Detailed Structural Analysis of a Self-Assembled Vesicular Amphiphilic NCN-Pincer Palladium Complex by Using Wide-Angle X-Ray Scattering and Molecular Dynamics Calculations

Article abstract of DOI:10.1002/chem.201603494

Wide-angle X-ray scattering experiments and all-atomistic molecular dynamics calculations were performed to elucidate the detailed structure of bilayer vesicles constructed by self-assembly of an amphiphilic palladium NCN-pincer complex. We found an excel

Full text of DOI:10.1002/chem.201603494

DOI: 10.1002/chem.201603494
Full Paper
&
Bilayer Vesicles |Hot Paper|
Detailed Structural Analysis of a Self-Assembled Vesicular
Amphiphilic NCN-Pincer Palladium Complex by Using Wide-Angle
X-Ray Scattering and Molecular Dynamics Calculations
Go Hamasaka,[a, b] Tsubasa Muto,[a, b] Yoshimichi Andoh,[c] Kazushi Fujimoto,[d] Kenichi Kato,[e]
Masaki Takata,[e] Susumu Okazaki,*[d] and Yasuhiro Uozumi*[a, b, f, g]
Abstract: Wide-angle X-ray scattering experiments and all-
atomistic molecular dynamics calculations were performed
to elucidate the detailed structure of bilayer vesicles con-
structed by self-assembly of an amphiphilic palladium NCN-
pincer complex. We found an excellent agreement between
the experimental and calculated X-ray spectra, and between
the membrane thickness determined from a TEM image and
that calculated from an electron-density profile, which indi-
cated that the calculated structure was highly reliable. The
analysis of the simulated bilayer structure showed that in
general the membrane was softer than other phospholipid
bilayer membranes. In this bilayer assemblage, the degree of
alignment of complex molecules in the bilayer membrane
was quite low. An analysis of the electron-density profile
shows that the bilayer assemblage contains a space through
which organic molecules can exit. Furthermore, the catalyti-
cally active center is near this space and is easily accessible
by organic molecules, which permits the bilayer membrane
to act as a nanoreactor. The free energy of permeation of
water through the bilayer membrane of the amphiphilic
complex was 12 kJmolÀ1, which is much lower than that for
phospholipid bilayer membranes in general. Organic mole-
cules are expected to pass though the bilayer membrane.
The self-assembled vesicles were shown to be catalytically
active in a Miyaura–Michael reaction in water.
Introduction
ported that amphiphilic molecules that contain a catalytically
active site self-assemble to form catalytically active self-assem-
bled vesicles that can be used to induce organic reactions in
water.[2–4] We recently reported the formation of self-assembled
vesicles of amphiphilic palladium pincer complexes and their
use in the catalysis of reactions in water (Figure 1).[5–7] The for-
mation of bilayer vesicles was shown to be essential for the ef-
ficient promotion of reactions in water. The promotion of the
As nanomaterials, bilayer vesicles constructed by the self-as-
sembly of amphiphilic molecules have received much attention
from a wide range of scientists.[1] In particular, it has been re-
[a] Dr. G. Hamasaka, Dr. T. Muto, Prof. Dr. Y. Uozumi
Institute for Molecular Science, Myodaiji, Okazaki 444-8787 (Japan)
E-mail: uo@ims.ac.jp
[b] Dr. G. Hamasaka, Dr. T. Muto, Prof. Dr. Y. Uozumi
SOKENDAI (The Graduate University for Advanced Studies)
Myodaiji, Okazaki 444-8787 (Japan)
[c] Dr. Y. Andoh
High-Performance Computation Section
Center for Computational Science, Nagoya University
Furo-cho, Chikusa-ku, Nagoya 464-8603 (Japan)
[d] Dr. K. Fujimoto, Prof. Dr. S. Okazaki
Department of Applied Chemistry
Faculty of Engineering, Nagoya University
Furo-cho, Chikusa-ku, Nagoya 464-8603, (Japan)
E-mail: okazaki@apchem.nagoya-u.ac.jp
[e] Dr. K. Kato, Prof. Dr. M. Takata
RIKEN SPring-8 Center, 1-1-1 Kouto, Sayo-cho, Sayo-gun 679-5148 (Japan)
[f] Prof. Dr. Y. Uozumi
RIKEN Center for Sustainable Resource Science, Wako 351-0198 (Japan)
E-mail: uo@riken.jp
[g] Prof. Dr. Y. Uozumi
JST-CREST and JST-ACCEL, Myodaiji, Okazaki 444-8787 (Japan)
Supporting information for this article can be found under
http://dx.doi.org/10.1002/chem.201603494.
Figure 1. Self-assembly of amphiphilic palladium NCN-pincer complex 1.
1291 ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2017, 23, 1291 –1298

Full Paper
reaction through the formation of a vesicular structure is ex-
plained in terms of the spontaneous concentration of organic
substrates in the hydrophobic region as a result of hydropho-
bic interactions, and the subsequent approach of the substrate
to the catalytic center (Figure 2). The organic transformation
nation of WAXS and MD calculations is one of the most useful
techniques for performing detailed structural analyses of self-
assembled nanoarchitectures.
Here we report a detailed analysis of the bilayer structure
generated by the self-assembly of a palladium NCN-pincer
complex conducted by means of WAXS experiments and MD
calculations. A good agreement between the experimental and
calculated X-ray spectra and between the membrane thickness
as determined from TEM images and that calculated from the
electron-density profile was obtained, and indicated that the
results of MD calculations are highly reliable. The calculation
showed that the alignment order of the complex molecules in
the bilayer structure was much lower than in pure phospholip-
id bilayer membranes. In addition, the MD calculations showed
that water molecules can enter the hydrophobic region of the
bilayer membrane. The calculation results also showed that or-
ganic substrates can be incorporated into the membranes, so
that bilayer vesicles can act as nanoreactors.
Figure 2. Schematic diagram of the catalytic system.
proceeds rapidly as a result of the presence of high concentra-
tions of the organic substrate near the catalytic center. To the
best of our knowledge, this study was the one of the earliest
successful examples of catalysis by using a self-assembled ar-
chitecture. Although microscopic analyses (transmission elec-
tron microscopy (TEM), scanning electron microscopy (SEM),
atomic-force microscopy, fluorescence microscopy, and confo-
cal laser scanning microscopy) of the vesicles showed that
these vesicles were spherical hollow structures, the detailed
molecular structure of the bilayer membrane remained unclear.
If the detailed bilayer structure could be clarified, we would be
able to discuss how organic transformations take place in bi-
layer membranes.
In the structural analysis of liposomes, solid-state 2H NMR
spectroscopy is a powerful tool for understanding the fluidity
of the membrane and the packing of the lipid molecules in
the membrane, among other aspects. Furthermore, molecular-
dynamics (MD) calculations for liposomes have shown a high
degree of consistency with the results of solid-state 2H NMR
spectroscopy on liposomes.[8]
Experimental Section
General information
Commercially available chemicals (purchased from Aldrich, TCI,
Kanto, Wako, Nakalai, or Alfa Aesar) were used without further pur-
ification, unless otherwise noted. NMR spectra were recorded by
using a JNM-A500 spectrometer (500 MHz for H, 125 MHz for 13C).
1
1
Chemical shifts for H NMR spectra are reported in d [ppm] refer-
enced to an internal tetramethylsilane standard. Chemical shifts for
13C NMR spectra are given relative to CDCl3 as an internal standard
1
(d=77.0 ppm). H and 13C NMR spectra were recorded in CDCl3 at
258C unless otherwise noted. EI mass spectra were recorded by
using an Agilent 5973N spectrometer attached to an Agilent
6890N gas chromatograph. Millipore water was obtained from
a Millipore Milli-Q Biocel A10 purification unit. Complex 1 and ve-
sicular 1vscl were prepared by using the reported methods.[5a,b]
Grazing-incidence-angle wide-angle X-ray scattering
The sample for grazing-incidence-angle WAXS (GIWAXS) measure-
ments was prepared as follows. A suspension of 1vscl in H2O was
dropped onto a silicon wafer and then air dried. High-resolution X-
ray scattering experiments were carried out at 258C by using a syn-
chrotron radiation X-ray beam with a wavelength of 0.80 Š on
beamline BL44B2[11] at SPring-8 (Hyogo, Japan). An angle of inci-
dence of 0.048 was used in the measurements. A large Debye–
Scherrer camera, 286.48 mm in length, was used with an imaging
plate as a detector and all diffraction patterns were obtained in
0.018 steps (2q). The duration of exposure to the X-ray beam was
60 min for 1 and 15 min for background.
MD calculations are also a powerful tool for estimating the
structure of membranes. In the last decade, as a result of the
increase in computational power and the development of so-
phisticated software, MD calculations have become a useful
tool for investigation of the properties of biomembranes. Fur-
thermore, the use of all-atomistic force-field parameters has
greatly improved the reliability of MD calculations in compari-
son with NMR spectroscopy measurements. Although many
MD calculation studies on phospholipid bilayers have been re-
ported, no study of the application of MD calculations to de-
termine the self-assembled bilayer structure of amphiphilic
transition-metal complexes has been reported.
Catalytic reaction with 1vscl in water
An aqueous suspension of 1vscl (1 mL, 2.6 mg, 2.4”10À3 mmol), cy-
clohex-2-ene-1-one (2, 11.5 mg, 0.12 mmol), and sodium tetraphe-
nylborate (3, 61.6 mg, 0.18 mmol) were placed in a vial, and the
mixture was agitated by shaking at 258C for 12 h. The reaction
was quenched with sodium chloride, then the mixture was extract-
ed with methyl tert-butyl ether (MTBE; 4”1.5 mL). The extracts
were combined, dried over Na2SO4, and filtered through silica gel
with MTBE as an eluent. After removal of the solvent, the product
A detailed structural analysis of self-assembled architectures
is important to elucidate their function. Wide-angle X-ray scat-
tering (WAXS) and MD calculation have been used in the struc-
tural analyses of self-assembled architectures (such as bilayer
vesicles or nanotubes).[9,10] In these reports, the structures of
assemblages at the molecular level were confirmed. A combi-
Chem. Eur. J. 2017, 23, 1291 –1298
www.chemeurj.org
1292
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
was dissolved in CDCl3 that contained 1,1,2,2-tetrachloroethane as
an internal standard. The chemical yield was determined by using
1H NMR spectroscopy.
the reported method.[17] The PINC, DODE, and TEG moieties were
connected by deleting extra hydrogen atoms, and the charges
were also transferred with charge adjustment of the connecting
atoms[17] to obtain pincer complex 1.
Data for 3-phenylcyclohexanone (4) [CAS: 20795-53-3]
MD calculation of the bilayer structure of complex 1
1H NMR (500 MHz, CDCl3): d=7.31–7.34 (m, 2H, PhH), 7.21–7.25 (m,
3H, PhH), 2.98–3.04 (m, 1H, CH-Ph), 2.57–2.62 (m, 1H, CH2), 2.50–
2.55 (m, 1H, CH2), 2.43–2.48 (m, 1H, CH2), 2.34–2.41 (m, 1H, CH2),
2.12–2.17 (m, 1H, CH2), 2.06–2.10 (m, 1H, CH2), 1.73–1.89 ppm (m,
2H, CH2); 13C NMR (125 MHz, CDCl3): d=210.9 (C=O), 144.3 (Ph-C
attached to cyclohexanone), 128.6 (PhC), 126.6 (PhC), 126.5 (PhC),
48.9 (CH attached to Ph), 44.7 (CH2), 41.1 (CH2), 32.7 (CH2),
25.5 ppm (CH2); MS: m/z: 174 [M]+.
An MD calculation for the self-assembled amphiphilic NCN-pincer
palladium complex bilayer membrane was performed to investi-
gate its detailed structure. The system investigated herein is sum-
marized in Table 1. In the bilayer membrane of the palladium NCN-
Table 1. Details of the system investigated in the present calculations.
T [K]
P [atm]
298.15
1.0
Computational methods
No. of complex molecules
No. of water molecules
Total number of atoms
128
15532
67972
Parameterizations for pincer complex 1
The all-atomistic CHARMM general force field (CGenFF)[12] was used
for pincer complex 1. Complex 1 was formed by connecting the
pincer unit (PINC), dodecane (DODE), and triethylene glycol (TEG)
residues (Figure 3). For the DODE and TEG residues, we used the
pincer complex, the number of complex molecules per leaflet was
64 and the number of water molecules was 15532.[19] The Len-
nard–Jones interaction was cut off at 1.2 nm by applying a switch-
ing function from 0.8 to 1.2 nm. The electrostatic interaction was
calculated by using the particle mesh Ewald (PME) method.[20] The
temperature (T) and the hydrostatic pressure (P) were held at
298.15 K and 1 atm, respectively, by using a combination of the
NosØ–Hoover chain and Parrinello–Rahman methods.[21] The equa-
tions of motion were numerically solved by using integrators
based on the RESPA[22,23] with a single time step of 2 fs. The
lengths of all chemical bonds relative to hydrogen atoms and of
the chemical and dummy bonds around the Pd atoms described
above were constrained by using the SHAKE/RATTLE/ROLL
method.[23] An MD calculation 130 ns long was performed by using
the PME version of our originally developed software MODYLAS.[24]
Convergence to the equilibrium state may be monitored by using
the membrane area per molecule. As shown in the Supporting In-
formation (Figure S2), it converged well within 60 ns. Trajectories
for the last 70 ns run have been used for detailed analyses.
Figure 3. The molecular structure of complex 1. Right: A model drawn by
using VMD[18] (C: cyan, H: white, O: red, N: blue, Cl: yellow, Pd: pink).
CGenEF parameters for hexane (HEXA) and poly(ethylene glycol)
monomer (PEGM) residues, respectively, as presented in the stan-
dard top_all36_cgenff.rtf and par_all36_cgenff.prm files. The pa-
rameters for the PINC residue were determined by noting that this
residue is constructed from the pincer backbone of complex 1 and
a benzene (BENZ) residue. The rigid TIP3P[13] model was used for
water molecules. The initial structure of the pincer backbone of
1 was constructed on the basis of the crystal structure of an NCN-
pincer complex.[14] The atom types of the PINC residue were trans-
ferred from BENZ, SCH2, and CLET residues in the CGenFF parame-
ters without palladium. The atom type of Pd was determined ac-
cording to the reported method.[15] The bond lengths, bond
angles, and dihedral angles of the PINC residue were determined
on the basis of these atom types and the optimized geometry. The
lengths of the chemical bonds (CÀPd, PdÀN1, PdÀN2, PdÀCl) and
the dummy bonds (CÀN1, N1ÀCl, ClÀN2, N2ÀC) were fixed by
using the SHAKE/RATTLE algorithm.[16] Consequently, Pd atoms
were fixed on the rigid rhomboidal plane formed by the C-N1-Cl-
N2 atoms.
Results and Discussion
Comparison of the experimental and simulated WAXS
spectra of 1vscl
To demonstrate the validity of our simulation result, we first com-
pared the experimental WAXS spectrum of 1vscl with the computed
spectrum of the calculated bilayer assemblage of complex 1. The
GIWAXS experiment for 1vscl gave an amorphous pattern spectrum
(Figure 4, blue line). A simulated WAXS spectrum was calculated by
using a reported method.[9a] The X-ray scattering intensity in the
Born approximation can be written as follows [Eq. (1)]:
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
*
+
2
X
j;m
1
I q /
fj q eiqÁr
ꢀ
ꢀ
ꢀ
ꢀ
j;m
in which rj,m is the position of the jth atom on the mth molecule
and the symbols <···> indicate an ensemble average. The overbar
indicates an average over all possible relative orientations of the
wave vector q with respect to the simulation cell, to take into ac-
count the fact that the experimental samples consist of vesicles,
This planar structure is consistent with the structure of palladium
NCN-pincer complexes in the solid state.[14] The charges on the
PINC residue were transferred from the pincer backbone and
BENZ, with charge adjustment of the linking atoms according to
Chem. Eur. J. 2017, 23, 1291 –1298
www.chemeurj.org
1293
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 2. Calculated structural and thermodynamic properties of the bilay-
er structure of complex 1, and the experimental membrane area per
lipid[25] and isothermal area compressibility of DPPC, POPC, and DMPC bi-
layers in the liquid-crystalline state.[26–28]
S”103 [Š2] a [Š2]
h [Š]
V”105 [Š3] cTS [m2 JÀ1
]
1[a]
DPPC
5.48Æ0.12 85.6Æ1.9
120Æ2 6.56Æ0.01 5.57Æ0.04
–
–
–
63.1
–
–
–
–
–
–
4.0
(323.15 K)[25]
64.3
(323.15 K)[26]
3.0–5.5
POPC
(303.15 K)[25]
59.9
(298.15 K)[27]
4.3Æ0.4
DMPC
(303.15 K)[25]
(302.15 K)[28]
[a] Errors in the first five quantities correspond to the standard deviation
over the average of seven 10 ns intervals. Errors for the last quantity rep-
resent the propagation error of the standard deviation of the variance,
<(SÀ<S>)2 >, and the standard deviation of the average, <S>, over
one 70 ns interval average.
