ARTICLE

pubs.acs.org/cm

Controlling Zeolitic Imidazolate Framework Nano- and Microcrystal

Formation: Insight into Crystal Growth by Time-Resolved In Situ Static
Light Scattering
Janosch Cravillon,† Roman Nayuk,‡ Sergej Springer,† Armin Feldhoff,§ Klaus Huber,*,‡ and
Michael Wiebcke*,†
†
Institut f€ur Anorganische Chemie, Leibniz Universit€at Hannover, Callinstr. 9, 30167 Hannover, Germany
‡
Department Chemie, Universit€at Paderborn, Warburger Str. 100, 33098 Paderborn, Germany
§
Institut f€ur Physikalische Chemie und Elektrochemie, Leibniz Universit€at Hannover, Callinstr. 3A, 30167 Hannover, Germany

bS Supporting Information
ABSTRACT: We report on a simple and straightforward
method that enables the rapid room-temperature production
of nanocrystals (finely tuned in size between ∼10 and 65 nm)
and microcrystals (∼1 μm) of the prototypical microporous
zeolitic imidazolate framework (ZIF) material ZIF-8. Control of
crystal size is achieved in a novel approach by employing an
excess of the bridging bidentate ligand and various simple
auxiliary monodentate ligands with different chemical function-
alities (carboxylate, N-heterocycle, alkylamine). The function of the monodentate ligands can be understood as a modulation of
complex formation and deprotonation equilibria during crystal nucleation and growth. Using time-resolved static light scattering,
the functioning of modulating ligands is monitored for the first time by in situ experiments, which offered significant insight into the
crystal growth processes. Formation of nanocrystals is characterized by continuous, comparatively slow nucleation and fast crystal
growth occurring on a time scale of seconds. Although nucleation and growth are not separated from each other, a significant
narrowing of the particle size distribution during early stages results in rather monodisperse nanocrystals, before broadening of the
particle size distribution occurs, as observed by complementary ex situ electron microscopy studies. Microcrystal growth is
dominated by a particle
monomer addition mechanism, but indications for the operation of a coalescence process during early
stages of growth have been also obtained. During later stages of microcrystal growth crystals change their shape from cubes to
rhombic dodecahedra. The prepared phase-pure ZIF-8 nanoscale materials exhibit good thermal stability in air and large surface
areas, which are comparable to those of large macrocrystals. Nanocrystal powders exhibit dual micro- and mesoporosity.
KEYWORDS: metal
organic framework, zeolitic imidazolate framework, nanomaterials, modulated synthesis, crystal growth,
in situ static light scattering

’ INTRODUCTION on their supply as nano- and microcrystals of well-defined size

Zeolitic imidazolate framework (ZIF) materials1 constitute a and shape, as is well-known for zeolites.7 Development of size-
new distinctive, rapidly developing subclass of crystalline porous and shape-controlled syntheses, in turn, benefits considerably
coordination polymers (PCPs) or metal
organic frameworks from a detailed understanding of the physicochemical funda-
(MOFs).2 For example, bulk ZIF materials3 and supported ZIF mentals of the crystallization processes.8
membranes and films4 have shown promising properties in the These important issues have only scarcely been addressed in
fields of CO2 capture and storage, separation of light gases, the field of porous MOFs. A limited number of carboxylate-based
sensing of vapors, and heterogeneous catalysis. The three- nanoscale MOFs have been prepared, for example, by reverse
periodic tetrahedral framework structures of ZIFs are con- microemulsion methods and microwave- or ultrasound-assisted
structed from bivalent metal cations and bridging substituted syntheses.9,10 A very attractive approach to controlling crystal
imidazolate anions and frequently possess a zeolite topology.1,5 size and shape that has been recently introduced is the coordina-
tion modulation method which employs an auxiliary mono-
Numerous ZIFs combine the attractive features of MOFs
dentate ligand that acts in competition to the multidentate
(diversity of framework structures and pore systems, large sur-
bridging ligand at surface-exposed metal centers of the forming
face areas, post-synthetically modifiable organic bridging ligands)
with high thermal and chemical stability.6 It is this combination
of properties which makes ZIFs very promising candidate Received: December 15, 2010
materials for many technological applications. On the other Revised: February 24, 2011
hand, properties and performance of porous materials rely much Published: March 18, 2011

r 2011 American Chemical Society 2130 dx.doi.org/10.1021/cm103571y | Chem. Mater. 2011, 23, 2130–2141
Chemistry of Materials ARTICLE

crystals.11,12 Such modulating ligands are usually monocarbox- centrifugation. Washing with fresh MeOH and centrifugation is repeated
ylates, having the same chemical functionality as the bridging three times. The product is dried at room-temperature under reduced
polycarboxylate ligands. The power of this method has been very pressure.
recently demonstrated by Kitagawa and co-workers12b who were Synthesis of ZIF-8 Microcrystals with Sodium Formate or
able to prepare [Cu3(btc)2] crystals (HKUST-1, btc = benzene- 1-Methylimidazole As a Modulating Ligand. Typically, 734.4
1,3,5-tricarboxylate) in the whole range from the small nanoscale mg (2.469 mmol) of Zn(NO3)2 3 6H2O is dissolved in 50 mL of MeOH.
(∼20 nm) to the microscale (∼2 μm). First insight into the A second solution is prepared by dissolving 810.6 mg (9.874 mmol) of
mechanisms of carboxylate-based MOF crystallization has been Hmim and 810.6 mg (9.874 mmol) of 1-methylimidazole in 50 mL of
obtained by a number of recent in situ investigations employing MeOH. The latter clear solution is poured into the former clear solution
static light scattering,11a,13 surface plasmon resonance spectro- under stirring with a magnetic bar. Stirring is stopped after combining
scopy,14 atomic force microscopy,15 and energy-dispersive X-ray the component solutions. After 24 h, the precipitate is recovered by
diffraction.16 Furthermore, the growth of nanorods of a carbox- filtration, washing with MeOH, and drying under reduced pressure.
ylate-based MOF by oriented attachment has been demonstrated Syntheses in the presence of sodium formate is made similarly (see
Supporting Information).
in a remarkable ex situ electron microcopy study.12a
Progress is still more limited for ZIFs. We have recently Synthesis of ZIF-8 Nanocrystals with n-Butylamine as a
Modulating Ligand. Typically, 734.4 mg (2.469 mmol) of Zn-
reported in a preliminary communication on a simple and rapid
(NO3)2 3 6H2O is dissolved in 50 mL of MeOH. A second solution is
room-temperature solution-based synthesis of ∼45 nm-sized
prepared by dissolving 810.6 mg (9.874 mmol) of Hmim and 0.975 mL
ZIF-8 nanocrystals with a rhombic dodecahedral shape and a
(9.874 mmol) of n-butylamine in 50 mL of MeOH. The latter clear
narrow size distribution.17 Such ZIF-8 nanocrystals have mean- solution is poured into the former clear solution under stirring with a
while been used to fabricate porous composite nanofibers by magnetic bar. Stirring is stopped after combining the component
electrospinning,18 supported membranes with random19 and solutions. After 24 h, the gel-like solid is recovered by centrifugation.
preferred crystal orientation for gas separation,20 thin films with Washing with fresh MeOH and centrifugation is repeated three times.
dual micro- and mesoporosity for selective adsorption and The product is dried at room-temperature under reduced pressure.
sensing of vapors,21 and capillary coatings for the chromato- Syntheses with variation of the molar ratios were made similarly.
graphic separation of alkanes,22 indicating the wide range of Methods of Characterization. Powder X-ray diffraction (XRD)
potential applications of nanoscale ZIF materials. Later, Thalla- at room temperature was performed in transmission mode on a STOE
pally and co-workers23 have also prepared ∼50 nm-sized ZIF-8 Stadi-P diffractometer using monochromatized Cu KR1 radiation of
nanocrystals in a similar approach but with the addition of an wavelength λ = 1.54059 Å. The instrumental peak broadening needed
organic polymer, claiming that the nanocrystals have a hexagonal for Scherrer analysis was determined using a silicon reference standard
shape. In addition, Li and co-workers24 have recently reported on (NIST 640c). For intensity data recording at variable temperatures in
spherical ZIF-7 nanocrystals as well as ZIF-7 nano- and micro- Debye
Scherrer geometry, the same diffractometer was equipped with
rods. All these successful syntheses of nano- and microscale ZIF a STOE high-temperature oven. Samples were filled into thin-walled
materials had to be developed empirically by exploratory syn- silica glass capillaries (diameter 0.5 mm) which were left unsealed.
thetic work, since a detailed understanding of the crystallization Small-angle X-ray scattering (SAXS) measurements were performed
processes is missing. in transmission mode on a RIGAKU system consisting of a microfocus
In order to gain a better understanding of the mechanisms of X-ray tube with Cu target and mirror optics (λ = 1.541 Å), a three-
ZIF crystallization and thereby put the synthesis of ZIF materials pinhole collimating system, and a 2D gas-filled multiwire detector.
on a more rational basis, we have combined systematic synthetic Colloidal solutions were filled into thin-walled glass capillaries
work with time-resolved experiments, namely, in situ static light (diameter 1.5 mm), while powder samples were kept between thin
scattering (SLS) and ex situ scanning electron microscopy Kapton foils. SANS Analysis software25 provided by the NIST Center for
(SEM). Herein, we demonstrate for the first time that various Neutron Research was used for model fitting of SAXS curves in
simple modulating monodentate ligands with different function- reciprocal space. Indirect Fourier transformation was applied to obtain
alities (sodium formate, 1-methylimidazole, n-butylamine) can pair distance distribution functions (PDDFs)26 in direct space from the
be used in rapid room-temperature syntheses to tune the size of SAXS data by using the progam GNOM.27
ZIF-8 crystals between ∼10 nm and 1 μm. Along with this, Thermogravimetry (TG) and difference thermal analysis (DTA)
significant insight is presented into ZIF-8 crystallization pro- measurements were performed simultaneously on a NETZSCH 429
cesses. We further demonstrate that the prepared nanoscale ZIF- Thermoanalyzer. Samples were filled into alumina crucibles and heated
in a flow of air with a ramp of 5
C 3 min
1 from room temperature up to
8 materials are easily activated and exhibit good thermal stability
1000
C.
in air as well as large surface areas.
Nitrogen physisorption isotherms were measured at
196
C on a
QUANTACHROME Autosorb1-MP volumetric instrument. Samples
’ EXPERIMENTAL SECTION were outgassed in vacuum at room temperature for at least 24 h before
the sorption measurements. Surface areas were estimated by applying
Materials. Commercially available chemicals were used without the Brunauer
Emmett
Teller (BET) equation. The Barrett
Joyner

further purification (see Supporting Information). All syntheses were Halenda (BJH) method was applied to determine mesopore size
performed at ambient conditions. distributions.
Synthesis of ZIF-8 Nanocrystals in the Absence of a Fourier-transform infrared (FT-IR) spectra were recorded on a
Modulating Ligand. Typically, 734.4 mg (2.469 mmol) of Zn- BRUKER Tensor 27 spectrometer using the attenuated total reflection
(NO3)2 3 6H2O and 810.6 mg (9.874 mmol) of 2-methylimidazole (ATR) technique.
(Hmim) are each dissolved in 50 mL of MeOH. The latter clear solution Scanning electron micrographs (SEM) were taken in secondary
is poured into the former clear solution under stirring with a magnetic electron contrast at an acceleration voltage of 2 keV using a JEOL
bar. Stirring is stopped after combining the component solutions. After JSM-6700F field-emission instrument. Samples were dispersed on a carbon
24 h, the solid is separated from the milky colloidal dispersion by sample holder. Transmission electron microscopy (TEM) micrographs

2131 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141

Chemistry of Materials ARTICLE

Figure 1. XRD patterns (a) simulated from crystal structure, (b) of

65 nm-sized nanocrystals prepared in the absence of a modulating
ligand, (c) of microcrystals prepared in the presence of sodium formate,
(d) of microcrystals prepared in the presence of 1-methylimidazole, and
(e) of 18 nm-sized nanocrystals prepared in the presence of n-butylamine.

were taken using a JEOL JEM-2100F-UHR field-emission instrument at

an acceleration voltage of 200 kV. The same instrument was used for
taking selected area electron diffraction (SAED) patterns. Samples were
collected with a syringe from the colloidal solutions and dispersed on a
copper-supported carbon film.
Time-Resolved in Situ Static Light Scattering (SLS). SLS
measurements were performed at 25
C with a home-built multi-angle
goniometer described by Becker and Schmidt.28 Cylindrical silica glass
cuvettes with a diameter of 25 mm served as scattering cells. The
goniometer was equipped with a He-Ne laser operating at a wavelength
of 632.8 nm. It enabled simultaneous recording of the scattering
intensity at 2 times 19 scattering angles arranged in pairs symmetrically
on both sides of the beam in an angular regime of 25.84
e θ e 143.13
.
Recording of an angular dependent curve was completed after 2 ms.
1000 successive recordings were added to form one measurement
requiring 2 s in total. The time interval between the start of the
successive measurements was 10 s. The component solutions (see
syntheses) were cleaned by passing the solutions through 0.20 μm
filters to remove dust particles and to combine them into the scattering
cell. Addition of the second component solution determined the starting
point (t = 0) of the experiment. Scattering curves were processed as the
Rayleigh ratio ΔRθ at variable scattering angle θ.29 The scattering curves
could be approximated by means of a Guinier plot,30 which enabled the
extraction from each scattering curve the accumulated weight-averaged
molar mass, MW, and the square root of the z-averaged squared radius of
gyration, Rg (details are provided in the Supporting Information).

’ RESULTS AND DISCUSSION Figure 2. (a) SEM image of 65 nm-sized nanocrystals prepared in the
absence of a modulating ligand, (b) drawings of a rhombic dodecahe-
Our approach to the size-controlled synthesis of nano- and dron in two orientations, and (c) size distribution of 65 nm-sized
microscale ZIF materials is an extension of the method we have nanocrystals (the line represents a Gaussian fit).
recently developed for the production of ∼45 nm-sized ZIF-8
nanocrystals. It employs an excess of the bridging 2-methylimi-
dazole (Hmim) ligand with respect to the Zn(II) source, Zn- method has been employed in the synthesis of carboxylate-based
(NO3)2 3 6H2O, in methanolic solutions.17 This simple method MOFs with modulating ligands with only the same carboxylate
works already well at room temperature without the need of any functionality as the bridging ligands.11,12 We note that modulat-
activation, for example, by conventional or microwave heating. ing ligands do not only act as competitive ligands at the metal
We have now added to such ZIF-8 synthesis solutions various centers but also as bases on the deprotonation of the bridging
monodentate ligands with different chemical functionalities ligands. Thus, they may affect crystal nucleation and growth via
(carboxylate, N-heterocycle, alkylamine). This is in contrast to both coordination and deprotonation equilibria. This combined
related previous work, where the coordination modulation effect has not been investigated and utilized before.
2132 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141
Chemistry of Materials ARTICLE

In the presence of formate or 1-methylimidazole as a mod-

ulating ligand at x = 4, pure-phase ZIF-8 microcrystals are
obtained (see XRD patterns in Figure 1c,d). SEM images
(Figure 3a,b) reveal a rhombic dodecahedral shape and a
remarkable narrow size distribution. There is no significant
difference between the products despite the quite different
nature of the modulating ligands (anionic carboxylate vs neutral
N-heterocyclic molecule). It should be noted that the narrow size
distribution is only obtained when the synthesis solutions are not
stirred. Stirring results in very broad size distributions, which may
be due to secondary nucleation caused by turbulences.31
When n-butylamine is added as the modulating ligand at x = 4,
nearly instantaneous formation of a solid is observed upon
combining the component solutions, and pure-phase ZIF-8
nanocrystals are recovered after 24 h (see XRD pattern in
Figure 1e). An average size of 18 nm is estimated from the
broadening of the Bragg reflections. TEM images (Figure 4a)
show roughly spherical particles being <20 nm in size, including
some isolated particles and particles with sharp edges. It should
be noted that the small ZIF nanocrystals are very sensitive to the
high energy of the electron beam of a TEM.17 SAED patterns
(Figure 4b) confirm that the particles are crystalline ZIF-8. The
nanocrystals can be redispersed in MeOH. According to a SAXS
analysis, the redispersed particles indicate the primary 18 nm-
sized nanoparticles. Yet, they now form small secondary aggre-
gates with a maximum size of ∼88 nm and an average radius of
gyration of Rg = 27 nm. A SAXS pattern along with the
corresponding pair distance distribution function (PDDF) is
Figure 3. SEM images of (a) microcrystals prepared in the presence of provided in Figure S3 of the Supporting Information. The
sodium formate and (b) microcrystals prepared in the presence of dispersions remain optically clear for a few hours, before a fine
1-methylimidazole. precipitate begins to form.
Then, we varied the molar ratios of the components in the
Synthesis of the Nano- and Microcrystals. We first inves- system Zn/Hmim/n-butylamine/MeOH. Table 1 lists the com-
tigated synthesis solutions with the total molar ratio Zn/Hmim/ positions of the studied synthesis solutions and the average sizes
L/MeOH = 1:4:x:1000 (L = modulating ligand: sodium for- of the obtained particles. XRD demonstrates that all products are
mate,1-methylimidazole, n-butylamine). The products were iso- pure-phase ZIF-8 (see Figure S5 of the Supporting Information).
lated as described in the Experimental Section after 24 h of Particle sizes were determined from the powder samples by XRD
reaction at room temperature. The synthesis without addition of (Scherrer’s equation) and SAXS (PDDF), thus providing in-
a modulating ligand (x = 0) yields pure-phase ZIF-8 nanocrystals formation on both the size of the crystalline domains (XRD) and
(compare the XRD patterns in Figure 1a,b) with an average the primary particles (SAXS). Compared to TEM, which is
particle size of 64 nm, as estimated from the broadening of the usually used to analyze primary particle size, SAXS has two
Bragg reflections by applying Scherrer’s equation. SEM images advantages: (i) SAXS probes a much larger sample volume,
(Figure 2a) reveal that the well-defined nanocrystals have a which is therefore representative for the whole sample, and (ii)
rhombic dodecahedral shape (see Figure 2b for comparison) SAXS is nondestructive to ZIF nanocrystals. The PDDF, ob-
with 12 exposed {110} faces. This is a special crystal form of the tained by indirect Fourier transformation of a SAXS curve, is a
crystallographic point group 43m (in line with the cubic space representation of the intraparticle distance distribution, and in
group I43m of crystalline ZIF-8).6 We note that rhombic the case of homogeneous monodisperse spherical particles
dodecahedra, when viewed directly on a {110} face, appear with exhibits a maximum close to the radius of the spheres.26 This
a hexagonal cross-section (see drawing on the left-hand side of behavior is preserved for powders consisting of spherical parti-
Figure 2b). Thus, care has to be taken not to mix up rhombic cles, where a maximum or shoulder is still observed at about the
dodecahedral nanocrystals on SEM or TEM images with a distance that corresponds to the radius of the primary particles.32
hexagonal shape, which may have happened in recent work23 Representative SAXS patterns along with the corresponding
(further comments on this point are given in the Supporting PDDF curves are displayed in Figure 5 (the remaining SAXS
Information). A statistical evaluation of 500 particles results in an and PDDF data are provided in Figure S4 of the Supporting
average size of 65 ( 13 nm (see size distribution in Figure 2c). Information). As can be seen from Table 1, the size values
The nanocrystals are redispersible in MeOH, yielding disper- determined by XRD and SAXS are in good agreement for the
sions of nearly nonaggregated particles from which sedimenta- products, confirming that the primary particles are crystalline
tion occurs only very slowly. SAXS analysis results in an average ZIF-8. In addition, it can be also seen from Table 1 that by varying
size for the redispersed particles of 88 nm. A SAXS pattern taken the compositions of the synthesis solutions, the nanocrystal size
from a dispersion along with a fitted curve obtained with a model can be tuned between 9 and 65 nm. It should be noted that the
of polydisperse spherical particles is provided in Figure S2 of the volume of the smallest nanocrystals (diameter 9 nm) corresponds
Supporting Information. to only 78 unit cells of ZIF-8 (cubic cell constant a = 1.7012 nm).6b
2133 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141
Chemistry of Materials ARTICLE

Table 1. Compositions of Synthesis Solutions in the System

Zn/Hmim/n-Butylamine/Methanol and Average Diameters
of ZIF-8 Nanocrystals As Determined by XRD and SAXS

composition nanocrystal diameter

Zn/Hmim/n-BuNH2/MeOH dXRD, nm dSAXS, nm

(1) 1:4:4:1000 18 17
(2) 1:2:4:1000 45 39
(3) 1:4:2:1000 10 9
(4) 1:4:4:5000 24 20
(5) 1:2:4:5000 55 40
(6) 1:4:2:5000 9 9
(7) 1:4:4:500 16 16
(8) 1:2:4:500 43 42
(9) 1:4:2:500 10 8

Figure 4. (a) TEM image and (b) SAED pattern of 18 nm-size

nanocrystals prepared in the presence of n-butylamine.
Figure 5. (a) Representative SAXS patterns and (b) inner parts of the
corresponding PDDF curves of nanocrystals prepared in the presence of
By employing an excess of the bridging bidentate ligand n-butylamine. Compositions of the synthesis solutions are indicated.
together with different modulating ligands we have not only
been able to achieve size control from the smallest nanoscale media, as an approximation for methanolic solutions). The list of
(∼10 nm) to the microscale but, in addition, could gain new pKa values includes an estimate for the deprotonation constant of
insight into the functioning of modulating ligands during MOF Hmim being bound to a Zn(II) cation, as taken from the work by
crystallization. In previous work,12 only the effect of modulators Kimura et al.33 Hereby, the following explanations can be given
on the coordination equilibria during nucleation and growth had for the syntheses with constant total molar ratios Zn/Hmim/L/
been considered. In order to rationalize the outcome of our MeOH = 1:4:4:1000. An excess of Hmim in reaction (i) yields, at
present syntheses we have to take deprotonation equilibria into the beginning of ZIF-8 formation, a high concentration of
consideration in addition to coordination equilibria. Scheme 1 [Zn(Hmim)mLn] species with m > n, which via deprotonation
summarizes in a simplified manner the basic reactions of ZIF-8 (ii) and ligand exchange (iii) results in a high nucleation rate and
formation that have to be considered, namely, (i) complex consequently in a small size of the final crystals (∼65 nm in our
formation, (ii) deprotonation, and (iii) ligand exchange. Due case). Only the more basic modulating ligands with pKa > 10.3
to the labile nature of Zn(II) complexes all equilibria are likely to can deprotonate the [Zn(Hmim)mLn] species and thereby
be attained fast. Unfortunately, a complete set of the relevant accelerate ligand exchange reactions, resulting in an even higher
complex formation constants is not available in the literature. We nucleation rate and consequently in a smaller final crystal size
can, however, base a first qualitative discussion on the deproto- (∼18 nm in our case). On the other hand, the less basic
nation constants of the conjugate acids of the ligands, which are modulating ligands with pKa < 10.3 cannot effectively deproto-
provided in Scheme 1 in the form of pKa values (for aqueous nate [Zn(Hmim)mLn] species but compete in reaction (i) with
2134 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141
Chemistry of Materials ARTICLE

Scheme 1. Basic Reactions of ZIF-8 Formation (Top) and

pKa Values for the Conjugate Acids of Relevant Ligands
(Bottom)a

a
Charges of species are omitted in the reaction equations. L denotes all
ligands that may be present: Auxiliary modulating ligands as well as
NO3
, H2O, and MeOH.

Hmim, resulting in a low concentration of [Zn(Hmim)mLn]

species with m > n, in a low nucleation rate and consequently in a Figure 6. Weight-averaged particle mass, MW, and radius of gyration,
large final crystal size (∼1 μm in our case). The modulating Rg, as they evolve with time for (a) the formation of 65 nm-sized
ligands also affect crystal growth (see below), but their influence nanocrystals in the absence of a modulating ligand, (b) the formation of
on nucleation appears to be more important. For a more detailed microcrystals in the presence of sodium formate, and (c) the formation
understanding, knowledge of the complex formation and depro- of microcrystals in the presence of 1-methylimidazole. Error bars are
tonation constants in methanol would be required. The results of shown for Rg values.
the syntheses in the Zn/Hmim/n-butylamine/MeOH system at
variable molar ratios (Table 1) suggest that, for targeting inconsistency can be interpreted as follows. Single particles grow
particular small nanocrystals, an excess of the bridging ligand is very fast compared to the time resolution of our experiment, and
the dominating factor (e.g., compare compositions 1, 2, and 3 in the individual growth becomes slow when a size close to Rg =
Table 1). Increasing the content of the n-butylamine base 20 nm is attained. The increase in the averaged particle mass Mw
obviously does not lead to a smaller particle size, because then (of a bimodal system composed of small building units and
its function as a competitive ligand comes into play (e.g., particles) is predominantly due to a continuous and compara-
compare compositions 1 and 3 in Table 1). tively slow nucleation, which increases the number of particles
Time-Resolved in Situ SLS and ex Situ SEM Investigations with time. The averaged squared size Rg2 and the averaged mass
of Nanocrystal Formation. Particle formation could be success- values MW are based on different averaging procedures. Whereas
fully monitored by SLS with synthesis solutions of composition Rg2 corresponds to the third moment, MW is the second moment
Zn/Hmim/MeOH = 1:4:1000, yielding 65 nm-sized nanocryst- of the particle mass distribution. Thus, the average square of the
als. From each scattering pattern the accumulated weight-aver- size weights larger particles stronger than the average mass does.
aged molar mass, MW, and the square root of the z-averaged As a consequence, Rg2 of an ever increasing ensemble of particles
squared radius of gyration, Rg, were extracted for the growing with similar size values approaches its final average value much
particles. The evolution with time of these parameters is dis- earlier than MW does.
played in Figure 6a. Nanoparticles with a radius of gyration of Rg After approximately 800 s the particle size and mass increases.
≈ 20 nm (corresponding to Rsphere = 26 nm) become detectable This is due to a loose agglomeration of the primary nanoparticles
for the first time after 130 s. The size of these particles which essentially keep their size, as evidenced by SEM investiga-
corresponds to approximately 3/4 of the size of the final tions (see below). As can be seen from Figure 7a, this agglom-
nanocrystals recovered after 24 h from the synthesis solution, eration stage is indicated by a steep upturn in the log(Rg) vs
and the particles are likely to be crystalline. During the following log(MW) plot. The analysis during the agglomeration stage is
670 s the total particle mass increases continuously, while the impaired by an increasing turbidity of the dispersions causing
average particle size remains essentially constant. This apparent multiple scattering. This results in mass values which increasingly
2135 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141
Chemistry of Materials ARTICLE

depart toward lower values, which prevents further quantitative This clearly proves that the steep increase in size and mass at
interpretation of the upturn. Note that, after the beginning of 800 s observed by SLS is an agglomeration of primary nano-
the agglomeration stage, SLS can give no information anymore crystals with Rg ≈ 20 nm. Importantly, the size distribution
of whether nucleation still further continues or stops after (Figure 8d) has become very narrow after 1 h. The nanoparticles
some time. with an age of 1 h are crystalline ZIF-8 and have a rhombic
For complementary SEM investigations, small amounts of dodecahedral shape (see SEM image in Figure 8b and the
liquid were taken from the synthesis solutions after 240 s (before detailed XRD and TEM characterization of similar nanocrystals
the agglomeration stage) and after 1 h (during the agglomeration presented elsewhere)17 as the final 65 nm-sized nanocrystals
stage). After 240 s, spherical particles with an average size of 42 ( recovered after 24 h (see SEM image in Figure 2a). Note that the
11 nm and some amorphous material adhering to the particles final nanocrystals (size 65 ( 13 nm) have a considerably broader
are observed (see SEM image in Figure 8a). The particle size size distribution (Figure 2c) as the nanocrystals with an age of
distribution (Figure 8c) is comparatively broad. This supports 1 h.
the above interpretation of the SLS data that slow, persistent Narrowing (“focusing”) of size distribution was first experi-
nucleation and fast crystal growth take place in the early stages of mentally observed for semiconductor nanoparticles34 and is
nanoparticle formation. After 1 h, the nanoparticles still have reported here for the first time for a MOF. In this case, however,
nearly the same size, 38 ( 4 nm (see SEM image in Figure 8b). the narrowing of the size distribution of the primary nanocrystals
at intermediate stages (after ∼1 h) and, in particular, the very
narrow size distribution at the size focusing point are rather
surprising findings considering the observation by SLS that slow
nucleation occurs together with fast growth at least for the first
800 s of the crystallization process. This means that nucleation
and growth are not well separated from each other. Separation of
both steps is usually believed to be necessary to obtain particles of
low polydispersity (LaMer mechanism of burst nucleation and
subsequent growth).8,34 The “focusing” may be explained by the
termination of the fast particle growth at a size of Rg ≈ 20 nm,
due to colloidal stabilization of these intermediate particles by
surface-coordinated neutral, nondeprotonated Hmim ligands
that are present in excess and a corresponding build-up of
positive surface charges and favorable interactions with the polar
Figure 7. log(Rg) vs log(MW) plot for (a) the formation of 65 nm-sized solvent. The effect of positive surface charges is supported
nanocrystals in the absence of a modulating ligand, (b) the formation of experimentally by a zeta potential of ξ = þ55 mV for ZIF-8
microcrystals in the presence of sodium formate, and (c) the formation nanocrystals redispersed in MeOH.17 After supersaturation has
of microcrystals in the presence of 1-methylimidazole. sufficiently decreased and nucleation ceased, all nanoparticles

Figure 8. SEM images of intermediate particles during the formation of 65 nm-sized nanocrystals in the absence of a modulating ligand (a) after 240 s
and (b) 1 h. Corresponding particle size distributions (c) after 240 s and (d) 1 h (the lines represent Gaussian fits).

2136 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141

Chemistry of Materials ARTICLE

terminate at nearly the same size, before “defocusing” of size

distribution takes place in later stages due to growth of larger
particles at the expense of smaller ones (Ostwald ripening).8,34
The observations clearly demonstrate that time is an important
parameter to be considered when targeting monodisperse MOF
nanocrystals.
Time-Resolved in Situ SLS and ex Situ SEM Investigations
of Microcrystal Formation. SLS measurements were success-
fully performed with synthesis solutions of composition Zn/
Hmim/L/MeOH = 1:4:4:1000 (L = formate and 1-methylimi-
dazole), yielding microcrystals. The systems behave quite differ-
ently from those yielding nanocrystals. As can be seen from
Figure 6b, in the case of microcrystal formation with formate as a
modulating ligand, SLS enables the first significant size evalua-
tion of particles (Rg ≈ 50 nm) after approximately 350 s. The
particle size rapidly increases with time, as does the total particle
mass. Since the correlation of Rg with time is almost linear
between 350 and 480 s, an estimate for the starting point of
particle formation after mixing (induction time, tind) has been
obtained as the intercept of the backward extrapolated curve with
the abscissa (tind = 280 s). At 480 s, when the particles have
grown to Rg ≈ 110 nm, an abrupt change is clearly seen in the
correlation of Rg with time, indicating a change in the mechanism
of particle growth (see below). Beyond that point, growth in size
and mass continues and could be safely monitored by SLS up to
950 s (Rg ≈ 220 nm). Thereafter, sedimentation of larger
particles was indicated by a beginning decrease of the scattering Figure 9. Normalized scattering intensity, P(q), vs rescaled scattering
intensity. vector modulus, q 3 Rg, for two intermediates during the formation of
Significant insight into the mechanism of particle growth can microcrystals in the presence of (a) sodium formate and (b) 1-methy-
be obtained from the exponent R of the power law relation limidazole. Theoretical curves for monodisperse spheres, coils, and rods
between Rg and MW, Rg ∼ MWR. On the log(Rg) vs log(MW) plot are shown for comparison.
displayed in Figure 7b, the slope before and after 480 s
corresponds to R = 0.29 and R = 0.15, respectively. Interestingly, curves for noninteracting particles of different morphologies,
these values are close to the values predicted recently by some of namely, monodisperse spheres,35 coils,36 and rods.37 To better
us29 for spherical particles that grow by coalescence (R = 1/3) or illustrate the shape selectivity, the scattering vector modulus
according to a monomer addition model (R = 1/6). The same q has been rescaled to the size of the particles according to u =
exponents are expected for particles with a cubic symmetry. The q 3 Rg. In this dimensionless representation, curves from self-
SEM investigations presented below support the assumption that similar structures fall on top of each other, if the structures differ
intermediate particles in the course of ZIF-8 growth are isome- in size only. For both modulating ligands, the individual experi-
trical. Thus, we may infer from the SLS data that ZIF-8 mental curves overlay and are close to the curve predicted for
microcrystals grow, under the conditions studied here, by two monodisperse spheres, suggesting that the particles are compact
consecutive mechanisms: (i) particle aggregation comparable to and isometrical. The first oscillation which is clearly seen in the
coalescence and (ii) particle
monomer attachment. experimental curves is a strong indication of a narrow size
A similar pattern of growth is observed in the case of distribution, since increasing polydispersity results in blurring
microcrystal formation with 1-methylimidazole as a modulating of such oscillations.13
ligand (see Figure 6c). The first significant particle size values (Rg Complementary SEM and TEM investigations were carried
≈ 50 nm) are detected by SLS after approximately 500 s (tind = out for the synthesis with 1-methylimidazole as a modulating
350 s). The particles continuously grow in size and mass until ligand. We were able to observe particles that are clear inter-
1200 s, corresponding to the last point shown in Figure 4c (Rg ≈ mediates in the course of ZIF-8 formation after approximately
180 nm). However, a careful inspection of the log(Rg) vs 600 s, which apparently belong to the particle
monomer growth
log(MW) plot displayed in Figure 7c reveals a bent curve, which process. The first particles seen on the SEM images have the
only enables the estimate of a final slope after 800 s correspond- shape of a cube ({100} crystal form) with rounded edges and a
ing to R = 0.19. Thus, we may infer form the SLS data that size of approximately 100 nm (see Figure 10a). After 2000 s, the
coalescence and monomer attachment simultaneously take place particles still have the same shape but have grown to a size of
during early stages of growth, while monomer attachment 400 nm (see TEM image in Figure 10b), revealing that the
dominates in later stages. relative growth rates of different faces are slowest along the Æ100æ
For larger particles (Rg > 100 nm), information on their shape directions. The sharp spots seen on the ED pattern in Figure 10c,
can be obtained from the SLS curves. Scattering patterns with the which is a view down a Æ100æ zone axis with fourfold rotational
intensity I(q) in its normalized form P(q) = I(q)/I(q = 0) are symmetry, demonstrate that the cube-shaped particles are well
displayed for selected intermediates of microcrystal formation in crystalline ZIF-8. On the TEM image in Figure 10b, a thin rim
the presence of formate and 1-methylimidazole in Figure 9a,b, around the nanocrystal with different electron density contrast is
respectively. The data are compared with theoretically predicted identified, which may, for example, indicate the presence of an

2137 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141

Chemistry of Materials ARTICLE

Figure 10. Intermediate particle during the formation of microcrystals in the presence of 1-methylimidazole: (a) SEM image of a particle after 600 s, (b)
TEM image of a particle after 2000 s, (c) ED pattern of the particle after 2000 s, (d) SEM image of particles after 3000 s, (e) SEM image of particles after
5000 s, and (f) SEM image of particles after 24 h.

amorphous growth layer. The size distribution of the nanocryst- The formation of microcrystals in the presence of formate
als revealed by the SEM image taken after 3000 s (Figure 10d) follows the same pattern of shape evolution from cubes with
appears to be rather narrow, in close agreement with the SLS data truncated edges to final rhombic dodecahedra (see the SEM
(oscillation in the scattering curve after 1145 s, Figure 9b). images provided in Figure S7 of the Supporting Information). It
Beyond 4000 s, the nanocrystals change their shape due to a should be noted here that SEM and TEM investigations on
change in the relative growth rates, which are now slowest along formate-containing systems are hampered much more seriously
Æ110æ directions. On the SEM image taken after 5000 s, 500 nm- by species other than real ZIF-8 intermediates (e.g., sodium
sized cubes with truncated edges that expose 6 {100} and 12 formate that deposits upon evaporating the dispersions prior to
{110} faces are seen (Figure 10e). The micrometer-sized crystals the SEM/TEM investigations) as is the case with 1-methylimida-
recovered after 24 h have a rhombic dodecahedral shape with zole-containing systems. Due to uncertainty in the identity of
exposed {110} faces (see SEM image in Figure 10f). species observed on SEM images taken during early stages of
2138 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141
Chemistry of Materials ARTICLE

ZIF-8 microcrystal formation (an example is provided in Figure S8

of the Supporting Information), we cannot up to now verify
the coalescence mechanism suggested by our SLS data. Before
unambiguous information can be obtained by SEM/TEM, specific
procedures for separating impurities have to be developed first.
From a coordination chemistry point of view, 1-methylimida-
zole is expected to be a stronger coordinating ligand to Zn(II)
than formate. In agreement with this, addition of 1-methylimi-
dazole results in a lower nucleation rate, as judged from the
longer induction time (350 vs 280 s), and a lower growth rate, as
indicated by the slower increase of the Rg and MW values with
time. In addition, the change of shape from cubes to rhombic
dodecahedra occurs at later times in the presence of 1-methyli-
Figure 11. (a) Nitrogen sorption isotherms at
196
C for the 18 nm-
midazole than of formate.
sized nanocrystals prepared in the presence of n-butylamine. Black
A comparison of the SLS data of the modulated and non- squares: adsorption branch; open dots: desorption branch.
modulated syntheses points to an interesting difference between
both types of crystallization processes. During the first 800 s, the Figure S9 of the Supporting Information) as well as variable-
apparent weight-averaged particle mass MW recorded in the temperature XRD patterns (see Figure S10 of the Supporting
absence of a modulating ligand exeeds considerably the MW Information). Since nearly no mass loss is seen on the TG curves
values measured in the presence of either modulating ligand. As before the onset of the exothermic decomposition of the organic
can be seen from Figures 6 and 7, the difference amounts to more bridging ligand, it is clear that solvent (MeOH) and modulator
than a magnitude. This is particularly intriguing as the particle molecules have already left the intracrystallite cavities during
size values are much smaller if modulating ligands are absent. workup (drying) after synthesis. Hence, no further solvent
However, in order to better judge this feature, the physical exchange and/or heat treatment procedures are necessary for
meaning and origin of the apparent mass values have to be activation of the materials.
briefly outlined. The values are directly taken from the intercepts On the N2 sorption isotherms of the 18 nm-sized nanocrystals
of the scattering curves (see Supporting Information) and thus (Figure 11), a first steep step at low relative pressure (p/p0 <
are proportional to the solid mass concentration times the 0.08) is seen, revealing that the nanocrystals are microporous.
apparent weight averaged solid mass. Now, this product can be The specific surface area estimated by the BET method amounts
looked at from two different perspectives:39 (i) since the solid to SBET = 1617 m2 3 g
1. The value compares well with those
concentration in g 3 L
1 is a constant determined by weight and is reported recently for ZIF-8 macrocrystals (SBET = 1630
not changing during the growth process, the intercept can be m2 3 g
1)6a and ∼30 nm-sized ZIF-8 nanocrystals (SBET =
considered to be directly proportional to the weight averaged 1696 m2 3 g
1).21 This confirms that the 18 nm-sized nanocryst-
mass including all species, that is, monomers and growing als are well crystalline, as already evidenced by XRD and SAED.
particles, and (ii) the monomers can to a good approximation The isotherms exhibit a second step at high relative pressure (p/
be considered as species with a negligible scattering contribution, p0 > 0.7) with an adsorption
desorption hysteresis loop of type
which means that the intercept increases via the concentration of H2.38 The step originates from interparticle mesopores, demon-
the generated and growing particles and/or via the growing strating the dual micro- and mesoporosity of the ZIF-8 nano-
particle mass (now excluding the monomer fraction). If we adopt cystal powders. The mesopore size distribution estimated by the
the latter perspective, the discrepancy in apparent mass values for BJH method is centered at ∼8 nm. Small crystal size and dual
the two growth processes is immediately understandable: In the porosity are of interest for the development of advanced adsor-
absence of modulating ligands, the increase of weight-averaged bents and catalysts with fast mass transport kinetics. Indeed, a
apparent mass values is due to a persisting nucleation during at significant increase of adsorption rates of nanocrystals compared
least the first 800 s. In the presence of either modulating ligand, to microcrystals was recently demonstrated for ZIF-8 (∼45 nm-
we observe a simultaneous growth of both the averaged size and sized nanocrystals)18 and the flexible MOF [Zn(ip)(bpy)]
the apparent mass values. Despite the much larger size values (CID-1, ip = isophthalate, byp = 4,40 -bipyridyl).10f
achieved in this case, the respective mass values are considerably
lower because the number of particles and with it the mass
concentration of particles are much lower in the latter case. This ’ CONCLUSION
is due to the fact that the extent of nucleation has to be much We have reported a novel synthetic approach to ZIF-8 nano-
lower and perhaps even ceases entirely before 800 s if modulating and microscale materials in which size-control between ∼10 nm
ligands are present. Thus, the SLS data provide clear direct and 1 μm is achieved by employing an excess of the bridging
evidence from in situ experiments that a function of modulating bidentate ligand and three simple auxiliary modulating ligands
ligands of comparatively low bacisity (pKa < 10.3 in our case) is that act as competitive ligands in coordination equilibria and
to slow down the nucleation rate, as proposed above and bases in deprotonation equilibria during nucleation and growth.
previously by Kitagawa and co-workers12b on the basis of Time-resolved in situ SLS and ex situ SEM/TEM investigations
systematic synthetic work. have provided insight into the functioning of modulating ligands.
Investigation of Thermal Stability and Porosity. All pre- Furthermore, the experiments have revealed that nanocrystal
pared ZIF-8 materials exhibit good thermal stability in air, up to formation is characterized by continuous, comparatively slow
250
C in the case of the smaller 18 nm-sized nanocrystals, and nucleation and fast crystal growth. A focusing of the nanocrystal
up to 300
C in the case of the larger nanocrystals and size distribution occurs with increasing time, resulting in a very
microcrystals. This is demonstrated by TG/DTA curves (see narrow size distribution at the size focusing point, and is followed
2139 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141
Chemistry of Materials ARTICLE

by defocusing of the size distribution at later stages of the growth. R.; Noro, S. Angew. Chem., Int. Ed. 2004, 43, 2334. (c) Ferey, G. Chem.
The growth of microcrystals takes place by a particle
monomer Soc. Rev. 2008, 37, 191.
attachment mechanism, and a change of crystal shape from cubes (3) (a) Banerjee, R.; Phan, A.; Wang, B.; Knobler, C.; Furukawa, H.;
to rhombic dodecahedra occurs in later stages of growth. O’Keeffe, M.; Yaghi, O. M. Science 2008, 319, 939. (b) Banerjee, R.;
Furthermore, indications for the occurrence of a coalescence Furukawa, H.; Britt, D.; Knobler, C.; O’Keeffe, M.; Yaghi, O. M. J. Am.
mechanism during early stages of the growth have been obtained. Chem. Soc. 2009, 131, 3875. (c) Li, K.; Olson, D. H.; Seidel, J.; Emge,
T. J.; Gong, H.; Zeng, H.; Li, J. J. Am. Chem. Soc. 2009, 131, 10368.
The prepared ZIF-8 materials are easily activated (due to the use (d) Morris, W.; Leung, B.; Furukawa, H.; Yaghi, O. K.; He, N.; Hayashi,
of MeOH as the solvent) and exhibit good thermal stability in air H.; Houndonougbo, Y.; Asta, M.; Laird, B. B.; Yaghi, O. M. J. Am. Chem.
as well as large surface areas, which are comparable to those of Soc. 2010, 132, 11006. (e) Chizallet, C.; Lazare, S.; Bazar-Bachi, D.;
large macrocrystals. Nanocrystal powders exhibit dual micro- and Bonnier, F.; Lecocq, V.; Soyer, E.; Quoineaud, A.-A.; Bats, N. J. Am.
mesoporosity. Chem. Soc. 2010, 132, 12365.
The novel synthetic strategy and insight into the crystal- (4) (a) Bux, H.; Liang, F.; Li, Y.; Cravillon, J.; Wiebcke, M.; Caro, J. J.
lization processes may help to put size- and shape-controlled Am. Chem. Soc. 2009, 131, 16000. (b) Li, Y.; Liang, F.; Bux, H.; Feldhoff,
syntheses of nano- and microscale bulk ZIF materials and A.; Yang, W.-S.; Caro, J. Angew. Chem., Int. Ed. 2010, 49, 548.
supported ZIF membranes and films on a more rational basis. (c) McCarthy, M. C.; Varela-Guerrero, V.; Barnett, G. V.; Jeong, H.-
Indeed, the recent successful synthesis of ZIF-7 nanocrystals by K. Langmuir 2010, 26, 14636. (d) Huang, A.; Bux, H.; Steinbach, F.;
employing an excess of the bridging benzimidazolate ligand4b Caro, J. Angew. Chem., Int. Ed. 2010, 49, 4958. (e) Lu, G.; Hupp, J. T. J.
Am. Chem. Soc. 2010, 132, 7832. (f) Aguado, S.; Canivet, J.; Farrusseng,
indicates that the synthetic principles reported here may be D. Chem. Commun. 2010, 46, 7999.
transferable to other ZIF and possibly even to other MOF (5) (a) Tian, Y.-Q.; Cai, C.-X.; Ji, Y.; You, X.-Z.; Peng, S.-M.; Lee,
systems. G.-H. Angew. Chem., Int. Ed. 2002, 41, 1384. (b) Wu, T.; Zhang, J.; Zhou,
While the manuscript was under reviewing two papers C.; Wang, L.; Bu, X.; Feng, P. J. Am. Chem. Soc. 2009, 131, 6111.
appeared in the literature which are related to the work (c) Tian, Y.-Q.; Yao, S.-Y.; Gu, D.; Cui, K.-H.; Guo, D.-W.; Zhang, G.;
presented herein: Pan et al.40 report the rapid synthesis of Chen, Z.-X.; Zhao, D.-Y. Chem.—Eur. J. 2010, 16, 1137.
ZIF-8 nanocrystals being between ∼50 and 85 nm in size. Their (6) (a) Park, K. S.; Ni, Z.; C^ote, A. P.; Choi, J. Y.; Huang, R.;
protocols differ from ours in the use of water (instead of Uribe-Romo, F. J.; Chae, H. K.; O’Keeffe, M.; Yaghi, O. M. Proc. Natl.
MeOH) as the solvent and a much larger excess of the bridging Acad. Sci. U.S.A. 2006, 103, 10186. (b) Huang, X.-C.; Lin, Y.-Y.; Zhang,
Hmim ligand (Hmim/Zn g 70:1). Venna et al.41 report an J.-P.; Chen, X.-M. Angew. Chem., Int. Ed. 2006, 45, 1557.
ex situ XRD and TEM study of the formation of ZIF-8 (7) Tosheva, L.; Valtshev, V. P. Chem. Mater. 2005, 17, 2494.
nanocrystals at room temperature in the presence of an excess (8) Park, J.; Joo, J.; Kwon, S. G.; Jang, Y.; Hyeon, T. Angew. Chem.,
Int. Ed. 2007, 46, 4630.
of the bridging Hmim ligand (Hmim/Zn = 8:1). They con-
(9) (a) Spokoyny, A. M.; Kim, D.; Sumrein, A.; Mirkin, C. A. Chem.
cluded that the crystallization process is nucleation-controlled.
Soc. Rev. 2009, 38, 1218. (b) Lin, W.; Rieter, W. J.; Taylor, K. M. L.
This is in general agreement with our findings by in situ SLS and Angew. Chem., Int. Ed. 2009, 48, 650. (c) McKinlay, A. C.; Morris, R. E.;
ex situ SEM. Horcajada, P.; Ferey, G.; Gref, R.; Couvreur, P.; Serre, C. Angew. Chem.,
Int. Ed. 2010, 49, 6260.
’ ASSOCIATED CONTENT (10) (a) Huang, L.; Wang, H.; Chen, J.; Wang, Z.; Sun, J.; Zhao, D.;
Yan, Y. Microporous Mesoporous Mater. 2003, 58, 105. (b) Qiu, L.-G.; Li,
bS Supporting Information. Details of syntheses, SLS mea- Z.-Q.; Wu, Y.; Wang, W.; Xu, T.; Jiang, X. Chem. Commun. 2008, 3642.
surements, comments on nanocrystal shape, SAXS/PDDF, (c) Taylor-Pashow, K. M. L.; Rocca, J. D.; Xie, Z.; Tran, S.; Lin, W. J. Am.
SEM, XRD, TG/DTA, variable-temperature XRD, nitrogen Chem. Soc. 2009, 131, 14261. (d) Horcajada, P.; Chalati, T.; Serre, C.;
physisorption, and FT-IR data (PDF).This material is available Gillet, B.; Sebrie, C.; Baati, T.; Eubank, J. F.; Heurtaux, D.; Clayette, P.;
free of charge via the Internet at http://pubs.acs.org. Kreuz, C.; Chang, J.-S.; Hwang, Y. K.; Marsaud, V.; Bories, P.-N.;
Cynober, L.; Gil, S.; Ferey, G.; Couvreur, P.; Gref, R. Nat. Mater. 2010,
9, 172. (e) Guillerm, V.; Gross, S.; Serre, C.; Devic, T.; Bauer, M.; Ferey,
’ AUTHOR INFORMATION G. Chem. Commun. 2010, 46, 767. (f) Tanaka, D.; Henke, A.; Albrecht,
K.; Moeller, M.; Nakagawa, K.; Kitagawa, S.; Groll, J. Nat. Chem. 2010,
Corresponding Author 2, 410.
*E-mail: [email protected] (M.W.), klaus. (11) (a) Hermes, S.; Witte, T.; Hikov, T.; Zacher, D.; Bahnm€uller,
[email protected] (K.H.). S.; Langstein, G.; Huber, K.; Fischer, R. A. J. Am. Chem. Soc. 2007,
129, 5324. (b) Horcajada, P.; Serre, C.; Grosso, D.; Boissiere, C.;
Perruchas, S.; Sanchez, C.; Ferey, G. Adv. Mater. 2009, 21, 1931.
’ ACKNOWLEDGMENT (12) (a) Tsuruoka, T.; Furukawa, S.; Takashima, Y.; Yoshida, K.;
The authors thank G. Platz for performing the physisorption Isoda, S.; Kitagawa, S. Angew. Chem., Int. Ed. 2009, 48, 4739. (b) Diring,
measurements. The work was supported by the Deutsche For- S.; Furukawa, S.; Takashima, Y.; Tsuruoka, T.; Kitagawa, S. Chem. Mater.
schungsgemeinschaft (DFG) within the frame of the Priority 2010, 22, 4531.
Progam 1362 (Porous Metal
Organic Frameworks) organized (13) Zacher, D.; Lin, J.; Huber, K.; Fischer, R. A. Chem. Commun.
2009, 1031.
by S. Kaskel. The authors gratefully acknowledge the funding by
(14) Shekhah, O.; Wang, H.; Zacher, D.; Fischer, R. A.; W€oll, C.
the DFG (Grants WI1156/2-1 and HU807/12-1). Angew. Chem., Int. Ed. 2009, 48, 5038.
(15) John, N. S.; Scherb, C.; Sh€o^aee, M.; Anderson, M. W.; Attfield,
’ REFERENCES M. P.; Bein, T. Chem. Commun. 2009, 6294.
(1) Phan, A.; Doonan, C. J.; Uribe-Romo, F. J.; Knobler, C. B.; (16) Millange, F.; Medina, M. I.; Guillou, N.; Ferey, G.; Golden,
O’Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2010, 43, 58. K. M.; Walton, R. I. Angew. Chem., Int. Ed. 2010, 49, 763.
(2) (a) Yaghi, O. M.; O’Keeffe, M.; Ockwig, N. W.; Chae, H. K.; (17) Cravillon, J.; M€unzer, S.; Lohmeier, S.-J.; Feldhoff, A.; Huber,
Eddaoudi, M.; Kim, J. Nature 2003, 423, 705. (b) Kitagawa, S.; Kitaura, K.; Wiebcke, M. Chem. Mater. 2009, 21, 1410.

2140 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141

Chemistry of Materials ARTICLE

(18) Ostermann, R.; Cravillon, J.; Weidmann, C.; Wiebcke, M.;

Smarsly, B. M. Chem. Commun. 2010, 47, 442.
(19) Venna, S. R.; Carreon, M. A. J. Am. Chem. Soc. 2010, 132, 76.
(20) Bux, H.; Feldhoff, A.; Cravillon, J.; Wiebcke, M.; Li, Y.; Caro, J.
Chem. Mater. 2011, accepted.
(21) Demessence, A.; Boissiere, C.; Grosso, D.; Horcajada, P.; Serre,
C.; Ferey, G.; Soler-Illia, G. J. A. A.; Sanchez, C. J. Mater. Chem. 2010,
20, 7676.
(22) Chang, N.; Gu, Z.-Y.; Yan, X.-P. J. Am. Chem. Soc. 2010,
132, 13645.
(23) Nune, S. K.; Thallapally, P. K.; Dobnalkova, A.; Wang, C.; Liu,
J.; Exarbos, G. J. Chem. Commun. 2010, 46, 4878.
(24) Li, Y.; Bux, H.; Feldhoff, A.; Li, G.-L.; Yang, W.-S.; Caro, J. Adv.
Mater. 2010, 22, 3322.
(25) Kline, S. R. J. Appl. Crystallogr. 2006, 39, 895.
(26) Glatter, O.; Kratky, O. Small Angle X-ray Scattering; Academic
Press: London, 1982.
(27) Konarev, P. V.; Petoukhov, M. V.; Volkov, V. V.; Svergun, D. I.
J. Appl. Crystallogr. 2006, 39, 277.
(28) Becker, A.; Schmidt, M. Macromol. Chem. Macromol. Symp.
1991, 50, 249.
(29) Liu, J.; Rieger, J.; Huber, K. Langmuir 2008, 24, 8262.
(30) Guinier, A.; Fournet, G. Small-Angle Scattering of X-Rays; Wiley:
New York, 1955.
(31) Mullin, J. W. Crystallization, 4th ed.; Butterworth-Heinemann:
Oxford, 2001.
(32) Rath, T.; Kunert, B.; Resel, R.; Fritz-Popovski, G.; Saf, R.;
Trimmel, G. Inorg. Chem. 2008, 47, 3014.
(33) Kimura, E.; Kurogi, Y.; Shionoya, M.; Shiro, M. Inorg. Chem.
1991, 30, 4524.
(34) Peng, X.; Wickham, J.; Alivisatos, A. P. J. Am. Chem. Soc. 1998,
120, 5343.
(35) Rayleigh, L. Proc. R. Soc. London, Ser. A 1914, 90, 219.
(36) Debye, P. J. Phys. Colloid Chem. 1947, 51, 18.
(37) Neugebauer, T. Ann. Phys. 1943, 434, 509.
(38) Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.;
Pierotti, R. A.; Rouquerol, J.; Siemieniewska, T. Pure Appl. Chem. 1985,
57, 603.
(39) Liu, J.; Pancera, S.; Boyko, V.; Shukla, A.; Narayanan, T.;
Huber, K. Langmuir 2010, 26, 17405.
(40) Pan, Y.; Liu, Y.; Zeng, G.; Zhao, L.; Lai, Z. Chem. Commun.
2011, 47, 2071.
(41) Venna, S. R.; Jasinski, J. B.; Carreon, M. A. J. Am. Chem. Soc.
2010, 132, 18030.

2141 dx.doi.org/10.1021/cm103571y |Chem. Mater. 2011, 23, 2130–2141