Inductive effects through Alkyl Groups - How long is long enough?

Article abstract of DOI:10.1002/ejoc.201300486

The influence of the length of alkyl substituents on various kinetic, thermodynamic, and spectroscopic properties of 4-(dialkylamino)pyridines has been determined by using a combination of experimental and theoretical methods. The chain-length dependence of these properties has subsequently been analyzed by using a quantitative model for through-bond inductive effects. Substituent effects of linear alkyl groups show a saturation effect beyond four carbon atoms for numerous properties of 4-(dialkylamino)pyridines as test substrates. These observations can be rationalized by using an increment-based method for the calculation of inductive substituent effects. Copyright

Full text of DOI:10.1002/ejoc.201300486

Job/Unit: O30486
/KAP1
Date: 16-07-13 10:53:07
Pages: 9
FULL PAPER
Lewis Base Catalysis
Substituent effects of linear alkyl groups
show a saturation effect beyond four car-
bon atoms for numerous properties of 4-
(dialkylamino)pyridines as test substrates.
These observations can be rationalized by
using an increment-based method for the
calculation of inductive substituent effects.
R. Tandon, T. A. Nigst,
H. Zipse* .......................................... 1–9
Inductive Effects through Alkyl Groups –
How Long is Long Enough?
Keywords: Through-bond interactions / Re-
action mechanisms / Lewis bases / Nitrogen
heterocycles / Kinetics
0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Job/Unit: O30486
/KAP1
Date: 16-07-13 10:53:07
Pages: 9
R. Tandon, T. A. Nigst, H. Zipse
FULL PAPER
DOI: 10.1002/ejoc.201300486
Inductive Effects through Alkyl Groups – How Long is Long Enough?
Raman Tandon,[a] Tobias A. Nigst,[a] and Hendrik Zipse*[a]
Keywords: Through-bond interactions / Reaction mechanisms / Lewis bases / Nitrogen heterocycles / Kinetics
The influence of the length of alkyl substituents on various
kinetic, thermodynamic, and spectroscopic properties of 4-
(dialkylamino)pyridines has been determined by using a
combination of experimental and theoretical methods. The
chain-length dependence of these properties has sub-
sequently been analyzed by using a quantitative model for
through-bond inductive effects.
pyridine-catalyzed acylations of alcohols, the benzoylation
Introduction
of pyridines, the reactions of N-benzoylpyridinium ions
The influence of alkyl groups of variable length on the
ground state properties of organic molecules and on their
reaction rates in unimolecular and bimolecular transforma-
tions has been discussed in terms of steric and electronic
substituent effects.[1] The latter may be divided further into
inductive (through-bond) and field (through-space) compo-
nents, however the strict separation of these components is
difficult. In recent efforts to develop nucleophilic catalysts
of enhanced reactivity and selectivity, we[2] and others[3] ex-
1
with benzylamine, on their H NMR properties, their theo-
retically calculated acetylation enthalpies, and on the elec-
tronic and geometric properties of the respective acylated
intermediates, to provide a quantitative answer.
Results
Aminopyridines 1a–h were synthesized by treatment of
plored the performance of alkyl-substituted aminopyridines 4-chloropyridine with the corresponding dialkylamines, fol-
of general formula 1, 2, and 3 in pyridine-catalyzed acyl- lowing established literature procedures (Scheme 1).[4]
ation reactions of tertiary alcohols (Figure 1). These com-
pounds are highly active nucleophilic catalysts with small
alkyl substituents such as R = methyl or ethyl, and the
question thus arises whether the catalytic efficiency can be
increased further by elongation of the alkyl groups.
Scheme 1. Synthesis of pyridine derivatives 1a–h.
The catalytic potential of these catalysts has been ex-
plored in the acetylation of tertiary alcohol 4 with acetic
anhydride as followed by 1H NMR spectroscopy
(Scheme 2). All reactions eventually proceed to full conver-
sion, and the rates of the reactions can thus be charac-
terized by the reaction half-life t1/2 using the approach de-
scribed previously (Table 1).[2b]
Figure 1. Structures of pyridine derivatives used as nucleophilic cat-
alysts.
In more general terms, this poses the question of the
chain-length dependence of inductive effects in alkyl sub-
stituents. We report here the effects of chain elongation of
the alkyl groups in 4-(alkylamino)pyridines on the rates of
In the benchmark reaction mentioned above, the cata-
lytic activity increases with the chain length of the alkyl
substituents on the amino group at the C4 position by a
factor of 2.6 from 4-(dimethylamino)pyridine (DMAP; 1a)
to 4-(di-n-octylamino)pyridine (1h), which has an activity
comparable to that of 4-pyrrolidinopyridine (PPY).[2d]
The best compromise between chain length and activity
is obtained with R = butyl, since elongation of the chain
from butyl to octyl does not lead to increased activities. The
[a] Department of Chemistry, LMU München,
Butenandtsrasse 5–13, 81377 München, Germany
E-mail: zipse@cup.uni-muenchen.de
Homepage: www.cup.uni-muenchen.de/oc/zipse
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201300486.
2
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O30486
/KAP1
Date: 16-07-13 10:53:07
Pages: 9
Inductive Effects through Alkyl Groups
pyridines 7. These latter intermediates are reacted in the
second model reaction with benzylamine 8. The kinetics of
both model reactions were followed by UV/Vis spectropho-
tometry under first-order conditions by monitoring the ab-
sorbances of the N-benzoylpyridinium chlorides 7a–h as de-
scribed previously.[5b,5c] For all reactions, monoexponential
increases or decays of the absorbances of 7a–h were ob-
served and the first-order rate constants kobs were obtained
by least-squares fitting of the exponential function A =
A0 (1 – e–k(obs)t) + C or A = A0 e–k(obs)t + C to the time-
dependent absorbances. Plots of kobs vs. the concentration
Scheme 2. Benchmark reaction monitored by 1H NMR spec-
troscopy in CDCl3.
Table 1. Half-lives (t1/2) for the pyridine-catalyzed reaction of 4
with acetic anhydride in CDCl3.
Catalyst
R
RЈ
t1/2 [min]
1a
1b
1c
1d
1e
1f
Me
Me
Et
nPr
nBu
nPent
nHex
nOct
Me
Et
Et
nPr
nBu
nPent
nHex
nOct
151.0Ϯ1.7[a]
110.3Ϯ0.2
74.9Ϯ0.6
65.5Ϯ0.4
58.7Ϯ1.6
60.8Ϯ0.3
61.0Ϯ0.2
59.1Ϯ0.5
1g
1h
[a] From ref.[2b]
results obtained for the unsymmetrically substituted pyr-
idine 1b blend rather well with those for the symmetrically
substituted aminopyridines, which indicates that the induc-
tive effects of the alkyl substituents are cumulative (Fig-
ure 2).
Scheme 3. Simplified mechanism for the DMAP-catalyzed acyl-
ation of alcohols.
Figure 2. Correlation of half-life times t1/2 (in min) with the length
of the N-alkyl chains of 1a–h.
N-Acylpyridinium ions have been proposed as the key
intermediates of the catalytic cycle for the acylation reac-
tions described above (Scheme 3).[5] In a first step, the nu-
cleophilic catalyst (here pyridine 1a) reacts with the acyl
donor in a fast pre-equilibrium to form N-acylpyridinium
ion II. In the rate-determining, second step the acylated cat-
alyst reacts with the alcohol to form ester IV together with
one equivalent of acid as a byproduct. To quantify the in-
fluence of the alkyl chain length on the first and the second
step of the catalytic cycle separately, reaction rates for the
two model reactions shown in Scheme 4 have been
measured. In the first of these reactions, pyridines 1 react
Scheme 4. Reaction of benzoyl chloride (6) with pyridines 1a–h and
aminolysis reactions of the formed N-benzoylpyridinium chlorides
with benzoyl chloride 6 to yield the respective benzoylated 7a–h with benzylamine (8) in acetonitrile.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3

Job/Unit: O30486
/KAP1
Date: 16-07-13 10:53:07
Pages: 9
R. Tandon, T. A. Nigst, H. Zipse
FULL PAPER
of the compound in excess (6 or 7) were linear and the sec-
It has been noted earlier that the stabilities of acylpyrid-
ond-order rate constants k2 (Table 2) were obtained as the inium ions serve as a quantitative indicator for the catalytic
slopes of these plots. The intercepts of the plots were close performance of the respective pyridines.[2b] Thus, relative
to zero, in line with the quantitative course of the reactions. acetylation enthalpies for 1a–h were calculated at the
MP2(FC)/6-31+G(2d,p)//B98/6-31G(d) level (in short:
Table 2. Second-order rate constants k2 for the reactions of benzoyl
chloride (6) with pyridines 1a–h and N-benzoylpyridinium chlor-
ides 7a–h with benzylamine (8) in acetonitrile at 20 °C.
MP2–5) according to the isodesmic reaction depicted in
Scheme 5.
1
k2(II) [m–1 –1
]
7
k2(III) [m–1 –1
]
1a
1b
1c
1e
1g
1h
3.00ϫ104 [a]
3.36ϫ104
4.06ϫ104
4.14ϫ104
4.13ϫ104
4.37ϫ104
7a
7b
7c
7e
7g
7h
4.78ϫ102 [b]
4.11ϫ102
3.78ϫ102
3.68ϫ102
3.53ϫ102
3.69ϫ102
[a] From ref.[5b] [b] From ref.[5c]
Scheme 5. Isodesmic reaction for the calculation of acetylation en-
thalpies of pyridines 1.
While the reactivities of pyridines 1a–h towards benzoyl
chloride (6) increase with the chain length of the alkyl
groups by a factor of 1.5, a decrease of the reactivities of
the N-benzoylpyridinium ions 7a–h by a factor of 1.3 was
observed. Similar to the pyridine-catalyzed acylation reac-
tion of 4, saturation of the inductive effects of the alkyl
groups are already observed at R = ethyl (Figure 3).
For catalysts 1a–d all possible conformers within a range
of 30 kJ/mol were taken into account. Conformers for cat-
alysts 1e–h were generated from those for 1d by adding the
required number of terminal CH2 groups in an all-trans
conformation (Figure 4). The Boltzmann distribution was
used to calculate the statistical weight of the individual con-
formers and averaged energies were used for the calculation
of the acetylation enthalpies (ΔH298,ave). Additionally, acet-
ylation enthalpies were computed by using only the energet-
ically best conformers of 1a–h together with the best con-
formers of acylpyridinium ion 10a–h (ΔH298,best), or those
conformers of 10a–h that match the best conformation of
the parent pyridines 1a–h (ΔH298,sel). How much these latter
data depend on the theoretical level was tested by calculat-
ing acylation enthalpies at the MP2–5 and also at the more
expensive G3(MP2)B3 level of theory (see the Supporting
Information). Table 3 shows a summary of the calculated
reaction enthalpies for catalysts 1a–h for the Boltzmann-
averaged conformers (second and third column) and the
best conformers (fourth and fifth column).
Figure 4. Best conformer of 1d from front view (left) and top view
(right).
The plots of Boltzmann-averaged and best acetylation
enthalpies vs. the number of C-atoms in the alkyl substitu-
ents in pyridines 1a–h show a steady increase in acylation
enthalpies with increasing alkyl chain length, which do not
become saturated even for octyl substituents (Figure 5).
Due to the very large number of conformers in 1e–h, a
significant part of the conformational space might not be
covered due to the restrictions made for the generation of
conformers. We note in passing that the trend observed here
Figure 3. Correlation of the second-order rate constants k2 for the
reactions of 1a–h with benzoyl chloride (6) with the length of the
N-alkyl chains of 1a–h (top) and that of the reactions of 7a–h with
benzylamine (8) with the length of the N-alkyl chains in 7a–h (bot-
tom).
4
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O30486
/KAP1
Date: 16-07-13 10:53:07
Pages: 9
Inductive Effects through Alkyl Groups
Table 3. Acetylation enthalpies for pyridines 1a–h according to
Scheme 5 for the Boltzmann-averaged and best conformers calcu-
lated at the MP2–5 level of theory in the gas phase and in CHCl3
solution.
Table 4. Acetylation enthalpies in CHCl3 for pyridines 1a–h calcu-
lated at the MP2–5 level according to Scheme 5 including only the
electrostatic (second column) or only the nonelectrostatic (third
column) component of the solvation energy.
Cat.
ΔH298,ave [kJ/mol]
ΔH298,best [kJ/mol]
Cat.
ΔH298,best [kJ/mol]
Nonelectrostatic only
Gas phase
CHCl3
Gas phase
CHCl3
Electrostatic only
1a
1b
1c
1d
1e
1f
–77.1
–81.6
–87.6
–90.1
–93.2
–94.1
–94.8
–96.7
–61.2
–63.3
–67.1
–66.1
–67.9
–67.8
–68.1
–68.5
–77.1
–81.9
–88.2
–90.6
–92.2
–93.2
–94.2
–95.0
–61.2
–63.6
–67.8
–66.1
–67.1
–67.6
–67.4
–68.2
1a
1b
1c
1d
1e
1f
–61.3
–63.0
–63.4
–65.5
–67.2
–67.3
–67.6
–67.5
–77.1
–81.2
–83.9
–88.9
–92.0
–93.1
–94.1
–94.9
1g
1h
1g
1h
Figure 6. Acetylation enthalpies in CHCl3 for pyridines 1a–h calcu-
lated at the MP2–5 level according to Scheme 5 including only the
electrostatic (squares) or only the nonelectrostatic (circles) compo-
nent of the solvation energy.
Figure 5. Correlation of acylation enthalpies in the gas phase
(circles) and CHCl3 solution (squares) for the isodesmic reaction
in Scheme 5 with the length of the N-alkyl chain of 1a–h (open
symbols: ΔH298,ave; closed symbols: ΔH298,best).
for selected conformers at the MP2–5 level is also found
Analysis of the structures of the best conformers of the
when using significantly more accurate G3(MP2)B3 ener- acetylated pyridines shows that inductive effects also impact
gies (see the Supporting Information). Recalculation of rel- the length of the amide bond, that is, the distance between
ative acylation enthalpies in the presence of the PCM con- the pyridine nitrogen and the carbonyl carbon atom (see
tinuum model for CHCl3 leads to significantly reduced acyl- the Supporting Information). Even though these changes
ation enthalpies for all pyridines studied here (Figure 5). are rather moderate in absolute terms, there is nevertheless
Moreover, the acylation enthalpies now level out at around a clear trend towards shorter bond lengths for increasingly
–68 kJ/mol starting with an alkyl chain length of n = 2, Lewis basic pyridines (Figure 7).
which is largely similar to the trend observed in the catalysis
experiments. This latter trend is found in an almost iden- trogen atoms in 10a–h shows that the charge varies with
tical manner for both ΔH298,ave and ΔH298,best the size of the alkyl substituents in the expected manner,
A closer look at the partial charge of the pyridinium ni-
.
To further trace the origin of this apparent “shielding” assuming slightly more negative values with longer alkyl
of the PCM solvent model on the calculated acylation en- substituents. The magnitude of these variations is, however,
thalpies, the solvation energies were recalculated by only very small and does not lend itself to quantitative analysis
considering the electrostatic or only the nonelectrostatic (for further details see the Supporting Information).
component of the overall solvation free energies (Table 4).
The above mentioned metrics test the influence of induc-
In the former case (electrostatics only) the calculated acyl- tive effects on the pyridine ring and its thermodynamic and
ation enthalpies in CHCl3 solution are numerically quite kinetic properties. How much of the inductive effect arrives
close to the data obtained for the full PCM model and dis- at the other (alkyl) terminus of the respective systems can
play a saturation effect beyond R = butyl, whereas no such be quantitatively assessed by using the 1H NMR resonances
observation can be made when using only the nonelectro- of the terminal methyl group protons. Figure 8 shows that
1
static component of the solvation energy (Figure 6). This the H NMR chemical shift of the terminal CH3 groups in
implies that the electrostatic component of the solvation en- pyridines 1a–h represents a rather sensitive probe for the
ergy is essential to match the experimentally observed pyridine substitution pattern. The reason for the observed
trends.
changes is primarily rooted in the distance between the
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5

Job/Unit: O30486
/KAP1
Date: 16-07-13 10:53:07
Pages: 9
R. Tandon, T. A. Nigst, H. Zipse
FULL PAPER
the quantitative expression [Equation (1)] for the calcula-
tion of σ*, the magnitude of the substituent effect depends
on the distance between each center in the substituent and
a point of reference ri, the covalent radius of the respective
atomic center Ri, and the electronegativity difference be-
tween the respective center and the point of reference Δχi.
(1)
The proper topological “point of reference” for the sys-
tems under study is undoubtedly the pyridine nitrogen atom
as the reaction center. Calculating σ* substituent param-
eters with respect to this center will be referred to as “model
1” in the following (Figure 9). Relay of substituent effects
between the nitrogen attached at the C4 position and the
pyridine nitrogen atom is, however, not well described by
using an inductive through-bond mechanism due to the
presence of the common π system. An alternative “model
2” is therefore explored in which the point of reference is
assumed to be the nitrogen atom holding the two alkyl sub-
stituents. Finally, a third model may be necessary to ratio-
nalize the results of the 1H NMR studies, in which the
points of reference are the hydrogen atoms of the terminal
CH3 groups (model 3).
Figure 7. Correlation of the N–C bond length r(N–C) [pm] of the
best conformers of 10a–h with the length of the N-alkyl chains in
10a–h.
probe nucleus (terminal hydrogen atom) and the signifi-
cantly more electronegative nitrogen atom to which the
alkyl substituent attaches.
Figure 9. Overview of models 1–3 used for the calculation of σ*
values for 1a–h according to Equation (1). The encircled atom illus-
trates the respective “center of reference” for each model.
Figure 8. Correlation of 1H NMR shifts of terminal CH3 groups
δ(1H,CH3) with the length of the N-alkyl chains in 1a–h.
The calculated σ* values (Table 5) can subsequently be
used together with a Hammett expression [Equation (2)] to
quantitatively describe the influence of alkyl chain length
on experimentally observed or theoretically calculated data.
Discussion
The impact of inductive effects in general[1] and espe-
cially that of alkyl groups[6] is not yet fully understood and
is thus still part of an ongoing discussion. This may be due
to the fact that the inductive effects of alkyl groups are
often considered to be rather small,[7] despite several exam-
ples that indicate the opposite.[8] We test here the suitability
of a model proposed by Galkin et al.[6] to describe the
chain-length-dependent properties of pyridines 1a–h in
quantitative terms. One of the basic assumptions made here
is that the increasingly large alkyl substituents influence the
pyridine system only through electronic (inductive) effects
and impart effectively no steric shielding on the reaction
center (the pyridine nitrogen atom). Under these condi-
tions, the effect of the dialkylamino substituent at the C4
position of the pyridine ring can be expressed with a single
substituent parameter σ*, the magnitude of which is de-
rived by using an inductive increment method. As shown in
kX
log
= ρσ*
(2)
kH
Table 5. Calculated σ* values for 1a–h for models 1–3.
σ*
Model 1
σ*
Model 2
σ*
Model 3
1a
1b
1c
1d
1e
1f
3.47
3.51
3.55
3.60
3.64
3.67
3.70
3.74
2.27
2.55
2.84
3.09
3.24
3.33
3.39
3.48
–2.23
–2.13
–2.01
–1.93
–1.89
–1.87
–1.87
–1.85
1g
1h
6
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O30486
/KAP1
Date: 16-07-13 10:53:07
Pages: 9
Inductive Effects through Alkyl Groups
Table 6 shows the R2 values together with the values for of the whole acyl group in 10a–h are too small to allow for
ρ using the σ* properties for model 2, with the exception of a quantitative analysis (see the Supporting Information).
1
the rationalization of the H NMR shifts, in which the σ*
The first group contains experimentally measured rate
constants, which were correlated in a Hammett expression
according to Equation (3).
values for model 3 were used.
Table 6. R2 values together with the values for ρ using the σ* prop-
erties for model 2, with exception of the 1H NMR correlation in
which the σ* values for model 3 were used.
log(kx) = ρσ*
(3)
R2
ρ
According to the values of the sensitivity parameter ρ,
which is the slope of the linear correlation between log(kx)
and σ*, the catalytic acylation reaction (I) shows the steep-
est increase with the most significant changes of inductive
effects (cf. Table 6, entry 1). This implies that the half-life
times for benchmark reaction (I) show the largest variation
in this particular dataset. Furthermore, reaction (I) is even
better represented by the increment-based calculation of in-
ductive effects (R2 = 0.9202, Table 6, entry 1) than the two
single-step model reactions (entries 2 and 3).
Entry
σ* (model 2)
1
2
3
log(k2)[a]
0.9202
0.8752
0.8262
0.33
0.12
0.09
log[k2(benzoylation)][b]
log[k2(aminolysis)][c]
4
5
6
7
8
9
ΔH298,ave gas phase[d]
ΔH298,ave (CHCl3)[e]
0.9935
0.8968
0.9809
0.7955
0.9599
–15.91
–5.76
–14.61
–5.07
–5.56
–15.26
ΔH298,best gas phase[f]
ΔH298,best (CHCl3)[g]
ΔH298,best only electrostatic[h]
ΔH298,best only nonelectrostatic[i] 0.9911
10
11
r(N–C)[j]
0.9925
0.9236
–0.67
–5.25
The second group of results contains energy data, which
were correlated according to Equation (4).
δ(1H,CH3)[k]
[a] Observed reaction rates in reaction (I) according to Scheme 2.
[b] Observed reaction rates for benzoylation reaction (II) according
to Scheme 4. [c] Observed reaction rates for aminolysis reac-
tion (III) according to Scheme 4. [d] Boltzmann-averaged ΔH298 at
MP2–5 level of theory: “MP2–5”: MP2/6-31+G(2d,p)//B98/6-
31G(d). [e] Boltzmann-averaged ΔH298 at MP2–5 level of theory
with PCM/UAHF/RHF/6-31G(d) solvation energies for chloro-
form. [f] Energetically best conformer used for the calculation of
ΔH298 at MP2–5 level of theory. [g] Energetically best conformer
used for the calculation of ΔH298 at MP2–5 level of theory with
inclusion of solvent effects as described above. [h] Acetylation en-
thalpies for 1a–h according to Scheme 5 including only the electro-
static term for the solvent contribution. [i] Acetylation enthalpies
for 1a–h according to Scheme 5 including only the nonelectrostatic
term for the solvent contribution. [j] N–C bond length of the pyr-
idine nitrogen and the carbonyl group in 10a–h. [k] 1H NMR shifts
of terminal CH3 groups in 1a–h, data displayed for model 3.
ΔH298 = ρσ*
(4)
The results displayed in Table 6 clearly illustrate that the
inclusion of solvent effects (in chloroform) gives signifi-
cantly inferior correlations with the tested model proposed
by Galkin et al. than the gas phase enthalpies themselves
(cf. entries 4 and 5, and entries 6 and 7). The acetylation
enthalpies in the gas phase are closely matched by the acet-
ylation enthalpies including only the nonelectrostatic sol-
vent contribution (cf. entries 6 and 9). This finding is not
only supported by the very similar values for R2 (R2
=
0.9809 vs. 0.9911), but also on the basis of the slope of the
correlation as indicated by ρ (ρ = –14.61 vs. –15.26).
The last group (Table 6, entries 10 and 11) contains data
that describe structural parameters and therefore should
not be correlated linearly according to the proposed in-
crement model by Galkin et al. Nevertheless, employing the
linear model used before for the acylation energies also for
this last dataset yields very good linear correlations for r(N–
C) (R2 = 0.9925) and δ(1H,CH3) (R2 = 0.9925, model 3 was
used according to the 1H NMR experiment). More complex
relationships between σ* and this last dataset were therefore
not tested.
In general, the observed saturation effects are best re-
flected by model 3, in which the impact of a change in in-
ductive effects is measured at the terminal methyl group
protons in 1a–h (for details see the Supporting Infor-
mation). This is most likely due to the well-marked satura-
tion effect of the calculated σ* values obtained in model 3
(see the Supporting Information). This method of analysis
1
is in accordance with the determination of the H NMR
resonance experiment used for the correlation in Figure 8
1
and therefore only used to interpret the H NMR shifts.
Nevertheless the chemical background of the other metrics
tested is different, and models 1 and 2 account for the sur-
veyed properties therein. Model 1 yields in any case much
lower R2 values than those of model 2 (see the Supporting
Conclusions
Numerous properties of (alkylamino)pyridines show sat-
Information). This may well be due to the large distance uration of substituent effects with alkyl group chain lengths
between the reaction center and the growing alkyl chain in beyond four carbon atoms. We believe this to be a general
model 1. Therefore model 2 will subsequently be used to result in situations lacking direct steric effects between alkyl
determine the respective R2 values displayed in Table 6 in substituents and the respective reaction centers. The fact
the discussion of the correlation analysis.
that all observed saturation phenomena could be explained
The data obtained in Table 6 can be divided into three in a more than satisfactory way by the aid of a simple in-
groups, as indicated by the dotted lines (entries 1–3, entries crement-based prediction of through-bond inductive effects
4–9 and entries 10 and 11). As already mentioned, the can be seen as strong support for this latter type of elec-
changes in charge of the pyridine nitrogen, as well as that tronic substituent effect.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7

Job/Unit: O30486
/KAP1
Date: 16-07-13 10:53:07
Pages: 9
R. Tandon, T. A. Nigst, H. Zipse
FULL PAPER
chromatography (silica; ethyl acetate/NEt3, 20:1) of the brown
crude mixture gave 1d (230 mg, 38%) as a pale-yellow solid. 1H
NMR (300 MHz, CDCl3): δ = 8.14 (dd, J = 5.0, 1.6 Hz, 2 H, Ha),
6.40 (dd, J = 5.0, 1.6 Hz, 2 H, Hb), 3.21 (t, J = 7.5 Hz, 4 H, Hc),
1.70–1.48 (m, 4 H, Hd), 0.91 (t, J = 7.4 Hz, 6 H, He) ppm. 13C
NMR (75 MHz, CDCl3): δ = 149.9 (Ca), 106.3 (Cb), 51.8 (Cc), 20.1
(Cd), 11.2 (Ce) ppm (in agreement with literature[10]).
Experimental Section
All air- and water-sensitive manipulations were carried out under a
nitrogen atmosphere using standard Schlenk techniques. Calibrated
flasks for kinetic measurements were dried in an oven at 120 °C for
at least 12 h prior to use and then assembled quickly while still hot,
cooled under a nitrogen stream, and sealed with a rubber septum.
All commercial chemicals were of reagent grade and were used as
received unless otherwise noted. CDCl3 was heated to reflux for at
least 1 h over CaH2 and subsequently distilled.
1H and 13C NMR spectra were recorded with Varian 300 or Varian
INOVA 400 instruments at room temperature. All 1H chemical
shifts are reported in ppm (δ) relative to TMS (0.00 ppm); 13C
chemical shifts are reported in ppm (δ) relative to CDCl3 (77.16
ppm). 1H NMR kinetic data were measured with a Varian Mercury
200 MHz spectrometer at 23 °C. HRMS spectra (ESI) were re-
corded with a Thermo Finnigan LTQ FT instrument. IR spectra
were measured with a Perkin–Elmer FTIR BX spectrometer
mounting an ATR technology. Analytical TLC were carried out
using aluminum sheets (silica gel Si 60 F254).
4-(Dibutylamino)pyridine (1e): To 4-chloropyridine hydrochloride
(1.00 g, 6.66 mmol) in an oven-4-(dried microwave vial was added
dibutylamine (2.49 mL, 14.6 mmol). After addition of pyridine
(2.68 mL, 33.3 mmol) the vial was closed with a septum cap and
the reaction mixture was heated for 2 h at 170 °C (200 W). The
brown residue was taken up in CH2Cl2 and washed with a saturated
solution of K2CO3 (ϫ3). The collected organic phase was dried
with MgSO4 and filtered. After evaporation of the solvent, the
crude mixture was purified two times by column chromatography
(silica; ethyl acetate/isohexane, 5:1) to give 1e (240 mg, 18%) as a
pale-brown oil. 1H NMR (300 MHz, CDCl3): δ = 8.16 (d, J =
3.3 Hz, 2 H, Ha), 6.43 (d, J = 3.3 Hz, 2 H, Hb), 3.26 (t, J = 7.6 Hz,
4 H, Hc), 1.68–1.44 (m, 4 H, Hd), 1.43–1.25 (m, 4 H, He), 0.95 (t,
J = 7.3 Hz, 6 H, Hf) ppm. 13C NMR (75 MHz, CDCl3): δ = 149.9
(Ca), 106.3 (Cb), 49.9 (Cc), 29.0 (Cd), 20.2 (Ce), 13.9 (Cf) ppm (in
agreement with literature[11]).
1H NMR kinetic measurements were performed by following a pre-
viously published procedure[2b] as described in the Supporting In-
formation. Procedures for measuring reaction kinetics using UV/
Vis spectroscopy can be found in the Supporting Information.
4-(Dipentylamino)pyridine (1f): To 4-chloropyridine hydrochloride
(2.00 g, 13.3 mmol) in a oven-dried pressure tube was added pyr-
idine (3.75 mL, 46.6 mmol). After addition of dipentylamine
(3.26 mL, 15.98 mmol) the reaction mixture was heated for 22 h at
185 °C. The brown crude mixture was taken up in CH2Cl2 and
washed with sat. K2CO3 solution. The collected organic phase was
dried with MgSO4, filtered, and the solvent was evaporated under
reduced pressure. After column chromatography (silica; ethyl acet-
ate/NEt3, 10:1) of the brown mixture, the product was distilled
three times at 140 °C (4 mbar) to give 1f (190 mg, 6%) as a pale-
4-(Ethylmethylamino)pyridine (1b): To 4-chloropyridine hydrochlo-
ride (1.50 g, 9.99 mmol) in an oven-dried pressure tube was added
Cs2CO3 (6.51 g, 20 mmol). After addition of N-methylethylamine
(4.29 mL, 50 mmol), [Pd(PPh3)4] (0.23 g, 0.20 mmol) and dist.
water (2 mL), the pressure tube was closed and heated for 72 h at
120 °C in an oil bath. The brown suspension was poured into
CH2Cl2, filtered, and extracted with distilled water (ϫ3). After dry-
ing over MgSO4, filtration and evaporation of the solvent, the
crude product was purified by column chromatography (silica;
CHCl3/MeOH, 10:1) to give 1b (870 mg, 64%) as a yellow liquid.
1H NMR (400 MHz, CDCl3): δ = 8.17 (dd, J = 5.0, 1.6 Hz, 2 H,
Ha), 6.45 (dd, J = 5.0, 1.6 Hz, 2 H, Hb), 3.38 (q, J = 7.1 Hz, 2 H,
Hd), 2.91 (s, 3 H, Hc), 1.12 (t, J = 7.1 Hz, 2 H, He) ppm. 13C NMR
(101 MHz, CDCl3): δ = 149.9 (Ca), 106.4 (Cb), 45.6 (Cd), 36.5 (Cc),
1
yellow viscous liquid. H NMR (400 MHz, CDCl3): δ = 8.15 (d, J
= 6.6 Hz, 2 H, Ha), 6.40 (d, J = 6.6 Hz, 2 H, Hb), 3.26 (t, J =
7.6 Hz, 4 H, Hc), 1.65–1.48 (m, 4 H, Hd), 1.44–1.19 (m, 8 H, He,
Hf), 0.90 (t, J = 7.1 Hz, 6 H, Hg) ppm. 13C NMR (75 MHz,
CDCl3): δ = 149.8 (Ca), 106.4 (Cb), 50.1 (Cc), 29.1 (Ce), 26.6 (Cd),
22.5 (Cf), 14.0 (Cg) ppm (in agreement with literature[12]).
11.2 (Ce) ppm. MS (EI): m/z (%) = 136 (30) [M+], 121 (100) [M+
–
CH3], 78 [C5H4N+]. HRMS (EI): m/z calcd. for C8H12N2 [M +
4-(Dihexylamino)pyridine (1g): In a pressure tube 4-chloropyridine
hydrochloride (2.00 g, 13.3 mmol) was suspended in pyridine
(3.75 mL, 46.6 mmol). After addition of dihexylamine (6.21 mL,
26.6 mmol) the pressure tube was closed and heated at 185 °C in
an oil bath. After 22 h reaction time the residue was dissolved in
CH2Cl2 and washed with a saturated solution of K2CO3 (ϫ3). Af-
ter drying the organic layer over MgSO4 and filtration, the solvent
was evaporated at reduced pressure. Column chromatography (sil-
ica; ethyl acetate/NEt3, 10:1) followed by distillation (130 °C,
10 mbar) gave 1g (570 mg, 16%) as a pale-yellow viscous liquid.
1H NMR (400 MHz, CDCl3): δ = 8.15 (dd, J = 5.0, 1.6 Hz, 2 H,
Ha), 6.40 (dd, J = 5.0, 1.6 Hz, 2 H, Hb), 3.26 (t, J = 7.5 Hz, 4 H,
Hc), 1.61–1.49 (m, 4 H, Hd), 1.39–1.18 (m, 12 H, He, Hf, Hg), 0.90
(t, J = 7.1 Hz, 6 H, Hh) ppm. 13C NMR (75 MHz, CDCl3): δ =
149.8 (Ca), 106.4 (Cb), 50.1 (Cc), 31.6 (Ce), 26.9 (Cd), 26.7 (Cf), 22.6
(Cg), 14.0 (Ch) ppm (in agreement with literature[11]).
H]+ 137.1037; found 137.0988. IR (ATR): ν = 2928 (w), 2850 (w),
˜
1594 (vs), 1514 (vs), 1466 (w), 1371 (vs), 1230 (vs), 1210 (vs), 1110
(w), 1080 (w), 998 (vs), 799 (vs), 810 (vs), 736 (s) cm–1.
4-(Diethylamino)pyridine (1c): To 4-chloropyridine hydrochloride
(1.50 g, 9.99 mmol) in a oven-dried pressure tube was added
Cs2CO3 (6.51 g, 20 mmol). After adding diethylamine (2.08 mL,
20 mmol) the pressure tube was closed and heated for 5 d at 170 °C
in an oil bath. The brown solution was poured into CH2Cl2, fil-
tered, and the solvent was evaporated. Column chromatography
(silica; ethyl acetate/NEt3, 20:1) followed by distillation of the
brown crude product gave 1c (220 mg, 20%) as a pale-yellow solid.
1H NMR (300 MHz, CDCl3): δ = 8.19 (dd, J = 5.0, 1.6 Hz, 2 H,
Ha), 6.47 (dd, J = 5.0, 1.6 Hz, 2 H, Hb), 3.37 (q, J = 10.5, 4.5 Hz,
4 H, Hc), 1.19 (t, J = 7.1 Hz, 6 H, Hd) ppm. 13C NMR (75 MHz,
CDCl3): δ = 149.9 (Ca), 106.2 (Cb), 43.7 (Cc), 12.3 (Cd) ppm (in
agreement with literature[9]).
4-(Dioctylamino)pyridine (1h): In a pressure tube, 4-chloropyridine
hydrochloride (2.00 g, 13.3 mmol) was suspended in pyridine
(3.75 mL, 46.6 mmol). After addition of dioctylamine (8.10 mL,
26.6 mmol) the pressure tube was sealed and heated at 160 °C in
an oil bath. After 18 h, the residue was dissolved in CH2Cl2 and
washed with a saturated solution of K2CO3 (ϫ3). After drying the
organic layer over MgSO4 and filtration, the solvent was evapo-
4-(Dipropylamino)pyridine (1d): To 4-chloropyridine hydrochloride
(1.00 g, 6.66 mmol) in an oven-dried pressure tube was added
Cs2CO3 (3.25 g, 9.99 mmol). After adding dipropylamine (0.46 mL,
3.33 mmol) the pressure tube was closed and heated for 3 d at
170 °C oil bath temperature. The warm, brown solution was poured
into CH2Cl2, filtered, and the solvent was evaporated. Column
8
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O30486
/KAP1
Date: 16-07-13 10:53:07
Pages: 9
Inductive Effects through Alkyl Groups
J. Phys. Org. Chem. 1999, 12, 283–288; d) O. Exner, M. Char-
rated under reduced pressure. Column chromatography on silica
(ethyl acetate/NEt3, 10:1) and on basic AlOx. (ethyl acetate/isohex-
ane, 1:10 Ǟ 10:1) gave 1h (610 mg, 14%) as a brown viscous liquid.
1H NMR (300 MHz, CDCl3): δ = 8.16 (dd, J = 5.0, 1.6 Hz, 2 H,
Ha), 6.40 (dd, J = 5.0, 1.6 Hz, 2 H, Hb), 3.33–3.14 (t, J = 7.7 Hz,
4 H, Hc), 1.72–1.41 (m, 4 H, Hd), 1.41–1.18 (m, 20 H, He, Hf, Hg,
Hh, Hi), 1.01–0.76 (m, J = 6.0 Hz, 3 H, Hj) ppm. 13C NMR
(75 MHz, CDCl3): δ = 149.9 (Ca), 106.3 (Cb), 50.1 (Cc), 31.8 (Ce),
29.4 (Cf), 29.3 (Cg), 27.0 (Cd), 26.9 (Ch), 22.6 (Ci), 14.1 (Cj) ppm.
MS (EI): m/z (%) = 318 (13) [M+], 219 (100) [C14H23N2], 121 (30)
[C7H9N2]. HRMS (EI): m/z calcd. for C21H38N2 [M+] 318.3035;
ton, V. Galkin, J. Phys. Org. Chem. 1999, 12, 289.
[2] a) V. D’Elia, Y.-H. Liu, H. Zipse, Eur. J. Org. Chem. 2011,
1527–1533; b) I. Held, E. Larionov, C. Bozler, F. Wagner, H.
Zipse, Synthesis 2009, 2267–2277; c) I. Held, P. von den Hoff,
D. S. Stephenson, H. Zipse, Adv. Synth. Catal. 2008, 350, 1891–
1900; d) I. Held, S. Xu, H. Zipse, Synthesis 2007, 1185–1196;
e) M. R. Heinrich, H. S. Klisa, H. Mayr, W. Steglich, H. Zipse,
Angew. Chem. 2003, 115, 4975–4977; Angew. Chem. Int. Ed.
2003, 42, 4826–4828.
[3] N. De Rycke, F. Couty, O. R. P. David, Chem. Eur. J. 2011, 17,
12852–12871 and references cited therein.
[4] C. Pedersen, Synthesis 1978, 844–845.
[5] a) S. Xu, I. Held, B. Kempf, H. Mayr, W. Steglich, H. Zipse,
Chem. Eur. J. 2005, 11, 4751–4757; b) T. A. Nigst, R. Tandon,
H. Mayr, H. Zipse, manuscript in preparation; c) T. A. Nigst,
H. Mayr, Eur. J. Org. Chem. 2013, 2155–2163.
[6] A. R. Cherkasov, V. I. Galkin, R. A. Cherkasov, Russ. Chem.
Rev. 1996, 65, 641–656.
found 318.3030. IR (ATR): ν = 2923 (vs), 2854 (s), 1593 (vs), 1512
˜
(vs), 1466 (s), 1371 (s), 1228 (s), 1104 (w), 986 (s), 799 (vs), 734
(w) cm–1.
Supporting Information (see footnote on the first page of this arti-
cle): Kinetic and thermodynamic data, together with 1H and 13C
NMR spectra for all catalysts are included.
[7] a) M. Charton, J. Org. Chem. 1979, 44, 903; b) M. Charton, J.
Am. Chem. Soc. 1977, 99, 5687.
[8] a) B. I. Istomin, V. A. Baranskii, Russ. Chem. Rev. 1982, 52,
223; b) M. I. Kabachnik, Russ. Chem. Rev. 1979, 48, 814.
[9] J. Jerchel, Chem. Ber. 1958, 91, 1266.
Acknowledgments
[10] C. Pedersen, Synthesis 1978, 844.
We thank Prof. Herbert Mayr for fruitful discussions and the Fonds
der Chemischen Industrie for financial support.
[11] D. J. Brunelle, D. A. Singleton, Tetrahedron Lett. 1984, 25,
3383.
[12] K. Ko, K. Nakano, S. Watanabe, Y. Ichikawa, H. Kotsuki, Tet-
rahedron Lett. 2009, 50, 4025–4029.
[1] a) O. Exner, J. Phys. Org. Chem. 1999, 12, 265–274; b) M.
Charton, J. Phys. Org. Chem. 1999, 12, 275–282; c) V. Galkin,
Received: April 4, 2013
Published Online: ᭿
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
9