Stereoselective Synthesis of 1,3-Diaminotruxillic Acid Derivatives: An Advantageous Combination of C-H-ortho-Palladation and On-Flow [2+2]-Photocycloaddition in Microreactors

Article abstract of DOI:10.1002/chem.201503742

The stereoselective synthesis of ε-isomers of dimethyl esters of 1,3-diaminotruxillic acid in three steps is reported. The first step is the ortho-palladation of (Z)-2-aryl-4-aryliden-5(4H)-oxazolones 1 to give dinuclear complexes 2 with bridging carboxylates. The reaction occurs through regioselective activation of the ortho-C-H bond of the 4-arylidene ring in carboxylic acids. The second step is the [2+2]-photocycloaddition of the C-C exocyclic bonds of the oxazolone skeleton in 2 to afford the corresponding dinuclear ortho-palladated cyclobutanes 3. This key step was performed very efficiently by using LED light sources with different wavelengths (465, 525 or 625 nm) in flow microreactors. The final step involved the depalladation of 3 by hydrogenation in methanol to afford the ε-1,3-diaminotruxillic acid derivatives as single isomers.

Full text of DOI:10.1002/chem.201503742

DOI: 10.1002/chem.201503742
Full Paper
&
Flow Microreactors |Hot Paper|
Stereoselective Synthesis of 1,3-Diaminotruxillic Acid Derivatives:
An Advantageous Combination of CÀH-ortho-Palladation and On-
Flow [2+2]-Photocycloaddition in Microreactors
Elena Serrano,[b] Alberto Juan,[c] Angel García-Montero,[a] Tatiana Soler,[d] Francisco JimØnez-
Mµrquez,[e] Carlos Cativiela,*[a] M. Victoria Gomez,*[c] and Esteban P. Urriolabeitia*[a]
Abstract: The stereoselective synthesis of e-isomers of di-
methyl esters of 1,3-diaminotruxillic acid in three steps is re-
ported. The first step is the ortho-palladation of (Z)-2-aryl-4-
aryliden-5(4H)-oxazolones 1 to give dinuclear complexes 2
with bridging carboxylates. The reaction occurs through re-
gioselective activation of the ortho-CÀH bond of the 4-aryli-
dene ring in carboxylic acids. The second step is the [2+2]-
photocycloaddition of the C=C exocyclic bonds of the oxa-
zolone skeleton in 2 to afford the corresponding dinuclear
ortho-palladated cyclobutanes 3. This key step was per-
formed very efficiently by using LED light sources with differ-
ent wavelengths (465, 525 or 625 nm) in flow microreactors.
The final step involved the depalladation of 3 by hydrogena-
tion in methanol to afford the e-1,3-diaminotruxillic acid de-
rivatives as single isomers.
Introduction
Truxillic and truxinic acid derivatives (Figure 1) have been
known for more than a century.[1] In 1888, Liebermann charac-
terized two isomers of truxillic acid among the decomposition
products of cocamine,[2a–c] an extract obtained by Hesse from
Erythroxylum Coca leaves in 1887.[2d] This type of compound
quickly attracted the interest of a broad range of scientists be-
cause of their outstanding pharmacological and biological
properties. Among many other properties, it is worth highlight-
ing their remarkable antinociceptive (reduction of the sensitivi-
ty to painful stimuli by blocking the transmission of electrical
Figure 1. Truxillic and truxinic acids.
signals)[3] and antimuscarinic[4] activities, even at low doses. In
a different biological target, a specific class of truxillic acids,
namely diaminotruxillic acid derivatives (Figure 1), have very
recently been identified as the only non-peptidic agonists of
GLP-1R (Glucagon-Like Peptide 1 Receptor), which is involved
in the treatment of type 2 diabetes mellitus.[5] As a conse-
quence, the high added value of these bioactive compounds is
evident.
[a] A. García-Montero, Prof. Dr. C. Cativiela, Dr. E. P. Urriolabeitia
Instituto de Síntesis QuímicayCatµlisis HomogØnea (ISQCH)
CSIC-Universidad de Zaragoza, Fac. Ciencias, Edificio D
Pedro Cerbuna 12, 50009 Zaragoza (Spain)
E-mail: cativiela@unizar.es
The properties and applications outlined above prompted
extensive research aimed at developing a simple and large-
scale method to synthesize truxillic acids and their derivatives.
In this respect, the best method for the preparation of truxillic
acid itself is the solid-state [2+2]-photodimerization of cinnam-
ic acid under irradiation with UV light (Figure 2a).[6] By using
the same approach, a retrosynthetic analysis leads to the con-
clusion that the [2+2]-photocycloaddition of the well-known
and synthetically accessible (Z)-4-aryliden-5(4H)-oxazolones
could provide a very plausible synthetic pathway (Figure 2b).[7]
Despite the chemical versatility of 5(4H)-oxazolones, access
to truxillic acid derivatives from oxazolones is still difficult and
is really quite limited. In this respect, the photochemical step is
the bottleneck in the process. Previous reports on the photoir-
radiation of free 5(4H)-oxazolones in solution cover a variety of
esteban@unizar.es
[b] Dr. E. Serrano
Centro Universitario de la Defensa, Academia General Militar
50090 Zaragoza (Spain)
[c] Dr. A. Juan, Dr. M. V. Gomez
Instituto Regional de Investigación Científica Aplicada (IRICA)
Universidad de Castilla La Mancha
Avenida Camilo JosØ cela s/n, 13071 Ciudad Real (Spain)
E-mail: MariaVictoria.Gomez@uclm.es
[d] Dr. T. Soler
Servicios Centrales Investigación, Universidad de Alicante
03690 Alicante (Spain)
[e] F. JimØnez-Mµrquez
E.T.S. Ingenieros Industriales, Universidad de Castilla-La Mancha
Avenida Camilo JosØ cela s/n, 13071 Ciudad Real (Spain)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201503742.
Chem. Eur. J. 2016, 22, 144 – 152
144
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
planes, point in the same direction and have an intermolecular
distance shorter than 4.2 Š) the process is not efficient.[10]
Other synthetic routes are available to prepare cyclobutanes
and these include the transition-metal-mediated CÀH function-
alization of preformed cyclobutanes.[11] Despite recent impres-
sive developments, these reactions are still synthetically com-
plex because they involve the C(sp3)ÀH bond, which is difficult
to activate, the need for precise chelating directing groups,
and they require careful control of the newly formed stereo-
centers.
Figure 2. a) Synthesis of truxillic acid by [2+2]-photocycloaddition of cin-
namic acid and b) proposed retrosynthesis of 1,3-diaminotruxillic acid deriv-
atives.
We report here an efficient method for the synthesis of 1,3-
diaminotruxillic acid derivatives based on a novel approach
(see Figure 3). The first step is the ortho-palladation of (Z)-2-
aryl-4-(het)arylidene-5(4H)-oxazolones. We recently studied this
reaction in depth in 4-arylidene systems and showed that this
process only takes place with carboxylic acids and is regiose-
lective at the ortho-CÀH bond of the arylidene ring.[12] In the
present work, the ortho-palladation of 4-hetarylidene-5(4H)-ox-
azolones was achieved with the same regioselectivity, thus
showing the broad scope of the process. The resulting dimeric
ortho-palladated complexes are especially suited to undergo,
in a second step, the [2+2]-photocycloaddition process be-
cause of their clamshell structure. This structure behaves as
a molecular template and locates the two C=C bonds in close
proximity, thus favoring their interaction.[12a] In this work we
present a very efficient methodology based on the irradiation
of samples in solution with LEDs as a light source combined
with flow microreactors.[13] In this way, yields of 100% were ob-
tained for the [2+2]-photodimerized products in reaction
times as short as 5–20 minutes by using LEDs as low-energy
input, long lifetime and inexpensive light sources. In addition,
the rigid environment in which the palladated oxazolones are
located during the photocycloaddition step promotes the ste-
reoselective formation of the cyclobutane ring and the synthe-
sis of the corresponding palladated truxillic acid derivatives as
a single isomer (e-truxillic) in solution. The liberation of the e-
1,3-diaminotruxillic acid derivatives occurs smoothly in a third
step after treatment of the [2+2]-photodimerized complexes
with H2 in methanol. Thus, this methodology overcomes all of
the previously noted problems for the photocycloaddition of
free oxazolones (low yields, long reaction times and the pres-
ence of several isomers) and affords e-1,3-diaminotruxillic acid
derivatives with a remarkably broad scope of application.
Figure 3. Previous work on the irradiation of oxazolones and comparison
with the work described herein.
processes (see Figure 3), which include the formation of diaze-
tidines by [2+2] photocycloaddition of the C=N bonds,[8] Z to E
isomerization[9] and, in only a single example, a [2+2]-photodi-
merization.[5d] The latter process takes place with very poor effi-
ciency, and long reaction times (3 days) and strong irradiation
(500 W) of the starting oxazolones are required to achieve Results and Discussion
a very low yield of the corresponding cyclobutanes (typically
ortho-Palladation of (Z)-2-aryl-4-hetaryliden-5(4H)-oxazo-
lones
10–15%).[5d] An additional problem is that the diaminotruxillic
acid species are obtained as a mixture of at least four different
isomers, which must be separated by tedious HPLC separations
before determination of their pharmacological and biological
properties.[5d] Solid-state irradiation of samples in the crystalline
state (as reported, for instance, for cinnamic acid) has been
used as a method to limit the number of isomers.[6,10] However,
these reactions are controlled by the crystal packing of the re-
acting molecules and, unless certain topochemical conditions
are met (the C=C double bonds must be located in parallel
The starting oxazolones 1 used in this work (see Figure 4) were
prepared by following reported methods (see the Supporting
Information for details). Complexes 2a–m have been reported
previously.[12]
The ortho-palladation of oxazolones 1n–t derived from het-
erocycles was carried out by treatment with Pd(OAc)2 (1:1
molar ratio) in CF3CO2H as solvent (see Figure 5). In most cases,
the reaction took place at room temperature in short reaction
Chem. Eur. J. 2016, 22, 144 – 152
www.chemeurj.org
145
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The characterization of 2n–t was carried out by using con-
ventional techniques. The analytical data are consistent with
a dinuclear structure. The NMR data suggest that CH bond ac-
tivation occurred at the heterocyclic ring to afford a six-mem-
bered palladacycle in all cases. Thus, the ortho-palladation of
4-hetarylidene oxazolones 1n–t shows the same regioselectivi-
ty as observed for the 4-arylidene systems 1a–m.[12b] In addi-
tion, the ease of palladation of the 3-position of the thiophene
ring in 1n, 1p, 1q, and 1r is remarkable because this is not
the “natural” reactivity of the thiophene ring. This seems to be
a consequence of the N-directing effect. Finally, we suggest for
complexes 2n–t a dimeric clamshell structure (Figure 5), with
a relative head-to-tail arrangement of the two palladated oxa-
zolones (C2 symmetry), because only a single peak is observed
in the 19F NMR spectrum and because of similarities with close-
ly related complexes that have fully determined X-ray diffrac-
tion structures.[12b]
Figure 4. Oxazolones employed as starting materials in this work.
[2+2]-Photocycloaddition processes in flow reactors
Previous work by our group showed that CDCl3 solutions of
some ortho-palladated complexes (2a, 2k, 2l) evolved to give
the product of a [2+2]-cycloaddition when exposed to sun-
light.[12a] This procedure has clear advantages over other re-
ported [2+2]-cycloadditions:[5d] 1) the reaction occurs with
complete conversion into a single compound and 2) full ste-
reoselectivity is obtained because of molecular constraints. The
[2+2] photocycloaddition of the ortho-palladated complexes
2a–m[12b] and 2n–s was studied. In the first step we screened
the range of reaction times required for complete conversion
(where possible) of 2a–s into 3a–s upon irradiation with sun-
light in CDCl3 solution. The results, which are collected in
Table 1, show that the times required for complete conversion
range from 28 to 480 h. It is clear that this is not an efficient
process. In some other cases the cycloaddition was either only
partial or was not observed at all. Given that the reactions
were performed in NMR tubes, only small amounts of samples
(milligrams) were obtained in each case and attempts to scale
up the process were unsuccessful. In addition, the free truxillic
acid derivatives were not obtained. In an effort to overcome
these problems and to define a highly efficient process with
the widest scope, a different methodology based on the com-
bined use of microreactors and flow chemistry was developed.
This approach has the advantages of both fields. On the one
hand, the high surface-to-volume ratio of microreactors allows
the effective transmission of light through the reaction mix-
ture, thus favoring diffusion and mass transfer to provide en-
hanced reaction efficiency and safety.[13c] On the other hand,
the use of continuous flow allows good control of reaction
conditions, which minimizes product photodegradation. The
general process is shown in Figure 6 and the results are sum-
marized in Table 1.
Figure 5. Regioselective ortho-palladation of (Z)-2-phenyl-4-hetarylidene-
5(4H)oxazolones 1n–t.
times and only in the cases of nitro- (1p) and benzothiophene
(1s) derivatives was heating necessary to achieve complete
conversion. The corresponding ortho-palladated dimers 2n–t
were obtained in very good yields as air-stable solids. In gener-
al, the ortho-palladation of hetarylidene-oxazolones 1n–t
occurs under milder reaction conditions and in shorter reaction
times than for the corresponding arylidene-oxazolones 1a–m
bearing the same substituents. For instance, ortho-palladation
of unsubstituted 4-benzylidene-oxazolone (1a) required heat-
ing at 758C for four hours to obtain 64% yield of the isolated
product 2a,[12b] whereas ortho-palladation of thiophene deriva-
tive 1n occurred at room temperature in only five minutes to
give 2n in 95% yield. The same conclusions can be drawn by
comparison between arylidene-oxazolones containing elec-
tron-releasing or electron-withdrawing groups and the corre-
sponding heterocyclic derivatives. Therefore, the electronic
nature of the ring is important for the outcome of the reaction,
but it is not critical, and favors palladation of the electron-rich
heterocyclic rings.
Our approach employed a microreactor that consisted of
a coil of fused silica capillary placed on top of a light source.
The total length of the capillary was 1.9 m (internal diameter
of 100 mm), which resulted in a reaction volume of 15 mL. The
light source consisted of a collection of 28 light-emitting
Chem. Eur. J. 2016, 22, 144 – 152
www.chemeurj.org
146
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 1. Optimization of the [2+2]-photocycloaddition of ortho-palladated complexes 2a–s.
Entry
2
t [h][a]
Conv. [%][b]
t [h][c]
Conversion [%][d]
Flow
Batch
white
blue
green
red
1
2
3
4
5
6
7
8
2a
2b
2c
2d
2e
2 f
2g
2h
2i
28
480
432
240
72
100
75
100
100
100
100
100
100
0
100
100
0
0
0
0.5
0.33
0.5
0.33
0.33
0.33
0.5
0.5
0.33
0.5
0.5
0.33
0.5
81 (red)
100 (blue)
100 (blue)
100 (blue)
98 (blue)
100 (blue)
100 (blue)
100 (blue)
70 (white)
56 (blue)
100 (blue)
34 (blue)
70 (white)
100 (blue)
0 (white)
22 (green)
80 (0.33)
100
–
–
–
88
70
–
100
52 (0.33)[e]
–
20
30
100
100
0
85
100
–
100
90
100
100
–
45
90
–
52
58
100
0
86
90
–
–
–
68
–
–
54
62
–
40
96
0
–
–
–
6
–
–
0
0
–
4
0
0
0
0
72
432
192
480
120
380
480
480
480
480
480
9
10
11
12
13
14
15
16
2j
2m
2n
2o
2p
2r
49
0.16
0.33
0.5
94 (0.33)[e]
0
0
0
20
2s
0
[a] Irradiation time using sunlight. [b] Conversion of substrate 2 into product 3 achieved during the time specified in [a]. [c] Irradiation time using LEDs as
the light source. [d] Conversion of substrate 2 into product 3 achieved during the time specified in [c]. [e] Irradiation time was 0.33 h instead of 0.16 h.
700 nm) and for different residence times (10, 20, and 30 min)
by simply changing the flow rate accordingly. The reaction
mixture (around 100 mL) was collected in a vial, transferred to
a 5 mm NMR tube and diluted with 400 mL of CDCl3. The prog-
1
ress of the reaction was monitored by H NMR spectroscopic
analysis of the reaction mixture at different times to determine
the maximum conversion achieved for each wavelength and
for each residence time. The reaction conditions that afforded
the best conversions in continuous flow were subsequently
transferred to batch mode by using a 5 mm NMR tube as the
reaction vessel. The mixture was directly illuminated by the
same system of LEDs.
It can be seen from the data presented in Table 1 that the
improvement in the efficiency of the reaction was outstanding
in two respects. Firstly, the acceleration experienced by reac-
tions that already took place under sunlight irradiation but re-
quired days (even weeks) to reach completion, which was the
case for most complexes. As a representative example, com-
pound 2a required 28 hours under sunlight for complete con-
version into 3a, as reported previously.[12a] On applying the
methodology developed here, the irradiation of 2a (entry 1)
Figure 6. [2+2]-Cycloaddition of ortho-palladated oxazolones.
under flow conditions with red LED lamps afforded a clean
diodes (LEDs) placed on a printed circuit board (PCB) (see the
Supporting Information for more details of the experimental
setup). The reaction volume was directly illuminated by the
PCB of LEDs. LEDs have recently become the energy saving
light source of choice in many applications[13c] and they are
available in a wide range of wavelengths.[14] Solutions of the
corresponding ortho-palladated derivatives 2a–s in CDCl3
(6 mm) were pumped through the capillary while being irradi-
ated with the LED printed board. The residence time was con-
trolled by the flow rate of a Harvard syringe pump. To optimize
the absorption of energy for each individual compound, four
different wavelengths were tested on each sample: blue
(465 nm), green (525 nm), red (625 nm), and white (405–
and virtually complete transformation of 2a into 3a as the
only product in reaction times as short as 30 minutes. It is
clear that, although the best result was obtained by using red
light, the compound is sensitive to changes in wavelength be-
cause high conversions (ꢀ80%) were also achieved with
white, blue, or green lights during the same period. On the
other hand, the enhancement of the reaction under the con-
tinuous flow regime is clear: 81% conversion in batch vs. 96%
in continuous flow with red light (625 nm).
Huge accelerations were also observed for 2c–h and 2m
(Table 1), for which complete conversions into the correspond-
ing photocycloaddition dimers 3c–h and 3m were achieved
on using blue light (465 nm) under either flow or batch condi-
Chem. Eur. J. 2016, 22, 144 – 152
www.chemeurj.org
147
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tions. In these cases (entries 3–8 and 11) the reaction times
were reduced from up to 18 days to barely 10–30 minutes. In
the case of 2j (entry 10), a lower conversion into 3j was ob-
tained but this remained as high as 90%. The transfer of the
optimized flow conditions to batch conditions did not lead to
significant changes, and complete conversions were also ob-
tained in minutes, again with the exception of 2j. Notably, the
progress of the reaction seems to be independent of the
nature and position of the substituents, because very similar
conversions were obtained in quite similar reaction times
when electron-donating (Me in 2d or 2e, OMe in 2b) or elec-
tron-withdrawing (Br in 2 f, Cl in 2g and 2h, NO2 in 2m) sub-
stituents were present in the ortho- or para-positions.
conditions also led to complete transformation into 3b in only
20 minutes. Therefore, not only is it possible to accelerate the
reaction, but the method also provides access to compounds
that cannot otherwise be made in a straightforward manner.
More impressive are the cases of complexes 2i and the
whole set of 4-hetarylidene derivatives 2n–s, because, in these
cases, conversion was not detected at all even after exposure
to sunlight for periods as long as 20 days (Table 1, entries 9
and 12–16). For complex 2i, 100% conversion into 3i was ach-
ieved on using white LED light under flow conditions in only
20 minutes, whereas 70% conversion was observed under
batch conditions with the same irradiation (entry 9). In addi-
tion, partial transformation of 2i into 3i was observed on
using blue (45%) and green (54%) LED lights, whereas the cy-
cloaddition did not take place on using red light. For com-
plexes 2n and 2o, partial conversions were observed under
both flow and batch conditions, with 2n proving to be more
sensitive to blue light (maximum conversion 52%) and 2o to
white light (maximum conversion 70%). Fortunately, better re-
sults were obtained for the 5-substituted thiophenes, regard-
less of the electron-releasing or electron-withdrawing nature of
the substituent. For example, 5-nitro derivative 2p underwent
the [2+2]-photocycloaddition with 100% conversion into 3p
on using blue light, under both flow and batch conditions,
with a reaction time of only 10 minutes. Excellent results were
also obtained when 2p was irradiated with white light (100%
conversion). In the case of the 5-methyl derivative 2r, 100%
conversion into 3r was observed after 20 minutes irradiation
with white light under flow conditions, but the reaction did
not proceed at all in batch mode. Finally, the presence of
a benzo moiety fused to the thiophene group seemed to deac-
tivate the system; in this case, only 22% conversion was ob-
served for 2s on using green light.
To assess the stability in solution of the photocycloaddition
dimers 3a, 3c–h, 3j, and 3m, further NMR measurements
were carried out on the corresponding solutions for several
hours after irradiation had finished. For the aforementioned
complexes, a reversal of the photocycloaddition was not ob-
served under those conditions because signals due to the
starting compounds were not detected during the monitoring
1
time. As a representative example (Figure S2), a set of H NMR
spectra for complex 3g taken every hour after irradiation had
finished is presented in the Supporting Information. The ab-
sence of new signals due to 2g in these spectra shows that
3g is stable in solution.
The marked decrease in the reaction time needed to reach
maximum conversions could be due to two factors; namely,
the selection of the wavelength and the large increase in the
surface to volume ratio, as is usual for flow conditions, in com-
parison with reaction conducted in a 5-mm NMR tube.[13–15] Al-
though it is difficult to quantify the contribution of each of
these factors to the substantial decrease in the time needed to
reach full conversion, it seems that a suitable choice of wave-
length is critical. In fact, once the reaction had been optimized
under flow conditions, it was possible to transfer the condi-
tions to a batch reactor without a detrimental effect on the
conversion. Note that several reactions proceeded with higher
conversions when working under continuous flow (2a, 2i, 2j,
and 2n) than under batch conditions. It is worth mentioning
that the optimization of the reaction conditions in continuous
flow is less time-consuming and less expensive than for the
batch conditions and this justified the approach followed.
The second remarkable improvement achieved with this
system concerns reactions that did not undergo complete
[2+2]-photocycloaddition upon irradiation with sunlight, or
those that did not proceed at all regardless of the exposure
time used. This is the case for complexes 2b, 2i, and 2n–s,
with some notable differences observed between them. In the
case of complex 2b, a mixture of 2b and 3b (1:3 molar ratio)
was obtained after exposure to sunlight for 20 days. This molar
ratio could not be improved upon further exposure. Notably,
irradiation of 2b with either blue (465 nm) or white LED lamps
under flow conditions with a residence time of only 20 mi-
nutes led to complete photocycloaddition and to 100% con-
version of 2b into 3b. A conversion of 90% was even ob-
served on using green light (525 nm) with the same residence
time. Moreover, irradiation of 2b with blue light under batch
The stability of the last set of complexes (namely 3b, 3i and
3n–s) in solution was assessed and it was found that the sta-
bility is somewhat limited. As mentioned above, the control of
1
the reverse reaction was evaluated by measuring the H NMR
spectra of the solutions of complexes of type 3 at regular in-
tervals for several hours after irradiation had finished; that is,
after maximum conversion had been attained (see examples in
the Supporting Information). In this way, it was established
that complex 3b reverted completely to 2b in approximately
12 hours (Figure S1 in the Supporting Information), whereas
complex 3i reverted into 2i in approximately 24 hours. This
finding is a drawback for this method because it limits to
some extent the scope of accessible 1,3-diaminotruxillic acid
derivatives. This reversal is more severe in the case of hetero-
cyclic derivatives 3n–o. As shown in Figures S3 and S4 in the
Supporting Information, the time required for complete rever-
sal of 3n to 2n is approximately 98 minutes, whereas that for
3o to 2o is around 24 hours. It is likely that the reversal of the
[2+2]-photocycloaddition is closely related to the long times
required for the reaction promoted by sunlight for these deriv-
atives.
The outstanding efficiency of the present system based on
LED irradiation and flow conditions in microreactors with high
surface-to-volume ratios is clear, because a significant improve-
Chem. Eur. J. 2016, 22, 144 – 152
www.chemeurj.org
148
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ment in the results was obtained in comparison to the use of
sunlight. Furthermore, the approach allows easy access to
compounds that are not attainable by other methodologies.
The only drawback found to date is the reversal of the photo-
cycloaddition in some cases. It seems that the nature of the
substituents has a clear influence on the outcome of the reac-
tion; however, it is difficult to establish correlations with the
present data.
were the mixtures 4d/5d and 4g/5g, which proved to be diffi-
cult to separate by chromatography even after exhaustive
screening of solvents. As a general rule, the hydrogenation tol-
erated well the presence of functional groups that are often
susceptible to hydrogenation, such as halogens or alkoxides. In
this respect, excellent results were obtained with substituents
such as 2-Cl (4g), 4-Cl (4h), 3’-OMe (4k), or 3’,4’-(OMe)2 (4l),
which remained intact at the end of the reaction. However, hy-
drogenation of the 4-Br derivative (3 f) occurred with concomi-
tant dehalogenation and 4a was obtained instead. The reac-
tion of 4-NO2 derivative 3m occurred with simultaneous depal-
ladation and hydrogenation of the nitro group to afford the di-
amino compound 4m’ (Figure 7).
The synthesis of compounds 4a–m’ as dimethyl esters
means that not only had depalladation taken place by hydro-
genation of the PdÀC bond, but that the resulting bis-oxazo-
lone had also reacted with the solvent (MeOH) to give the
methanolysis product by opening of the oxazolone ring. This
dependence of the reaction product on the nature of the sol-
vent was confirmed by performing the hydrogenation of com-
plex 4a in ethanol. In this case, the corresponding ethyl ester
derivative 6a was obtained. This finding is advantageous be-
cause it paves the way for the modular synthesis of 1,3-diami-
notruxillic acid esters that are substituted not only on the
phenyl rings, initially in the 4- and 2-positions, but also in the
carboxylate and the amido functions. Not surprisingly, when
the hydrogenation was carried out in aprotic solvents, regard-
less of their polarity (toluene, CH2Cl2 or THF), decomposition
was observed but definite products could not be identified
from the crude mixtures.
Liberation and characterization of e-truxillic acid derivatives
Once the experimental conditions for the [2+2]-photocycload-
ditions had been optimized and good yields of the photodi-
merized complexes 3a–s had been obtained, we focused on
the synthesis of palladium-free 1,3-diaminotruxillic acid esters.
To achieve this goal, depalladation of complexes of type 3 was
attempted. For this task, representative examples were select-
ed from cases in which the photocycloaddition had been
shown to be irreversible, with the aim of avoiding the forma-
tion of byproducts derived from the presence of type-2 com-
plexes. Several methods can be used to induce depalladation,
but we found that the best performance was achieved by
treatment of solutions of 3a–m in methanol with molecular
hydrogen. The hydrogenation process (see Figure 7) took place
The formation of the phenylalanine derivatives 5 is particu-
larly striking; however, it is not clear at which point they are
formed. Firstly, these compounds were not detected in all
cases. Secondly, the stability in solution of the type-3 com-
plexes was studied for several hours and it was confirmed that
reversal to type-2 products did not occur in detectable
amounts during these periods, which are longer than the time
required for hydrogenation. For instance, the evolution of 3g
in solution was evaluated for 10 hours (Figure S3) and evi-
dence for a retro-[2+2] reaction and formation of starting com-
plex 2g was not found. However, the hydrogenation of 3g
produced a mixture of 4g and 5g, as described above. There-
fore, it seems unlikely that the presence of phenylalanines 5
should be due to the depalladation of complexes 2, formation
of the corresponding oxazolone 1 and hydrogenation of the
oxazolone. Another possibility is the reversal of the diamino-
truxillic acid derivatives 4 to the corresponding dehydrophen-
ylalanines and subsequent hydrogenation in the presence of
Pd0 to give the phenylalanine. This also appears to be unlikely
because the stability of representative examples of the truxillic
acid derivatives (4a, 4d, 4e, and 4m’) was checked and evi-
dence of further evolution of these species in solution was not
found.
Figure 7. Depalladation of representative examples and liberation of alkyl
esters of the 1,3-diaminotruxillic acids.
under mild conditions (room temperature, pH2 =1 atm) and in
reaction times ranging from 30 minutes to 2 hours. After the
reaction, simple filtration of the resulting suspension removed
the Pd0, and evaporation to dryness of the resulting solution
afforded a mixture of two compounds. The major components
of these mixtures correspond to the methyl esters of 1,3-diami-
notruxillic acid derivatives 4a–m’ (Figure 7), which were ob-
tained in approximately 40% yield. The minor components of
the mixtures were the phenylalanine derivatives 5a–m’, which
were formed in small amounts, with typical yields of 5–10%.
In general, the truxillic acid esters 4 can be easily separated
from phenylalanines 5 by standard column chromatography
using ethyl acetate/n-hexane mixtures. The only exceptions
Characterization of the truxillic acid derivatives was carried
out by using standard spectroscopic and analytical methods
(see the Supporting Information). Concerning the optical activi-
ty of these derivatives, the measurement of the optical rota-
Chem. Eur. J. 2016, 22, 144 – 152
www.chemeurj.org
149
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tion of a mixture of 4a/5a gave a value of zero. This is expect-
ed bearing in mind that 4a is achiral and that the hydrogena-
tion takes place in the absence of chiral sources and, therefore,
5a must be obtained as the racemic mixture.
It can clearly be observed from the structure of the diamino-
truxillic acid derivative that it is the e-isomer, which is consis-
tent with the original assignment of isomers by Stoermer and
BachØr.[17] Thus, the two carboxylate groups in positions 1 and
3 are in a cis disposition, as are the two phenyl rings in posi-
tions 2 and 4, whereas the phenyl rings and the carboxylate
units are in mutual trans positions. This situation is clearly the
result of a head-to-tail syn-dimerization of two (Z)-oxazolones,
which is the same arrangement that was found for the ortho-
palladated cycloadducts 3.[12a] This finding indicates that con-
formational changes had not occurred during the hydrogena-
tion. The cyclobutane core is not planar and has a slightly
folded structure. The out-of-plane dihedral angles C(1)-C(2)-
C(3)-C(4) and C(2)-C(3)-C(4)-C(1) have values of À21.8(3)8 and
21.9(3)8, which are similar to values found for other cyclobu-
tane rings with these types of substituents.[5d,18] The two
phenyl rings in the 2- and 4-positions are rotated by around
908 with respect to one another, and similar rotations were
found in closely related structures.[5d] The phenylalanine unit
5a forms strong H-bonds with the truxillic acid derivative 4a,
which is probably one of the reasons for the easy co-crystalli-
zation. The H-bonds are established between the two NH pro-
tons of the truxillic acid derivative 4a, which behaves as a H-
donor, and the carbonyl oxygen of the amido group of the
phenylalanine fragment 5a as a H-acceptor. The parameters
that define these H-bonds are: For the N(1)ÀH(1A)···O(9) H-
bond N(1)···O(9)=3.110(5) Š, H(1A)···O(9)=2.27 Š and N(1)-
H(1A)-O(9)=164.08, whereas for the N(2)-H(2A)···O(9) H-bond
N(2)···O(9)=3.024(5) Š, H(2A)···O(9)=2.17 and N(2)-H(2A)-O(9)=
169.68, both corresponding to H-bonds of moderate strength,
as defined by Steiner.[19] Finally, the internal parameters for
bond lengths and angles are as one would expect, with some
deviations with respect to reported values for similar bonding
situations.[5d,20]
Compounds 4a and 5a were also characterized by X-ray dif-
fraction methods. Attempts to grow single crystals of each iso-
lated compound 4 were unsuccessful. In contrast, crystals that
were suitable for diffraction were obtained by crystallization of
the crude mixture of truxillic acid derivative 4a and the corre-
sponding phenylalanine 5a; that is, the mixture obtained just
after hydrogenation and removal of the Pd. The difficulties en-
countered in crystallizing each separate member were remark-
able given that the mixture 4a/5a crystallized easily. Two fig-
ures of the co-crystallized mixture 4a/5a are provided in the
Supporting Information, along with a detailed analysis of the
structure. The structure of the e-diaminotruxillic acid derivative
4a is shown in Figure 8.
Figure 8. X-ray crystal structure of e-diaminotruxillic acid derivative 4a. Se-
lected lengths [Š] and angles [8]: N(1)ÀC(19): 1.349(4), N(1)ÀC(4): 1.437(4),
N(2)ÀC(28): 1.361(4), N(2)ÀC(2): 1.458(4), O(1)ÀC(17): 1.198(4), O(3)ÀC(19):
1.240(4), O(4)ÀC(26): 1.203(4), O(6)ÀC(28): 1.217(4), C(1)ÀC(5): 1.506(5), C(1)À
C(4): 1.563(4), C(1)ÀC(2): 1.567(5), C(2)ÀC(26): 1.540(5), C(2)ÀC(3): 1.568(4),
C(3)ÀC(11): 1.499(5), C(3)ÀC(4): 1.559(4), C(4)ÀC(17): 1.534(5), C(19)-N(1)-C(4):
121.4(3), C(28)-N(2)-C(2): 119.8(3), C(5)-C(1)-C(4): 122.1(3), C(5)-C(1)-C(2):
122.6(3), C(4)-C(1)-C(2): 88.7(3), N(2)-C(2)-C(26): 110.5(3), N(2)-C(2)-C(1):
112.3(3), C(26)-C(2)-C(1): 113.9(3), N(2)-C(2)-C(3): 118.6(3), C(26)-C(2)-C(3):
112.9(3), C(1)-C(2)-C(3): 86.7(2), C(11)-C(3)-C(4): 122.6(3), C(11)-C(3)-C(2):
125.7(3), C(4)-C(3)-C(2): 88.8(3), N(1)-C(4)-C(17): 111.7(3), N(1)-C(4)-C(3):
115.0(3), C(17)-C(4)-C(3): 113.3(3), N(1)-C(4)-C(1): 114.2(3), C(17)-C(4)-C(1):
113.5(3), C(3)-C(4)-C(1): 87.2(3).
Conclusions
The stereoselective synthesis of a wide variety of 1,3-diamino-
truxillic acid derivatives, as the e-isomers, has been accom-
plished starting from (Z)-4-(het)aryliden-2-aryl-5(4H)-oxazolones
by an optimized three-step method. This procedure involves
the regioselective ortho-palladation of the oxazolones, the
[2+2]-photocycloaddition of the ortho-palladated complexes
induced by irradiation with an LED light source, either in mi-
croreactors under continuous flow conditions or under batch
conditions, and the liberation of the truxillic acid derivatives as
dialkyl esters by hydrogenation in alcoholic media. The reac-
tion has a wide scope for 4-arylidene-oxazolones and tolerates
substituents of different electronic nature at different positions
of both the 4-arylidene and the 2-aryl rings. In all cases, the re-
action takes place in the range of hours and, because of the
templating effect of the ortho-palladation, it is stereoselective
to the e-isomer. All of these characteristics (wide scope, short
reaction times, and full stereoselectivity) represent improve-
ments on the present availability of these types of compounds,
which are of high added value because of their pharmacologi-
cal activity. Limitations in the scope of the reaction are due to
The mixture 4a/5a crystallized in the orthorhombic space
group P212121, with one molecule of e-diaminotruxillic (4a)
and one N-benzoylphenylalanine derivative (5a) (i.e., one unit
of the mixture) in the asymmetric part. This means that only
one enantiomer of phenylalanine 5a had crystallized together
with the achiral 4a. Given that the value of the optical rotation
of the mixture 4a/5a was zero, this unexpected spontaneous
resolution could be explained as being due to a “racemic con-
glomerate”.[16] Given the absence of heavy atoms, it was not
possible to determine unambiguously the absolute configura-
tion of the Ca of the phenylalanine moiety.
Chem. Eur. J. 2016, 22, 144 – 152
www.chemeurj.org
150
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(100.61 MHz, CDCl3): d=20.97 (CH3), 56.40 (CH), 70.35 (C), 122.27
(C), 126.96 (CH), 127.20 (C), 128.00 (CH), 129.47 (CH), 130.89 (CH),
131.66 (CH), 136.07 (C), 136.10 (CH), 137.58 (C), 168.54 (C),
174.07 ppm (C); 19F NMR (376.50 MHz, CDCl3): d=À75.22 ppm; IR
(nujol): n˜ =1837 (C=O), 1655 cmÀ1 (C=N); MS (MALDI+): m/z (%):
368.0 (21.4) [M/2ÀCF3COOÀ]+; elemental analysis calcd (%) for
C38H24F6N2O8Pd2·1/3CDCl3 (1003.25): C 45.85, H 2.44, N 2.79; found:
C 45.83, H 2.43, N 2.75.
the reversal of the photochemical process. This aspect requires
improvement and is under investigation.
The methodology developed here combines the advantages
of microreactor devices with a continuous flow regime. It ena-
bles reproducibility, process reliability, and a rapid preparation
with minimum workup for a wide collection of photocycload-
dition dimers. The high surface-to-volume ratio typical of mi-
croreactors (resulting in short diffusion paths) in combination
with the optimum wavelength enhances the reaction in terms
of reaction yields and reaction times when compared with
direct irradiation with sunlight. This approach also enables
access to several photocycloaddition dimers that cannot be
obtained otherwise. The method described here permits rapid
experimentation and scale-up, thus shortening the time from
research to development and production, which is an interest-
ing field considering the properties and applications of these
truxillic acid derivatives. Scalability of this kind of reaction is
simply a matter of pumping and irradiating the starting materi-
al with LEDs continuously through the microreactor. Once
again, microreactor technology has enabled the development
of a green and sustainable process with lower energy de-
mands.[21]
Synthesis of e-1,3-diaminotruxillic derivatives: The synthesis of
4a is described here; data for 4d–m’ are collected in the Support-
ing Information. A yellow solution of 3a (0.204 g, 0.22 mmol) in
MeOH (20 mL) was stirred under a H2 atmosphere for 30 min.
When the reaction was complete, the black suspension was filtered
through a pad of Celite and the resulting colorless solution was
evaporated to dryness to afford a mixture of aminotruxillic acid de-
rivative 4a and phenylalanine 5a (6.7:1 molar ratio), which were
separated by silica gel chromatography (ethyl acetate/n-hexane,
40:30). The first colorless band collected corresponded to com-
pound 5a. Further elution using the same eluent allowed the isola-
tion of aminotruxillic acid derivative 4a. Compound 4a crystallized
with a molecule of ethyl acetate, the presence of which was deter-
1
mined by H NMR spectroscopy. The crystals were used for analyti-
cal purposes.
Compound 4a: Yield: 46.3 mg (37%); white solid; 1H NMR
(400.13 MHz, CDCl3): d=3.90 (s, 3H; OCH3), 5.29 (s, 1H; CH), 7.30 (t,
3
3J=7.3 Hz, 2H; H3’’, H5’’), 7.32 (m, 1H; H4’’), 7.37 (t, J=7.8 Hz, 2H;
H3’, H5’), 7.45 (t, 3J=7.8 Hz, 1H; H4’), 7.52 (d, 3J=7.3 Hz, 2H; H2’’,
H6’’), 7.57 (d, 3J=7.8 Hz, 2H; H2’, H6’), 7.92 ppm (br. s, 1H; NH);
13C{1H} NMR (100.61 MHz, CDCl3): d=50.83 (CH), 53.51 (CH3), 62.81
(C), 126.98 (CH), 127.80 (CH), 128.61 (CH), 128.70 (CH), 128.80 (CH),
131.79 (CH), 133.03 (C), 133.79 (C), 167.24 (CON), 172.87 ppm (CO2);
IR (nujol): n˜ =3283 (NH), 1735, 1660 cmÀ1 (C=O); MS (ESI+): m/z:
563.2 [M+H]+; elemental analysis calcd (%) for C34H30N2O6·EtOAc
(650.3): C 70.14, H 5.89, N 4.31; found: C 70.02, H 5.86, N 4.41.
Compound 5a: Yield: 13.7 mg (11%); pale-yellow solid; 1H NMR
(400.13 MHz, CDCl3): d=3.26 (m, 2H; CH2), 3.77 (s, 3H; OCH3), 5.10
(m, 1H; CH), 6.59 (br. s, 1H; NH), 7.14 (m, 2H; Hm, CH2Ph), 7.07–
7.50 (m, 3H; Ho +Hp, CH2Ph), 7.43 (t, 3J=7.5 Hz, 2H; Hm, COPh),
7.50 (t, 3J=7.4 Hz, 1H; Hp, COPh), 7.72 (d, 3J=7.5 Hz, 2H; Ho,
COPh); IR (nujol): n˜ =3318 (NH), 1735, 1637 cmÀ1 (C=O); MS (ESI+):
m/z: 284.4 [M+H]+.
Experimental Section
General methods: See the Supporting Information for full details.
Synthesis of ortho-palladated complexes by CÀH bond activa-
tion—typical procedure: The synthesis of 2n is described here;
data for 2o–t are collected in the Supporting Information.
Pd(OAc)2 (205 mg, 0.91 mmol) was added to a solution of 1n
(232.1 mg, 0.91 mmol) in CF3CO2H (5 mL) and the resulting mixture
was stirred at RT for 5 min. During this time an orange solid pre-
cipitated. Distilled water (50 mL) was added to the resulting sus-
pension and stirring was maintained for a further 15 min. The re-
sulting orange precipitate was filtered off, washed with water (3”
20 mL), dried in vacuo, and characterized as 2n. Yield: 420.4 mg
(97%). 1H NMR (400.13 MHz, CDCl3): d=6.87 (d, 3J=5.1 Hz, 1H;
3
3
SC4H2), 7.47 (t, J=7.7 Hz, 2H; H3’, H5’), 7.57 (s, 1H; H7’’), 7.61 (t, J=
7.5 Hz, 1H; H4’), 7.72 (d, 3J=5.1 Hz, 1H; SC4H2), 7.93 ppm (d, 3J=
7.3 Hz, 2H; H2’, H6’); 13C{1H} NMR (100.61 MHz, CDCl3): d=120.6 (C),
122.6 (C), 127.2 (C), 128.6 (CH), 129.4 (CH), 131.0 (CH), 133.2 (CH),
133.6 (CH), 134.8 (CH), 137.3 (C), 160.6 (C), 168.0 ppm (C) (13C sig-
nals assigned to the CF3CO2 ligand were not observed); 19F NMR
(376.50 MHz, CDCl3): d=À74.75 ppm; IR (nujol): n˜ =1789 (C=O),
1662 (C=C), 1625 cmÀ1 (C=N); elemental analysis calcd (%) for
C32H16F6N2O8Pd2S2: C 40.57, H 1.70, N 2.96, S 6.77; found: C 40.61, H
1.61, N 2.87, S 6.73.
X-ray crystallography: CCDC 1401532 (4a/5a) contains the sup-
plementary crystallographic data for this paper. These data are pro-
vided free of charge by The Cambridge Crystallographic Data
Centre. For specific experimental details, see the Supporting Infor-
mation.
Acknowledgements
Funding by the Ministerio de Economia y Competitividad
(MINECO) (Spain, Projects CTQ2011-22589, CTQ2011-22410 and
CTQ2013-40855-R, CTQ2014-54987-P) and Gobierno de
Aragón-Fondo Social Europeo (Spain, groups E40 and E97) is
gratefully acknowledged. M.V.G. thanks the Ministerio de Econ-
[2+2]-Photocycloaddition of ortho-palladated complexes—typi-
cal procedure: The synthesis of 3d is described here; data for 3a–
s are collected in the Supporting Information. Optimized proce-
dure using a continuous flow microreactor and LEDs as the light
source: A solution of ortho-palladated 2d (6 mm in CDCl3) was
pumped through a fused silica capillary with a residence time of
20 min. The capillary coil was irradiated during this time with a col-
lection of blue LEDs (465 nm) mounted on a PCB. The irradiated
solution was collected in a small vial. Evaporation of the solvent
omia
y Competitividad (Spain) for funding her contract
through the Ramón y Cajal program and Parque Científico y
Tecnológico de Castilla-La Mancha.
1
gave 3d in quantitative yield. H NMR (400.13 MHz, CDCl3): d=2.22
(s, 3H; CH3), 5.36 (s, 1H; H7’’), 6.76–6.84 (m, 2H; H5’’, H6’’), 6.98 (d,
Keywords: CÀH activation
palladium · photochemistry
· lactones · microreactors ·
3J=7.6 Hz, 1H; H4’’), 7.69 (t, 3J=7.7 Hz, 2H; H3’, H5’), 7.78 (t, 3J=
3
7.9 Hz, 1H; H4’), 9.16 ppm (d, J=7.7 Hz, 2H; H2’, H6’); 13C{1H} NMR
Chem. Eur. J. 2016, 22, 144 – 152
www.chemeurj.org
151
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[1] A. Pictet in The Vegetable Alkaloids, With Particular Reference to Their
Chemical Constitution, Wiley, New York, 1913, p. 236.
[2] a) C. Liebermann, Ber. Dtsch. Chem. Ges. 1888, 21, 2342; b) C. Lieber-
mann, Ber. Dtsch. Chem. Ges. 1888, 21, 3196; c) C. Liebermann, Ber.
Dtsch. Chem. Ges. 1888, 21, 3372; d) O. Hesse, Pharm. Ztg. 1887, 32,
407.
gaard, K. A. Joergensen, J. Am. Chem. Soc. 2008, 130, 12031; j) J. S. Fisk,
R. A. Mosey, J. J. Tepe, Chem. Soc. Rev. 2007, 36, 1432; k) C. Cativiela,
M. D. Díaz-de-Villegas, Tetrahedron: Asymmetry 2007, 18, 569; l) C. E. Au-
gelli-Szafran, A. T. Sakkab, T. Yasunaga, U.S. Pat. Appl. Publ. US
20040180943 A1, 2004.
[8] D. Lawrenz, S. Mohr, B. Wendländer, J. Chem. Soc. Chem. Commun.
1984, 863.
[3] Antinociceptive activity: a) A. Sergeiko, V. V. Poroikov, L. O. Hanus, V. M.
Dembitsky, Open Med. Chem. J. 2008, 2, 26; b) W. T. Berger, B. P. Ralph,
M. Kaczocha, J. Sun, T. E. Balius, R. C. Rizzo, S. Haj-Dahmane, I. Ojima,
D. G. Deutsch, PLoS One 2012, 7, e50968; c) M. Nakamura, T. Yoshizawa,
Y.-M. Chi, T. Nohara, S. Sakurada, European Patent Application
EP1118324A1, 2001; d) Y.-M. Chi, M. Nakamura, X.-Y. Zhao, T. Yoshizawa,
W.-M. Yan, F. Hashimoto, J. Kinjo, T. Nohara, S. Sakurada, Biol. Pharm.
Bull. 2006, 29, 580; e) Y.-M. Chi, M. Nakamura, X.-Y. Zhao, T. Yoshizawa,
W.-M. Yan, F. Hashimoto, J. Kinjo, T. Nohara, S. Sakurada, Biol. Pharm.
Bull. 2006, 29, 489; f) Y.-M. Chi, M. Nakamura, T. Yoshizawa, X.-Y. Zhao,
W.-M. Yan, F. Hashimoto, J. Kinjo, T. Nohara, S. Sakurada, Biol. Pharm.
Bull. 2005, 28, 1776; g) S. X. Liu, H. Z. Jin, L. Shan, H. W. Zeng, B. Y. Chen,
Q. Y. Sun, W. D. Zhang, Bioorg. Med. Chem. Lett. 2013, 23, 2207; h) M.
Kaczocha, S. T. Glaser, D. G. Deutsch, Proc. Natl. Acad. Sci. USA 2009,
106, 6375.
[9] a) I. Funes-Ardoiz, M. Blanco-Lomas, P. J. Campos, D. Sampedro, Tetrahe-
dron 2013, 69, 9766; b) M. Blanco-Lomas, I. Funes-Ardoiz, P. J. Campos,
D. Sampedro, Eur. J. Org. Chem. 2013, 6611; c) M. Blanco-Lomas, P. J.
Campos, D. Sampedro, Org. Lett. 2012, 14, 4334; d) K. Kubo, M. Koshiba,
H. Hoshina, T. Sakurai, Heterocycles 1998, 48, 25 (ab-dehydrophenylala-
nines).
[10] a) M. Ichikawa, M. Takahashi, S. Aoyagi, C. Kibayashi, J. Am. Chem. Soc.
2004, 126, 16553; b) A. G. Jarvis, H. A. Sparkes, S. E. Tallentire, L. E.
Hatcher, M. R. Warren, P. R. Raithby, D. R. Allan, A. C. Whitwood, M. C. R.
Cockett, S. B. Duckett, J. L. Clark, I. J. S. Fairlamb, CrystEngComm 2012,
ˇ
14, 5564; c) T. Runcevski, M. Blanco-Lomas, M. Marazzi, M. Cejuela, D.
Sampedro, R. E. Dinnebier, Angew. Chem. Int. Ed. 2014, 53, 6738; Angew.
Chem. 2014, 126, 6856.
[11] Selected contributions dealing with application of CÀH functionalization
logic to cyclobutane synthesis: a) W. R. Gutekunst, P. S. Baran, J. Am.
Chem. Soc. 2011, 133, 19076; b) F. FrØbault, N. Maulide, Angew. Chem.
Int. Ed. 2012, 51, 2815; Angew. Chem. 2012, 124, 2869; c) W. R. Gute-
kunst, R. Gianatassio, P. S. Baran, Angew. Chem. Int. Ed. 2012, 51, 7507;
Angew. Chem. 2012, 124, 7625; d) R. Liu, M. Zhang, T. P. Wyche, G. N.
Winston-McPherson, T. S. Bugni, W. Tang, Angew. Chem. Int. Ed. 2012, 51,
7503; Angew. Chem. 2012, 124, 7621; e) R. Parella, B. Gopalakrishnan,
S. A. Babu, J. Org. Chem. 2013, 78, 11911; f) W. Gutekunst, P. S. Baran, J.
Org. Chem. 2014, 79, 2430. A review: g) Y. J. Hong, D. J. Tantillo, Chem.
Soc. Rev. 2014, 43, 5042. Photoinduced electron transfer: h) M. Riener,
D. A. Nicewicz, Chem. Sci. 2013, 4, 2625.
[12] a) D. Roiban, E. Serrano, T. Soler, I. Grosu, C. Cativiela, E. P. Urriolabeitia,
Chem. Commun. 2009, 4681; b) G.-D. Roiban, E. Serrano, T. Soler, G.
Aullón, I. Grosu, C. Cativiela, M. Martínez, E. P. Urriolabeitia, Inorg. Chem.
2011, 50, 8132.
[13] a) B. P. Mason, K. E. Price, J. L. Steinbacher, A. R. Bogdan, D. T. McQuade,
Chem. Rev. 2007, 107, 2300; b) A. Yavorskyy, O. Shvydkiv, N. Hoffmann,
K. Nolan, M. Oelgemçller, Org. Lett. 2012, 14, 4342; c) Y. Su, N. J. W.
Straathof, V. Hessel, T. Nçel, Chem. Eur. J. 2014, 20, 10562.
[14] a) U. Megerle, R. Lechner, B. Kçnig, E. Riedle, Photochem. Photobiol. Sci.
2010, 9, 1400; b) J. W. Tucker, Y. Zhang, T. F. Jamison, C. R. J. Stephen-
son, Angew. Chem. Int. Ed. 2012, 51, 4144; Angew. Chem. 2012, 124,
4220.
[15] a) J. P. Knowles, L. D. Elliot, K. I. Booker-Milburn, Beilstein J. Org. Chem.
2012, 8, 2025; b) S. B. Ötvçs, F. Fülçp, Catal. Sci. Technol. 2015, 5, 4926;
c) K. Gilmore, P. H. Seeberger, Chem. Rec. 2014, 14, 410; d) Y. Nishiyama,
R. Mori, K. Nishida, H. Tanimoto, T. Morimoto, K. Kakiuchi, J. Flow Chem.
2014, 4, 185; e) K. Terao, Y. Nishiyama, H. Tanimoto, T. Morimoto, M. Oel-
gemçller, T. Morimoto, J. Flow Chem. 2012, 2, 73.
[4] Antimuscarinic activity: M. Lysíkovµ, K. Fuksovµ, T. Elbert, J. Jakubík, S.
ˇ
Tucek, Br. J. Pharmacol. 1999, 127, 1240.
[5] Aminotruxillic derivatives as the only non-peptidic agonists of GLP-1R:
a) D. Wootten, E. E. Savage, F. S. Willard, A. B. Bueno, K. W. Sloop, A.
Christopoulos, P. M. Sexton, Mol. Pharmacol. 2013, 83, 822; b) G.-B. Ge,
C.-Z. Ai, W.-B. Hu, J. Hou, L.-L. Zhu, G.-Y. He, Z.-Z. Fang, S.-C. Liang, F. Y.
Wang, L. Yang, Eur. J. Pharm. Sci. 2013, 48, 360; c) N. Guan, W. Gao, M.
He, M. Zheng, X. Xu, X. Wang, M.-W. Wang, Anal. Biochem. 2012, 423,
61; d) Q. Liu, N. Li, Y. Yuan, H. Lu, X. Wu, C. Zhou, M. He, H. Su, M.
Zhang, J. Wang, B. Wang, Y. Wang, D. Ma, Y. Ye, H.-C. Weiss, E. R. F.
Gesing, J. Liao, M.-W. Wang, J. Med. Chem. 2012, 55, 250; e) M. He, N.
Guan, W.-W. Gao, Q. Liu, X.-Y. Wu, D.-W. Ma, D.-F. Zhong, G.-B. Ge, C. Li,
X.-Y. Chen, L. Yang, J.-Y. Liao, M. W. Wang, Acta Pharm. Sin. 2012, 33,
148; f) M. He, H. Su, W. Gao, S. M. Johansson, Q. Liu, X. Wu, J. Liao, A. A.
Young, T. Bartfai, M.-W. Wang, PLoS One 2010, 5, e14205; g) F. Lin, R.
Wang, J. Mol. Model. 2009, 15, 53; h) M.-W. Wang, Q. Liu, L. Lin, Y.
Zhang, L. Zhu, Y. Shu (Faming Zhuanli Shenqing), Chinese Patent CN
101274927 A 20081001, 2008; i) H. Su, M. He, H. Li, Q. Liu, J. Wang, Y.
Wang, W. Gao, L. Zhou, J. Liao, A. A. Young, M.-W. Wang, PLoS One
2008, 3, e2892; j) M.-W. Wang, D. Chen, N. Li, Q. Liu, J. Liao, PCT Int.
Appl. WO 2008067711 A1, 2008; k) M.-W. Wang, N. Li, Q. Liu, L. Lin, Y.
Zhang, J. Liao, PCT Int. Appl. WO 2008067709 A1, 2008; l) D. Chen, J.
Liao, N. Li, C. Zhou, Q. Liu, G. Wang, R. Zhang, S. Zhang, L. Lin, K. Chen,
X. Xie, F. Nan, A. A. Young, M.-W. Wang, Proc. Natl. Acad. Sci. USA 2007,
104, 943; m) M.-W. Wang, N. Li, G. Wang, X. Xie, Q. Liu, C. Zhou, F. Nan,
PCT Int. Appl. WO 2006136101 A1, 2006.
[6] a) H. I. Bernstein, W. C. Quimby, J. Am. Chem. Soc. 1943, 65, 1845;
b) G. M. J. Schmidt, J. Chem. Soc. 1964, 2014; c) G. M. J. Schmidt, Pure
Appl. Chem. 1971, 27, 647; d) V. Enkelmann, K. B. Wagener, J. Am. Chem.
Soc. 1993, 115, 10390; e) S. M. Hein, J. Chem. Educ. 2006, 83, 940; f) M.
Khan, G. Brunklaus, V. Enkelmann, H.-W. Spiess, J. Am. Chem. Soc. 2008,
130, 1741.
[16] H. D. Flack, G. Bernardinelly, Acta Crystallogr. Sect. A 1999, 55, 908.
[17] a) R. Stoermer, F. BachØr, Ber. Dtsch. Chem. Ges. 1922, 55, 1860; b) R.
Stoermer, F. BachØr, Ber. Dtsch. Bot. Ges. 1924, 57, 15.
[7] a) C. Cativiela, M. D. Díaz-de-Villegas in The Chemistry of Heterocyclic
Compounds, Vol. 60 (Ed.: D. C. Palmer), Wiley, New York, 2004, p. 129;
b) F. W. Hartner Jr., in Comprehensive Heterocyclic Chemistry II, Vol. 3 (Ed.:
I. Shinkai), Pergamon, Oxford, UK, 1996, p. 261; c) “Oxazoles”: Y. S. Rao,
R. Filler in The Chemistry of Heterocyclic Compounds; Vol. 45 (Ed.: I. J.
Turchi), Wiley, New York, 1986, p. 363; d) G. V. Boyd in Comprehensive
Heterocyclic Chemistry, Vol. 6 (Ed.: K. T. Potts), Pergamon, Oxford, UK,
1984, pp. 177–234; e) T. Cleary, J. Brice, N. Kennedy, F. Chavez, Tetrahe-
dron Lett. 2010, 51, 625; f) C. Cativiela, M. OrdóÇez, Tetrahedron: Asym-
metry 2009, 20, 1; g) N. M. Hewlett, C. D. Hupp, J. J. Tepe, Synthesis
2009, 2825; h) R. A. Mosey, J. S. Fisk, T. L. Friebe, J. J. Tepe, Org. Lett.
2008, 10, 825; i) S. Cabrera, E. Reyes, J. Alemµn, A. Milelli, S. Kobbel-
[18] a) K. Gnanaguru, N. Ramasubbu, K. Venkatesan, V. Ramamurthy, J. Org.
Chem. 1985, 50, 2337; b) R. S. Harvey, B. V. McNally, C. J. Timmons, S. C.
Wallwork, Cryst. Struct. Commun. 1974, 3, 747.
[19] T. Steiner, Angew. Chem. Int. Ed. 2002, 41, 48; Angew. Chem. 2002, 114,
50.
[20] F. H. Allen, O. Kennard, D. G. Watson, L. Brammer, A. G. Orpen, R. Taylor,
J. Chem. Soc. Perkin Trans. 2 1987, S1.
[21] J. Yoshida, H. Kim, A. Nagaki, ChemSusChem 2011, 4, 331.
Received: September 17, 2015
Published online on November 24, 2015
Chem. Eur. J. 2016, 22, 144 – 152
www.chemeurj.org
152
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim