Annelated pyridines as highly nucleophilic and Lewis basic catalysts for acylation reactions

Article abstract of DOI:10.1002/chem.201204452

New heterocyclic derivatives of 9-azajulolidine have been synthesized and characterized with respect to their nucleophilicity and Lewis basicity. The Lewis basicity of these bases as quantified through their theoretically calculated methyl-cation affinities correlate well with the experimentally measured reaction rates for addition to benzhydryl cations. All newly synthesized pyridines show exceptional catalytic activities in benchmark acylation reactions, which correlate only poorly with Lewis basicity or nucleophilicity parameters. A combination of Lewis basicity with charge and geometric parameters in the framework of a three-component quantitative structure-activity relationship (QSAR) model is, however, highly predictive. Copyright

Full text of DOI:10.1002/chem.201204452

FULL PAPER
DOI: 10.1002/chem.201204452
Annelated Pyridines as Highly Nucleophilic and Lewis Basic Catalysts for
Acylation Reactions
Raman Tandon,[a] Teresa Unzner,[a] Tobias A. Nigst,[a] Nicolas De Rycke,[b]
Peter Mayer,[a] Bernd Wendt,[c] Olivier R. P. David,[b] and Hendrik Zipse*[a]
Abstract: New heterocyclic derivatives
of 9-azajulolidine have been synthe-
sized and characterized with respect to
their nucleophilicity and Lewis basicity.
The Lewis basicity of these bases as
quantified through their theoretically
calculated methyl-cation affinities cor-
relate well with the experimentally
measured reaction rates for addition to
benzhydryl cations. All newly synthe-
sized pyridines show exceptional cata-
lytic activities in benchmark acylation
reactions, which correlate only poorly
with Lewis basicity or nucleophilicity
parameters. A combination of Lewis
basicity with charge and geometric pa-
rameters in the framework of a three-
component quantitative structure–ac-
tivity relationship (QSAR) model is,
however, highly predictive.
Keywords: acylation · nucleophilic
catalysis
·
organocatalysis
·
structure–activity relationships
Introduction
The catalytic potential of donor-substituted pyridines in ac-
ACHTUNGTRENNUNGylation reactions is well established, since the reports on 4-
N,N-dimethylaminopyridine (DMAP; 1) by Litvinenko et al.
in 1967 and by Steglich et al. in 1969.[1–3] A small improve-
ment in catalytic activity was soon after documented for 4-
(pyrrolidinyl)pyridine (PPY; 2),[4] but it took until 2003 that
a larger increase in activity could be realized with annelated
pyridines such as 9-azajulolidine (3).[5] Pyridine derivative 3
is a powerful organocatalyst not only suitable for acylation
reactions, but also for other transformations such as the aza-
Morita–Baylis–Hillman reaction.[6] Further development of
this class of catalysts through modification of the attached
ring systems thus seems desirable, but faces significant syn-
thetic hurdles. We, therefore, explore here the potential of
the respective di- and trinitrogen-substituted systems 4 and
5 (Figure 1).
Figure 1. Structures of pyridine derivatives based on the DMAP motif
used as nucleophilic organocatalysts.
Selected members of this family were recently character-
ized with respect to their affinity towards carbocation elec-
trophiles.[7] These data show that the additional amino-sub-
stituents attached to the 3- and 5-position of pyridines 4 and
5 further increase the Lewis basicity relative to the all-
carbon analogue 3. Furthermore, for some derivatives of 5 it
was recently shown in kinetic studies that the reaction rates
for addition to benzhydryl cations correlate with theoretical-
ly calculated Lewis basicity.[8] We now show how the Lewis
base properties of the tricyclic catalysts 3–5 depend on their
particular substituent pattern, and also test the catalytic ac-
tivity of these compounds using the acylation of 1-ethinylcy-
clohexanol as a benchmark reaction.
[a] R. Tandon, T. Unzner, Dr. T. A. Nigst, Dr. P. Mayer,
Prof. Dr. H. Zipse
Department of Chemistry, LMU Mꢀnchen
Butenandtsrasse 5-13, 81377 Mꢀnchen (Germany)
Fax : (+49)89-2180-77738
E-mail: zipse@cup.uni-muenchen.de
[b] Dr. N. De Rycke, Dr. O. R. P. David
Institut Lavoisier—Versailles, UMR 8180,
Universitꢁ de Versailles St-Quentin-en-Yvelines
45 avenue des Etats-Unis, 78035, Versailles (France)
Results and Discussion
[c] Dr. B. Wendt
Synthesis and structure: The synthesis of catalysts of general
structure 4 follows the same strategy used to access pyridine
derivatives 6.[9,10] This involves initial condensation of 3,4-di-
Certara, Martin-Kollar-Straße 17, 81829 Mꢀnchen (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201204452.
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

aminopyridine (7) with glyoxal and subsequent bromination
at C5-position of the pyridine ring with N-bromosuccinimide
(NBS) in acetonitrile at room temperature. Subsequent
Heck reaction of 9 with ethyl acrylate worked best on large
scale with minimal solvent in a pressure tube. The reduction
and immediate cyclization of 10 to core structure 11 was ac-
complished using palladium on charcoal in 88% yield. Cata-
lysts 4a–c can all be synthesized starting from 11. Synthesis
of 4a was accomplished through reduction with LiAlH4/
AlCl3 to furnish 96% of 12 and subsequent Eschweiler–
Clark reaction providing N-methylated product 4a in 61%
yield (Scheme 1). It should be noted that the acidic reaction
Scheme 2. Synthesis of catalysts 5a–e. Previously obtained yields by
David et al.[8] are given in brackets.
originally proposed route (Scheme 2),[8] minor modifications
in workup procedures and the choice of reducing agent
were introduced in the synthesis of 5b and c. Changes in-
clude the use of microwave irradiation (instead of conven-
tional heating) in the respective acylation reactions, which
reduces reaction times for the acylation of 14 from six hours
(5 f) or three days (15c) to just one hour (for both systems)
without any detriment of yield. Final reduction of the bis-
AHCTUNGRTENaNGUN mides 15 was accomplished with LiAlH4/AlCl3 (instead of
BH3·SMe2 used by David et al.) in just one hour at ꢀ108C
yielding 5b and c in comparable yields (Scheme 2). Follow-
ing the same strategy additional derivatives of 5 bearing
strongly electron-donating groups in the aryl ring (as in 5d)
as well as electron-withdrawing groups (as in 5e) were syn-
thesized in order to investigate different electronic effects
on the catalyst reactivity. The acylation of 14 with four dif-
ferent acyl donors and subsequent reduction of the amide
intermediates proceeded smoothly and in good to excellent
yields in all cases, except for catalyst 5d. Due to the high ba-
sicity of this latter compound the final product isolation and
purification steps after reduction of amide 15d could only
be executed with substantial loss of material.
Scheme 1. Synthesis of compounds 4a–c from 3,4-diaminopyridine (7)
It should be noted at this point that purification of com-
pounds of type 4 and 5 requires repeated chromatography
on basic aluminum oxide[11] (Brockmann 3) until analytically
pure material was obtained (as determined by 1H NMR
spectroscopy). Such material solidified on standing and was
further purified through recrystallization from ethyl acetate
under nitrogen and dried in high vacuum for at least eight
hours and stored in an exsiccator in the dark at ꢀ48C. The
evaporation of solvent in all synthetic and purification steps
of 4 and 5 was performed in a rotatory evaporator operated
under nitrogen and at a water bath temperature not exceed-
ing 308C. The catalytic activity of 4 and 5 depends signifi-
cantly on their purity, and deviations from the purification
protocol outlined above may thus lead to significant reduc-
tion of reaction rates. The Lewis basicity of amino-substitut-
ed pyridines depends on the alignment of the substituent ni-
trogen lone pair orbital with the pyridine p system. Structur-
and glyoxal.
conditions used under Eschweiler–Clark conditions are sig-
nificantly more effective than alkylation under basic condi-
tions due to the intermediate protonation (and thus deacti-
vation) of the pyridine nitrogen atom. Catalysts 4b and c
were synthesized by microwave-induced acylation of the
amino group of 11 with the appropriate acid chloride in pyr-
idine. The acylated species 13b (96%) and 13c (82%) were
obtained in good to excellent yields. Subsequent reduction
of both amide groups in 13 with LiAlH4/AlCl3 furnished 4b
(81%) and 4c (56%) in just one hour reaction time.
Catalysts 5a–e were synthesized from the 3,4,5-triamino-
pridine derivative 14, whose efficient synthesis was recently
described by David et al.[8] Although catalyst 5a was synthe-
sized in 77% yield by reductive amination according to the
&
2
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Annelated Pyridines for Acylation Reactions
FULL PAPER
al parameters connected to this orbital interaction are the 4-
and accompanying variations in the pyramidalization of the
central nitrogen atom occurs with rather little increase in
energy. From all structures analyzed it appears that catalysts
4a and 5d (see Supporting Information) provide the best or-
bital alignment between the nitrogen lone-pair electrons
and pyridine p-system.
ꢀ
N C bond length distNꢀ (the distance between the pyridine
carbon atom at the 4-position and the nitrogen atom attach-
C
ed to the same position) as well as the degree of pyramidali-
ꢀ
zation of the nitrogen substituent (d(abcd)). Shorter N C
bond lengths and smaller pyramidalization angles are com-
monly considered to reflect better orbital alignment. These
parameters are shown in Figure 2 in the crystal structures of
pyridines 3, 4a, and 5c as depicted in front (left) and top
(right) view.
Cation affinity data: In order to match these structural prop-
erties to actual Lewis basicities, methyl-cation affinities
(MCA) and acylation enthalpies DHac as defined in Scheme 3a
Scheme 3. Definitions of a) methyl-cation affinities and b) isodesmic
acetyl-transfer reaction.
and b have been collected in Table 1 for most of the synthe-
sized pyridines. The definition chosen here for methyl-
cation affinities follows the mass-spectrometric definition of
the gas-phase proton affinity as recently applied to a larger
number of Lewis bases.[12,7] The acylation enthalpies as de-
fined in Scheme 3b reflect relative acetyl-cation affinities
between the selected Lewis base and pyridine as the refer-
ence system. This latter definition was also used in previous
studies matching the Lewis basicities of pyridines with the
respective catalytic activity in acylation reactions. Both
measures of Lewis basicity have been quantified here at the
MP2/6-31+GACTHNUGRTNEUNG(2d,p)//B98/6-31G(d) level of theory in the gas
phase. How solvent effects impact the Lewis basicity has
been tested for the acylation enthalpies using the polariza-
ble-continuum model (PCM) in chloroform at the RHF/6-
31G(d) level with UAHF radii.
ꢀ
Figure 2. Crystal structures of 3, 4a, and 5c. 4-N C bond lengths (distNꢀ
pm) and pyramidalization angles d(abcd) of crystal structure and best com-
/
C
puted conformer.
All three types of affinity data indicate quite clearly, that
all tricyclic catalysts displayed in Figure 1 are stronger Lewis
bases than the parent DMAP. The triaminopyridine 5d is
the strongest Lewis base, irrespective of the choice of Lewis
basicity indicator.
ꢀ
The 4-N C bond lengths in the crystal structures of
mono-, di-, and triaminopyridines are actually identical
within experimental uncertainty (137.5ꢁ0.2 pm), whereas
the structures optimized at B98/6-31G(d) level (best con-
former each) show slightly larger bond lengths in all cases.
For the pyramidalization angles (d(abcd)) the best conformers
of 3 and 4a reflect the experimental results rather well. For
5c the best conformer shows a 108 higher pyramidalization
angle as compared to the crystal structure. For this latter
system a second conformer exists only 1 kJmolꢀ1 higher in
energy with a deformation angle of 10.78, which is closely
similar to the crystal structure. This implies that conforma-
tional reorientation of the annelated six-membered rings
Nucleophilicity: In how far the theoretically calculated
Lewis basicity correlates well with reaction rates for reac-
tion with reference electrophiles was next tested for addi-
tions to benzhydrylium cations. This type of addition reac-
tion provides the basis for Mayrꢃs comprehensive nucleo-
AHCTUNGTERGpNUNN hilACHTNURTGENNUNGiACHUTNGTRENNUGcity scale according to the linear free energy relation-
ship [Eq. (1)].
log k2 ð20 ꢂCÞ ¼ sNðN þ EÞ
ð1Þ
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

H. Zipse et al.
Table 1. Half-life times and affinity numbers for the catalysts synthe-
sized.
[c]
Cat. MCA
[kJmolꢀ1
DHac
DHac/solv MP2-5/ N (sN)[b]
t1/2
U
]
MP2-5[a]
solv[a]
ACHTUNGTNER[NUNG min]
[kJmolꢀ1
]
[kJmolꢀ1
ACHTUNGTERNNUNG ]
10 mol% catalyst loading
5d 661.3
5c 636.8
ꢀ163.7
ꢀ106.7
ꢀ98.3
–
38.4ꢁ1.0
65.4ꢁ3.1
ꢀ141.3
17.69
(0.57)[e]
16.81
(0.60)[e]
–
16.65
(0.58)[e]
–
5b 621.6
ꢀ120.6
ꢀ87.8
44.2ꢁ1.5
4c
5a
620.8
618.7
ꢀ123.4
ꢀ118.0
ꢀ90.2
ꢀ89.2
34.3ꢁ0.2
38.0ꢁ1.5
4b 613.3
ꢀ113.5
ꢀ111.0
ꢀ102.3
ꢀ87.5
ꢀ86.8
ꢀ82.2
13.8ꢁ0.4
17.9ꢁ0.9
14.7ꢁ0.5[g]
4a
3
611.0
602.7
–
15.60
(0.68)
–
14.99
(0.68)
14.95
(0.67)[e,f]
–
5e
2
596.8
590.1
ꢀ101.5
ꢀ87.5
ꢀ69.5
ꢀ67.6
227.3ꢁ0.5
67.0ꢁ0.1[h]
Figure 3. Correlation of acylation enthalpies (DHac/solv) with N-parame-
ters.
1
581.2
ꢀ77.1
ꢀ61.2
151.0ꢁ1.7[h]
18
5 f
–
–
ꢀ90.1[d]
ꢀ71.6
ꢀ55.0[d]
ꢀ48.2
878.1ꢁ59.9
ꢃ2880[i]
15.39
Scheme 3b the nucleophilicity parameters of 2 and 3 were
determined in acetonitrile (Figure 3).
(0.60)[e]
Figure 4 shows that nucleophilic reactivities of pyridines
1–3,5 correlate quite well (with R2 =0.9022) with acylation
enthalpies in chloroform and even better with the gas phase
acylation enthalpies (R2 =0.9702) (see Supporting Informa-
tion).
3 mol% catalyst loading
4b 613.3
ꢀ113.5
ꢀ111.0
ꢀ102.3
ꢀ87.5
ꢀ86.8
ꢀ82.2
–
–
45.9
57.4
46.7
4a
3
611.0
602.7
15.60
(0.68)
[a] Levels of theory: “MP2-5”: MP2/6-31+G
ACHTUNGTNER(NUNG 2d,p)//B98/6-31G(d), “MP2-
5/solv”: MP2/6-31+G(2d,p)//B98/6-31G(d) with PCM/UAHF/RHF/6-
ACHTUNGTRENNUNG
31G(d) solvation energies for chloroform. [b] Nucleophilicity parameters
in acetonitrile according to equation (1). [c] Half-life times (min) for ac-
ACHTUNGTRENNUNGylation reaction (Scheme 5). [d] Extrapolated (for details see the Sup-
porting Information). [e] Data from [8]. [f] New measurements: 15.51
(0.62)[14]. [g] Remeasured and in line with published data (15ꢁ0.1)[10]
.
[h] Data from [10]. [i] Extrapolated (13% conversion after 12 h).
The rate constants for nucleophile/electrophile combina-
tion reactions k2 were determined as described in the Sup-
porting Information. The k2 values were combined with the
solvent independent electrophilicity parameter E for differ-
ent cations (see Supporting Information) to yield the solvent
dependent nucleophilicity parameter N (nucleophilicity) and
sN (nucleophile-dependent sensitivity) as shown in Scheme 4
and Figure 3. These parameters are available for a variety of
pyridines in several solvents such as CH2Cl2 and CH3CN.[13]
In order to compare the nucleophilic reactivities of 1–5
with the corresponding acylation enthalpies according to
Figure 4. Determination of nucleophilicity parameters N and sN (nucleo-
phile-dependent sensitivity) through combination of rate constants k2
with electrophilicity parameter E for different cations.
We can thus conclude that the kinetic data for single-step
addition to cationic electrophiles are related to the thermo-
dynamic data for this structurally quite homogenous set of
nucleophiles.
Catalytic activity: The catalytic potential of the synthesized
catalysts has subsequently been explored in the acetylation
of tertiary alcohol 16 (Scheme 5). The reactions were fol-
lowed by 1H NMR spectroscopy in CDCl3 as the solvent.
All reactions eventually proceed to full conversion and the
Scheme 4. Reactions of pyridines with benzhydrylium cations in CH3CN
at 208C.
&
4
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Annelated Pyridines for Acylation Reactions
FULL PAPER
Scheme 5. Acylation reaction in CDCl3.
rates of the reactions can thus be characterized by the reac-
tion half-life t1/2 using an approach described previously.[10]
As expected from the calculated Lewis basicity and the ex-
perimentally measured N-parameters, all annelated pyridine
derivatives except 5e and 5 f are more reactive than DMAP
(1) (Table 1, last column).
Figure 6. Conversion–time plots for catalysts 4b, 3, and 4a at 3 mol%
catalyst loading.
The substantial influence of conformational fixation on
the reactivity is illustrated by the 60-fold decrease in activity
going from catalyst 3 to 18
for these systems were therefore redetermined at 3% cata-
lyst loading. From the conversion data shown in Figure 6
and reaction half-lives in Table 1 one can see that catalysts
4b and 3 are equally active under these conditions. With re-
action half-lives of 46–47 min they are more than three
times faster at 3% catalyst loading than DMAP (1) at 10%
loading and thus the most active catalysts among the highly
nucleophilic pyridines. With a half-life of 57.4 min, catalyst
4a is still somewhat less active than the 3 and 4b, but still
more active at 3 mol% than PPY (2) at 10 mol%.
The unexpectedly moderate catalytic activity of the
strongly Lewis basic 3,4,5-triaminopyridines 5 raises the
question whether the reaction mechanism is the same for all
aminopyridines studied here. One of the key questions con-
cerns the dependence of the reaction rate on the catalyst
concentration. Earlier studies had established a clear first-
order dependence for aminopyridines such as DMAP (1)
for benchmark reaction (Scheme 5).[15] Using catalyst con-
centrations between 2.5–10 mol% a first-order dependence
can also be established for the triaminopyridine 5c
(Figure 7), which is one of the most Lewis basic candidates
of series 5. This excludes the possibility that a second cata-
lyst molecule acts as general base in the rate-limiting step.
(compare Table 1, the structure
of catalyst 18 is depicted in
Figure 5). In catalyst 18, formal-
ly even more inductive effects
than in the annelated 3 are
Figure 5. Conformationally un-
restricted catalyst 18.
present, but without any confor-
mational fixation.
A comparison of the catalytic
activities of the annelated cata-
lyst systems shows that the conformationally constrained 4-
aminopyridine 3 and the 3,4-diaminopyridines 4 are signifi-
cantly more active than the 3,4,5-triaminopyridines
5
(Table 1). This is in remarkable contrast to the rather high
Lewis basicity and N-parameters determined for this latter
class of nucleophiles. The optimum balance between catalyt-
ic activity and nucleophilicity is thus achieved in pyridines 3
and 4, in which one or two nitrogen atoms are incorporated
into conformationally-restricted ring systems. The most
active catalyst is the newly synthesized catalyst 4b, which
displays an even shorter half-life time than the most active
catalyst 6b in previous studies[10] and is also slightly more
active than 9-azajulolidine (3). These catalysts are more
than ten times more active than DMAP (1) (151 min) and
thus very potent acyl-transfer reagents. It is of note that in
series 4, ethyl-substituted compounds are slightly more effi-
cient than methyl derivatives, whereas in the 3,4,5-triamino-
pyridine series the ethyl-substituted compound 5b is less ef-
fective than the methyl-substituted 5a. The benzylic com-
pound 4c is, in comparison, much less active than the alkyl-
substituted compounds 4a and b. This ordering is also found
for the 3,4,5-triaminopyridines 5a–c. Comparison of the
benzylic compounds 5c–e shows that electron-donating
groups shorten the reaction half-life whereas electron-with-
drawing groups decrease the catalytic activity tremendously
(e.g., 5e has a half-life time of 227 min whereas that of 5c
amounts to 65 min). Reaction times are very short at 10%
catalyst loading for some of the annelated pyridine deriva-
tives, making precise measurements and an accurate com-
parison of the most active systems difficult. Reaction rates
Figure 7. Variation of the reaction rate as a function of the concentration
of catalyst 5c in acylation reaction shown in Scheme 5.
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

H. Zipse et al.
The rather small intercept of the correlation line with the
ordinate axis also indicates a negligible background reaction
and the reaction mechanism thus seems to be identical to
that for DMAP (1).
How annelated catalysts 5 respond to changes in the
steric demand of the acylating reagent was subsequently
tested by reacting alcohol 16 with isobutyric anhydride
(Scheme 6). These studies were performed at slightly elevat-
Figure 8. Conversion–time plots for catalysts 2, 3, and 5b for acylation re-
actions (Scheme 5) in dichloromethane.
Scheme 6. Isobutyrylation of alcohol 16 in chloroform at 408C.
Table 3. Half-life times for catalysts 2, 3 and 5b for reactions (Scheme 5)
in dichloromethane (t
1/2ACHTUNTGERNNUG(CD2Cl2)) and chloroform (t1/2ACHTUNGTERN(NUGN CDCl3)), and the
ratio of these kinetic parameters.
ed temperatures (408C) in deuterochloroform (Scheme 4).
As reported earlier for simple aminopyridines such as PPY
(2), the rates of this latter transformation are much slower
than that of the benchmark reaction shown in Scheme 5
(Table 2). The ratio of half-life times for these two transfor-
Cat.
t
G
t
G
Ratio
3
5b
2
14.7ꢁ0.5
44.2ꢁ1.5
1.42
0.96
1.61
108.3ꢁ0.9
[a] Half-life time (min). [b] Data from [10].
Table 2. Comparison of half-life times[a] for benchmark reactions shown
lation between Lewis basicity and nucleophilicity parame-
ters on one hand and catalytic performance in acylation re-
actions at different catalyst loadings in the other hand is
thus clearly not an issue of solvent polarity.
The catalytic activities for the benchmark reaction shown
in Scheme 5 correlate poorly with the calculated acylation
enthalpies (R2 =0.2733, Figure 9). This is in contrast to earli-
er studies for less Lewis basic catalysts,[18] in which signifi-
cantly better correlations were found. Could the nature of
the one-step reaction underlying the quantification of Lewis
basicity and nucleophilicities (as described above) in com-
parison to the multistep nature of catalytic processes be the
in Scheme 5 (t1/2ACHTUNGTRENNUNG(S 5)) and Scheme 6 (t1/2ACHUTNGTREN(NGUN S 6)) and the ratio of these two
reactivity parameters.
Cat.
t
G
t
G
Ratio
3
5b
2
67[b]
14.7ꢁ0.5
44.2ꢁ1.5
67ꢁ0.1[c]
151ꢁ1.7[c]
4.56
2.56
2.55
2.00
114
171[b]
302[b]
1
[a] Half-life time (min). [b] Data from [9a]. [c] Data from [10]
mations displays a distinct dependence on the absolute cata-
lyst activity in that the most active catalyst 3 reacts most
sensitively to the steric demand of the acylating agent.
Previous studies have clearly established that Lewis base
catalyzed acylations work best in apolar organic solvents
such as toluene, but can be quite slow in polar organic sol-
vents.[4,16] We therefore examined acylation reaction
(Scheme 5) in different solvents using commercially availa-
ble 2 as the catalyst. Among all tested solvents the reaction
was found to be fastest in CDCl3 (for details see Supporting
Information). Also, earlier studies by Han et al. indicate
that catalysts 3 and 5b react with comparable rates with 16
according to Scheme 5 in CD2Cl2.[17] As these results differ
from our findings in CDCl3 (Table 1) we have now reinvesti-
gated benchmark reaction (Scheme 5) in CD2Cl2.
The results obtained at 10% catalyst loading (Figure 8,
Table 3) indicate that reactions can indeed be slightly slower
in CD2Cl2 than in CDCl3, but that the magnitude of the sol-
vent effect depends on the nature of the catalyst. In both
solvents, however, we find that catalyst 3 is clearly more ef-
fective than the 3,4,5-triaminopyridine 5b. The lack of corre-
Figure 9. Correlation of acylation enthalpies (DHac/solv) with kinetic data
for acylation reactions (Scheme 5) for 1–3 (circles), 4a–c (squares), 5a–e
(triangles).
&
6
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Annelated Pyridines for Acylation Reactions
FULL PAPER
Table 4. Comparison of QSAR parameters of annelated catalysts 3, 4a,
and 5a.
reason for this discrepancy? As noted in previous studies,
none of the ground state descriptors for the pyridine nitro-
gen atom (that is: the reaction center) has significant predic-
tive value.[18]
The possibility to predict catalytic rates using a multipara-
meter QSAR model was therefore tested, starting from an
initial set of eight different parameters (see Supporting In-
formation for details). The most important descriptors could
be identified as the acylation enthalpies including solvation
terms in chloroform (DHac/solv, in kJmolꢀ1), the charge of the
Cat.
DHac/solv
[kJmolꢀ1
qortho
distNꢀ
[ꢄ]
ln
exptl
(1/t1/2
)
ln
calcd
(1/t1/2
)
residual
ꢀ
H
C
N
]
3
4a
5a
ꢀ82.20
ꢀ86.80
ꢀ89.20
0.2108
0.2122
0.2137
1.3884
1.3939
1.3997
ꢀ2.6879
ꢀ2.8848
ꢀ3.6376
ꢀ2.9648
ꢀ3.0668
ꢀ3.3741
0.2769
0.1820
ꢀ0.2635
catalysts and their acylated intermediates, lies in the poten-
tial to further optimize the catalytic activity of pyridine cata-
lysts.
ortho-hydrogen atom of the free catalyst (qortho H, in units of
ꢀ
elemental charge, e) and the bond length (distNꢀC, in ꢄ) be-
tween the carbon atom at the 4-position and the pyridine ni-
trogen atom attached to the same position. A QSAR model
based on these three parameters for catalysts 1–5 (11 struc-
tures) constructed with Sybyl X 2.0 yields good correlations
between actual and predicted catalytic activity (R2 =0.9320,
Q2 =0.7880, see Figure 10). The QSAR Equation (2) con-
Conclusion
In conclusion we synthesized several new derivatives of 9-
azajulolidine (3) and proved their structures by X-ray analy-
sis. All new derivatives are strong Lewis bases relative to
the parent DMAP (1) system. The Lewis basicity correlates
rather well with reaction rates for addition to cationic elec-
trophiles, but not with rate data for the catalytic acylation of
tertiary alcohol 16. In qualitative terms this implies that cat-
alysts with greater Lewis basicity will eventually slow down
catalytic processes due to the increasing difficulty of detach-
ing the product electrophiles. This general observation has
recently also been made by Christmann et al. in the Lewis
base catalyzed isomerization of (Z)-allylic alcohols.[19]
Experimental Section
All air and water sensitive manipulations were carried out under a nitro-
gen atmosphere using standard Schlenk techniques. Calibrated flasks for
kinetic measurements were dried in the oven at 1208C for at least 12 h
prior to use and then assembled quickly when still hot, cooled under a ni-
trogen stream and sealed with a rubber septum. All commercial chemi-
cals were of reagent grade and were used as received unless otherwise
noted. Acetonitrile (Acros>99.9%, extra dry), was purchased and used
without further purification. CDCl3 was refluxed for at least one hour
over CaH2 and subsequently distilled. 1H and 13C NMR spectra were re-
corded on Varian 300 or Varian INOVA 400 and 600 machines at room
temperature. All 1H chemical shifts are reported in ppm (d) relative to
TMS (0.00); 13C chemical shifts are reported in ppm (d) relative to
Figure 10. Experimental versus predicted values for ln
ACHTUGNTNREN(UNG 1/t1/2) for the
three-parameter QSAR model involving DHac/solv, qortho H and distNꢀ
.
C
ꢀ
tains the acylation enthalpies DHac/solv as a dominant term
(term 2), more negative acylation enthalpies implying faster
reaction rates. The ortho-hydrogen charge and the N C dis-
tance parameters, in contrast, enter the QSAR equation as
term 3 and 4 with negative sign, implying slower reaction
rates for catalysts with more positively charged ortho-hydro-
ꢀ
1
CDCl3 (77.16). H NMR kinetic data were measured on a Varian Mercu-
ry 200 MHz spectrometer at 238C. HRMS spectra (ESI-MS) were carried
out using a Thermo Finnigan LTQ FT instrument. IR spectra were meas-
ured on a Perkin–Elmer FT-IR BX spectrometer mounting ATR technol-
ogy. Reactions utilizing microwave technology were conducted in a CEM
Discover Benchmate microwave reactor (model nr. 908010). Analytical
TLC was carried out using aluminum sheets with silica gel Si 60 F254. 9-
Azajulolidine (3) was obtained from TCI China (CAS.nr.: 6052–72–8),
purity:>97.0% (GC).
ꢀ
gen atoms and larger N C bond lengths.
Reaction kinetics were followed using 1H NMR spectroscopy and evalu-
ated according to a previously published method[10] as described in detail
in the Supporting Information.
As is readily seen from the data collected in Table 4 for
catalysts 3, 4a, and 5a, the last two terms are those relevant
for predicting slower reactions for catalysts of higher Lewis
basicity such as 5a.
The value of this QSAR model, which is exclusively
based on parameters available from theoretical data for the
The measurements of N-parameters were carried out according to the
method described by Mayr et al.[20,13b]
Detailed procedures for the synthesis of all new compounds and catalysts
have been compiled in the Supporting Information together with the re-
quired analytic data. CCDC-633500 (3a), CCDC-914973 (4a), CCDC-
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

H. Zipse et al.
914974 (5c), CCDC-914975 (5d) and CCDC-914976 (5e) contain the
supplementary crystallographic data for this paper. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
[12] a) C. Lindner, B. Maryasin, F. Richter, H. Zipse, J. Phys. Org. Chem.
2010, 23, 1036–1042; b) Y. Wei, G. N. Sastry, H. Zipse, J. Am. Chem.
Soc. 2008, 130, 3473–3477; c) Y. Wei, T. Singer, H. Mayr, G. N.
Sastry, H. Zipse, J. Comp. Chem. 2008, 29, 291–297.
[13] a) H. Mayr, B. Kempf, A. R. Ofial, Acc. Chem. Res. 2003, 36, 66–77;
b) H. Mayr, M. Patz, Angew. Chem. 1994, 106, 990–1010; Angew.
Chem. Int. Ed. Engl. 1994, 33, 938–957; c) H. Mayr, M. Patz, M. F.
Gotta, A. R. Ofial, Pure Appl. Chem. 1998, 70, 1993–2000; d) H.
Mayr, A. R. Ofial in Carbocation Chemistry (Eds.: G. A. Olah,
G. K. S. Prakash), Wiley, New York, 2004, pp. 331–358; e) H. Mayr,
A. R. Ofial, Pure Appl. Chem. 2005, 77, 1807–1821; f) R. Lucius, R.
Loos, H. Mayr, Angew. Chem. 2002, 114, 97–102; Angew. Chem.
Int. Ed. 2002, 41, 91–95; g) M. Baidya, S. Kobayashi, F. Brotzel, U.
Schmidhammer, E. Riedle, H. Mayr, Angew. Chem. 2007, 119,
6288–6292; Angew. Chem. Int. Ed. 2007, 46, 6176–6179; h) H.
Mayr, J. Bartl, G. Hagen, Angew. Chem. 1992, 104, 1689–1691;
Angew. Chem. Int. Ed. Engl. 1992, 31, 1613–1615.
[14] T. A. Nigst, J. Ammer, H. Mayr, J. Phys. Chem. A 2012, 116, 8494–
8499.
[15] S. Xu, I. Held, B. Kempf, H. Mayr, W. Steglich, H. Zipse, Chem.
Eur. J. 2005, 11, 4751–4757.
[16] A. Sakakura, K. Kawajiri, T. Ohkubo, Y. Kosugi, K. Ishihara, J. Am.
Chem. Soc. 2007, 129, 14775–14779.
[17] a) S. Singh, G. Das, O. V. Singh, H. Han, Org. Lett. 2007, 9, 401–
404; b) S. Singh, G. Das, O. V. Singh, H. Han, Tetrahedron Lett.
2007, 48, 1983–1986.
Acknowledgements
The authors gratefully acknowledge valuable comments and suggestions
by Prof. H. Mayr (Dept. Chemistry), and by Dr. J. Pabel and Dr. T. Wein
(Dept. Pharmacy).
[1] a) L. M. Litvinenko, A. I. Kirichenko, Dokl. Akad. Nauk SSSR Ser.
Khim. 1967, 176, 97–100; b) W. Steglich, G. Hçfle, Angew. Chem.
1969, 81, 1001; Angew. Chem. Int. Ed. Engl. 1969, 8, 981.
[2] a) G. Hçfle, W. Steglich, H. Vorbrꢀggen, Angew. Chem. 1978, 90,
602–615; Angew. Chem. Int. Ed. Engl. 1978, 17, 569–583.
[3] Review: N. De Rycke, F. Couty, O. R. P. David, Chem. Eur. J. 2011,
17, 12852–12871.
[4] A. Hassner, L. R. Krepski, V. Alexanian, Tetrahedron 1978, 34,
2069–2076.
[5] M. R. Heinrich, H. S. Klisa, H. Mayr, W. Steglich, H. Zipse, Angew.
Chem. 2003, 115, 4975; Angew. Chem. Int. Ed. 2003, 42, 4826–4828.
[6] C. Lindner, R. Tandon, Y. Liu, B. Maryasin, H. Zipse, Org. Biomol.
Chem. 2012, 10, 3210–3218.
[18] E. Larionov, F. Achrainer, J. Humin, H. Zipse, ChemCatChem. 2012,
4, 559–566.
[7] C. Lindner, R. Tandon, B. Maryasin, E. Larionov, H. Zipse, Beilstein
J. Org. Chem. 2012, 8, 1406–1442.
[19] D. Kçnning, W. Hiller, M. Christmann, Org. Lett. 2012, 14, 5258–
5261..
[8] N. De Rycke, G. Berionni, F. Couty, H. Mayr, R. Goumont, O. R. P.
David, Org. Lett. 2011, 13, 530–533.
[9] a) I. Held, S. Xu, H. Zipse, Synthesis 2007, 1185–1196; b) H. Zipse,
I. Held, Bayer Material Science LLC, US 2008/0176747 A1, 2008.
[10] I. Held, E. Larionov, C. Bozler, F. Wagner, H. Zipse, Synthesis 2009,
2267–2277.
[20] H. Mayr, T. Bug, M. F. Gotta, N. Hering, B. Irrgang, B. Janker, B.
Kempf, R. Loos, A. R. Ofial, G. Remennikov, H. Schimmel, J. Am.
Chem. Soc. 2001, 123, 9500–9512. For a comprehensive listing of nu-
cleophilicity parameters N, sN and electrophilicity parameters E see
http://www.cup.uni-muenchen.de/oc/mayr/DBintro.html.
[11] The catalyst was found to bind strongly to silica gel and was difficult
to recover.
Received: December 13, 2012
Published online: && &&, 0000
&
8
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Annelated Pyridines for Acylation Reactions
FULL PAPER
A special relationship: Highly nucleo-
philic and Lewis basic derivatives of 9-
azajulolidine have been synthesized
and a good correlation between Lewis
basicity and nucleophilicity has been
obtained. The catalytic performance of
these compounds in acylation reactions
of sterically hindered alcohols is found
to be more complex and quantitative
prediction of reaction rates requires a
three-component quantitative struc-
ture–activity relationship (QSAR)
model including Lewis basicity, struc-
tural, and charge distribution parame-
ters (see figure).
Lewis Base Catalysis
R. Tandon, T. Unzner, T. A. Nigst,
N. De Rycke, P. Mayer, B. Wendt,
O. R. P. David, H. Zipse* ... &&&& — &&&&
Annelated Pyridines as Highly Nucleo-
philic and Lewis Basic Catalysts for
Acylation Reactions
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!