вход по аккаунту


Crystalline Porous Microspheres Made from Amino Acids by Using Polymer-Induced Liquid Precursor Phases.

код для вставкиСкачать
Crystal Engineering
Crystalline, Porous Microspheres Made from
Amino Acids by Using Polymer-Induced Liquid
Precursor Phases**
Sebastian Wohlrab, Helmut Clfen,* and
Markus Antonietti
Control of the crystallization processes is one of the most
important techniques in applied colloid chemistry. With a
controlled preparation of crystallites of defined size, the
kinetic aspects of solubility (e.g. of pharmaceuticals), crystal
superstructures, texture-controlled flow properties, filling
degrees, and mechanical properties can be influenced. A
high reproducibility of the chosen procedure is necessary to
ensure the quality of industrial products.
As nucleation and growth are very sensitive processes,
crystallization is usually controlled by addition of nucleation
agents, stabilizers, or other ternary components. The use of
special cosolvents,[1] low molecular additives, surfactants, and
functional polymers is regularly reported (for recent reviews,
see refs. [2–4]). Recent progress in the field of morphosynthesis indicates that there are non-classical pathways of
crystallization via colloidal intermediates and subsequent
mesoscale transformation.[3, 5–7] In this process, crystalline
structures are constructed by assembly and/or transformation
from larger units (instead of by the addition of single
molecules). Both Addadi et al.[8] and Clfen and Mann[3]
reviewed independently the role of amorphous nanoparticles
in bio- and biomimetic mineralization. The experimental
verification and role of similar intermediates in the sizedirected formation of CaCO3 particles is a question of
industrial relevance.[9, 10] Taden et al. analyzed a model
system based on dye nanodroplets and found extraordinarily
ordered dye crystals formed by mesoscale transformation.[7]
Such colloidal intermediates are a manifestation of the
Ostwald rule of stages, which describes structural growth in
the practically relevant kinetic control limit.[11]
High-concentration liquid intermediates seem to be
especially promising for rapid crystallization leading to
objects with a controlled outer shape. Gower et al. were the
first to describe the fine droplets of a polymer-induced liquid
precursor phase (PILP) for a CaCO3/polypeptide system,[12, 13]
the subsequent crystallization of which leads to bent, corrugated, and curved crystal superstructures. A special feature of
[*] Dr. S. Wohlrab, Dr. H. Clfen, Prof. Dr. M. Antonietti
Max-Planck-Institute of Colloids and Interfaces
Department of Colloid Chemistry
Research Campus Golm, 14424 Potsdam (Germany)
Fax: (+ 49) 331-567-9502
[**] The support of this work by the Max-Planck-Society is gratefully
acknowledged. We thank Mrs. M. Barth for experimental assistance.
Dr. Helmut Schlaad is acknowledged for the help with interpreting
the 1H NMR spectroscopy results.
Supporting information for this article is available on the WWW
under or from the author.
Angew. Chem. Int. Ed. 2005, 44, 4087 –4092
DOI: 10.1002/anie.200462467
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
the PILP phases is that they allow easy morphosynthesis of
novel, biorelevant mineral structures through the shaping of a
liquid on the micrometer scale so that crystallization occurs
without requiring mass transport over longer distances.[14, 15]
Herein we extend this approach to polar organic molecules by studying the formation of PILP phases consisting of
charged amino acids and oppositely charged polyelectrolytes
in ethanol/water mixtures. While addition of ethanol to an
aqueous solution of the pure amino acid leads to well shaped
single crystals, stable PILP microdroplets are formed spontaneously by adding an oppositely charged polyelectrolyte to
the starting solution. These precursor phases can subsequently be crystallized to build novel crystalline porous
microspheres of the component amino acids.
The method presented herein is an addition to the
available approaches for the preparation of porous microspheres, such as solvothermal methods,[16] sol–gel methods,[17]
preparation in commercially available microspherical silica
gels,[18] the use of microemulsion droplets as liquid templates
to prepare hollow inorganic microspheres,[16, 19–21] and
dynamic emulsification of oil droplets containing metal
oxide precursors which through a reaction at the oil–water
interface give porous star-shaped shells.[22]
Amino acids are practically insoluble in organic media, so
their aqueous solutions undergo fast crystallization upon
addition of a water-miscible organic solvent.[23] The charged
amino acids employed, glutamic acid, aspartic acid, lysine,
and histidine, crystallize according to this general scheme.
Figure 1 a shows the result of a default experiment with d,lglutamic acid, and is representative for the other amino acids.
Irregular elongated crystals with a size of 50 to 400 mm are
obtained. As a result of the very fast precipitation process, no
defined crystal faces or well-shaped edges are formed,
although the macroscopic facetted crystal implies a singlecrystalline nature.
A completely different behavior is obtained by addition of
different amounts of branched polyethyleneimine (PEI, Mw =
600 g mol 1) corresponding to polymer:amino acid mass ratios
of approximately 2:1, 1:1, 1:2, and 1:10 to the d,l-glutamic
acid starting solution. The addition of ethanol results in
turbidity, but no precipitation of crystals. Light microscopy
investigations indicate the formation of amorphous liquid
droplets (Figure 1 b), similar to the PILP phases reported by
Owing to their size and the small refractive index differences to the surrounding solvent, the existence of a PILPphase is sometimes hard to observe. Anionic or cationic dyes,
which interact with the polymers in the PILP phase, can be
used for staining, such as the anionic dye amaranth for the
PILP of d,l-glutamic acid and polyethyleneimine. Amaranth
is highly soluble in ethanol and in water, but in the charged
PILP-phase it acts as an ionophile leading to high extraction
The process is illustrated in Figure 2. A 0.01 wt %
amaranth solution in ethanol (I, red solution) is added to a
d,l-glutamic acid solution (II, 10 g L 1, clear solution) with
1 wt % PEI600. A turbid pink emulsion is formed, and
microscopy reveals liquidlike red droplets in a transparent
medium. Centrifugation (3500 rpm for 30 min) of the emul-
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 1. a) d,l-glutamic acid crystals made by precipitation crystallization after adding 10 mL ethanol to 1 mL of a saturated solution (at
20 8C) of d,l-glutamic acid, b) PILP formation under the same conditions in the presence of 1 wt % PEI (Mw = 600 g mol 1). Scale bars
a) 50 mm, b) 20 mm.
Figure 2. a) I contains ethanol + 0.01 % amaranth, II contains 10 g L
d,l-glutamic acid + 1 % PEI600, b) PILP formation after mixing of I
and II, c) forced sedimentation and demixing by centrifugation
(3500 rpm min 1; 30 min) d) colored highly viscous PILP after
sion results in a phase-separated binary system, where the
dyed PILP forms a liquid viscous phase at the bottom
(Figure 2 c). This phase is very viscous as demonstrated in
Figure 2 d and subsequent crystallization leads to an imprint
Angew. Chem. Int. Ed. 2005, 44, 4087 –4092
of the centrifuge tube, the opposite side of the crystal has a
microporous appearance (see Supporting Information). This
experiment clearly demonstrates the existence of a liquid
amino acid and polymer-rich phase in chemical equilibrium
with its environment.
The droplets consist of amino acid bound to the polymer
with a certain amount of water, while a surface excess of the
charged polymers stabilizes the droplets. FTIR and 1H NMR
spectroscopic analyses qualitatively revealed a high amount
of protonated PEI in the d,l-glutamic acid PILPs but through
quantitative 1H NMR spectroscopic analysis it was not
possible to determine the exact concentration (data not
shown). According to the altered 1H NMR chemical shifts for
all d,l-glutamic acid signals, the entire d,l-glutamic acid in
the PILP interacts with PEI, thus indicating electrostatic as
well as hydrogen-bonding interactions exist. This result was
also supported by the FTIR data and pH dependent experiments (data not shown). For all polymer/amino acid pairs at
different mass ratios, investigations indicate the formation of
PILPs in the range of 1 to 5 mm. The PILPs have a long time
stability (several months), but vanish immediately when the
surrounding ethanol is evaporated. Experimental parameters,
such as mode of mixing, stirring (at 0, 400, 1200 rpm), or
ageing, had no influence on the particle size distribution.
To investigate the interacting forces and in particular the
influence of charge, the zeta potentials of pure d,l-glutamic
acid crystals where measured. Additionally, PEI600 was added
to the crystals in a crystal:polymer ratio of 1:1 to give a
reference value for polymer-coated colloids. While crystals of
d,l-glutamic acid show a negative zeta potential of around
19 mV, crystals with adsorbed PEI have a positive potential
of around 27 mV (Figure 3).
All PILPs made of d,l-glutamic acid and PEI600 gave a
positive zeta potential of around 18 mV, indicating that the
droplets are indeed electrostatically stabilized by an outer
excess layer of adsorbed PEI. This finding also reflects the
high PEI amount in the PILPs found by FTIR and 1H NMR
Figure 3. Zeta potentials for different particles: pure d,l-glutamic acid
crystals (c); d,l-glutamic acid crystals plus PEI600 (b); positive
zeta potential of a PILP made of d,l-glutamic acid and PEI600 (g).
Angew. Chem. Int. Ed. 2005, 44, 4087 –4092
The three other polyelectrolyte/amino acid pairs examined also show comparable PILP formation. For the basic
amino acids l-lysine and l-histidine, polyacrylic acid, PAA2000
(Mw = 2000 g mol 1), was employed to form the PILPs.
Crystallization does not occur in any of the systems containing
a low amount of amino acid and a high amount of polymer. A
charge-coupled complex between the polyelectrolyte and the
amino acid counterions is presumably inhibited from crystallizing, very similar to many structures formed by ionic self
To promote crystallization from the PILPs, the amino acid
concentration and thus the driving force towards crystallization has to be increased, while keeping the polymer concentration at the former level. The high amino acid concentrations of 10 wt % can only be realized by dissolving the amino
acid at 60 8C, the PILPs formed at this temperature are then
slowly cooled to room temperature to drive crystallization. As
shown by light and polarized-light microscopy, the PILP is
still liquid directly after cooling. However, after 1 h crystallization is underway for all amino acid/polymer pairs and
concentrations, and within 24 h, the vast majority of droplets
have crystallized, but a certain amount of PILP droplets are
still present. After one week the crystallization process is
completed. The resulting crystalline superstructures are
spherical and have diameters between 10–100 mm.
The final crystalline structures are therefore much larger
than the primary droplets. It is meanwhile a quite common
observation that the colloidal stability is lowered throughout
the crystallization process, and that the first-formed structures
grow by attraction and aggregation of the primary amorphous
species.[3, 7] Figure 4 a shows the light-microscopy image of the
so-formed spherical crystal superstructures of d,l-glutamic
acid. The typical spherical shape is a superstructure made of
primary nanocrystals which in polarization microscopy results
in the so called “Brewster cross” (Figure 4 b).[25] This result
indicates at a perfect radial alignment of the nanocrystals.
Scanning electron microscopy (SEM) investigations
reveal that the spherical superstructure is porous, with the
primary nanocrystals having a platelet like appearance with a
diameter of 200 nm and thickness of 10–20 nm (Figure 4 c and
d). As d,l-glutamic acid usually crystallizes as needles, this
shape speaks for a morphology control provided by the
cationic polyelectrolyte, which stabilizes a distinct face. A
double hydrophilic block copolymer was recently reported to
exert a similar influence on the crystallization of d,lalanine.[26] To learn more about the microparticle composition, we performed FTIR experiments, these indicated a
negligible amount of PEI in the d,l-glutamic acid microcrystals (data not shown). These results were confirmed by
H NMR spectroscopy data (data not shown) which revealed
a low amount of protonated PEI around the 1H NMR
detection limit in the microparticles. Wide-angle X-ray
scattering (WAXS) revealed domain sizes of d,l-glutamic
acid monohydrate > 100 nm for all directions except along
the [232] axis which was found to be 20 nm in agreement with
the nanoplate thickness in Figure 4 d. The formation of these
exposed and initially negatively charged (Cerius2 calculation,
not shown) nanoplatelet (232) faces maybe the result of faceselective polymer adsorption. This idea is supported by
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 4. d,l-glutamic acid crystals through growth from a concentrated PILP precursor phase. The PILP formation was induced by adding 10 mL
ethanol to 1 mL of a saturated solution (at 60 8C) of d,l-glutamic acid in the presence of 1 wt % PEI (Mw = 600 g mol 1), a) light microscopy,
b) polarized light microscopy, c) SEM overview of some crystalline superstructures indicating porosity, d) SEM image with a higher magnification
to resolve the pore architecture and the construction principle of the spheres. Scale bars: a,b) 100 mm, c) 10 mm, d) 200 nm.
H NMR spectroscopy which revealed a notable portion of
d,l-glutamic acid with a different chemical shift to the pure
crystal. This difference was also found in the PILP and
indicates interaction with cationic PEI (data not shown).
The porous character is a memory of the excess water and
polyelectrolyte in the parent PILP droplets. Therefore it is
important to dry the crystals at 50 8C under vacuum to ensure
a water-free product. The porosity is then determined by
nitrogen sorption experiments using the BET method. For the
samples prepared with polymer concentrations of 1 wt %, a
BET-surface area of approximately 13 m2 g 1 is found which is
independent of the initial PILP droplet size. An estimate of
the surface area based on the SEM images would result in
slightly higher values and suggests the possibility that parts of
the structural core are not accessible. BET measurements for
l-lysine particles reveal a specific surface of 17 m2 g 1, very
similar to the value for the d,l-glutamic acid.
In the case of the basic amino acids, crystallization again
leads to spherical aggregates of smaller crystals (Figure 5). In
this case, we have chosen to use the chiral l-enantiomorphs.
Again they show a completely different growth pattern to that
of the corresponding pure solutions (default experiments
shown in the insets of Figure 5), that is, the added polyacrylic
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
acid also modifies the morphology of the primary nanocrystals. Flat crystals, with a thickness of around 50 nm (lhistidine) and about 30 nm (l-lysine), come together to form
porous spherical particles with high surface areas. The
platelets however have larger lateral extensions than in the
d,l-glutamic acid case, resulting in a snowball-like morphology of the final crystalline superstructures. Again, it is
interesting to see that the crystals are—whenever possible—
radially aligned, so that the distance between the platelets can
be seen.
The aspartic acid is effectively blocked from crystallization by the PEI coordination even at elevated concentrations,
this is because of its low solubility in water (4 g L 1 at 20 8C
compared to 10 g L 1 for d,l-glutamic acid at 20 8C), thus
other conditions have to be explored for this system.
In conclusion, precipitation of charged amino acids in the
presence of oppositely charged polyelectrolytes leads in all
the examined cases to rather unusual colloidal structures, the
resulting liquid droplets are made of the corresponding amino
acid complexes in water. At lower supersaturations these
supramolecular “water-in-ethanol” emulsions are stable
against crystallization and Ostwald ripening for several
months. This prolonged stability against crystallization is
Angew. Chem. Int. Ed. 2005, 44, 4087 –4092
crystallization into desired morphologies or crystalline coatings,[13] as well as into micro- and nanostructures made of
polar organic molecules. The particles have in addition an
appropriate size, shape, and porosity for chromatographic
applications, where beads made of the enantio-pure amino
acids are especially promising.
Experimental Section
Figure 5. SEM images of a) l-histidine (scale bar 20 mm), b) l-lysine
(scale bar 3 mm) crystal superstructures formed through growth from
their PILPs. Insets: light micrographs of the default experiments of the
as-prepared crystals precipitated without polymer (scale bars 250 mm).
most likely a result of the organic molecule crystals used
which have a lower lattice energy than the inorganic ionic
crystals investigated previously.[13–15] A kinetic effect can not
be ruled out. By increasing the amino acid concentration, it is
possible to use this polymer-induced liquid precursor state as
a basis for crystallization. After some days, spherical, porous
crystal superstructures with 10–100 mm diameter and a radial
alignment of the primary nanocrystals are obtained. If the
crystallization is repeated in 0.1n HCl or 0.1n NaOH, the
electrostatic influence of the polymer on the resulting
structures can be revealed. Although PILPs were obtained
in every case, porous microcrystals were only obtained at
acidic to neutral pH values for the acidic d,l-glutamic acid,
and neutral to basic pH values for the basic l-histidine and llysine.
The described crystallization can also be understood as a
special case of the crystallization by “drowning-out”, “salting
out”, or “solventing out”.[27] During protein crystallization
similar effects can be observed when “liquid protein” phases
are built.[28] While this effect is undesirable during protein
crystallization, application of such precursors for crystal
morphogenesis is very promising. Since the crystallization
proceeds by the Ostwald rule of stages, including a liquid
precursor step, the drowning-out process now becomes more
attractive for crystal engineering. By using liquid intermediates for further crystallization, it is possible to direct the
Angew. Chem. Int. Ed. 2005, 44, 4087 –4092
All polymers and amino acids were purchased from Aldrich (purity:
99 % for d,l-glutamic acid, 98 % for l-histidine and d,l-aspartic acid,
and > 97 % for l-lysine), and used without further purification.
Double distilled water was used for the preparation of the solutions.
As the precipitation agent, HPLC grade ethanol (Merck) was used.
PILP formation: The oppositely charged polymer (0.1, 0.5, 1.0,
and 2.0 wt % with respect to the water amount) was added to a
solution of the amino acid (1 mL, 1 wt %; or in case of a lower
solubility product, the saturated solution) at 20 8C in double distilled
water. Branched polyethyleneimine (PEI600 ; Mw = 600 g mol 1) was
used as the oppositely charged polymer for the negatively charged
amino acids d,l-glutamic acid (isoelectric point (IEP) = pH 4.20) and
d,l-aspartic acid (IEP = pH 4.25). Polyacrylic acid (PAA2000 ; Mw =
2000 g mol 1) was employed to form PILPs of the positively charged
amino acids l-lysine (IEP = pH 8.88) and l-histidine (IEP =
pH 6.78). Ethanol (10 mL) was added to this mixture with stirring.
The developed PILP phases could be observed by light microscopy as
droplets with a diameter of 1–5 mm, depending on mechanical
agitation. These low-concentration PILPs are stable for at least
6 months, both against crystallization and colloidal agglomeration.
Crystallization from the PILPs: To stimulate crystallization from
the PILPs, the concentration of amino acid has to be increased, which
was done by elevating the temperature to 60 8C. The oppositely
charged polymers (0.1, 0.5, 1.0, and 2.0 wt %) were added to the
solution of the amino acid (1 mL of a 10 wt % solution, or in case of a
lower solubility, the saturated solution) in double distilled water.
Ethanol (10 mL) was added with stirring and the solution was allowed
to cool to room temperature. Liquid PILP droplets could be observed,
that crystallized within 24 h as indicated by light microscopy. The
crystals were isolated from the fluid and washed three times with
ethanol, then dried at 50 8C under vacuum.
Light microscopy, polarized light microscopy, and high-resolution
scanning electron microscopy (HRSEM) were applied to all samples.
Light microscopy in solution establishes that the SEM micrographs
show real structures and not drying artifacts resulting from the sample
preparation. The SEM measurements were performed on a LEO 1550
GEMINI. Light microscopy images were taken in solution with an
Olympus BX50 microscope. BET isotherms were measured with a
Micromeretics Tristar 3000. Zeta potentials were measured with a
Zetasizer HS3000 (Malvern), which determines the zeta potential
from the speed of the particle movement in an electric field. Crystal
and PILP containing dispersions in ethanol with and without PEI600
were investigated according to the conditions of the crystallization
Received: October 29, 2004
Revised: March 19, 2005
Published online: June 1, 2005
Keywords: amino acids · crystal engineering · glutamic acid ·
porous microsphere · superstructures
[1] M. Lahav, L. Leiserowitz, Chem. Eng. Sci. 2001, 56, 2245 – 2253.
[2] L. A. Estroff, A. D. Hamilton, Chem. Mater. 2001, 13, 3227 –
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
[3] H. Clfen, S. Mann, Angew. Chem. 2003, 115, 2452 – 2468;
Angew. Chem. Int. Ed. 2003, 42, 2350 – 2365.
[4] F. C. Meldrum, Int. Mater. Rev. 2003, 48, 187 – 224.
[5] N. Jongen, P. Bowen, J. Lematre, J. C. Valmalette, H. Hofmann,
J. Colloid Interface Sci. 2000, 226, 189 – 198.
[6] R. L. Penn, J. F. Banfield, Geochim. Cosmochim. Acta 1999, 63,
1549 – 1557.
[7] A. Taden, K. Landfester, M. Antonietti, Langmuir 2004, 20,
957 – 961.
[8] L. Addadi, S. Raz, S. Weiner, Adv. Mater. 2003, 15, 959 – 970.
[9] D. Horn, J. Rieger, Angew. Chem. 2001, 113, 4460 – 4492; Angew.
Chem. Int. Ed. 2001, 40, 4331 – 4361.
[10] J. Rieger, Tenside Surfactants Deterg. 2002, 39, 221 – 225.
[11] W. Ostwald, Z. Phys. Chem. 1897, 22, 289 – 330.
[12] L. A. Gower, D. A. Tirrell, J. Cryst. Growth 1998, 191, 153 – 160.
[13] L. B. Gower, D. J. Odom, J. Cryst. Growth 2000, 210, 719 – 734.
[14] M. J. Olszta, E. P. Douglas, L. B. Gower, Calcif. Tissue Int. 2003,
72, 583 – 591.
[15] M. J. Olszta, D. J. Odom, E. P. Douglas, L. B. Gower, Connect.
Tissue Res. Suppl. 1 2003, 44, 326 – 334.
[16] H. Hu, B. Yang, X. Liu, R. Zhang, Y. Qian, Inorg. Chem.
Commun. 2004, 7, 563 – 565.
[17] B. Zhang, B. Chen, K. Shi, S. He, X. Liu, Z. Du, Appl. Catal. B
2003, 40, 253 – 258.
[18] Y. Tsurita, M. Nogami, J. Ceram. Soc. Jpn. 2001, 109, 992 – 999.
[19] D. Walsh, B. Lebeau, S. Mann, Adv. Mater. 1999, 11, 324 – 328.
[20] D. Walsh, S. Mann, Nature 1995, 377, 320 – 323.
[21] S. Schacht, Q. Huo, I. G. Voigt-Martin, G. D. Stucky, F. Schth,
Science 1996, 273, 768 – 771.
[22] D. Volkmer, S. Tugulu, M. Fricke, T. Nielsen, Angew. Chem.
2003, 115, 60 – 64; Angew. Chem. Int. Ed. 2003, 42, 58 – 61.
[23] R. C. Weast, CRC Handbook of Chemistry and Physics, 65th ed.,
CRC, Boca Raton, FL, 1984/85.
[24] C. F. J. Faul, M. Antonietti, Adv. Mater. 2003, 15, 673 – 683.
[25] H. Freund, Handbuch der Mikroskopie in der Technik,
Umschauverlag, Frankfurt a.M., 1957.
[26] S. Wohlrab, N. Pinna, M. Antonietti, H. Clfen, Chem. Eur. J.
2005, 11, 2903 – 2913.
[27] J. W. Mullin, Crystallisation, Butterworth Heinemann, Woburn,
2001, p. 333.
[28] Y. G. Kuznetsov, A. J. Malkin, A. McPherson, J. Cryst. Growth
2001, 232, 30 – 39.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 4087 –4092
Без категории
Размер файла
513 Кб
acid, polymer, porous, using, crystalline, induced, amin, microspheres, made, precursors, liquid, phase
Пожаловаться на содержимое документа