Kinetic precipitation of solution-phase polyoxomolybdate followed by transmission electron microscopy: A window to solution-phase nanostructure

Article abstract of DOI:10.1002/chem.200305468

This study aimed to elucidate the structural nature of the polydisperse, nanoscopic components in the solution and the solid states of partially reduced polyoxomolybdate derived from the {Mo₁₃₂} keplerate, {(Mo)Mo ₅}₁₂-{Mo

Full text of DOI:10.1002/chem.200305468

FULL PAPER
Kinetic Precipitation of Solution-Phase Polyoxomolybdate
Followed by Transmission Electron Microscopy:
A Window to Solution-Phase Nanostructure
Yan Zhu,[a] Arthur Cammers-Goodwin,*[a] Bin Zhao,[b] Alan Dozier,[c] and
Elizabeth C. Dickey[d]
Abstract: This study aimed to elucidate
the structural nature of the polydis-
perse, nanoscopic components in the
solution and the solid states of partially
reduced polyoxomolybdate derived
particle size distributions were snap
shots of the solution phase. The mother
Nanoparticles of coprecipitate were
more stable than solids derived solely
from partially reduced molybdate. The
TEM features of all material analyzed
lacked facets on the nanometer length
scale; however, the structures diffract-
ed electrons and appeared to be
defect-free as evidenced by Moirÿ pat-
terns in the TEM images. Moirÿ pat-
terns and size-invariant optical densi-
ties of the features in the micrographs
suggested that the molybdate nanopar-
ticles were vesicular.
liquor of the preparation of the {Mo132
}
keplerate after three days revealed
large species (r=20 30 nm) in the co-
precipitate, whereas {Mo132} keplerate
redissolved in water revealed small
species (3 7 nm) in the coprecipitate.
from
the
{Mo132
}
keplerate,
{(Mo)Mo5}12-{Mo2 acetate}30. Designer
tripodal hexamine-tris-crown ethers
and nanoscopic molybdate coprecipi-
tated from aqueous solution. These mi-
crocrystalline solids distributed particle
radii between 2 30 nm as assayed by
Keywords: electron microscopy
¥
molecular recognition ¥ nanostruc-
tures ¥ phase transitions ¥ polyan-
ions
transmission
electron
microscopy
(TEM). The solid materials and their
Introduction
molybdates waited two centuries to be elucidated by the
solid-state studies of M¸ller and co-workers.
Polyoxometalates of early transition metals are structurally
diverse.[1] In particular, physical and chemical properties of
polyoxomolybdate has intrigued investigators for many
years.[2] However, structural details of polymeric polyoxo-
Arguably, the most intriguing oxometalate species are var-
ious closed-surface derivatives.[3] These structures highlight
the fact that the closed surface is composed of discrete subu-
nits. In the case of the {Mo132} keplerate,[4] the repeating
structural motif is the pentagonal subunit, (Mo)Mo5, linked
at the edges by {Mo2 acetate} to form an Ih symmetric clus-
ter. These nanoscale species carve out inner space with in-
teresting internal solvent structure and they open the possi-
bility of chemical reactions inside the closed surface.[5] Struc-
tural diversity in material related to {Mo132} keplerate in-
cludes discrete, nano-meso scale toroids and spheres arising
from hierarchical assembly,[6] and the reduction of symme-
try.[7] The diverse morphologies reported for the solid states
invite speculation that media-dependent solution-state equi-
libria also involve multiple sizes and morphologies. The cur-
rent work focuses on nanoscopic variants of the {Mo132} kep-
lerate. This polymeric distribution of anions will be referred
to as polypent-Mo2.
[a] Dr. Y. Zhu, Prof. Dr. A. Cammers-Goodwin
University of Kentucky, Chemistry Department
Lexington, KY 40506-0055 (USA)
Fax : (+1)859323-1069
E-mail: a.cammers@uky.edu
[b] B. Zhao
University of California Riverside
Department of Chemistry, Riverside, CA 92521-0403 (USA)
[c] Dr. A. Dozier
University of Kentucky, Chemical and Materials Engineering
A004 ASTeCC Bldg, Lexington, KY 40506-0286 (USA)
[d] Prof. Dr. E. C. Dickey
Pennsylvania State University
Department of Materials Science and Engineering, 195 MRI Bldg.
University Park, PA 16802 (USA)
Weight-average, size-average, and multi-modal distribu-
tions of particle sizes are available from dynamic laser light
scattering (DLS) techniques. The large extinction coeffi-
cients of the {Mo132} keplerate and related structures hinder
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author. This includes full
field views of the partial TEM images presented in the Figures.
Selected NMR spectra of 1 and 2.
2421
Chem. Eur. J. 2004, 10, 2421 2427
DOI: 10.1002/chem.200305468
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

FULL PAPER
DLS sizing of polypent-Mo2 particles. However, enough
data are available to conclude that polypent-Mo2 in water is
more size-disperse than polypent-Mo2 in other solvent sys-
tems. DLS techniques gave good results with the relatively
transparent, aqueous Fe derivative, of which the {Fe30Mo72}
keplerate is the smallest, discrete structure with a closed sur-
face. The DLS studies indicate two-size regimes in solution
and imply hollow structures for the nanoscale Fe Mo poly-
oxometalate species.[8]
Ensemble sizing techniques such as DLS give averages of
sizes, whereas non-ensemble techniques such as electron mi-
croscopy (EM) can highlight the properties of individual
structures such as particle morphology and composition.[9]
SEM (scanning) and TEM (transmission) have been applied
to solids derived from polypent-Mo2.[8a,10] The current study
corroborates these results and adds new insights into the
nature of polypent-Mo2 solution states.
The current work develops a protocol for the kinetic pre-
cipitation of polypent-Mo2 with chelating agents 1 and 2 to
produce ppt1 and ppt2 (Figure 1). The preservation of solu-
tion-state conformation was also posited for an oligomeric
polyelectrolyte. In that study, kinetic entrapment on surfaces
followed by atomic force microscopy probed solution-phase
folding.[11] The current study is also related to work involv-
ing complex formulations of polypent-Mo2 material with sur-
factant in which the inorganic material forms ordered
phases with less aggressive additives.[12]
podal structures related to 1 were less successful. This larger
study will appear in doctoral thesis format.
To invoke kinetic precipitation, the phase transition from
the solution state to the solid state must occur faster than
structural changes in the material. From previous work, the
dynamic structure in the solution state of chemically related
species easily satisfies these conditions.[8a,13] The solubility of
polypent-Mo2 decreases with increasing ionic strength, pre-
sumably due to the destruction of the hydration shell.[14]
Likewise, electrostatic interactions between the tripodal che-
lating agents (1 or 2) and polypent-Mo2 should have cooper-
atively destroyed the hydration shell and led to an insoluble
polypent-Mo2 complex. The amines in 1 and 2 protonate
+
below pH 7. Crown ethers associate with H3O+, NH4 or
K+ and thereby can take on positive charges. In any sur-
face-bound state, the positively charged benzocrown ethers
in 1 would have to be proximal. Precedent exists for cation-
associated crown ether moieties interacting favorably in the
solid state.[15]
Results and Discussion
Material redissolved from the crystallization-driven prepara-
tion of the {Mo132} keplerate[3a] and material from the
mother liquor was used in this study. In the TEM surveys in
Figures 2 and 3, individual {Mo132} keplerate species (r=
1.3 nm) probably merged with the granularity of the micro-
graphs. Thus, it was much easier to image the larger, less
popular species on which this study focused. The nanoscopic
species derived from the mother liquors of the preparation
(Figure 4A) were unambiguously larger than were those de-
rived from the redissolved keplerate material (Figures 2 and
3). Thus, the time scale of coprecipitation was shorter than
the solution-phase enlargement of nanoscopic species.
Samples for TEM analysis were prepared by three meth-
ods. These are described in detail in the first part of the ex-
perimental section. TEM analysis repeatedly revealed nano-
scopic spherical features in ppt1 whereas micrographs of
ppt2 were devoid of features with radii greater than 4 nm
(compare Figures 2A, 3A, and 3B with 2B).
Ppt1 formed within seconds whereas the super-sized
structures of aqueous state polypent-Mo2 require two to
three days to evolve. When a chemically related species is
prepared fresh, small angle X-ray scattering (SAXS) does
not detect particles in the solution phase with radii greater
than 5 nm. After the material is allowed to stand for two
days, SAXS analysis indicates the evolution of particles with
sizes in the r ~20 nm range.[8a] One parameter that would
have made kinetic precipitation of polypent-Mo2 impossible
would have been a fast chemical process that would have re-
moved large polypent-Mo2 particles from the distribution.
The slow forward rate process for the evolution of nano-
structured species guarantees a slower reverse process for
the decomposition of the nanoscopic species. Slow assembly
of nanoscopic species chemically related to the {Fe30Mo72}
keplerate has also been recently reported by Liu.[13]
Figure 1. Top: tripodal molecules used to trap polypent-Mo2; bottom:
schematic representation of kinetic precipitation of polypent-Mo2. The
circles represent solvent; the triangles represent tripodal molecules 1;
and the large sphere represents nanoscale polypent-Mo2 aqueous species.
Kinetic solution state structure is preserved in the solid.
In the current study, large ppt1 species derived from the
mother liquor of the preparation of {Mo132} keplerate popu-
lated the micrographs. However, small species were found
in ppt1 generated from dissolved, crystalline {Mo132} kepler-
ate. These two facts indicated that the distributions of parti-
cle sizes and morphologies of ppt1 as revealed by TEM gen-
erated repeatable snap shots of polypent-Mo2 in solution.
Diamines related to crown 1 and to hexamine 2 did not suc-
ceed in trapping the polypent-Mo2 solution state. Other tri-
Coprecipitates ppt1 and ppt2, derived from polypent-Mo2
and 1 or 2, were very insoluble; titration of polypent-Mo2(aq)
2422
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2421 2427

Solution-Phase Nanostructure
2421 2427
The large features in ppt1 decomposed over the course
of 3 4 weeks into featureless material by TEM. The decom-
posed material resembled the Mo-containing material in
Figure 2B (black arrow). In a few micrographs, the restruc-
turing of the spherical features might have been caught on
camera. Figure 2C shows a micrograph containing a rare el-
liptical feature that is approximately twice as long as it is
wide (33î17 nm). With microscopy of lower resolution, this
hypothesis for the observation of asymmetric transition
structures was offered sixty years ago.[8a]
Species larger than those derived from the dissolution of
crystalline Mo132 were observed in solids derived from the
mother liquor of the preparation of Mo132 (Figure 4A C).
These were imaged as ppt1 and as polypent-Mo2, which
linked the observations of nanostructure to polypent-Mo2
and not to synergy between polypent-Mo2 and 1.
Figure 2. Micrographs of ppt1 (A) and ppt2 (B) formed upon addition of
1 or 2 respectively to polypent-Mo2 in a 20:1 ratio. Analysis of the solid
confirmed the 20:1 ratio of 1 (C, H, N elemental analysis) to total Mo by
inductively coupled plasma atomic emission.[1b] Micrographs A and B
have identical scale and magnification. Solids ppt1 and ppt2 were similar
in appearance. Preparations of the solids for TEM were identical. Grey
and white arrows indicate the lacey carbon substrate and voids respec-
tively. The black arrows indicate material that contained Mo by energy
dispersive X-ray spectroscopy. Ellipsoidal features were also located (C).
These were probably structures in transition.
Figure 3. Micrographs from Method 1. Enclosures for measurement are
drawn around features in A).
with excess 1 or 2 left little polypent-Mo2 in solution detect-
able by UV at 455 nm, e=1.85î105 mÀ1 cmÀ1 [1b] Tripodal 1 in
0.1m KCl became soluble below pH 6 as determined by si-
multaneously decreasing the pH and monitoring the absorb-
ance of the liquid at 290 nm. A titration monitored at
455 nm showed that polypent-Mo2, 4.0î10À9 m,[1b] irreversi-
bly decomposed above pH 6. In contrast, ppt1 did not dis-
solve after agitation in water from pH 1 11 at room temper-
ature.
Figure 4. Solids from nanoscopic species in the mother liquor after 2 4 d.
A) ppt1. B) TEM at magnification 120k of polypent-Mo2 by Method 2,
dry preparation. C) TEM of polypent-Mo2 by Method 3, wet preparation.
D) Energy dispersive spectra of polypent-Mo2 (top) and ppt1 (bottom).
The Cu grid produced the Cu peaks.
2423
Chem. Eur. J. 2004, 10, 2421 2427
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

A. Cammers-Goodwin et al.
FULL PAPER
Ppt1 and polypent-Mo2 differed in elemental content by
energy dispersive X-ray spectra (EDS). Material absorbs
high-energy electrons and releases X-rays with frequencies
and intensities semi-quantitatively characteristic of elemen-
tal composition.[16] Figure 4D displays two representative
EDS spectra of polypent-Mo2 (top) and ppt1 (bottom). The
relative amount of Mo versus lighter elements in polypent-
Mo2, was lower than in ppt1. The benzocrown moiety in 1
presumably sequestered potassium from solution as evi-
denced by its detection in ppt1.
The features in the micrographs of ppt1 produced ordered
spot diffraction patterns, signaling a microcrystalline lattice
in these objects. The diffraction pattern shown in Figure 5A
had Bragg lattice spacing 1.1, 1.8, 2.2 and 4.2 ä. Most of the
spacings in the spot diffraction pattern in Figure 5A were
likely produced from high-Miller index phenomena, through
the Mo-lattice of one or more nanoscale species. However,
with the aid of image processing software. Examples of the
counting/measuring process are shown in Figures 2C and 3A
in which boundaries were drawn around the features. In the
image analysis, ellipses were mathematically fitted to the
closed curves and evaluated statistically in terms of size, and
circularity.
The nanostructures in the micrographs of polypent-Mo2 in
the absence of tripodal molecules were not as circular as
were those of ppt1; compare Figure 4A C with Figures 2A
and 3. Quantitatively, the ratio of anisotropic to spherical
structures was higher in solid polypent-Mo2 than in ppt1 re-
gardless of the source of the material. Figure 5B is a distri-
bution of particle morphologies by the index function: asym-
metry=(major axisÀminor axis)/(average width). The struc-
tural anisotropy of the polypent-Mo2 solid states probably
corresponded to their decreased stability compared to ppt1.
Asymmetry in these structures could have been due to the
loss of solvent from the interior of the vesicles.
À
lattice spacing of 4.2Æ0.4 ä matched Mo Mo distances (~
3.8 ä) in the X-ray structure of the {Mo}132 keplerate.[17]
The micrographs were used to produce particle size distri-
butions by measuring and counting the particles in the field
Figure 6A indicates that the TEM-derived particle size
distributions of the dissolved crystalline material were
skewed toward the size the smallest discrete closed struc-
ture, the keplerate. The resolution of these electron micro-
graphs is ~2 3 nm, approximately the diameter of {Mo132
}
keplerate. Analyses of two samples of ppt1 in Figure 6 at
magnification 130k and 210k showed magnification-inde-
pendent size distributions of polypent-Mo2.
Figure 6. A) Magnification-independent distributions of particle sizes.
B) Conversion of the data in A to mass distributions. The shaded line at
left is the TEM resolution limit for particle selection and measurement.
These graphs are not equilibrium distributions. The data sets generated
at magnification 130 and 210k were identical within experimental error.
Figure 5. A) A spot diffraction pattern of a feature in Figure 2A that indi-
cated that the Mo-atom lattice was intact in the superstructure. The su-
perimposed dashed lines are a diffraction pattern produced by an Au cal-
ibration standard to determine the camera constant of the TEM. Seg-
ments X1-X5 correspond to lattice spacings 1.8, 1.1, 1.7, 2.2, 4.1Æ0.4 ä,
respectively. B) Plots the number of particles as a function of a unit-less
asymmetry index: asymmetry = (long axisÀshort axis)/(average diame-
ter) is the deviation from circularity of the feature normalized by its aver-
age size. When the index is zero, the feature is a perfect circle. One lot
was analyzed at magnification 130 and 210k; the asymmetry did not
differ at these two magnifications.
Most experimental results scale with the mass distribution
of the material instead of the number distribution of parti-
cles. For example, larger particles scatter light more effi-
ciently, skewing the measurement toward larger values. An
argument is presented below for hollow polypent-Mo2 struc-
2424
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2421 2427

Solution-Phase Nanostructure
2421 2427
tures. Therefore, conversion of size distribution to mass dis-
tribution should employ the formula for the surface area of
a sphere. The ith population element in the distribution is
expressed as: P(r)i =4pr2 ni/N. The mass distribution, thus
derived, is shown in 6B. The noise in the heavy region of the
mass distribution is understandable when one considers that
one large particle out of hundreds raised the graph off the
zero line.
with Liu×s light scattering studies of solution states of nano-
scopic species of {Fe30Mo72} keplerate.[8,13]
Synthesis: The tripodal molecules used in this study were ef-
ficiently assembled with a Sonogishira coupling protocol to
assemble chains on the aromatic hub.[18] The final step at-
Particles derived from the solution phase (material adsor-
bed onto TEM grids) of the polypent-Mo2 preparation[3a]
were larger than were those derived from the soluble crys-
talline product. Compare average particles sizes in ppt1 of
the three graphs in Figure 6A, rav =7Æ3, 8Æ4, and 5Æ2 nm
to the average sizes of two lots of the polypent-Mo2 materi-
al, rav =22Æ11 and 32Æ7 nm and a sample of ppt1 derived
from the preparative solution phase, rav =20 Æ5 nm. Even
though nanoscopic species found in ppt1 derived from the
mother liquor were larger than were those found in ppt1 de-
rived from Mo132, they were no less symmetrical. The size
distributions of the nanoscopic features in Figure 4A (r~
25 nm) were probably representative of a solution phase
near equilibrium conditions. The less symmetric polypent-
Mo2 species in Figure 4B C probably deformed because of
loss of internal solvent molecules under the TEM vacuum.
Neither their size nor morphology should be interpreted to
reflect the solution state polypent-Mo2.
The features in the micrographs above were hollow by
two arguments presented in this paragraph and the one that
follows. The morphologies of the polypent-Mo2 in Figure 4B
and C, and ppt1 in Figures 2, 3 and 4A appear to be contin-
uous and not constructed from units of {Mo}132 keplerate.
This smooth construction is exceptionally visible in Fig-
ure 4C which shows isolated nanoscopic features with Moirÿ
patterns suggesting wave interference produced by two latti-
ces each consisting of one or multiple layers. Isolated, solid,
objects would not have produced these Moirÿ patterns.
In general, the optical densities of micrographic features
scale with the atomic weight and the number of atoms en-
countered by the electron beam. Heavier and more numer-
ous atoms scatter more electrons which gives rise to darker
images. This maxim is most pertinent to objects in micro-
graphs that weakly diffract electrons, the optical densities of
which have weak angular dependencies on the electron
beam. Figure 7 indicates that doubling the mass through
which the electron beam passed detectably changed the op-
tical density of the images. However, the optical densities of
the features versus the radii of the features in the micro-
graphs above were essentially constant in the TEM images.
This effect is striking in Figure 4B, a micrograph with ob-
jects 4x larger than those in Figure 7, however, similar
changes in optical densities in Figures 4B and 7 were noted
when objects overlapped. Objects possessing radii from 3 to
30 nm were detected in the micrographs of this study. If the
features were structurally homologous from surface to core,
this range would have corresponded to a 1000-fold increase
in mass. Optical densities independent of mass can only be
met if the features in the TEM were either hollow or flat.
Spherical structures with Mo at the surface are the best in-
terpretation of the data. Hollow structures are in accordance
Figure 7. The weak dependence of optical density of objects on size sug-
gested hollow structures. Imaged electrons scatter through one sphere in
region 1 and two spheres in region 2. The grey scale value of region 1
is 13 versus 48 for region 2 (0 = black, 255 = white). The same effect
was observed in polypent-Mo2 in the absence of 1 (see Figure 4B).
tached a benzocrown ether to the three terminal alkynes.
The esoteric conditions shown in Figure 8 for the coupling
were necessary to obtain a respectable yield. Sonogishira
couplings with electron-rich aromatic substrates are known
to cause difficulties; there is only one recent example of suc-
cess with the commercial bromobenzocrown ether.[19] The
tri-functionalization of compound 2 required an efficient
process afforded by the iodo derivative.[20] For further de-
tails see the Experimental Section.
Figure 8. Synthetic overview: a) propargylbromide, NEt3, THF; b) nBuLi/
THF, then TBDMSCl, À788C; c) 1,3,5-tribromobenzene, [Pd(PPh3)4],
CuI, nBuNH2, reflux, 72 h; d) TBAF/THF; e) 4-iodobenzocrown ether,
[Pd2(dba)3], tris-(2-furyl)phosphine, piperidine/DMF.
Conclusion
TEM studies of precipitates suggested smoothly constructed
nanoscopic species in aqueous solution of polypent-Mo2 in-
stead of aggregates of {Mo}132 keplerate. The independence
of optical density on the size of the features in the micro-
graphs, the lack of facets and uniform Moirÿ patterns in iso-
lated objects indicated polysize-disperse, hollow polypent-
Mo2 in aqueous solution. The colloidal behavior of poly-
pent-Mo2 probably involves aggregation processes between
nanoscale species. Effort was invested to avoid aggregation
in this study with various levels of success.
2425
Chem. Eur. J. 2004, 10, 2421 2427
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

A. Cammers-Goodwin et al.
FULL PAPER
N,N’-Di-(2-propynyl)-4,4’-trimethylenedipiperidine (4): Neat propargyl
bromide (4.41 mL, 49.5 mmol) was added to commercial 3, 4,4’-trimethy-
lenedipiperidine (5.08 g, 24.1 mmol) in THF (50 mL). After the addition
of Et3N (14.9 mL), the resulting emulsion was stirred vigorously at 258C
for 20 h. Diethyl ether, 10% HClaq 50 mL each were added to the reac-
tion mixture, and the phases were separated. The aqueous phase was
made basic (2m NaOH, 30 mL) and extracted with diethyl ether (3î
60 mL). The organic phase was dried over MgSO4. The solvent was
evaporated and the residue separated on silica gel flash chromatography
(40 60% EtOAc/hexane gradient elution) gave a light yellow oil pure by
NMR (2.90 g, 42%). 1H NMR (400 MHz, CDCl3, 258C, TMS): d=1.20
1.34 (m, 12H; CH, CH2 and axial piperidine CH2), 1.68 1.72 (m, 4H;
equatorial piperidine CH2), 2.15 2.21 (m, 4H; axial CH2-N), 2.24 (t,
4J(H,H)=2.4 Hz, 2H; =CH), 2.86 2.90 (m, 4H; equatorial CH2-N), 3.30
(d, 4J(H,H)=2.4 Hz, 4H; propargyl CH2); 13C NMR (100 MHz, CDCl3):
d=23.8, 32.2, 35.1, 36.6, 47.2, 52.6, 73.1, 78.9; MS (70 eV, EI): m/z (%):
285 (11) [M +ÀH], 247 (100) [M +ÀC3H3]; elemental analysis calcd (%)
for C19H30N2 (286.5): C 79.66, H 10.56, N 9.78; found: C 79.27, H 10.88,
N 9.69.
This study developed a protocol to trap solution state
structures of polypent-Mo2. Future studies will involve at-
tempts to characterize equilibria of nanoscale polypent-Mo2
in solution and will include attempts to control distributions
of sizes.
Experimental Section
ImageJ, Java freeware for image processing and statistical analysis, from
(NIH, USA): http://rsb.info.nih.gov/ij/ was used to analyze image files
generated from the TEM studies.
General synthetic methods: All reactions were carried out under Ar or
N2. THF was predried over CaH2 and distilled from sodium/benzophe-
none. DMF was distilled from CaH2 and stored over 4 ä molecular
sieves under nitrogen. All the other reagents were used as received from
commercial sources. 1H NMR and 13C NMR spectra were recorded at
400 and 100 MHz, respectively.
N-(tert-Butyldimethylsilyl-2-propynyl)-N’-(2-propynyl)-4,4’-trimethylene-
dipiperidine (5): n-Butyllithium (14.6 mL, 29.2 mmol) was added drop-
wise to a THF solution of 4 (8.35 g, 29.2 mmol, À788C, in 100 mL THF)
and kept cold for 30 min followed by dropwise addition of TBDMSCl
(4.40 g, 29.2 mmol in 40 mL THF). The vessel was allowed to warm to
258C overnight. The solvent was evaporated and the resulting slurry was
dispersed in biphasic diethyl ether and water; the phases were separated
and the aqueous phase was extracted with ether, the combined organic
phase was dried over MgSO4. Flash chromatography on silica gel (20
60% EtOAc/hexane gradient) gave the title compound as a colorless oil
(4.67 g, 40%). 1H NMR (400 MHz, CDCl3, 258C, TMS): d=0.10 (s, 6H;
Si-CH3), 0.94 (s, 9H; C-CH3), 1.20 1.34 (m, 12H; CH, CH2 and axial pi-
peridine CH2), 1.68 1.72 (m, 4H; equatorial piperidine CH2), 2.13 2.20
Solids for TEM analysis: All solids analyzed by TEM were derived from
the published preparation of the {Mo132} keplerate.[3a] N2H4¥H2SO4 (0.8 g,
6.1 mmol) was added to a 250 mL aqueous solution of (NH4)6Mo7O24¥4-
H2O (5.6 g, 4.5 mmol) and ammonium acetate (12.5 g, 162.2 mmol) and
stirred for 10 min. Aqueous acetic acid (50% vol, 83 mL) was added and
the preparation stood in an open flask at 208C for four days. A crystal-
line reddish-brown material was filtered off. The dark filtrate was saved
for further study. The crystalline material was washed with 90% ethano-
laq, absolute ethanol, diethyl ether, and finally air-dried at 258C. The IR,
Raman and UV/Vis spectral data for the resulting material corresponded
with the literature.[3a]
4
(m, 4H; axial CH2-N), 2.23 (t, J(H,H)=2.3 Hz, 1H; =CH), 2.83 2.90 (m,
4H, equatorial CH2-N), 3.28 (d, 4J(H,H)=2.3 Hz, 2H; propargyl CH2),
3.33 (s, 2H; propargyl CH2); 13C NMR (100 MHz, CDCl3): d=À4.6, 16.5,
23.9, 26.1, 26.1, 31.9, 32.1, 35.1, 36.5, 36.6, 47.2, 48.1, 52.3, 52.6, 73.1, 73.2,
78.8; MS (70 eV, EI): m/z (%): 400 (12) [M +], 361 (77) [M +ÀC3H3], 343
(25) [M +ÀC4H9], 285 (41) [M +ÀC6H15Si], 247 (100) [M +ÀC9H17Si]; ele-
mental analysis calcd (%) for C25H44N2Si (400.7): C 74.93, H 11.07, N
6.99; found: C 74.87, H 11.17, N 7.07.
Ppt1: A precipitate formed immediately when material obtained above
(3.0 mg) in deionized H2O (2.0 mL) was prepared and immediately
mixed with tripodal compound 1 (4.0 mg) in H2O (3.0 mL, 0.1m KClaq
,
pH~3). Centrifugation followed by air-drying at 258C gave ppt1; ppt2
was obtained analogously by using molecule 2.
Ppt1’: Approximately 700 mL of the mother liquor of the published pro-
cedure[3a] were transferred to a 15 mL plastic centrifuge tube, and diluted
to 1 mL with deionized water to give a reddish brown solution. A precipi-
tate formed immediately upon mixing this solution with tripodal com-
pound 1 (4.0 mg) dissolved in H2O (3.0 mL, 0.1m KClaq, pH~3). Centri-
fugation followed by air-drying at 258C gave ppt1’.
1,3,5-Tris-[N’-(tert-butyldimethylsilyl-2-propynyl)-4,4’-trimethylenedipi-
peridino-N-(2-propyn-3-yl)]benzene (6): 1,3,5-Tribromobenzene (0.99 g,
3.145 mmol) and 5 (4.155 g, 10.38 mmol) was dissolved in n-butylamine
(40 mL). The resulting solution was treated with [Pd(PPh3)4] (0.363 g,
0.314 mmol) and CuI (0.12 g, 0.630 mmol) and the solution was heated
under reflux for 72 h. After cooling to 258C, the solvents were evaporat-
ed and the residue was extracted into EtOAc, the organic phase was
washed with water and brine and dried over MgSO4. Flash chromatogra-
phy on silica gel (gradient elution: 50 100% EtOAc/hexane followed by
Sample preparation for TEM analysis
Method 1: Ppt1 (~1 mg) was dispersed in water in a small vial by sonicat-
ing for 30 min (Fisher Ultrasonic cleaner FS9) while cooling with ice. A
drop of the dispersion was placed on a Lacey carbon copper grid (Lacey
Carbon Type-A, Ted Pella, Inc.). After soaking the grid for 2 min, filter
paper was used to absorb moisture. The grid was allowed to air dry at
258C and was subjected to high vacuum for 3 h before TEM analysis.
TEM samples of ppt1’ and ppt2 were prepared analogously. The materi-
als were examined using an Electron Microscope (JEOL JEM-2000FX)
and JEOL JEM-2010F.
Method 2: The solid material filtered from the published procedure[3a]
was crushed gently in weighing paper between the fingers. A Lacey
carbon grid was sprinkled with a small amount of the dry material for
TEM analysis.
1
2 8% MeOH/CHCl3) give a yellow oil (3.0 g, 75%). H NMR (400 MHz,
CDCl3, 258C, CDCl3): d=0.11 (s, 18H; Si-CH3), 0.94 (s, 27H; C-CH3),
1.22 1.38 (m, 36H; CH, CH2 and axial piperidine CH2), 1.71 1.75 (m,
12H; equatorial piperidine CH2), 2.25 (dd, 3J(H,H)~11.5, 2J(H,H)=
11.5 Hz, 6H; axial CH2-N), 2.30 (dd, 3J(H,H)~11.5, 2J(H,H)=11.5 Hz,
6H; axial CH2-N), 2.91 (d, 2J(H,H)=11.5 Hz, 6H; equatorial CH2-N),
2
2.97 (d, J(H,H)=11.5 Hz, 6H; equatorial CH2-N), 3.40 (s, 6H; propargyl
CH2), 3.51 (s, 6H; propargyl CH2), 7.42 (s, 3H; CH aromatic); MS
(MALDI-TOF): m/z (%): 1275 (100) [M ++H].
1,3,5-Tris-[N’-(2-propynyl)-4,4’-trimethylenedipiperidino-N-(2-propyn-3-
yl)]benzene (2): Compound 6 (2.28 g, 1.79 mmol) in 20 mL THF at 08C,
was treated with TBAF (6.5 mL, 1m, dropwise); the reaction mixture was
stirred at 08C for 10 min, and then stirred at room temperature for 4 h.
After quenching with NH4Claq, extracting into CHCl3, and washing with
water, the CHCl3 phase was dried over Na2SO4. Flash chromatography
on silica gel (gradient elution: 3 8% MeOH/CHCl3) yielded a hygro-
scopic residue (1.4 g, 84%). 1H NMR (400 MHz, CDCl3, 258C, TMS):
d=1.21 1.36 (m, 36H; CH, CH2 and axial piperidine CH2), 1.68 1.74 (m,
12H; equatorial piperidine CH2), 2.13 2.20 (m, 12H; axial CH2-N), 2.23
Method 3: Approximately 10 mL fresh filtrate[3a] was placed on a Lacey
carbon copper grid. After 2 min, the solution was filtered and the grid
was allowed to air dry at 258C. The grid was examined by TEM.
Electron microscopy: The camera constant (CC) of the TEM was deter-
mined by adjusting the electron microscope to the same settings for the
acquisition of the diffraction pattern in Figure 5 and creating a standard
ring diffraction pattern Gold on ™Holey∫ Carbon Film, Ted Pella Inc.
product no. 613. The four rings in this sample correspond to four known
lattice spacings: CC=r(ring)n î(dn) for n=1 4, CC=48.5Æ0.6 mmä.
The patterns were superimposed and the lattice spacings in the nano-
structure of ppt1 were measured from the lengths of segments Xn:
d(Xn)=CC/Xn.
4
2
(t, J(H,H)=2.4 Hz, 3H; =CH), 2.87 (d, J(H,H)=11.5 Hz, 6H; equatori-
al CH2-N), 2.94 (d, 2J(H,H)=11.5 Hz, 6H; equatorial CH2-N), 3.28 (d,
4J(H,H)=2.4 Hz, 6H; propargyl CH2), 3.47 (s, 6H; propargyl CH2), 7.40
2426
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2421 2427

Solution-Phase Nanostructure
2421 2427
(s, 3H; CH aromatic); 13C NMR (100 MHz, CDCl3): d=23.9, 32.3, 32.4,
35.2, 35.2, 36.66, 36.68, 47.2, 48.0, 52.7, 53.0, 72.8, 79.2, 83.5, 86.2, 123.7,
134.2; MS (MALDI-TOF): m/z (%): 932.5 (100) [M+H]+ ; elemental
analysis calcd (%) for C63H90N6¥2H2O (967.5): C 78.21, H 9.79, N 8.69;
found: C 78.47, H 9.76, N 8.80.
[3] a) A. M¸ller, E. Krickemeyer, H. Bˆgge, M. Schmidtmann, F.
Peters, Angew. Chem. 1998, 110, 3567; Angew. Chem. Int. Ed. 1998,
37, 3359; b) A. M¸ller, J. Meyer, E. Krickemeyer, C. Beugholt, H.
Bˆgge, F. Peters, M. Schmidtmann, P. Kˆgerler, M. J. Koop, Chem.
Eur. J. 1998, 4, 1000.
[4] The prepartion and crystallization of this material is an undergradu-
ate exercise. L. Cronin, E. Diemann, A. M¸ller, in Inorganic Experi-
ments (Ed.: J. D. Woollins), Wiley-VCH, Weinheim, 2003, p. 340.
[5] A. M¸ller, E. Krickemeyer, H. Bˆgge, M. Schmidtmann, B. Botar,
M. O. Talismanova, Angew. Chem. 2003, 115, 2131; Angew. Chem.
Int. Ed. 2003, 42, 2085.
[6] a) A. M¸ller, S. S. Q. N. Sarkar, H. Bogge, M. Schmidtmann, S.
Sarkar, P. Kogerler, B. Hauptfleisch, A. X. Trautwein, V. Schune-
mann, Angew. Chem. 1999, 111, 3435; Angew. Chem. Int. Ed. 1999,
38, 3238; b) A. M¸ller, P. Kˆgerler, H. Bˆgge, Struct. Bonding 2000,
96, 203.
1,3,5-Tris-[N’-(4’-benzo-18-crown[6])-2-propynyl)-4,4’-trimethylenedipi-
peridino-N-(2-propyn-3-yl)]benzene (1): [Pd2(dba)3] (7.9 mg, 8.6 mmol)
and CuI (1.6 mg, 8.4 mmol) and tris-(2-furyl)phosphine (4.1 mg,
17.6 mmol) were added to a dry, 5 mL septum-capped round flask, which
was then sparged with argon and charged with dry DMF (0.5 mL). Neat
piperidine (32 mL, 323 mmol) and 4’-iodobenzo-18-crown[6] (124 mg,
283 mmol, dissolved in 1.5 mL DMF) were added via syringe to the stir-
red reaction mixture. The resulted mixture stirred for 15 min at 258C,
then compound 2 (80 mg, 86 mmol, dissolved in 1 mL DMF) was added
dropwise via syringe in a period of 20 min. The whole reaction mixture
was stirred at 258C. for 24 h. Then the resulted solid was filtered, the sol-
vent was concentrated and the residue was purified by alumina column
chromatography (gradient elution: EtOAc followed by 2 5% MeOH/
[7] A. M¸ller, E. Diemann, S. Q. N. Shah, C. Kuhlmann, M. C. Letzel,
Chem. Commun. 2002, 440.
[8] a) A. M¸ller, E. Diemann, C. Kuhlmann, W. Eimer, C. Serain, T.
Tak, A. Knˆchel, P. K. Pranzas, Chem. Commun. 2001, 1928; b) T.
Liu, J. Am. Chem. Soc. 2002, 124, 10942.
[9] R. Xu, Particle Characterization: Light Scattering Methods Particle
Technology Series (series Ed.: B. Scarlett), Kluwer, 2001.
[10] a) F. B. J. Schinner, L. F. Audrieth, S. T. Gross, D. S. McClellan, L. J.
Seppi, J. Am. Chem. Soc. 1942, 64, 2543; b) T. Liu, Q. Wan, Y. Xie,
C. Burger, L.-Z. Liu, B. Chu, J. Am. Chem. Soc. 2001, 123, 10966.
[11] A. Kiriy, G. Gorodyska, S. Minko, W. Jaeger, P. Stepµnek, M.
Stamm, J. Am. Chem. Soc. 2002, 124, 13454.