Figure 4. Comparison of the experimental and simulated XRD spectra (blue
line: experimental XRD spectrum; red line: simulated XRD spectrum). The
error bars represent the standard deviations estimated from one 70 ns inter-
val average.
2
1 < SÀ < S >
kBT < S >
>
cTS
3
which we approximated as nonoriented bilayers. The number of
wave vectors was over 106, which enabled us to calculate I(q) up
to q=8.5 ŠÀ1 with very high accuracy.
in which <···> represents the time average over the final 70 ns.
Table 2 also lists the calculated values of cST . In membrane 1, the
calculated value of cTS was 5.57 m2 JÀ1 at 298.15 K. This value is
greater than that of experimental pure bilayer membranes of dipal-
mitoylphosphatidylcholine (DPPC), 1-palmitoyl-2-oleoylphosphati-
dylcholine (POPC), and 1,2-dimyristoylphosphatidylcholine (DMPC)
in their liquid-crystalline states (Table 2).[26–28] This result indicates
that the complex bilayer is laterally softer than the pure lipid bilay-
ers.[29]
The quantity fj(q) is the form factor of atom j and it is computed
by using the formula in Equation (2):
4
h
ꢁ
ꢂ i
X
2
q
4 p
f q
c
ai exp Àbi
2
i
1
in which c, ai, and bi are the Cromer–Mann coefficients for the
given atomic species.
To check the structure of the calculated bilayer membrane, a snap-
shot of the simulated bilayer structure is shown in Figure 5. The
order of alignment of the complex molecule was low.
The resulting spectrum consisted of the spectrum of the bilayer as-
semblage of complex 1 and the spectrum of the water molecules.
To obtain the calculated spectrum of the bilayer assemblage of
complex 1, it was necessary to subtract the spectrum of water mol-
ecules from the spectrum of the MD calculation result for the total
system.[9a] An MD calculation for the bulk water system (52080
water molecules) was also carried out under identical conditions to
those used for the bilayer membrane assemblage of complex
1 with water molecules. This gave a calculated X-ray spectrum for
the water molecules, which we subtracted from the calculated
spectrum for the total system to give the desired spectrum for the
bilayer membrane of complex 1 (Figure 4, red line).
The simulated WAXS spectrum (Figure 4, red line) shows excellent
agreement with the experimental WAXS spectrum of 1vscl. This indi-
cates that the simulated bilayer structure is highly reliable.
Structural properties of membranes of 1vscl
Next, we analyzed the structural properties of the membrane in
detail. Table 2 lists the geometric properties of the MD unit cell
and the lipid bilayer averaged over the final 70 ns. S is the mem-
brane area, a is the area per complex molecule, and h and V are
the height and volume of the unit cell, respectively. The error
given in Table 2 corresponds to the standard deviation over an
average of seven 10 ns intervals. The calculated membrane area
per complex molecule is (85.6Æ1.9) Š2, which is much greater than
that of phospholipid bilayers in the liquid-crystalline state (60–
65 Š2).[25]
Figure 5. Snapshot of the equilibrated bilayer membrane structure of palla-
dium NCN-pincer complex 1 (C: cyan, H: white, O: red, N: blue, Cl: yellow,
Pd: pink), drawn by using VMD.[18]
The fluctuation in S gives the isothermal area compressibility (cST )
of the bilayers by applying the expression in Equation (3).
Chem. Eur. J. 2017, 23, 1291 –1298
www.chemeurj.org
1294
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
in which q is the angle between the CÀH vector and the z axis.
The quantities SCD and SCH are essentially the same. Figure 7 shows
the calculated value of SCH as a function of the carbon number of
the dodecyl and TEG chains. The values of SCH for dodecyl and TEG
chains are very small compared with the corresponding values for
the acyl tails of phospholipids that form bilayers in the liquid-crys-
talline state (0.20–0.25 in the middle of acyl chains; <0.05 at the
chain terminal).[8]
Electron-density profile
To check the position of the atoms in the bilayer membrane, the
distribution of atoms along the bilayer normal (z axis) and the re-
sultant electron-density profile were investigated. Figure 6 shows
the electron-density profile along the z axis (1e(z)) averaged over
the final 70 ns, in which the origin of z is taken to be the center of
mass of the bilayer. The distribution was almost symmetric with re-
spect to z=0 Š for this system.
Figure 6. Calculated electron-density profiles (black line: total, orange line:
complex, aqua-blue line: water, red line: pincer backbone, blue line: dodecyl
group, green line: TEG group, pink line: palladium atom, purple line: chlor-
ine atom). The point at z=0 is the center of mass of the bilayer. Error bars
represent the standard deviations estimated from the average of seven
10 ns intervals.
Figure 7. The calculated order parameter of CÀH vectors of the dodecyl
chains (top right) and TEG chains (bottom right). The error bars represent
the standard deviations estimated from the average of seven 10 ns intervals.
The distribution of the complex ranged from À30 to +30 Š, which
shows that the thickness of the membrane was about 60 Š
(Figure 6, orange line). This calculated membrane thickness is con-
sistent with the TEM observations (Figure 1). The pincer backbone
(Figure 6, red line) was located between the hydrophobic dodecyl
group region (Figure 6, blue line) and the hydrophilic TEG group
region (Figure 6, green line). Interestingly, palladium (Figure 6, pink
line) faced both the hydrophilic and hydrophobic regions. Further-
more, a number of water molecules were observed in the inner hy-
drophobic region of the bilayer membrane (Figure 6, aqua-blue
line). Therefore, hydrophobic organic molecules also can exist in
the inner region of the bilayer assemblage. If external organic mol-
ecules are present in the hydrophilic or hydrophobic regions in the
bilayer, these molecules can easily reach the palladium atoms. Or-
ganic transformations can, therefore, be catalyzed by our vesicular
nanocomposite (1vscl) in water.
To elucidate the origin of the low value of SCH, we also estimated
the order parameter for PdÀC bonds (SPdC) as a measure of the
alignment order of the central pincer framework. The calculated
value of SPdC was (0.32Æ0.01), which is much smaller than the
order parameter for the long axes of DPPC molecules in a fluid
phospholipid bilayer (0.7 to 0.8).[8] That is, the order of alignment
in the central pincer framework in the bilayers of the amphiphilic
complex is quite low. Therefore, the very small values of SCH in the
side chains derive from low-order alignment of complex molecules
in the bilayers. This observation is consistent with the snapshot of
the bilayer structure (Figure 5). However, the SCH values for C9, C10,
C11, and C12 in the dodecyl chains were relatively high. These re-
sults show that the bilayer membrane is weakly supported by a rel-
atively high-order alignment of the terminal 1-methyl group and
the three methylene groups near the terminus of the dodecyl
chains (Figure 8).
Order Parameter of the CÀH Vector in the Dodecyl and TEG
Chains
Radial distribution function between palladium atoms
The order parameter of the CÀH vector (SCH) is a measure of the
2
Intermolecular metallophilic interactions are known to be one of
the driving forces for the construction of supramolecular struc-
tures.[30]
order of the dodecyl and TEG chains. In general, the H NMR spec-
troscopy experiment with deuterium-substituted acyl tails permits
the determination of the order parameter of the CÀD vector (SCD
)
from the width of the quadrupole splitting of the target CÀD
bonds. In the MD calculations, the order parameter was calculated
by using the expression in Equation (4).
Intermolecular PdÀPd interactions (typical PdÀPd contact length=
3.16–3.37 Š) have been observed in solid-state planar Pd com-
plexes.[31] To check for the presence of PdÀPd interactions in self-
assembled bilayer assemblage 1vscl, we investigated the radial dis-
tribution function between the palladium atoms (Figure 9). The
first peak was observed at 7.7 Š, which is much longer than the
ꢃ
ꢄ
1
2
SCH
3 cos2 q À 1
4
Chem. Eur. J. 2017, 23, 1291 –1298
www.chemeurj.org
1295
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Electrostatic potential of the membrane
To discuss the effect of the electrostatic properties on the permea-
bility of the bilayer membrane toward organic molecules, we ana-
lyzed the electrostatic potential of the membrane by the reported
method.[32] Figure 10 shows the electrostatic potentials of the
membrane along the z axis. The electrostatic potential at the
center of the membrane was 0.7 V. The potential value of the bilay-
er membrane of complex 1 is similar to that of bilayer membranes
of pure DPPC or POPC.[33] From the viewpoint of the electrostatic
properties of the membrane, the permeability of our bilayer assem-
blage is, therefore, similar to that of bilayer membranes of DPPC or
POPC.
Figure 8. Membrane model (red: four terminal C atoms in the dodecyl chain,
blue: remaining eight C atoms in the dodecyl chain; pink rectangle: pincer
framework, green: TEG chain).
Figure 10. Electrostatic potentials (blue line: total; red line: complex; green
line: water). The potential was set to zero at the edge of the simulation box.
The point at z=0 is the center of mass of the bilayer. Error bars represent
the standard deviations estimated from the average of seven 10 ns intervals.
Membrane permeation free-energy profile of water
Figure 11 shows the membrane permeation free-energy profile of
water along the z axis, DG(z), obtained from the number-density
profile of water molecules along the z axis, 1w(z), by using Equa-
tion (5).
ꢅ
ꢆ
1w z
1w;0
DG z
ÀkBT ln
5
in which the basis of DG(z) was chosen for the bulk region of
water with a number density of 1w,0. The free-energy barrier to
water permeation through our bilayer membrane was 12 kJmolÀ1
.
Figure 9. Radial distribution function between the palladium atoms. Error
bars represent the standard deviations estimated from the average of seven
10 ns intervals.
In contrast, the free-energy barrier to permeation of water through
pure phospholipid bilayer membranes in the liquid-crystalline state
is over 25 kJmolÀ1 [34]
The free energy of permeation of water
.
through our bilayer membrane is, therefore, much lower than for
permeation through pure phospholipid bilayer membranes. At
298.15 K, exp((À11 kJmolÀ1)/kBT)ꢀ0.012, which is much greater
than exp((À25 kJmolÀ1)/kBT)ꢀ0.00004. Therefore, water molecules
tend to permeate through the bilayer membrane of complex 1.
The reason for this lower energy barrier is still unclear. We specu-
late that the quite low order of alignment of the complexes in the
bilayer membrane (see above) might provide space for the perme-
typical range for PdIIÀPdII contact lengths (3.16–3.37 Š), which indi-
cates that PdÀPd interactions do not occur in this bilayer assem-
blage.
Chem. Eur. J. 2017, 23, 1291 –1298
www.chemeurj.org
1296
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
when vesicular 1vscl was used as the catalyst (Table 3, entries 3–8).
These results indicate that vesicular 1vscl disassembles or dissolves
in organic solvents to give catalytically less active monomeric 1,
whereas in water, vesicular 1vscl is present and operates as an effi-
cient catalyst.
Conclusion
We performed a GIWAXS experiment and all-atomistic MD calcula-
tion for a self-assembled vesicular palladium NCN-pincer complex.
We found an excellent agreement between the experimental and
calculated X-ray spectra and between the membrane thickness de-
termined from a TEM image and that calculated from an electron-
density profile, which indicated that the simulated structure was
reliable. An analysis of the simulated bilayer structure showed that
the membrane was relatively softer than phospholipid bilayer
membranes in general. In this bilayer assemblage, the alignment
order of complex molecules in the bilayer assemblages was quite
low. The analysis of the electron-density profile showed that the bi-
layer assemblage contained a space through which organic mole-
cules could exit. The catalytically active center is close to this space
and easily accessible by organic molecules. The bilayer membrane
structure can, therefore, act as a nanoreactor. In relation to the
electrostatic properties of bilayer membrane complex 1, the poten-
tial of the bilayer membrane of complex 1 is similar to that of bi-
layer membranes of pure phospholipids. In contrast, the free-
energy barrier to permeation of water through the bilayer mem-
brane of complex 1 is much lower than that for pure phospholipid
bilayer membranes. Therefore, water molecules tend to permeate
through the bilayer membrane of complex 1. This analytical result
suggests that organic molecules can also pass through the bilayer
membrane of complex 1 and diffuse to the catalytically active
center. Finally, we demonstrated the catalytic activity of self-assem-
bled vesicular 1vscl. Vesicular 1vscl catalyzed the Miyaura–Michael re-
action of cyclohex-2-en-1-one with sodium tetraphenylborate in
water to give 3-phenylcyclohexanone in 83% yield. In contrast,
when non-self-assembled 1amps was used as the catalyst, the prod-
uct was obtained in only 7% yield. The formation of a bilayer struc-
ture is, therefore, essential for the efficient promotion of the reac-
tion in water.
Figure 11. Membrane permeation free-energy profile of water. The point at
z=0 is the center of mass of the bilayer. Error bars represent the standard
deviations estimated from the average of seven 10 ns intervals.
ation of water molecules through the bilayer membrane. From this
analytical result, it is expected that organic molecules can pass
through the bilayer membrane of complex 1 and diffuse to the
catalytically active center.
The Miyaura–Michael reaction of cyclohex-2-en-1-one with
sodium tetraphenylborate
To demonstrate the catalytic activity of self-assembled vesicular
1
vscl, we performed the Miyaura–Michael reaction[35] for cyclohex-2-
en-1-one (2) with sodium tetraphenylborate (3) in the presence of
self-assembled vesicular 1vscl in water (Table 3). The reaction pro-
ceeded smoothly to give 3-phenylcyclohexanone (4) in 83% yield
(Table 3, entry 1). Conversely, when non-self-assembled complex
1amps was used as the catalyst, product 4 was obtained in only 7%
yield (Table 3, entry 2). The formation of the self-assembled archi-
tecture is, therefore, essential for efficient promotion of the reac-
tion. We also carried out the reaction in various organic solvents,
and the yield of 3-phenylcyclohexanone (4) was 6% or less even
Table 3. The Miyaura–Michael reaction of cyclohex-2-en-1-one (2) with
sodium tetraphenylborate (3).[a]
Acknowledgements
The synchrotron radiation experiments were performed at
beamline BL44B2 at SPring-8 with the approval of RIKEN (pro-
posal nos. 20090083, 20100016, and 20140078). The calcula-
tions were performed at the Research Center for Computation-
al Science, Okazaki, Japan. This research project was supported
by the JST-CREST and JST-ACCEL programs.
Entry
Catalyst
Solvent
Yield [%][b]
1
2
3
4
5
6
7
8
1vscl
1amps
1vscl
1amps
1vscl
1amps
1vscl
H2O
H2O
toluene
toluene
THF
THF
83
7
0
0
1
1
4
6
Keywords: molecular dynamics · pincer complexes · self-
assembly · vesicles · wide-angle X-ray scattering
[1] a) D. D. Lasic, Trends Biotechnol. 1998, 16, 307–321; b) A. Mueller, D. F.
O’Brien, Chem. Rev. 2002, 102, 727–757; c) E. Busseron, Y. Ruff, E.
Moulin, N. Giuseppone, Nanoscale 2013, 5, 7098–7140.
1,4-dioxane/H2O (10:1)
1,4-dioxane/H2O (10:1)
1amps
[2] For reviews of catalytic reactions with vesicles, see: a) D. M. Vriezema,
M. C. Aragon›s, J. A. A. W. Elemans, J. J. L. M. Cornelissen, A. E. Rowan,
R. J. M. Nolte, Chem. Rev. 2005, 105, 1445–1489; b) M. Raynal, P. Balles-
ter, A. Vidal-Ferran, P. W. N. M. van Leeuwen, Chem. Soc. Rev. 2014, 43,
1734–1787.
[a] Reaction conditions: 2 (11.5 mg, 0.12 mmol), 3 (61.6 mg, 0.18 mmol),
catalyst (2.6 mg, 2.4”10À3 mmol), solvent (1 mL), 258C, 12 h. [b] Deter-
mined by using 1H NMR spectroscopy with an internal standard
(Cl2CHCHCl2).
Chem. Eur. J. 2017, 23, 1291 –1298
www.chemeurj.org
1297
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[3] For selected recent examples of catalytic applications of self-assembled
architectures in water, see: a) M. S. Goedheijt, B. E. Hanson, J. N. H. Reek,
P. C. J. Kamer, P. W. N. M. van Leeuwen, J. Am. Chem. Soc. 2000, 122,
1650–1657; b) D. M. Vriezema, P. M. L. Garcia, N. S. Oltra, N. S. Hatzakis,
S. M. Kuiper, R. J. M. Nolte, A. E. Rowan, J. C. M. van Hest, Angew. Chem.
Int. Ed. 2007, 46, 7378–7382; Angew. Chem. 2007, 119, 7522–7526;
c) G. Delaittre, I. C. Reynhout, J. J. L. M. Cornelissen, R. J. M. Nolte, Chem.
Eur. J. 2009, 15, 12600–12603; d) L. Qin, L. Zhang, Q. Jin, J. Zhang, B.
Han, M. Liu, Angew. Chem. Int. Ed. 2013, 52, 7761–7765; Angew. Chem.
2013, 125, 7915–7919; e) J. P. Patterson, P. Cotanda, E. Z. Kelley, A. O.
Moughton, A. Lu, T. H. Epps, III, R. K. O’Reilly, Polym. Chem. 2013, 4,
2033–2039; f) M. C. M. van Oers, L. K. E. A. Abdelmohsen, F. P. J. Rutjes,
J. C. M. van Hest, Chem. Commun. 2014, 50, 4040–4043.
[4] For selected reviews of self-assembled metal–organic-framework cata-
lyst systems, see: a) J. K. M. Sanders, Chem. Eur. J. 1998, 4, 1378–1383;
b) P. A. Gale, Philos. Trans. R. Soc. London Ser. A 2000, 358, 431–453;
c) M. Fujita, K. Umemoto, M. Yoshizawa, N. Fujita, T. Kusukawa, K. Birad-
ha, Chem. Commun. 2001, 509–518; d) C. H. M. Amijs, G. P. M. van Klink,
G. van Koten, Dalton Trans. 2006, 308–327.
[5] a) G. Hamasaka, T. Muto, Y. Uozumi, Angew. Chem. Int. Ed. 2011, 50,
4876–4878; Angew. Chem. 2011, 123, 4978–4980; b) G. Hamasaka, T.
Muto, Y. Uozumi, Dalton Trans. 2011, 40, 8859–8868; c) G. Hamasaka, Y.
Uozumi, Chem. Commun. 2014, 50, 14516–14518.
[6] a) F. Sakurai, G. Hamasaka, Y. Uozumi, Dalton Trans. 2015, 44, 7828–
7834; b) G. Hamasaka, F. Sakurai, Y. Uozumi, Tetrahedron 2015, 71,
6437–6441.
[7] G. Hamasaka, Y. Uozumi, Chem. Lett. 2016, 45, 1244–1246.
[8] L. S. Vermeer, B. L. de Groot, V. RØat, A. Milon, J. Czaplicki, Eur. Biophys. J.
2007, 36, 919–931.
[20] U. Essmann, L. Perera, M. L. Berkowitz, T. Darden, H. Lee, L. G. Pedersen,
J. Chem. Phys. 1995, 103, 8577–8593.
[21] a) S. NosØ, Mol. Phys. 1984, 52, 255–268; b) W. G. Hoover, Phys. Rev. A
1985, 31, 1695–1697; c) G. J. Martyna, M. L. Klein, M. Tuckerman, J.
Chem. Phys. 1992, 97, 2635–2643; d) M. Parrinello, A. Rahman, Phys.
Rev. Lett. 1980, 45, 1196–1199.
[22] M. Tuckerman, B. J. Berne, G. J. Martyna, J. Chem. Phys. 1992, 97, 1990–
2001.
[23] G. J. Martyna, M. E. Tuckerman, D. J. Tobias, M. L. Klein, Mol. Phys. 1996,
87, 1117–1157.
[24] Y. Andoh, N. Yoshii, K. Fujimoto, K. Mizutani, H. Kojima, A. Yamada, S.
Okazaki, K. Kawaguchi, H. Nagao, K. Iwahashi, F. Mizutani, K. Minami, S.
Ichikawa, H. Komatsu, S. Ishizuki, Y. Takeda, M. Fukushima, J. Chem.
Theory Comput. 2013, 9, 3201–3209.
[25] N. Kucerka, M.-P. Nieh, J. Katsaras, Biochim. Biophys. Acta Biomembr.
2011, 1808, 2761–2771.
[26] J. F. Nagle, S. Tristram-Nagle, Biochim. Biophys. Acta Biomembr. 2000,
1469, 159–195.
[27] H. Binder, K. Gawrisch, J. Phys. Chem. B 2001, 105, 12378–12390.
[28] W. Rawicz, K. C. Olbrich, T. Mclntosh, D. Needham, E. Evans, Biophys. J.
2000, 79, 328–339.
[29] For recent examples of self-assembled bilayer membrane architectures
of polypeptides, see: a) C. Cai, J. Lin, T. Chen, X. Tian, Langmuir 2010,
26, 2791–2797; b) K. T. Nam, S. A. Shelby, P. H. Choi, A. B. Marciel, R.
Chen, L. Tan, T. K. Chu, R. A. Mesch, B.-C. Lee, M. D. Connolly, C. Kisielow-
ski, R. N. Zuckermann, Nat. Mater. 2010, 9, 454–460; c) Z. Zhuang, C.
Cai, T. Jiao, J. Lin, C. Yang, Polymer 2014, 55, 602–610; d) R. V. Mannige,
T. K. Haxton, C. Proulx, E. J. Robertson, A. Battigelli, G. L. Butterfoss, R. N.
Zuckermann, S. Whitelam, Nature 2015, 526, 415–420.
[9] a) M. Sega, G. Garberoglio, P. Brocca, L. Cantœ, J. Phys. Chem. B 2007,
111, 2484–2489; b) M. Sega, R. Vallauri, P. Brocca, L. Cantœ, S. Melchion-
na, J. Phys. Chem. B 2007, 111, 10965–10969; c) T. T. Mills, G. E. S.
Toombes, S. Tristram-Nagle, D.-M. Smilgies, G. W. Feigenson, J. F. Nagle,
Biophys. J. 2008, 95, 669–681.
[10] H. Xu, A. K. Das, M. Horie, M. S. Shaik, A. M. Smith, Y. Luo, X. Lu, R. Col-
lins, S. Y. Liem, A. Song, P. L. A. Popelier, M. L. Turner, P. Xiao, I. A. Kin-
loch, R. V. Ulijn, Nanoscale 2010, 2, 960–966.
[30] a) J. W. Steed, J. L. Atwood, Supramolecular Chemistry, 2nd ed., Wiley,
Chichester, 2009, pp. 36–37; b) P. Pyykkç, Chem. Rev. 1997, 97, 597–
636; c) M. J. Katz, K. Sakai, D. B. Lenznoff, Chem. Soc. Rev. 2008, 37,
1884–1895.
[31] For selected examples of intermolecular PdÀPd contacts, see a) K. J.
Miller, J. H. Baag, M. M. Abu-Omar, Inorg. Chem. 1999, 38, 4510–4514;
b) D. Eddings, C. Barnes, N. Gerasimchuk, P. Durham, K. Domasevich,
Inorg. Chem. 2004, 43, 3894–3909; c) N. Kumari, R. Parajapati, L. Mishra,
Polyhedron 2008, 27, 241–248; d) C. Yue, F. Jiang, Y. Xu, D. Yuan, L.
Chen, C. Yan, M. Hong, Cryst. Growth Des. 2008, 8, 2721–2728; e) S.
Akine, H. Nagumo, T. Nabeshima, Inorg. Chem. 2012, 51, 5506–5508.
[32] a) S. E. Feller, R. W. Pastor, A. Rojnuckarin, S. Bogusz, B. R. Brooks, J. Phys.
Chem. 1996, 100, 17011–17029; b) K. Tu, M. L. Klein, D. J. Tobias, Bio-
phys. J. 1998, 75, 2147–2156.
[11] K. Kato, H. Tanaka, Adv. Phys. X 2016, 1, 55–80.
[12] K. Vanommeslaeghe, E. Hatcher, C. Acharya, S. Kundu, S. Zhong, J. Shim,
E. Darian, O. Guvench, P. Lopes, I. Vorobyov, A. D. MacKerell, Jr., J.
Comput. Chem. 2010, 31, 671–690.
[13] S. R. Durell, B. R. Brooks, A. Ben-Naim, J. Phys. Chem. 1994, 98, 2198–
2202.
[14] K. Takenaka, M. Minakawa, Y. Uozumi, J. Am. Chem. Soc. 2005, 127,
12273–12281.
[33] J. B. Klauda, R. M. Venable, A. D. MacKerell Jr., R. W. Pastor, Curr. Top.
Membr. 2008, 60, 1–48.
[15] a) V. Yeguas, P. Campomanes, R. López, N. Díaz, D. Suµrez, J. Phys. Chem.
B 2010, 114, 8525–8535; b) T. W. Hambley, Inorg. Chem. 1998, 37,
3767–3774.
[16] G. J. Martyna, D. J. Tobias, M. L. Klein, J. Chem. Phys. 1994, 101, 4177–
4189.
[17] J. D. Evanseck, M. J. Field, S. Fisher, J. Gao, H. Guo, S. Ha, D. Joseph-Mc-
Carthy, L. Kuchnir, K. Kuczera, F. T. K. Lau, C. Mattos, S. Michnick, T. Ngo,
D. T. Nguyen, B. Prodhom, W. E. Reiher III, B. Roux, M. Schlenkrich, J. C.
Smith, R. Stote, J. Straub, M. Watanabe, J. Wiórkiewicz-Kuczera, D. Yin,
M. Karplus, J. Phys. Chem. B 1998, 102, 3586–3616.
[18] W. Humphrey, A. Dalke, K. Schulten, J. Mol. Graphics 1996, 14, 33–38.
[19] The initial states of the simulation are shown in the Supporting Infor-
mation.
[34] a) S.-J. Marrink, H. J. C. Berendsen, J. Phys. Chem. 1994, 98, 4155–4168;
b) T. Sugii, S. Takagi, Y. Matsumoto, J. Chem. Phys. 2005, 123, 184714;
c) K. Shinoda, W. Shinoda, M. Mikami, J. Comput. Chem. 2008, 29, 1912–
1918.
[35] For recent reviews of the Miyaura–Michael reaction, see: a) N. Miyaura,
Synlett 2009, 2039–2050; b) N. Miyaura, Bull. Chem. Soc. Jpn. 2008, 81,
1535–1553.
Manuscript received: July 24, 2016
Accepted Article published: October 13, 2016
Final Article published: December 14, 2016
Chem. Eur. J. 2017, 23, 1291 –1298
www.chemeurj.org
1298
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim