Silver (I) activated quaternization of tertiary amines by alkyl iodides: Overall analysis coupling homogeneous and heterogeneous processes

Article abstract of DOI:10.1016/j.molcata.2011.12.031

Kinetic data for the silver (I) activated homogeneous and heterogeneous quaternization of tributylamine by alkyl iodides in toluene is presented. Silver iodide was used as a solid catalyst. Solution and surface parameters were obtained applying the multistep kinetic model, previously proposed by Santos and Barbosa. A scrutiny of the derived quantities evidences competing structural and electronic effects. The solution reaction is dominated by structural factors while electronic effects govern the surface process. An overall analysis considering data from triethylamine systems, earlier investigated, allowed the establishment of a parallelism between the homogeneous and heterogeneous processes. A molecular level study involving size, shape and orientation of the chemical species on the surface, lead to estimates of interfacial layer thicknesses. A combination of surface parameters superficial reacting monolayer thicknesses and, the parallelism between homogeneous and heterogeneous catalysis consented the evaluation of "volumetric surface rate constants" which are directly comparable with their solution counterparts.

Full text of DOI:10.1016/j.molcata.2011.12.031

Journal of Molecular Catalysis A: Chemical 356 (2012) 106–113
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Silver (I) activated quaternization of tertiary amines by alkyl iodides: Overall
analysis coupling homogeneous and heterogeneous processes
M. Soledade C.S. Santos, Ester F.G. Barbosa∗
Centro de Química e Bioquímica, DQB, Faculdade de Ciências, Universidade de Lisboa, 1749-016 Lisboa, Portugal
a r t i c l e i n f o
a b s t r a c t
Article history:
Kinetic data for the silver (I) activated homogeneous and heterogeneous quaternization of tributylamine
by alkyl iodides in toluene is presented. Silver iodide was used as a solid catalyst. Solution and surface
parameters were obtained applying the multistep kinetic model, previously proposed by Santos and
Barbosa. A scrutiny of the derived quantities evidences competing structural and electronic effects. The
solution reaction is dominated by structural factors while electronic effects govern the surface process.
An overall analysis considering data from triethylamine systems, earlier investigated, allowed the estab-
lishment of a parallelism between the homogeneous and heterogeneous processes. A molecular level
study involving size, shape and orientation of the chemical species on the surface, lead to estimates of
interfacial layer thicknesses. A combination of surface parameters superficial reacting monolayer thick-
nesses and, the parallelism between homogeneous and heterogeneous catalysis consented the evaluation
of “volumetric surface rate constants” which are directly comparable with their solution counterparts.
© 2012 Elsevier B.V. All rights reserved.
Received 21 October 2011
Accepted 31 December 2011
Available online 10 January 2012
Keywords:
Homogeneous and heterogeneous catalysis
Silver (I) activated amine quaternization
Interfacial molecular thickness
Linking bidimensional and tridimensional
workspaces
1. Introduction
complexes and surfaces, have been addressed by Yermakov and
co-workers [9] as well as in our laboratory [19–21].
Weak intermolecular forces dictate how and why molecules
fit together, commanding binding, solubilization, reaction path-
ways, stereochemistry and molecular recognition phenomena.
Over the last forty years, considerable efforts have been made,
aiming a molecular understanding of chemical and biochemical
processes such as homogeneous, heterogeneous and biological
catalysis. Major progresses in surface and colloid chemistry were
of utmost importance, making relevant contributions to the devel-
opment of experimental, simulation and theoretical methods.
Several approaches, seeking specific localization and alignment of
molecules, imparted by surfaces or media confinements, have been
attempted [1–7]. Medium constraints have been accomplished by
synthesized active sites [8,9]; functionalized porous solids [10,11]
nanocages [3,4], zeolites or supramolecular aggregates [12,13].
The role of physical barriers imposed by structures, emphasiz-
ing a non bystander role, has been frequently evidenced by Rebek
Jr or Fujita and collaborators among others, while exploring the
entrapment of molecules within cages [3,5,14–18]. Likewise, other
surface immobilization strategies, involving supported metal clus-
ters, nanoparticles or metal complexes, have also been successfully
accomplished [8–11]; in particular the interactions between metal
The catalyzed quaternization of silver (I) coordinated tertiary
amines, in the presence of an insoluble silver salt, has been inves-
tigated in our laboratory. Other strategies to accelerate nitrogen
quaternization, a very slow process, may be mentioned, namely
the synthesis by means of resin bound tertiary amines [22], within
deep cavitands [16] or even by the so-called “one pot reactions”
involving amides as starting material [23].
The relevance of quaternary ammonium salts synthesis has been
reinforced by the expanding range of applications, beyond the tra-
ditional uses as surfactants or phase transfer catalysts, in view of
potential uses in battery technology, ionic liquids and drug devel-
opment [4,23,24].
Our previous studies on the joint homogeneous–heterogeneous
catalysis of triethylamine quaternization by alkyl halides [19–21]
disclosed a complex process involving free and adsorbed coordi-
nated triethylamine and alkyl halides. Herein coordinated silver
nitrate acts as a structural promoter for the nucleophilic attack of
the halide in solution and on the surface, due to a more favorable
surface arrangement of the reacting amine and halide molecules.
The extension of these studies to the silver promoted quaterniza-
tion of tributylamine by alkyl iodides, apart from further testing
the previously proposed model [20], enables a broader discus-
sion of free and coordinated amine homogeneous–heterogeneous
processes. Furthermore, the parallelism between solution and sur-
face processes consents the establishment of a molecular level link
between simultaneous homogeneous and heterogeneous process,
occurring within a surface compartment of monolayer thickness.
∗
Corresponding author. Tel.: +351 21 7500896; fax: +351 21 7500088.
E-mail addresses: mssantos@fc.ul.pt (M. Soledade C.S. Santos),
ester.barbosa@fc.ul.pt (E.F.G. Barbosa).
1381-1169/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2011.12.031

M. Soledade C.S. Santos, E.F.G. Barbosa / Journal of Molecular Catalysis A: Chemical 356 (2012) 106–113
107
An approach which evidences how the specific surface imparted
molecular confinement drives the heterogeneously catalyzed pro-
cess.
2. Experimental
2.1. Reagents and solutions
The nominal purity of solvent, Toluene BDH Aristar grade, and
reagents, Tributylamine BDH-gpr, Ethyl iodide BDH-gpr grade and
Butyl iodide BDH-gpr grade, was checked by G.C., being 99.98% for
the solvent Toluene, 94.47% for Tributylamine, 72.30% for Ethyl iodide
and 99.75% for Butyl iodide. Accordingly, Toluene and Butyl iodide
were used without further purification.
Ethyl iodide showed the yellowish color characteristic of pho-
tochemical decomposition [25]. Successive washings with dilute
sodium thiosulfate solutions removed the free iodine. The final
solution was dried with anhydrous calcium chloride, and finally
fractionally distilled over sodium wire [25] leading to a distillate
with 99.75% purity by G.C.
Tributylamine BDH-gpr, was treated with acetic anhydride and
fractionally distilled, refluxed over potassium hydroxide and finally
fractionally distilled from barium oxide through a column packed
with Pyrex glass rings, under a reduced pressure of 1 kPa, as
described by Barbosa and Lampreia [26], and showed a final purity
of 99.98%.
Scheme 1. Schematic representation of the molecular model proposed to interpret
the silver iodide catalyzed quaternization of coordinated trialkylamines by alkyl
iodides in toluene.
3. Results and discussion
Silver nitrate gpr grade (purity ≥ 99.8%) was supplied by BDH
and was used without further purification. Silver iodide was synthe-
sized before each run according to a procedure formerly described
[19–21,27–29]. The polycrystalline solid obtained was predomi-
nantly composed of ˇ-AgI and had a BET (N2) specific surface area
of 0.407 0.003 m2 g−1 [29].
The amine, ethyl and butyl iodide solutions in toluene were pre-
pared, by weight, in volumetric flasks and under a light N2 flux as
previously described [19–21].
3.1. Model framework
A brief description of the kinetic model used to analyze the
kinetic data is outlined below, emphasizing exclusively the major
landmarks for the sake of simplicity. The reaction Scheme 1, pro-
posed earlier [19], to characterize the overall solution and surface
process was considered once again. kQ and kQꢀ are the rate con-
stants for the solution and surface catalyzed quaternization of the
coordinated amine by the alkyl iodide, kM and kMꢀ the solution and
surface rate constants for the Menshutkin reaction, kC the rate con-
stant for the competitive solution reaction between silver nitrate
and the ethyl iodide and K1 = k1/k−1 and K2 = k2/k−2 are the equilib-
rium constants for the formation of the 1:1 and 1:2 silver–amine
complexes.
The total consumption of the reactants, alkyl iodide, tri-
alkylamine and silver nitrate, includes solution and surface
components, the overall rate equations were obtained neglecting
the contribution of the terms referring to the Menshutkin reactions
which are very slow (10−6 10−3 M s−1), the alkylation reactions of
the sterically hindered silver amine 1:2 complex, and the surface
term of the reaction between free silver nitrate and alkyl iodide,
that was disregarded versus its solution counterpart.
2.2. Kinetic runs
The kinetic experiments were performed at 25.0 0.05 ◦C, under
constant stirring, resorting to previously described procedures and
experimental set up.
The kinetic runs were monitored collecting reaction mixture
samples and determining by potentiometric back titration the
amine and the silver content. This procedure ensured that the reac-
tion under study was interrupted, in the collected sample mixtures,
and simultaneously interferences in both analytical methods were
avoided [19].
Amine titration: The amine concentration in sample reaction
mixtures was determined by potentiometric back-titration, of an
excess of perchloric acid (≈0.01 M) with standardized sodium
hydroxide (≈0.04 M) because the acid–base titration, in anhydrous
acetic acid, proved to be inappropriate due to the interference of
silver ions [19].
Silver titration: The silver concentration in solution was deter-
mined by potentiometric back titration of an excess of standardized
potassium iodide with silver nitrate in view of the difficulties asso-
ciated with silver titration, in the presence of solids [19].
In Table 1 are summarized the initial conditions of the experi-
mental kinetic runs used to obtain the relevant kinetic parameters
for the reactions studied.
Accordingly
d AgNO
d RI
d Et N
[
]
[
]
[
]
3
3
total
total
total
∼
=
>
(1)
dt
dt
dt
The corresponding differential rate expressions, containing
solution and interfacial contributions were scrutinized in previous
papers [20,21] and will not be repeated here. As before, integration
of the complex differential equations was possible over extended
periods of time, where the free amine concentration in solution
remained almost constant [20,21], leading to
ꢀ
ꢁ
ꢀ
ꢁ
e − ꢀx(t)
a − x(t)
e − ꢀx(tac
a − x(tac
)
)
ln
= ln
ꢂ
ꢃ
kC
AgI
−
kQ
+
(ꢀa − e)Q2(t − tac
)
(2)
Bu3N + Bu3NAg+NO3− + RI−→ RBu3N+NO3−↓ + AgI↓ + RBu3N+I↓−
K1[Et3N]free

108
M. Soledade C.S. Santos, E.F.G. Barbosa / Journal of Molecular Catalysis A: Chemical 356 (2012) 106–113
Table 1
Set of initial experimental conditions of the kinetic runs used to calculate the kinetic and adsorption parameters.
Initial conditions
Run
Reagents
[Bu3NAgNO3] (M)
[RI] (M)
m0 (g)
AgI
20a
21
22a
23a
24
25
26
27
28
29
30
31
[Bu3N]/M = 0.03911[AgNO3]/M = 0.03368
[Bu3N]/M = 0.03919[AgNO3]/M = 0.03796
[Bu3N]/M = 0.04011[AgNO3]/M = 0.03404
[Bu3N]/M = 0.03872[AgNO3]/M = 0.01885
[Bu3N]/M = 0.03938[AgNO3]/M = 0.01689
[Bu3N]/M = 0.03803[AgNO3]/M = 0.03384
[Bu3N]/M = 0.03803[AgNO3]/M = 0.03241
[Bu3N]/M = 0.03857[AgNO3]/M = 0.03267
[Bu3N]/M = 0.04032[AgNO3]/M = 0.02951
[Bu3N]/M = 0.03938[AgNO3]/M = 0.02870
[Bu3N]/M = 0.03884[AgNO3]/M = 0.02347
[Bu3N]/M = 0.03938[AgNO3]/M = 0.01566
0.03368
0.03781
0.03365
0.01818
0.01627
0.03356
0.03210
0.03238
0.02897
0.02819
0.02310
0.01508
EtI
EtI
EtI
EtI
EtI
EtI
EtI
BuI
BuI
BuI
BuI
BuI
0.04135
0.03990
0.04135
0.04007
0.04007
0.15708
0.15708
0.03944
0.03926
0.03911
0.03911
0.03911
0.00
0.75
1.50
0.00
1.52
0.00
1.56
0.00
0.74
1.58
0.00
1.54
a
Experimental data set presented previously [19], and reanalyzed according to the model formerly extended [20].
with tac signaling the onset of the reaction period where the
∼
free amine concentration remains constant ([Bu N] [Bu N]
=
3
3
5%), a is the corresponding (t = tac) alkyl iodide concentration,
e = [Bu3NAgNO3]t − ꢀ[AgNO3]tac and Q2 includes the underly-
ing dependenciesaoc n the catalyst mass and adsorption coefficients
(ϕ = bBu N/bRI).
3
kQꢀ asm
·
V
bBu
cmono
cmono
Bu NAgNO
RI
3
NAgNO
kC
3
3
3
Q2 = kQ
+
+
K1[Bu3N]free
(1/bRI) + [RI]t + ϕ[Bu3N]free
ac
(3)
Likewise, the integrated rate equation for the overall amine con-
sumption, valid within the same period, t ≥ tac, is
ꢀ
ꢁ
ꢀ
ꢁ
c − ꢀx1(t)
g − x1(t)
c − ꢀx1(tac
g − x1(tac
)
)
ln
= ln
− kQ (ꢀg − c)Q1(t − tac
)
(4)
where c is the concentration of coordinated amine at tac, g is [RI]t
−
Fig. 1. Plots of Eqs. (2) and (4) for kinetic runs 23 and 30.
ac
[AgNO3]t and Q1, is a coefficient containing exclusively parame-
ac
ters dependent on the quaternization reaction namely kQ , kꢀ and
Q
3.2.1. Side reaction between soluble silver nitrate and alkyl
halides
m
kQꢀ asm
V
bBu
cmono
cmono
Bu NAgNO
RI
3
NAgNO
Silver activated quaternization is accompanied by a competitive
side reaction of the alkyl halide with free silver nitrate. Estimates
of kC for the reaction of free silver nitrate with ethyl iodide and
butyl iodide in toluene were obtained independently, resorting
to two different approaches. The first method consists in the use
of expressions kC = K1[Bu3N](Q2 − Q1) or kC = K1[Bu3N](Q4 − Q3),
obtained resorting to Eqs. (3) and (5). The second methodol-
ogy involves extrapolation to zero mass of Q2 and Q1 or Q4 and
3
3
3
Q2 = kQ
+
·
(5)
(1/bRI) + [RI]t + ϕ[Bu3N]free
ac
An identical reasoning, for experimental runs conducted under
an excess of alkyl halide, leads to coefficients Q3 and Q4 which have
been presented previously [20].
Plots of the left hand side of Eq. (2) and (4) versus (t − tac) show
linear behavior, over extended time periods, with high correlation
coefficients(r2 > 0.99), substantiating the model and allowing the
determination of Qn coefficients under various experimental con-
ditions.
Q3 where Q1(m→0) = Q3(m→0) = kQ, thus kC = K1[Bu3N](Q2n(m→0)
−
Q2n−1(m→0)), see Eqs. (3) and (5) and Figs. 2 and 3.
3.2. Evaluation of kinetic and adsorption parameters
Fig. 1 presents the plots obtained for runs 23 and 30 of the reac-
tion between silver coordinated tributylamine with ethyl iodide
and butyl iodide. The same behavior was observed for all other
runs, substantiating the model.
The experimental coefficients Q1 and Q2, as well as the cor-
responding coefficients Q3 and Q4 obtained in the presence of
an excess of alkyl halide [20], are presented in Table 2. All data
was further analyzed in terms of subsets, obtained under identi-
cal experimental conditions. The application of the kinetic model
enables the calculation of the rate constants for the solution quater-
nization (kQ) and competitive reactions (kC), as well as the estimate
of surface term kꢀ bBu
associated with the surface catalyzed
NAgNO
3
quaternization (Table 3 and3Figs. 2 and 3).
Q
Fig. 2. . Dependence of Q2 and Q4 (runs 20–26 and 27–31) on the average mass of
solid during the reaction period under study (t ≥ tac).

M. Soledade C.S. Santos, E.F.G. Barbosa / Journal of Molecular Catalysis A: Chemical 356 (2012) 106–113
109
Table 2
Calculated Q1, Q2, Q3 and Q4 coefficients as well as the corresponding free amine concentration for the reaction period under study (t ≥ tac).
Run
tac (s)
[Bu3N] ꢁ (M)
[RI] ꢁ (M)
Q1 ꢁ (M−1 s−1
)
Q2 ꢁ (M−1 s−1
)
Q3 ꢁ (M−1 s−1
)
Q4 ꢁ (M−1 s−1
)
20
21
22
23
24
25
26
27
28
29
30
31
3373
5032
1364
841
1431
1470
1348
5332
2821
1615
1770
1081
0.028 0.002
0.024 0.001
0.028 0.001
0.032 0.001
0.034 0.001
0.0317 0.0008
0.033 0.001
0.0241 0.0009
0.027 0.001
0.023 0.001
0.029 0.0005
0.032 0.002
0.0165 0.0003
0.0145 0.0003
0.0194 0.0003
0.0270 0.003
0.0278 0.0003
0.1261 0.0003
0.1281 0.0003
0.0198 0.0003
0.0204 0.0003
0.0232 0.0003
0.0242 0.0003
0.0327 0.0003
0.00176 0.00007
0.00223 0.00006
0.00244 0.0001
0.00158 0.00008
0.00211 0.00006
0.0158 0.0002
0.0163 0.001
0.0181 0.0003
0.0155 0.0003
0.0178 0.0008
0.0014 0.0003
0.0019 0.0004
0.0145 0.0006
0.0166 0.0004
0.00059 0.00004
0.00071 0.00006
0.00092 0.00003
0.00058 0.00002
0.00089 0.00004
0.0062 0.0001
0.006 0.001
0.0068 0.0002
0.0057 0.003
0.0061 0.0005
The kC values contained in Table 3 are all in agreement
within the experimental error. Nonetheless a closer look at both
approaches shows the latter one leads to more reliable estimates,
an outcome expected in view of the multiple experimental runs
required by this procedure. Hence, from this set of kinetic runs,
the proposed values for kc are 2.8 0.1 × 103 mol−1 dm3 s−1 and
1.02 0.07 × 103 mol−1 dm3 s−1 for the reactions of AgNO3 + EtI,
and AgNO3 + BuI, in toluene, respectively.
fit, as all other quantities involved in the ratio are known ([RI]t
,
ac
[Bu3N] and bBu N) or related (ϕ = bBu N/bRI).
3
3
Â
EtI/[EtI]tac (excess halide)
∼
=
Constant
(6)
1/((1/bEtI) + [EtI]t + ϕ[Bu3N])
ac
(1/bBuI) + [BuI]t + ϕ[Bu3N]
ꢀ
ac
∼
Constant
(6ꢀ)
and
=
1/((1/bBuI) + [BuI]t + ϕ[Bu3N])
ac
These independent evaluations result in bEtI = 19
4
3.2.2. Solution and surface catalyzed quaternization of silver (I)
coordinated tributylamine by alkyl halides
22 < bBuI <30, confirming our previous educated guesses for both
halides [21].
Solution quaternization reaction rate constants, kQ, for trib-
utylamine with ethyl iodide and butyl iodide were determined
by extrapolation to m = 0, Fig. 3, of coefficients Q1 and Q3,
presented in Table 3. Best values, obtained by the weighted
average of the extrapolated values are presented in Table 4,
being 1.54 0.06 × 10−3 mol−1 dm3 s−1 for Bu3N–AgNO3 + EtI and
5.3 0.2 × 10−4 mol−1 dm3 s−1 for Bu3N–AgNO3 + BuI.
These estimates allow the calculation of “best values” for
kQꢀ bBu
from the slopes of Eqs. (3) and (5), consid-
NAgNO
3
3
ering as = 1.08 m2 g−1
,
cmono
= 3.1 × 10−6 mol m-2 [26] and
Bu
N
3
bEtI = 19 and bBuI = 22, and lead to 1.6 × 106 m2 mol−1 s−1 and
6.7 × 105 m2 mol−1 s−1 for the surface quaternization with ethyl
iodide and butyl iodide respectively, values which are included in
Table 4.
The surface contribution to the quaternization reaction is evi-
denced by the mass dependent plots (Figs. 2 and 3) of Q1, Q2, Q3
and Q4, in accordance with Eqs. (3) and (5), where the dependency
on the various adsorption quantities namely monolayer cover-
age, cmono, and adsorption coefficients, bi, for all species involved
(i = Et3N, RI and Et3NAgNO3) is explicit. For both amines adsorption
coefficients and monolayer coverage were previously determined
in our laboratory using the immersion technique. For the alkyl
halides and the silver amine coordination compounds no data
is available in the literature in this solvent; however as previ-
ously demonstrated [21], an adequate selection of experimental
conditions provides indirect estimates of the halide adsorption
coefficients magnitude (Eq. (6)). In the studies, involving Bu3N and
BuI, Eq. (6ꢀ), a general form of Eq. (6), was applied. The slope ratios
of Qn coefficients, namely Q3/Q1 or Q1/Q1, associated with various
4. Extended discussion of joint homogeneous and
heterogeneous processes
The joint homogeneous–heterogeneous silver activated amine
quaternization by alkyl halides is complex, involving various
adsorption and solution equilibrium, as well as several kinetic steps
on the surface and in solution. Having in mind a better understand-
ing of the overall process, data obtained in this work was combined
with previous studies on solution and surface quaternization of
reactions Et3N–AgNO3 + EtI and Et3N–AgNO3 + BuI [19–21].
4.1. The side reaction between free silver nitrate and alkyl halides
The reaction between free silver nitrate and alkyl halides
is always present while studying amine quarternization. The
rate constants for the side reaction, determined here, for
the tributylamine system, were 2.8 0.1 × 103 mol−1 dm3 s−1 for
EtI and 10.2 0.7 × 102 mol−1 dm3 s−1 for BuI. Larger values
have been reported earlier [21], for the same reactions, in
the presence of triethylamine, 6.4 0.6 × 103 mol−1 dm3 s−1 and
concentrations ([RI]t , [Bu3N]), allow an estimate of bRI by iterative
ac
5.5 0.2 × 103 mol−1 dm3 s−1
.
The rate constants for the side reactions between silver nitrate
and alkyl halides, determined in the presence of triethylamine are
much larger than those reported here for the same reaction. These
differences may be assigned to the unaccounted contributions of
the corresponding surface catalyzed reactions, kCꢀ , in the mathe-
matical development of the model. Nonetheless disregarding this
contribution has no effect on the parameters [kQ , kꢀ ] for the quat-
Q
ernization reactions.
The most reliable estimates of kC will correspond to
larger adsorption competition between free amine and
Fig. 3. Dependence of Q1 and Q3 (runs 20–26 and 27–31) on the average mass of
solid during the reaction period under study (t ≥ tac).
a

110
M. Soledade C.S. Santos, E.F.G. Barbosa / Journal of Molecular Catalysis A: Chemical 356 (2012) 106–113
alkyl halide, tributylamine rather than triethylamine data
meet this condition. Consequently the best estimates for
the solution rate constants of the side reactions AgNO3 + RI,
included in Table 4, are 2.8 0.1 × 103 mol−1 dm3 s−1 for EtI and
10.2 0.7 × 102 mol−1 dm3 s−1 for BuI.
Another interesting point is signaled while comparing the var-
ious rate constants for the side reactions. Larger decreases on kC
values, for the reactions with butyl iodide, indicate a bigger adsorp-
tion coefficient for this halide, a conclusion in agreement with the
one obtained previously resorting to an independent approach [21].
4.2. Solution and surface catalyzed quaternization of silver (I)
coordinated tertiary amines by alkyl halides
Values previously obtained for the solution and surface quater-
nization reactions Et3N–AgNO3 + EtI and Et3N–AgNO3 + BuI [20,21]
were also included in Table 4 to allow a full comparison. These
results show silver coordination significantly lowers the activa-
tion energy for solution amine quaternization, resulting in rate
constants increases of 103 times. Moreover in the presence of sol-
ubilized silver nitrate a reduction of 2.2 in the rate constants, for
tertiary amine quaternization by EtI [19], on going from triethy-
lamine to tributylamine, is observed against a 5.3 times reduction
in the absence of silver indicating a significant decrease on the tran-
sition state stereochemical constraints (Tables 4 and 5). The role
of silver as a carbon halogen bond cleavage activator is also evi-
denced by the smaller decrease on rate constants ratios (from 5.8
to 2.8) for the reaction of Et3N + RI, on going from EtI to BuI, in the
presence of silver nitrate (Table 4).
An overall analysis of the solution quaternization reactions in
the presence of solubilized silver nitrate evidences a slightly larger
decrease in the quaternization rate constants, associated with a
two carbon increase in the halide chain length (2.8–2.9), than with
the same chain length increase on the tertiary amine (2.2–2.4).
Data indicating that, in the homogeneous pathway, the role of sol-
ubilized silver nitrate is more significant in terms of minimizing
transition state stereochemical constraints than as a carbon halo-
gen bond electrophilic activator.
In terms of the surface reaction, silver–amine coordination is
crucial for the amine attack on a next neighbor halide. In fact quater-
nization by uncoordinated amine leads to inhibition in the presence
of solids [30], while silver coordination accomplishes the lifting of
the amine lone electron pair one atomic layer away from the sur-
face, rendering adequate stereochemical positioning for the surface
reaction.
As before, a thorough study of the surface catalyzed quaterniza-
tion reactions analysis can be made resorting to herein reported
and previously published data [20,21]. The values for kQꢀ bEt
NAgNO
3
were recalculated, in order to optimize comparisons between dif3-
ferent experimental data sets. The same adsorption coefficients for
the alkyl halides, namely bEtI = 19 and bBuI = 22, were considered
and the final values compiled in Table 4
The decrease in the experimental ratios associated with an
increase of the alkyl halide chain length are larger for the quat-
ernization of coordinated triethylamine (3.2) than for coordinated
tributylamine (2.5) indicating smaller stereochemical constraints
on the latter amine (Table 4). This apparently inconsistent result
is in accordance with previously predicted minor deviations from
spherical towards barrel type geometry for adsorbed tributylamine
[29].
The relationships between reactions of the same alkyl
halide with either amine are not straightforward as the ratios
kQꢀ bEt
/kQꢀ bBu
entail products of surface reactions
NAgNO
NAgNO
3
3
3
3
rates by adsorption coefficients. Surface rate constants, kQꢀ , could
not be explicitly obtained, due to the unavailability of adsorption

M. Soledade C.S. Santos, E.F.G. Barbosa / Journal of Molecular Catalysis A: Chemical 356 (2012) 106–113
111
Table 4
Rate constants for all reactions associated with amine quaternization in the presence and absence of solubilized silver nitrate and consequences of halide alkyl chain length
increase on all processes.
R3N + RI
R3N–AgNO3 + RI
kC (M−1 s−1
ꢀ
k
(EtI)
k
k
(EtI)
k
k
(EtI)
(BuI)
k
(EtI)
C
(BuI)
Q
Q
kM (M−1 s−1
)
)
kQ (M−1 s−1
)
kQꢀ bR
(m2 mol−1 s−1
)
NAgNO
3
M
ꢀ
3
k
(BuI)
k
(BuI)
M
C
Q
Q
Et3N + EtI
Et3N + BuI
Bu3N + EtI
Bu3N + BuI
3.3 × 10−6
5.6 × 10−7
6.2 × 10−7
–
6.4 × 103
5.5 × 103
6 × 102
2 × 102
3.46 × 10−3
1.244 × 10−3
1.54 × 10−3
5.3 × 10−4
2 × 10−4
2 × 10−6
6 × 10−5
2 × 10−5
5.0 × 106
1.5 × 106
1.6 × 106
6.7 × 105
1.2 0.2
2.8 0.3
2.8 0.2
2.9 0.3
5.8
–
3.2
2.5
2.8 × 103 1 × 102
1.0 × 103 7 × 101
Table 5
Effect of trialkylamine chain length on the rate constants for all reactions associated with amine quaternization in the presence of solubilized silver nitrate, and comparative
analysis of the results obtained by the strategies used to estimate bEt
3 /bBu
NAgNO
.
NAgNO
3
3
3
k'QbR NAgNO
k'Q Et3N
k'Q Bu3N
k'QbR NAgNO hmono
kQ
kQ' bEt
kQ' bBu
bEt NAgNO
bEt NAgNO
3 3
k
kQ (Et N)
3
C (Et N)
3
3
3
NAgNO
hmono
3
3
3
3
R3N-AgNO3+RI
bBu NAgNO
bBu NAgNO
k
C (Bu3 N)
(Bu3 N)
kQ
kQ
3
3
3
3
NAgNO
/103m-1
/m
1.43×109 a
1.09×109 a
1.43×109 a
1.09×109 a
9.54×10-10
1.4×103 b
1.2 ×103 b
1.4 ×103 b
1.2 ×103 b
Et3NAgNO3+EtI
Bu3NAgNO3+EtI
Et3NAgNO3+BuI
Bu3NAgNO3+BuI
2.3 0.3
5.4 0.8
2.2 0.2
2.4 0.1
3.1
2.3
1.8
1.10×10-9
1.11×10-9
1.11×10-9
1.24
1.31
2.3
a
See Fig. 4.
See Fig. 5.
b
coefficients for silver–amine complexes, bR
3 . Nevertheless
approximated by the quotient of the second stepwise forma-
tion constants for the corresponding silver–amine complexes
NAgNO
3
the ratio bEt
3 /bBu
NAgNO
3 , can be indirectly attained following
NAgNO
3
two different approach3es:
in solution K2(Et N)/K2(Bu
= 1.59 [31].
N)
3
3
(i) Plots of kꢀ bR
versus kQ were drawn for experiments of
NAgNO
the sameQam3ine an3d variable alkyl halide (Fig. 4), and the zero
intercepts allowed the calculation of 1.31 for the ratio of the
silver–amine complexes adsorption coefficients.
The higher basicity of triethylamine in toluene [29,31] is a
premise setting 1.59 as an upper limit for the corresponding ratio
of adsorption coefficients, a hypothesis in conformity with the pre-
ceding guesstimate, 1.31, which will be used from here on in this
discussion. Once a reliable estimate for the ratio of the silver amine
complexes adsorption coefficients is obtained, one may calculate
the ratio of the surface reaction rates involving the same halide and
different amines. Taking 1.31 as a trustworthy ratio 2.3 is obtained
for the ratio of the quaternization reactions of silver coordinated
(ii) Another independent reasoning, focused on the silver coordi-
nation sphere, was also pursued. Within this framework the
parallelism between the exchange of an electron donating sol-
vent molecule by a second amine molecule in solution, and the
adsorption of the coordinated amine, involving the substitu-
tion of a solvent molecule by a surface iodide, was examined.
amines (kꢀ bEt
/kQꢀ bBu
3 ) with ethyl iodide and 1.8 for
NAgNO
NAgNO
3
3
the corresQponding reactions3with butyl iodide, values which were
This assumption means the ratio bEt
3 /bBu
NAgNO
may be
NAgNO
3
3
3
included in Table 5.
A comparison between the ratios of solution and surface reac-
tions evidences that surface packing constraints are efficiently
reduced (2.2–2.3) or even surpassed (2.4–1.8) by the concomitant
carbon halogen bond activation of silver ions, either solubilized or
on the solid surface.
These results demonstrate the decrease on the surface quat-
ernization reaction rates, with increasing chain length, is less
accentuated for changes in alkyl halide chain length thus evidenc-
ing that apart from the usefulness of silver coordination in terms
of amine positioning, silver activation of the carbon–halide bond
stands as the driving force in the surface catalyzed pathway. This
reinforced electrophilic effect on surface pathways has its costs in
terms of the parallel reaction, which is noteworthy for the longer
alkyl halides, and entails a reduction of overall quaternization reac-
tion yield.
The pursuit of
a
broader understanding of these
homogeneous–heterogeneous systems led to a molecular level
inspection of the surface monolayer where the surface reaction
occurs. The surface monolayer thickness, hmono, is determined
by the nature, size and molecular geometry of the adsorbed
entities involved in the reaction. The thickness of a monolayer
Fig. 4. Dependence of kQꢀ bR
alkylamines.
on kQ for the quaternization of coordinated tri-
NAgNO
3
3

112
M. Soledade C.S. Santos, E.F.G. Barbosa / Journal of Molecular Catalysis A: Chemical 356 (2012) 106–113
5. Conclusions
This is the last of a series of studies for a homo/heterogeneous
processes, developed in our laboratory, on the quaternization of
tertiary amines by alkyl iodides in toluene. In particular, the role of
added silver ions in solution and solid AgI was investigated for this
model system.
The new data presented here confirm that the multiple step
model, built in preceding work, its mathematical description and
development, with a careful analysis of errors, generates consis-
tent and coherent results. The combination of different parameters,
obtained for all systems explored, establishes silver ions act as
structural promoter, minimizing stereochemical constraints, and
as carbon halogen bond cleavage activators. These conclusions are
consistent with a balance dominated by structural factors in the
solution process which are reinforced and exceeded by electronic
effects in the surface pathway.
The perfect parallelism between homogeneous and heteroge-
neous silver (I) activated quaternization of tertiary amines by alkyl
iodides, in this model system, supported a molecular level analysis
of the heterogeneous process. This strategy allowed the identifi-
Fig. 5. Relation between kQꢀ bEt
3 hmono and kQ for the quaternization of coordi-
NAgNO
nated trialkylamines consideri3ng bEt
3 hEt
NAgNO
3 /bBu
NAgNO
3 hBu
NAgNO
3 3
= 1.16.
NAgNO
3
3
3
cation of a surface compartment of monolayer thickness, hmono
,
of adsorbed alkyl halide was approximated by the length of a
◦
determined by size, shape and orientation of the chemical species
on the surface. “Volumetric surface rate constants” were estimated
and directly compared with their solution counterparts, provid-
ing a clear-cut link between the homogeneous and heterogeneous
processes. Finally, it is believed this model system and molecu-
lar analysis is a contribution for future approaches of complex
homo/heterogeneous systems.
∼
=
linear hydrocarbon lCH
1.57 + 1.27 n A plus a covalent
(CH
)
n
3
2
iodide diameter [32,33] rendering hmonoEtI = 8.51 × 10−10 m and
hmonoBuI = 1.11 × 10−9 m. For coordinated amines, adsorbed
on AgI, monolayer thicknesses were calculated resorting
to silver ionic diameter [32] and to the adsorbed tertiary
amine diameters, obtained from experimental molar volumes
and considering
molecules over
a
hexagonal close-packed arrangement of
bidimensional surface [29,32], leading to
a
Acknowledgments
monolayer thicknesses of hmonoEt
= 9.54 × 10−10 m and
NAgNO
3
3
hmonoBu
= 1.10 × 10−9 m.
NAgNO
3
3
This work was funded by National Funds (PEst-
OE/QUI/UI0612/2011) through FCT – Fundac¸ ão para a Ciência
e a Tecnologia.
These quantities consent the transformation of surface rate
constants, kQꢀ , into “volumetric surface rate constants” (a new desig-
nation, found appropriate for this quantity) for the reaction within
a tridimensional space on the surface, enabling a clear cut identifi-
cation of the surface role.
References
Plots of kꢀ bR
3 hmono versus kQ, presented in Fig. 4, dis-
NAgNO
3
Q
[1] K.I. Zamaraev, in: J.M. Thomas, K.I. Zamaraev (Eds.), Perspectives in Catalysis,
Blackwell Scientific Publications, London, 1992, pp. 35–37.
[2] G.A. Somorjai, M.C. Yang, Top. Catal. 24 (2003) 61–72.
[3] M. Yoshizawa, Y. Takeyama, T. Okano, M. Fujita, J. Am. Chem. Soc. 125 (2003)
3243–3247.
play different slopes, dependent on the adsorption coefficients of
the tertiary amine silver complexes, bR 3 . All the data may
NAgNO
3
be drawn to a common line (Fig. 5) upon the introduction of the
constant factor 1.16, thus enabling the calculation of 1.24 for the
ratio between adsorption coefficients, a value providing additional
validation for 1.31 previously proposed in (i).
The slope of the superimposed linear plots, presented in Fig. 5,
evidences clearly the catalytic effect of the AgI surface on the silver
coordinated quaternization of tertiary amines reaction. However
the precise value for the “volumetric surface rate constants” cannot
[4] J.J. Davis, Chem. Commun. 28 (2005) 3509–3513.
[5] M.P. Schramm, R.J. Hooley, J. Rebek Jr, J. Am. Chem. Soc. 129 (2007) 9773–9779.
[6] G.A. Somorjai, J.Y. Park, Angew. Chem. Int. Ed. 47 (2008) 9212–9228.
[7] G.A. Somorjai,.H. Frei, J.Y. Park, J. Am. Chem. Soc. 131 (2009) 16589–16605.
[8] A. Vazquez-Zavala, J. Garcia-Gomez, A. Gomez-Cortes, Appl. Surf. Sci. 167
(2000) 177–183.
[9] B.L. Moroz, V.A. Semikolenov, V.A. Likholobov, Y.I. Yermakov, J. Chem. Soc.
Chem. Commun. (1982) 1286–1287.
[10] K.H. Park, S. Uk Son, Y.K. Chung, Chem. Commun. (2003) 1898–1899.
[11] J. Guzman, B.C. Gates, Dalton Trans. (2003) 3303–3318.
[12] V. Ramamurthy, J. Shailaja, L.S. Kaanumalle, R.B. Sunoj, J. Chandrasekhar, Chem.
Commun. (2003) 1987–1999.
be determined as values for bR
are not available. Never-
NAgNO
3
theless, bearing in mind amine3electron donation will lead to a
reduction in the adsorption coefficient reported for silver nitrate
in water [34], and also considering the magnitude estimated here
for alkyl halides adsorption coefficients, a broad boundary may be
set, consenting the proposal of a tenth magnitude for the silver
amine complex adsorption coefficient on silver iodide in toluene.
Accordingly one may conclude that “volumetric surface rate con-
stants” are 102 times higher than the corresponding solution rate
constants.
[13] M.S. Masar, N.C. Gianneschi, C.G. Oliveri, C.L. Stern, S.T. Nguyen, C.A. Mirkin, J.
Am. Chem. Soc. 129 (2007) 10149–10158.
[14] R.J. Hooley, J. Rebek Jr., J. Am. Chem. Soc. 127 (2005) 11904–11905.
[15] B.W. Purse, J. Rebek, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 10777–10782.
[16] B.W. Purse, A. Gissot, J. Rebeck Jr., J. Am. Chem. Soc. 127 (2005) 11222–11223.
[17] R.J. Hooley, H. Van Anda,.J. Rebek Jr., J. Am. Chem. Soc. 129 (2007) 13464–13473.
[18] K. Nakabayashi, M. Kawano, M. Yoshizawa, S. Ohkoshi, M. Fujita, J. Am. Chem.
Soc. 126 (2004) 16694–16695.
[19] M. Soledade C.S. Santos, Ester F.G. Barbosa, J. Mol. Catal. A: Chem. 160 (2000)
293–313.
This rate increase was observed with AgI, an extremely low sur-
face area solid (1.08 m2 g−1 [29]), chosen to ensure the presence of
Ag+ and I− surface sites, and warranting an identity between par-
allel solution and surface processes, thus enabling the comparison
between the two reaction pathways. In view of the results, obtained
for this model system, much larger rate enhancements are foreseen
for higher surface area solids.
[20] M. Soledade C.S. Santos, Ester F.G. Barbosa, J. Mol. Catal. A: Chem. 197 (2003)
73–90.
[21] M. Soledade C.S. Santos, Ester F.G. Barbosa, J. Mol. Catal. A: Chem. 306 (2009)
82–88.
[22] K.S. Cameron, J.R. Morphy, Z. Rankovic, M. York, J. Comb. Chem. 4 (2002)
199–203.
[23] J. Ropponen, M. Lahtinen, S. Busi, M. Nissinen, E. Kolehamainen, K. Rissanen,
New J. Chem. 28 (2004) 1426–1430.
[24] V. Calo, A. Nacci, A. Monopoli, A. Fanizzi, Org. Lett. 4 (2002) 2561–2563.

M. Soledade C.S. Santos, E.F.G. Barbosa / Journal of Molecular Catalysis A: Chemical 356 (2012) 106–113
113
[25] D.D. Perrin, W.L.F. Armarego, D.R. Perrin, Purification of Laboratory Chemicals,
2nd Ed., Pergamon Press, Oxford, 1982, pp. 165.
[30] E.F.G. Barbosa, M. Spiro, J. Chem. Soc. Chem. Commun. (1977)
423–424.
[26] E.F.G. Barbosa, I.M.S. Lampreia, Can. J. Chem. 64 (1986) 387–393.
[27] H.J. van den Hul, J. Lyklema, J. Am. Chem. Soc. 90 (1968) 3010–3015.
[28] E.W. Sidebottom, W.A. House, M.J. Jaycock, J. Chem. Soc. Faraday Trans. 72
(1976) 2709–2721.
[31] M. Soledade C.S. Santos, E.F.G. Barbosa, M. Spiro, J. Chem. Soc. Faraday Trans.
84 (1988) 4439–4449.
[32] R.D. Shannon, Acta Crystallogr. A32 (1976) 751–767.
[33] A.Y. Meyer, D. Farin, D. Avnir, J. Am. Chem. Soc. 108 (1986) 7897–7905.
[34] J. Austin, O. Ibrahim, M. Spiro, J. Chem. Soc. 6 (1969) 669–674.
[29] M.S.C.S. Santos, E.F.G. Barbosa, J. Phys. Chem. B 102 (1998) 6040–6048.