(80 °C, CH3CN)
(23 °C, CH3CN)
(23 °C, CH3CN)
(23 °C, CH3CN)
(20 °C, CH2Cl2)
(20 °C, CH2Cl2)
(20 °C, CH2Cl2)
-
-
1 e
2 e
T (0.51)
1
9
,4-dihydronaphthalene (-0.07, 1.03)
2.8
-
1
2
2
3
1
2
1
1
-2 e
,10-dihydroanthracene (-0.86, 0.92)
4.8 × 10
1.0 × 10
5.0 × 10
5.7 × 10
1.5 × 10
1.3 × 10
1.1 × 10
3.2 × 10
1.4 × 10
f
cycloheptatriene (0.52, 0.97)
HSiEt3 (3.64, 0.65)
1.8 (20 °C, CH2Cl2)
2 g
1.2 × 10 (20 °C, CH2Cl2)
2 g
HSiMe2Ph (3.27, 0.73)
HSiBu3 (4.45, 0.64)
1,4-cyclohexadiene (0.09, 0.98)
1,4-cyclohexadiene (0.09, 0.98)
H3SiPh (0.25, 0.67)
2.1 × 10 (25 °C, CH2Cl2)
2 g
2.4 × 10 (20 °C, CH2Cl2)
e
(
(
3-Cl)T (1.06)
3-Cl)3T (1.99)
2.3 (20 °C, CH2Cl2)
1 e
5.4 × 10 (20 °C, CH2Cl2)
e
4.6 (20 °C, CH2Cl2)
a
From Scheme 2. From ref 8 or 12. Calculated according to eq 2. d From ref 25. e From ref 26. From ref 27. g From ref 28; for HSiEt3: ∆H ) 29.3
b
c
f
q
-
1
q
-1
-1
q
-1
q
-1
-1
kJ mol , ∆S ) -105 J K mol ; for HSiBu3: ∆H ) 26.4 kJ mol , ∆S ) -109 J K mol
.
Table 7. Approximate Nucleophilicity Parameters N and s of
subjected to a correlation analysis on the basis of eq 2. When
calculating the E parameters of these electrophiles by a least
squares minimization of the deviations between observed and
calculated rate constants (Table S3), the N and s values of Table
Nucleophiles from Reactions with Ritchie’s Electrophiles (Scheme
a
2)
nucleophile (solvent)
MeOH (methanol)
N
s
6.02b
8.70
9.19
9.26
9.81
1.01b
0.68
0.60
0.66
0.63
0.77
0.70
0.63
0.82
0.64
0.59
0.74
0.65
0.57
0.78
0.61
0.61
0.66
0.56
0.57
0.90
0.78
0.81
4
, which were derived from reactions with reference electro-
CF3CH2NH2 (water)
philes, were kept as fixed parameters. In Scheme 2 only those
E parameters are listed that could be derived from kinetics at
an ionic strength of I < 0.1.
Though calculated and experimental rate constants for reac-
tions with tropylium ions often differ by 1 order of magnitude
-
CN (water)
NH3 (water)
CH3ONH2 (water)
H2NCH2CONHCH2CO2 (water)
H2NCH2CO2Et (water)
-
10.28
10.28
10.33
10.37
10.83
11.02
11.15
11.69
11.81
12.23
12.24
12.45
12.48
12.66
12.77
13.59
15.62
15.82
NCCH CH NH (water)
2
2
2
H2NCH2CH2NH2 (water)
PhNHNH2 (water)
(Table S3), the E parameter derived for the parent tropylium
ion from reactions with n-nucleophiles (-3.63, Scheme 2)
-
BH3CN (water)
differs only slightly from that derived from the reactions of this
-
H2NCH2CO2 (water)
8
electrophile with the reference π-nucleophiles (-3.72). One
n-BuNH2 (water)
MeOCH2CH2NH2 (water)
can, therefore, expect that all E parameters for tropylium ions
given in Scheme 2 can be used for calculating reactivities of
these electrophiles toward all types of nucleophiles, i.e., also
for reactions with alkenes, arenes, or hydrides.
-
BH4 (water)
EtNH2 (water)
H NNH (water)
2
2
N-benzyldihydronicotinamide (water)
+
-
Me3N CH2CH2O (water)
The applicability of the E parameters of tritylium ions is more
limited, however, since the sensitivity of bulky reagents toward
variation of the steric requirements of the reaction partner will
be large. Because our approach, like Ritchie’s, does not
explicitly treat steric effects, we have recommended that
-
HCtCCH2O (water)
-
CH3O (methanol)
-
HOCH2CH2S (water)
MeO2CCH2CH2S (water)
-
a
The N and s parameters listed here are those from Table S5 and refer
5
,8
reactions of bulky reagents should not be treated with eq 2.
to correlations covering at least 3 orders of magnitude in rate constants.
These N and s parameters are less accurate than those in Table 4 because
of the indirect evaluation. b Based on first-order rate constants.
The satisfactory agreement between calculated (eq 2) and
experimental rate constants in Table S3 indicates, however, that
reactivities of tritylium ions toward n-nucleophiles can generally
be reproduced by eq 2, in accord with Ritchie’s previous work.
When the E parameters of tritylium ions given in Scheme 2
are used to calculate rate constants of hydride abstractions,
however, kcalc is usually somewhat larger than kobs, indicating
that the transition states of hydride transfer reactions have higher
steric requirements than the reactions of carbocations with
n-nucleophiles (Table 6). This result is surprising in view of
the almost linear C-H-X arrangement in the corresponding
transition states.29 However, since the deviation between kcalc
and kobs in Table 6 rarely exceeds 1 order of magnitude, it is
possible to combine all E parameters presented in Scheme 2
with the N and s parameters of hydride donors8,12 for estimating
the rates of hydride transfer reactions.
As expected, the reactions of π-nucleophiles with tritylium
ions are considerably slower than predicted by eq 2 (Table S4).
For the reactions of tritylium ions with 1-methoxy-2-methyl-
8
1
-(trimethylsiloxy)propene (N ) 9.00, s ) 0.98), the rate
constants calculated by eq 2 are 4-6 orders of magnitude higher
3
0
than experimentally observed by Fukuzumi. Since the large
steric demand of both reagents enforces a reaction at the
p-position of a phenyl group at tritylium, the steric effect for
attack at the tertiary carbenium center is even larger than derived
3
0
from the ratio kcalc/kobs. Furthermore, Fukuzumi reported rate
constants for the reactions of 1-ethoxy-1-(triethylsiloxy)ethene
and tert-butyldimethylsiloxy-1-ethoxyethene that are 2 orders
of magnitude smaller than the prediction of eq 2 for reactions
of tritylium ions with the less nucleophilic 1-phenoxy-1-
(
25) McDonough, L. M. Ph.D. Thesis, University of Washington, Seattle WA,
1
960.
(
(
(
26) Lang, G. Dissertation, Ludwig-Maximilians-Universit a¨ t M u¨ nchen, 1998.
27) M u¨ ller, K.-H. Dissertation, Technische Hochschule Darmstadt, 1997.
28) Chojnowski, J.; Fortuniak, W.; Sta n´ czyk, W. J. Am. Chem. Soc. 1987, 109,
7
776-7781.
(29) W u¨ rthwein, E.-U.; Lang, G.; Schappele, L. H.; Mayr, H. J. Am. Chem.
Soc. 2002, 124, 4084-4092.
(30) Fukuzumi, S.; Otera, J.; Ohkubo, K. J. Org. Chem. 2001, 66, 1450-1454.
J. AM. CHEM. SOC.
9
VOL. 125, NO. 1, 2003 293

A R T I C L E S
Minegishi and Mayr
Scheme 4. Comparison of the Nucleophilic Reactivities of
Scheme 3
n-Nucleophiles with Typical π-Nucleophiles, Hydride Donors, and
Carbanions from Refs 8 or 9c
8
(
trimethylsiloxy)ethene (N ) 8.23, s ) 0.81). These examples
demonstrate that reactions of tritylium ions with π-nucleophiles
cannot be described by eq 2, and we explicitly advise not using
the E parameters of tritylium ions listed in Scheme 2 for
calculating reactivities toward π-nucleophiles.
The previously mentioned suitability of the N and s param-
eters for the n-nucleophiles in Table 4 for calculating their
reactivities toward benzhydrylium, tritylium, tropylium, and
xanthylium ions indicates that differential steric effects are not
important in the reactions of O-, S-, and N-nucleophiles with
these carbocations. The nucleophilicity parameters N and s of
many additional n-nucleophiles (and hydride donors) can,
therefore, be derived from the published rate constants of their
reactions with tritylium, tropylium, and xanthylium ions and
the corresponding E parameters given in Scheme 2. As shown
in Table S5 and the plots in Figure S1 (Supporting Information,
pp S18-S19), 33 nucleophiles have been characterized in this
way. Because of the uncertainty of slopes for “short” correlation
lines, only the N and s parameters of 23 of these nucleophiles
for which rate constants over more than three log k units were
a From Table 4. b From Table 7. c The solvent was CH Cl , unless noted
otherwise: H2O (W), DMSO (D), methanol (M).
2
2
which would be impossible if all reactions would be treated
equally. Inclusion of reactivity data for more reactive carboca-
3
1
tions is in progress.
Conclusion
We have demonstrated that the problem with “constant
selectivity relationships” reported by Ritchie in 1986 is pre-
dominantly caused by the widely deviating slope parameter s
of water. By employing eq 2 instead of eq 1, we can describe
all reactions of tritylium, tropylium, and xanthylium ions with
n-nucleophiles with a single set of parameters. Since eq 2 has
previously been demonstrated to hold for the reactions of
carbocations with π- and σ-nucleophiles, it has now become
possible to combine kinetic data from different sources and
create a nucleophilicity scale that directly compares n-, π-, and
σ-nucleophiles (Scheme 4).
available, have been listed in Table 7.
Readers not familiar with our recent papers7
,8,12
may be
wondering why the stepwise procedure summarized in Scheme
has been used to determine the electrophilicity parameters of
3
Scheme 2 and the nucleophilicity parameters of Table 7 instead
of subjecting all available rate constants for the reactions of
nucleophiles with carbocations to a single correlation analysis.
8
As previously discussed in detail, only the unequal treatment
of data from different sources allows us to systematically extend
our reactivity scales without the necessity to continuously revise
the entire sets of electrophilicity and nucleophilicity parameters.
It is thus possible to employ preliminary numbers in some cases,
which can be replaced by more reliable data at a later stage
without affecting the “established” parameters. This procedure
also allows us to define reactivity parameters that can only be
Though the data of Table 4 indicate a fair correlation between
N and s for n-nucleophiles, the situation becomes more
complicated when π-nucleophiles and hydride donors are
(
31) (a) Cozens, F. L.; Mathivanan, N.; McClelland, R. A.; Steenken, S. J. Chem.
Soc., Perkin. Trans. 2 1992, 2083-2090. (b) McClelland, R. A. Tetrahedron
1996, 52, 6823-6858. (c) Cozens, F. L.; Kanagasabapathy, V. M.;
McClelland, R. A.; Steenken, S. Can. J. Chem. 1999, 77, 2069-2082. (d)
Pezacki, J. P.; Shukla, D.; Lusztyk, J.; Warkentin, J. J. Am. Chem. Soc.
1999, 121, 6589-6598.
+
used for certain types of reactions (e.g., E parameters of Ar3C
for their reactions with n-nucleophiles and hydride donors),
294 J. AM. CHEM. SOC.
9
VOL. 125, NO. 1, 2003

A General Reactivity Scale for Nucleophiles
A R T I C L E S
included. It is evident that most π-nucleophiles are characterized
by higher values of s than n-nucleophiles, even when compounds
of similar N values are compared. As a consequence, more
reactive carbocations (harder electrophiles) will show a relative
preference for π-nucleophiles over n-nucleophiles compared to
less reactive carbocations (softer electrophiles). Since alkoxides
and amines are considered as hard bases in contrast to alkenes
and arenes (soft bases), we must conclude that the Hard Soft
Acid Base Principle32 is not useful for describing our correla-
tions.
According to a recent theoretical analysis,13 slope parameters
of approximately 0.67, as found for most nucleophiles in this
investigation, are indicative of constant intrinsic barriers within
a reaction series. A more detailed analysis considering absolute
intrinsic barriers of these reactions4b,33 is now needed to reveal
the physical background of these correlations.
Acknowledgment. Dedicated to Professor Meinhart H. Zenk
on the occasion of his 70th birthday. Financial support by the
Deutsche Forschungsgemeinschaft and the Fonds der Chemis-
chen Industrie is gratefully acknowledged. We thank Dipl.-
Chem. R. Loos for the investigation of n-propylamine in
dimethyl sulfoxide.
Supporting Information Available: Correlation of electro-
phile-nucleophile reactions, details of the kinetic experiments,
and absorption maxima of benzhydryl cations. This material is
available free of charge via the Internet at http://pubs.acs.org.
JA021010Y
(
33) Richard, J. P.; Amyes, T. L.; Toteva, M. M. Acc. Chem. Res. 2001, 34,
(32) Pearson, R. G. Chemical Hardness; Wiley-VCH: Weinheim, 1997.
981-988.
J. AM. CHEM. SOC.
9
VOL. 125, NO. 1, 2003 295