[12] a) S. Palarz, B. Smarsly, M. Antonietti, ChemPhysChem 2001, 457;
b) D. G. Kurth, P. Lehmann, D. Volkmer, H. Cˆlfen, M. J. Koop, A.
M¸ller, A. Du Chesne, Chem. Eur. J. 2000, 6, 385; c) D. Volkmer, A.
Du Chesne, D. G. Kurth, H. Schnablegger, P. Lehmann, M. J. Koop,
A. M¸ller J. Am. Chem. Soc. 2000, 122, 1995.
CHCl3), which yielded
a yellow hygroscopic solid (120 mg, 75%).
1H NMR (400 MHz, CDCl3, 258C, TMS): d=1.18 1.34 (m, 36H; CH,
CH2 and axial piperidine CH2), 1.66 1.80 (m, 12H; equatorial piperidine
CH2), 2.14 2.22 (m, 12H; axial CH2-N), 2.95 (t, 2J(H,H)=11.8 Hz, 12H;
equatorial CH2-N), 3.45 (s, 6H; propargyl CH2), 3.47 (s, 6H; propargyl
CH2), 3.69 (s, 12H; O-(CH2)2-O), 3.70 3.73 (m, 12H; CH2-O), 3.76 3.78
(m, 12H; CH2-O), 3.90 3.93 (m, 12H; CH2-O), 4.12 4.16 (m, 12H; CH2-
O), 6.78 (d, 3J(H,H)=8.28 Hz, 3H; CH aromatic benzocrown), 6.95 (d,
4J(H,H)=2.00 Hz, 3H; CH aromatic benzocrown), 7.01 (dd, 3H,
3J(H,H)=8.28 Hz, 4J(H,H)=2.00 Hz, 3H; CH aromatic benzocrown),
7.40 (s, 3H; CH aromatic); 13C NMR (100 MHz, CDCl3): d=23.9, 32.4,
35.25, 35.30, 36.7, 48.0, 48.2, 53.0, 53.1, 69.0, 69.5, 70.8, 70.9, 83.5, 83.6,
84.8, 86.2, 113.5, 115.9, 117.2, 123.7, 125.3, 134.2, 148.4, 149.2; MS
(MALDI-TOF): m/z (%): 1863 (88) [M+H]+, 1885 (100) [M+Na]+ ; ele-
mental analysis calcd (%) for C111H156N6O18¥4H2O (1934.6): C 68.92, H
8.13, N 4.34; found: C 69.02, H 8.26, N 4.42.
[13] T. Liu, J. Am. Chem. Soc. 2003, 125, 312.
[14] A. M¸ller, M. Koop, H. Bogge, M. Schmidtmann, C. Beugholt,
Chem. Commun. 1998, 1501.
[15] K. C. V. Sharma, A. Clearfield, J. Am. Chem. Soc. 2000, 122, 1558.
[16] a) J. C. Russ, Fundamentals of energy dispersive X-ray analysis, But-
terworths, 1984; b) B. Dziunikowski, Energy dispersive X-ray fluo-
rescence analysis, Elsevier, 1989.
Acknowledgement
[17] When the ring in the Au standard corresponding to Miller index
[1.1.1] was used, the lattice spacing was 3.8 ä. The X-ray structure
of {Mo132} keplerate (CSD-410097, Fachinformationszentrum Karls-
ruhe, CRYSDATA@FIZ-Karlsruhe.DE) has Mo Mo distances
3.80 ä (180î), 3.30 ä (60î) and 2.61 ä (30î).
The authors thank Gerald Thomas of the Center for Applied Energy Re-
search at the University of Kentucky for ICP analysis and the National
Science Foundation, USA CHE0111578 for supporting this work. TEM
and NMR support acknowledge NSF (USA), EPS-9874764 and
CHE997841.
[18] a) K. Sonogashira in Metal-Catalyzed Cross-Coupling Reactions
(Eds.: F. Diederich, P. J. Stang), Wiley-VCH, New York, 1998,
p. 203; b) K. Sonogashira in Comprehensive Organic Synthesis (Eds.:
B. M. Trost, I. Flemming), Pergamon, New York, 1991, p. 521.
[19] W.-S. Xia, R. H. Schmehl, C.-J. Li, Tetrahedron 2000, 56, 7045.
[20] S. Kajigaeshi, T. Kakinami, M. Moriwaki, M. Watanabe, S. Fujisaki,
T. Okamoto, Chem. Lett. 1988, 795.
[1] a) M. T. Pope, in Comprehensive Coordination Chemistry, Vol. 3
(Eds.: G. Wilkinson, R. D. Gillard, J. A. McCleverty), Pergamon
Press, 1987, pp. 1023 1058; the {Mo132} keplerate is the smallest
structure with a closed surface found in solid states of PPMo2AcO;
{Mo132} keplerate was used as a convenient basis for the concentra-
tion of molybdate in these studies.
[2] a) A. M¸ller, J. Meyer, E. Krickemeyer, E. Diemann, Angew. Chem.
1996, 108, 1296; Angew. Chem. Int. Ed. 1996, 35, 1206; b) P. Sou-
chay, Polyanions et polycations, Oauthier-Villars, 1963; c) W. D.
Treadwell, Y. Schaeppi, Helv. Chim. Acta 1946, 29, 771; d) O.
Glemser, G. Lutz, Z. Anorg. Allg. Chem. 1951, 264, 17.
Received: August 20, 2003 [F5468]
Published online: March 22, 2004
2427
Chem. Eur. J. 2004, 10, 2421 2427
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim