Self-assembled aromatic molecules as efficient organic structure directing agents to synthesize the silicoaluminophosphate SAPO-42 with isolated Si species

Article abstract of DOI:10.1021/acs.chemmater.5b00337

The use of self-assembled aromatic molecules through π-π interactions has allowed the preparation of the silicoaluminophosphate (SAPO) form of the LTA, SAPO-42, with controlled Si content as isolated Si sites in the framework, and high thermal stability. Different SAPO-42 zeotypes can be synthesized with different acidity, morphology, and crystal size, depending on the selected quinolinium derived aromatic molecule as OSDA and the amount of fluoride content in the synthesis gels.

Full text of DOI:10.1021/acs.chemmater.5b00337

Article
pubs.acs.org/cm
Self-Assembled Aromatic Molecules as Efficient Organic Structure
Directing Agents to Synthesize the Silicoaluminophosphate SAPO-42
with Isolated Si Species
́
Raquel Martínez-Franco, Angel Cantín, Alejandro Vidal-Moya, Manuel Moliner,* and Avelino Corma*
Instituto de Tecnología Química (UPV-CSIC), Universidad Politec
Valencia, 46022, Spain
́
nica de Valencia, Consejo Superior de Investigaciones Científicas,
ABSTRACT: The use of self-assembled aromatic molecules through π−π
interactions has allowed the preparation of the silicoaluminophosphate (SAPO)
form of the LTA, SAPO-42, with controlled Si content as isolated Si sites in the
framework, and high thermal stability. Different SAPO-42 zeotypes can be
synthesized with different acidity, morphology, and crystal size, depending on the
selected quinolinium derived aromatic molecule as OSDA and the amount of
fluoride content in the synthesis gels.
It appears that the synthesis of ITQ-29 zeolite is a very
1. INTRODUCTION
interesting example of the correlation between the shape and
LTA is a crystalline molecular sieve, whose structure is formed
by large spherical cavities (∼1.14 nm) interconnected by three-
dimensional small pores (∼0.4 nm).1 Since its discovery in
1961 by Barrer and Denny,2 low silica LTA zeolite (Si/Al ratios
lower than 2) has become an important industrial material with
application as a drying agent or cation exchanger.3 Researchers
have endeavored to rationalize the synthesis of LTA in an
attempt to expand the number of applications of this zeolite to
catalysis or gas separations. One of the big challenges was to
increase the Si/Al ratio of the LTA zeolite to improve acidity,
hydrothermal stability, and hydrophobicity. In this sense,
Corma et al. proposed the use of bulky organic structure
directing agents (OSDAs) formed by self-assembly of two
aromatic organic molecules (MTPQ, see Figure 1) through
π−π interactions, to synthesize LTA in a wide range of
framework Si/Al ratios, even in the pure silica form (ITQ-29).4
size of the OSDA and the framework cavity.5
In addition to the silicate and silicoaluminate forms of LTA,
this material can also be prepared under other chemical
compositions, such as aluminophosphate (AlPO-42)6 and
silicoalumiphosphate (SAPO-42).7 Silicoaluminophosphates
(SAPOs) can be applied as heterogeneous acid catalysts since
the insertion of isolated framework Si atoms by selective
isomorphic substitution of Si by P atoms in their crystalline
structures generates medium-strong Brønsted acid sites.8
Unfortunately, the ideal Si distribution as isolated Si species
is not always accomplished, and large Si domains in the
framework, named “Si islands”, can also be present in SAPOs.9
In general, SAPOs with “Si islands” show lower acidity and
hydrothermal stability than similar SAPOs with better Si
distributions. For the particular case of SAPO-42 with LTA
structure, very high contents of Si (SiO2 between 15 and 35 wt
%) were always required for its synthesis.10 This results in
materials with large domains of “Si islands” and limited acidity
and hydrothermal stability.
Very recently, Moscoso11 and Schmidt et al.12 have
succeeded in the synthesis of SAPO-42 with low silicon
content (SiO2 less than 10 wt %). The OSDAs used for the
preparation of these low-silica SAPO-42 materials are a
combination of tetraethylammonium fluoride and diethanol-
amine11 and a bulky and complex triquad,12 respectively. In
both cases, characterization describing silicon distribution and
Figure 1. Aromatic organic molecules used as OSDAs for the synthesis
of SAPO-42 materials: (a) 2,2-dimethyl-2,3-dihydro-1H-benzo[de]-
isoquinoline-2-ium (DDBQ) and (b) 4-methyl-2,3,6,7-tetrahydro-
1H,5H-pyrido [3.2.1-ij] quinolinium (MTPQ).
Received: January 27, 2015
Revised: March 25, 2015
Published: March 27, 2015
© 2015 American Chemical Society
2981
DOI: 10.1021/acs.chemmater.5b00337
Chem. Mater. 2015, 27, 2981−2989

Chemistry of Materials
Article
filtered, and dried, providing 18.3 g (69%) of the desired ammonium
salt as a white powder.
acidity of these SAPO-42 has not been reported. Moreover,
SAPO-42 materials prepared using the bulky triquad OSDA
were not stable after calcination treatments.12
The spectral data. 1H NMR (DMSO-d6, 300.0 MHz): δH 8.0 (d, J =
9 Hz, 2H, Ar−H), 7.6 (t, J = 9 Hz, 2H, Ar−H), 7.5 (d, J = 9 Hz, 2H,
Ar−H), 5.0 (s, 4H, 2×CH2), and 3.2 (s, 6H, 2×NCH3). 13C NMR
(DMSO-d6, 75.5 MHz): δC 132.1, 128.6, 126.6, 124.9, 124.8, 124.3
(Ar−C), 63.4 (2×CH2), and 51.1 (2×NCH3). Anal. Calcd for
C14H16IN: C, 51.71; H, 4.96; N, 4.31. Found: C, 52.2; H, 5.17; N,
4.35.
In past years, we have worked on the synthesis of small pore
SAPOs with large cavities and controlled silicon distributions as
catalysts for the selective catalytic reduction (SCR) of NOx.13
Particularly, we have shown that the use of bulky self-assembled
aromatic molecules as OSDAs (MTPQ, see Figure 1) has
allowed the efficient synthesis of the small pore SAPO STA-6
with an excellent Si distribution, so the presence of large Si-rich
domains was avoided.14 Having in mind that the pure silica
LTA zeolite was achieved using the same bulky self-assembled
aromatic molecule (MTPQ, see Figure 1), it could be
envisioned that related OSDAs would show adequate proper-
ties to allow the synthesis of the small pore SAPO-42 zeotype
with proper Si distributions.
Herein, we describe the synthesis of SAPO-42 using two
aromatic organic molecules (MTPQ and DDBQ, see Figure 1),
which are able to form soluble dimers in the synthesis media
through π−π interactions. Self-assembled paired MTPQ
molecules allow the synthesis of SAPO-42 only when fluoride
anions are present in the synthesis media, while self-assembled
paired DDQB molecules are able to direct the crystallization of
SAPO-42 only under fluoride-free conditions. We will also
show that the presence of fluoride anions in the synthesized
SAPO-42 zeotypes with LTA structure clearly influences the
framework silicon distribution and, consequently, their acid
properties.
To obtain the hydroxide form of the organic cation, 9 g of the
iodide salt was dissolved in water, and thus, 30 g of resin Dower SBR
was added. The mixture was maintained overnight under stirring, and
after filtration, the hydroxide form of the organic cation was obtained.
4-Methyl-2,3,6,7-tetrahydro-1H,5H-pyrido[3.2.1-ij]quinolinium
(MTPQ). The synthesis of MTPQ has been described previously.4 To
prepare the corresponding hydroxide form of the organic cation, 8.52 g
of the iodide salt was dissolved in water, and 27 g of resin Dower SBR
was added. The mixture was maintained under stirring overnight. The
solution was filtered, and the hydroxide form of the organic cation was
obtained.
2.2. Synthesis of the Silicoaluminophosphate Materials. In a
typical zeotype synthesis procedure, the required amount of
orthophosphoric acid (85 wt %, Aldrich) was added to the aqueous
solution of the hydroxide form of the OSDA. Then, the alumina
source (75 wt %, Condea) was introduced, keeping the gel under
stirring for 5 min. A silica source (Ludox AS40, 40 wt %, Aldrich) was
also introduced into the synthesis gel, leaving the mixture under
stirring for 20 min. If required, HF was finally added to the gel, and the
resultant mixture was stirred for 20 min. The gel was transferred to a
Teflon-lined stainless steel autoclave with a free volume of 3 mL and
heated at the appropriate temperature under static conditions for 5
days. Crystalline products were filtered and washed with abundant
water and dried at 100 °C overnight. The samples were calcined in a
tube furnace following the next temperature program: the samples
were heated at 600 °C using a 1 °C/min ramp under flowing N2. The
temperature was held for 8 h under flowing air, and finally, the samples
were cooled to room temperature under flowing N2.
2. EXPERIMENTAL SECTION
2.1. Synthesis of the Organic Structure Directings Agents
(OSDAs). 2,2-Dimethyl-2,3-dihydro-1H-benzo[de]isoquinoline-2-
ium (DDBQ). A total of 21.7 g (109.6 mmol) of commercial 1,8-
naphthalic anhydride was heated under reflux in an aqueous solution
of N-methylamine (40 wt %) for 72 h. The mixture was cooled at
room temperature, and the white solid obtained was filtered and dried
under a vacuum to give N-methyl-1,8-naphthalimide (23.1 g, 100%).
The spectral data were fully coincident with those described in the
literature.15
2.3. Characterization. Powder X-ray diffraction (PXRD)
measurements were performed with a multisample Philips X’Pert
diffractometer equipped with a graphite monochromator, operating at
45 kV and 40 mA, and usig Cu Kα radiation (λ = 0.1542 nm).
The chemical analyses were carried out in a Varian 715-ES ICP-
Optical Emission spectrometer, after solid dissolution in HNO3/HCl/
HF aqueous solution. The organic content of the as-made materials
was determined by elemental analysis performed with a SCHN
FISONS elemental analyzer. Thermogravimetrical analysis was
performed using a Mettler Toledo thermo-balance.
1H NMR (CDCl3, 300.0 MHz): δH 8.5 (d, J = 9 Hz, 2H, Ar−H), 8.1
(d, J = 9 Hz, 2H, Ar−H), 7.7 (t, J = 9 Hz, 2H, Ar−H), and 3.5 (s, 6H,
2xNCH3). 13C NMR (CDCl3, 75.5 MHz): δC 164.5 (2×CO), 133.9,
131.6, 131.2, 126.9, 122.6 (Ar−C), and 27.0 (NCH3).
Then, 23.4 g (111.0 mmol) of N-methyl-1,8-naphthalimide was
added in small portions to a stirred solution of 14.8 g (388.0 mmol) of
LiAlH4 and 100.0 mL of AlCl3 THF complex (0.5 M in THF; 50.0
mmol) in 600 mL of anhydrous THF, cooled in an ice bath, and
maintained under N2. The mixture was stirred for 30 min at room
temperature before being refluxed for 7 h, and then, it was stirred at
room temperature overnight. The temperature was decreased with an
ice bath, and the excess of LiAlH4 was quenched by the slow addition
of 10 mL of H2O. After 30 min, the formed salts were filtered and
washed with THF and ethyl acetate, providing 14.9 g (73%) of the
desired amine as yellow oil after evaporation under a vacuum. The
spectral data were fully coincident with those described in the
literature.15
The morphology of the samples was studied by scanning electron
microscopy (SEM) using a JEOL JSM-6300 microscope and by field
emission scanning electron microscopy (FESEM) using a ZEISS Ultra-
55 microscope.
MAS NMR spectra were recorded at room temperature with a
Bruker AV 400 spectrometer. 29Si NMR spectra were recorded with a
spinning rate of 5 kHz at 79.459 MHz with a 55° pulse length of 3.5 μs
and repetition time of 180 s. 29Si chemical shift was referenced to
tetramethylsilane. 19F was measured at 376.28 MHz using a Bruker
probe with 2.5-mm-diameter zirconia rotors spinning at 25 kHz. The
19F spectra were collected using pulses of 4.5 μs corresponding to a flip
angle of Π/2 rad and a recycle delay of 100 s to ensure the complete
recovery of the magnetization.
1H NMR (CDCl3, 300.0 MHz): δH 7.7 (d, J = 6 Hz, 2H, Ar−H), 7.4
(t, J = 6 Hz, 2H, Ar−H), 7.2 (d, J = 6 Hz, 2H, Ar−H), 3.9 (s, 4H,
2×CH2), and 2.6 (s, 6H, 2×NCH3). 13C NMR (CDCl3, 75.5 MHz):
δC 133.5, 133.1, 127.9, 126.1, 125.6, 121.9 (Ar−C), 58.7 (2×CH2),
and 45.6 (NCH3).
NH3-TPD experiments were carried out in a Micromeritics 2900
apparatus. A calcined sample (100 mg) was activated by heating to 400
°C for 2 h in an oxygen flow and for 2 h in an argon flow.
Subsequently, the samples were cooled to 176 °C, and NH3 was
adsorbed. The NH3 desorption was monitored with a quadrupole mass
spectrometer (Balzers, Thermo Star GSD 300 T) while the
temperature of the sample was ramped at 10 °C/min under a helium
flow. Total ammonia adsorption was measured by repeated injection of
calibrated amounts of ammonia at 176 °C until saturation. Ammonia
Finally, 50.0 mL (803.2 mmol) of methyl iodide was slowly added
to a solution of 14.9 g (81.4 mmol) of 2-methyl-1,3-dihydrobenz-
[d,e]isoquinoline in 75 mL of MeOH. The mixture was stirred for 1
week, and no precipitate was observed. The solution was concentrated
to dryness, and the resulting solid was washed with acetonitrile,
2982
DOI: 10.1021/acs.chemmater.5b00337
Chem. Mater. 2015, 27, 2981−2989

Chemistry of Materials
Article
desorption was recorded by means of the mass 15, since this mass is
less affected by the water desorbed.
UV−vis spectra were obtained with a PerkinElmer (Lambda 19)
spectrometer equipped with an integrating sphere with BaSO4 as a
reference.
Steady-state photoluminescence measurements were performed in a
Photon Technology International (PTI) 220B spectrofluorimeter
having Xe arc lamp light excitation and a Czerny−Turner
monochromator, coupled to a photomultiplier. The solid samples
were pressed between two windows of Suprasil quartz cuvettes with a
path length of 0.01 mm and placed at a 45° angle to both the
excitation and emission monochromators. All measurements were
carried out at room temperature.
smaller than those of SAPO-42 (11.4 × 11.4 Å, see LTA cavity
in Figure 2a). Thus, the preferential structure directing effects
of the π-stacked dimers of MTPQ toward the STA-6 cavities
forced us to design new OSDAs that favor the stabilization of
the LTA cavity. Having in mind the different size of both
cavities, we have used aromatic molecules slightly larger than
MTPQ as potential OSDAs to stabilize the LTA cavities and,
consequently, to favor the crystallization of SAPO-42. In this
sense, a bulky aromatic molecule containing naphthalene
groups, such as 2,2-dimethyl-2,3-dihydro-1H-benzo[de]-
isoquinoline-2-ium (DDBQ, see Figure 1a), has been prepared
to study its templating directing role for the synthesis of
silicoaluminophosphate materials.
Textural properties were determined by N2 adsorption isotherms
measured at 77 K with a Micromeritics ASAP 2020.
For this purpose, an initial experimental design was
established to study the following synthesis variables: P/Al,
Si/(Al+P) and H2O/(Al+P). The values for P/Al, Si(Al+P),
and H2O/(Al+P) molar ratios were [1, 0.9, 0.8], [0, 0.052,
0.11], and [30, 50], respectively. The syntheses were carried
out at 175 °C for 5 days under static conditions in a fluoride-
free system.
The silicon-free aluminophosphate form of LTA, AlPO-42,
was achieved under very diluted synthesis conditions (H2O/
TO2 = 50, see experiment LTA-1 in Figure 3a), but the PXRD
pattern of this sample reveals the presence of a small amount of
a dense phase as an impurity (see LTA-1 in Figure 4a). If
silicon is introduced in the synthesis mixture under similar
diluted gels, the dense phase is obtained as a pure crystalline
phase (see Figure 3a). However, the preparation of gels under
more concentrated conditions (H2O/TO2 = 30) allows the
crystallization of the pure SAPO-42 material when the Si/(Al
+P) is fixed at 0.11 (see LTA-2 in Figure 3a), and a mixture of
an unstable lamellar phase and SAPO-42 is obtained when the
Si amount is decreased (Si/TO2 = 0.05, see phase diagram in
Figure 3a).
3. RESULTS AND DISCUSSION
3.1. Synthesis of SAPO-42 Material Using DDBQ
Organic Molecule as OSDA. Since the high silica LTA
zeolite (ITQ-29) has been synthesized using bulky self-
assembled dimers formed by the interaction of two MTPQ
molecules (see Figure 1) through π−π interactions,4 it could be
expected that this aromatic molecule would direct the
crystallization of the SAPO form of the LTA, SAPO-42.
However, under the typical SAPO conditions, MTPQ allows
the selective crystallization of the STA-6 material instead of the
SAPO-42.14 Despite the similarities between STA-6 and SAPO-
42 structures, both present small pores and large cavities, STA-
6 cavities (9.6 × 9.6 Å, see SAS cavity in Figure 2b) are slightly
The PXRD pattern of LTA-2 confirms the absence of other
crystalline impurities, but some amount of amorphous material
is also present in the sample, as reveals the slight curvature of
the PXRD pattern background (see LTA-2 in Figure 4a). The
existence of the amorphous phase in the LTA-2 sample is
confirmed by scanning electron microscopy (SEM), where the
presence of small crystallites with irregular morphologies
Figure 2. Large cavities present in the SAPO-42 (A) and STA-6 (B)
structures.
Figure 3. Phase diagrams achieved using DDBQ as OSDA at 175 °C for 5 days under static conditions.
2983
DOI: 10.1021/acs.chemmater.5b00337
Chem. Mater. 2015, 27, 2981−2989

Chemistry of Materials
Article
cm3/g, see Table 2). In contrast, SAPO-42 containing a mole
fraction of Al close to 0.5 (see LTA-4 in Table 1) shows a
significantly higher micropore volume (∼0.27 cm3/g, see Table
2), which is comparable to the values reported for the high-
silica LTA polymorphs.4
The amount of the organic occluded within the LTA-4
material after the crystallization procedure was determined by
thermogravimetric analysis (TGA). As seen in Table 3, the total
weight loss between 200 and 750 °C, which is assigned to the
organic moieties occluded within the microporous molecular
sieves, was almost 22 wt %. On the other hand, elemental
analysis of the as-prepared LTA-4 material reveals a C/N ratio
of 12.5. It is important to note that the theoretical C/N value
for the DDBQ molecule is 14, indicating that most of the
aromatic molecules entrapped within the micropore structure
(almost 90%) remain stable after zeotype crystallization (see
Table 1). Additionally, the stability of the aromatic OSDA
molecules within the as-prepared LTA-4 has also been studied
by 13C NMR spectroscopy. As shown in Figure 6, the solid
state 13C MAS NMR spectrum of the as-prepared LTA-4
material shows similar chemical shifts to those of the liquid 13
C
NMR spectrum of the organic DDBQ molecule, confirming the
stability of the aromatic molecules occluded in the crystalline
solids.
The number of DDBQ molecules per LTA cavity in the as-
prepared LTA-4 material can be calculated from the TGA
results (see LTA-4 in Table 3) and the structural information
on the LTA framework; i.e., each unit cell has one cavity, and it
is composed of 24 T atoms. Thus, it was calculated that it
should be almost two DDBQ molecules per cavity for LTA-4
(see Table 1). This result would indicate that DDBQ aromatic
molecules could act as paired-OSDAs during the crystallization
of the SAPO-42 materials. However, further characterization is
required to confirm or reject the paired nature of the aromatic
DDBQ molecules.
Figure 4. PXRD patterns of LTA materials in their as-prepared (a)
and calcined (b) forms. For comparison purposes, the simulated
PXRD pattern of the pure silica form of LTA has been added.
In our original work on the synthesis of the high silica LTA
zeolite, the state of aggregation of similar aromatic molecules
both in the synthesis gels and final crystalline products was
proven by fluorescence spectroscopy.4 An intense shift of the
emission band toward higher wavelengths in the fluorescence
spectrum was observed and was related to π−π interactions of
the aromatic rings of the supramolecular self-assembled
molecule.4 Here, the photoluminescence study with DDBQ
has been performed at the excitation wavelength of 284 nm,
since the UV−vis spectrum of DDBQ in aqueous solution
shows its maximum at this wavelength (see Figure 7a). Then,
aqueous solutions of DDBQ with two extreme concentrations
(1.10−4 and 1 M) have been prepared as a control for the
spectroscopic study. The highly diluted aqueous solution of
DDBQ (1.10−4 M) is expected to be mostly formed by
monomeric species, whereas the concentrated aqueous solution
of DDBQ (1 M) should favor the presence of dimeric self-
assembled species. It is important to note that the 1 M
concentration of the aqueous solution of DDBQ is comparable
to the real concentration of the aromatic organic molecules in
the synthesis gels (see Figure 3). The fluorescence spectrum of
the diluted solution shows a main fluorescence band centered
at 325 nm, and the concentrated solution shows a broad
fluorescence band ranging from 400 to 440 nm (see Figure 8).
The fluorescence spectrum of the as-prepared LTA-4 material
clearly shows the presence of a broad band between 400 and
450 nm (see Figure 8c), similarly to the spectrum of the
concentrated aqueous solution of DDBQ (see Figure 8b).
characteristic of amorphous materials is observed (see Figure
5a).
To improve the crystallinity of the SAPO-42, a small amount
of preformed crystals of SAPO-42 were introduced as seeds in
the synthesis gels (∼5 wt %). Two different Si contents (Si/
TO2 = 0.08, 0.11) and gel concentrations (H2O/TO2 = 30, 50)
have been selected to study the effect of the SAPO-42 seeds.
Two SAPO-42 materials, LTA-3 and LTA-4, have been
obtained when the synthesis gels are more concentrated (see
phase diagram in Figure 3b). The PXRD patterns of LTA-3 and
LTA-4 materials indicate high crystallinity and the absence of
impurities (see Figure 4a). Both samples present similar crystal
sizes of ∼1 μm with irregular rhombohedral morphology (see
Figures 5b and 5c).
Chemical analyses on LTA-3 and LTA-4 reveal a similar
silicon content in both samples [Si/(Al+P) ∼ 0.1]. However, a
significant difference in the aluminum content (0.66 and 0.54
for LTA-3 and LTA-4, respectively, see Table 1) is clearly
observed in these two SAPO-42 samples. Theoretically, SAPOs
should show Al mole fractions close to 0.50, because the
preferential isomorphic substitutions involve Si and P atoms.
Thus, the high Al mole fraction observed for LTA-3 (∼0.66)
could be explained by the presence of octahedral Al2O3 species
that have not been inserted in framework positions during the
synthesis. Moreover, the large amount of octahedral Al2O3
species in extra-framework positions could also explain the
relatively low micropore volume observed for LTA-3 (∼0.12
2984
DOI: 10.1021/acs.chemmater.5b00337
Chem. Mater. 2015, 27, 2981−2989

Chemistry of Materials
Article
Figure 5. SEM images of the different SAPO-42 materials: LTA-2 (a), LTA-3 (b), LTA-4 (c), LTA-5 (d), and LTA-6 (e).
a
Table 1. Chemical and Elemental Analysis of SAPO-42 Materials Synthesized Using DDBQ and MTPQ as OSDAs
b
b
b
sample
Si
P
Al
Si/(Al+P)
wt % N
wt % C
(C/N)real
OSDA/cavity
LTA-3
LTA-4
LTA-5
LTA-6
0.11
0.10
0.04
0.05
0.23
0.36
0.46
0.43
0.66
0.54
0.50
0.52
0.12
0.10
0.04
0.05
1.61
1.51
1.65
17.34
16.39
16.87
12.5
12.6
11.9
2.1
2.2
2.3
a
b
The estimated standard deviation (esd) of the analytical measures for Si, P, and Al is 0.002, and for wt % N and wt % C is 0.005. Mole fractions
(Si + Al + P = 1).
Table 2. Physico-chemical Properties of the Synthesized
SAPO-42 Materials
Table 3. Thermogravimetric Analysis (TGA) of the As-
Prepared SAPO-42 Materials
a
acidity concentration
(mmol NH3/g)
loss of weight
sample
<200 °C
200−750 °C
BET
microp.
area
microp.
volume
(cm3/g)
LTA-3
LTA-4
LTA-5
LTA-6
surface area
sample
(m2/g)
(m2/g)
weak
mild
strong
4.6
2.2
3.3
21.6
22.9
23.7
LTA-3
LTA-4
LTA-5
LTA-6
384
575
618
609
245
558
614
589
0.12
0.27
0.30
0.29
0.270
0.172
0.042
0.000
0.000
0.272
1.432
0.000
0.086
a
The estimated standard deviation (esd) of the analytical measures for
Thus, it can be proposed that aromatic DDBQ molecules
within the as-synthesized materials are forming self-assembled
dimers through π−π interactions, and these supramolecular
assemblies are acting as bulky OSDAs.
BET surface area and the micropore is 5 and for micropore volume is
0.01.
2985
DOI: 10.1021/acs.chemmater.5b00337
Chem. Mater. 2015, 27, 2981−2989

Chemistry of Materials
Article
similar SAPO-42 materials with low Si content using bulky
triquads as OSDAs shows that these materials were not stable
after the calcination treatments.12
Si Distribution and Surface Acidity. 29Si MAS NMR
spectroscopy has been used to investigate the coordination of
Figure 6. Solid 13C CP MAS NMR of as-prepared LTA-4 (top) and
liquid 13C NMR of DDBQ OSDA (bottom).
Figure 7. UV−vis spectra of the aromatic DDBQ (a) and MTPQ (b)
OSDAs in aqueous solution.
Figure 9. 29Si MAS NMR spectra of SAPO-42 materials.
the Si species in the LTA-4 material. As seen in Figure 9, the
29Si MAS NMR spectrum mainly exhibits a strong band at −93
ppm, which has been assigned to Si(4Al) species or isolated Si
atoms,16 a shoulder peak centered at −97, which has been
assigned to Si(1Si3Al), and finally, a minor band centered at
−110 ppm, which corresponds to undesired silicon islands (less
than 10%).15 The large presence of isolated silicon species,
which are selectively formed by the isomorphic substitution of
Si4+ atoms by P5+ atoms in the zeolitic framework, must provide
strong acidity to the SAPO. Thus, in order to evaluate the
acidic properties of the LTA-4 material, the calcined sample has
been characterized by temperature-programmed desorption of
ammonia (NH3-TPD). The different desorption temperatures
indicate the different acidic strength of the acid sites present in
the material. Indeed, two well-defined desorption peaks can be
found in the LTA-4 sample (see Figure 10). The first peak
centered at 200 °C corresponds to weak acid sites, mainly
attributed to P−OH hydroxyl groups,17 and the second peak
centered at 330 °C would indicate the presence of strong acid
sites, which would correspond to the acidic protons associated
with the presence of isolated silicon species in the zeolitic
framework.17 The amount of strong acid sites is much higher
than weak acid sites, as could expected by the high silicon
distribution achieved within the LTA-4 material (see 29Si MAS
NMR in Figure 9). The areas under the desorption temperature
peaks allow the quantification of each acid site, giving 1.43 and
0.27 mmol NH3/g for strong and weak acid sites, respectively
(see LTA-4 in Table 2).
Figure 8. Photoluminescence spectra (λex = 284 nm) of (a) diluted
aqueous solution of DDBQ at 1.10−4 M, (b) concentrated aqueous
solution of DDBQ at 1 M, and (c) as-prepared LTA-4 material.
To our delight, the crystalline structure of the LTA-4
material remains stable after calcination in air at 580 °C, as
reveals the PXRD pattern (see LTA-4_calc in Figure 4b). This
is interesting because a recent report describing the synthesis of
2986
DOI: 10.1021/acs.chemmater.5b00337
Chem. Mater. 2015, 27, 2981−2989

Chemistry of Materials
Article
While changes in synthesis conditions may help to direct the
synthesis toward the desired SAPO-42, we decided first to use
the slightly smaller aromatic molecule MTPQ that can also fit
well the cavity size and shape, as OSDA to attempt the
synthesis of the SAPO-42 under fluoride conditions. We
decided to use this OSDA on the basis that previous work
showed that paired self-assembled MTPQ molecules for the
synthesis of silicoaluminophosphate materials under fluoride-
free conditions result in the crystallization of the STA-6
zeotype.14 However, since D4R cages are not present in the
crystalline framework of STA-6 but in the LTA, it could be
expected that the introduction of fluoride anions in the
synthesis gel would favor the crystallization LTA with respect
to STA-6.
The initial synthesis conditions selected to attempt the
synthesis of the SAPO-42 under fluoride conditions using
MTPQ as OSDA were [P/Al = 0.9, Si/(Al+P) = 0.05, MTPQ/
(Al+P) = 0.5, HF/(Al+P) = 0.5, H2O/(Al+P) = 10] at 175 °C
for 5 days. The resultant solid shows the characteristic PXRD
pattern of SAPO-42 (see LTA-5 in Figure 4a), confirming the
preferential directing effects of fluoride anions toward zeotypes
containing D4R cages. The average crystal size for the LTA-5
sample is ∼4 μm (see Figure 5d), which is much larger than the
crystal sizes obtained previously for other SAPO-42 materials
synthesized under fluoride-free conditions (see Figure 5b and
c).
The elemental analysis of the as-prepared LTA-5 solid reveals
that the C/N molar ratio is very close to 13 (see Table 1),
indicating that almost 90% of the MTPQ molecules remain
intact after the SAPO-42 crystallization. The amount of
aromatic molecules per cavity was calculated, as above, from
the thermogravimetric analysis (see Table 3) and the structural
information on SAPO-42. It results to be ∼2 OSDAs/cavity
(see Table 1). Then, as before, the photoluminescence study of
control MTPQ aqueous solutions and as-prepared LTA-5 solid
has been performed at the excitation wavelength of 265 nm
(see UV−vis spectrum in Figure 7b). As can be seen in Figure
11, the as-prepared LTA-5 material shows a broad band in the
photoluminescence spectrum at 400−440 nm that can be
assigned to the presence of self-assembled MTPQ dimers
through π−π interactions of the aromatic rings. In conclusion,
fluorescence and elemental analyses confirm the bulky paired
nature of the aromatic MTPQ molecules acting as OSDA.
The crystallinity of the LTA-5 material remains intact after
calcination at 580 °C in the air (see the PXRD pattern of LTA-
5_calc in Figure 4b). Moreover, N2 adsorption of calcined
LTA-5 gives high micropore volume (0.30 cm3/g) and
micropore area (614 m2/g), revealing good quality of the
SAPO-42 material prepared.
Figure 10. Temperature-programmed desorption of ammonia of
calcined SAPO-42.
3.2. Synthesis of SAPO-42 Material Using MTPQ
Organic Molecule as OSDA in Fluoride Medium. In
addition to the organic directing effects of the paired-OSDAs to
fill and stabilize the large cavity of the LTA structure, the
introduction of fluoride anions in the synthesis media as
inorganic structure directing agents can influence the
nucleation and crystallization processes involved in the
synthesis of the LTA. It is well-known that fluoride anions
are able to stabilize some small cages present in zeolites, as for
instance double-4-rings (D4Rs),18 which are also present in the
LTA structure. Indeed, the synthesis of the high-silica form of
the LTA (ITQ-29) requires the combination of bulky paired
OSDAs (MTPQ) and fluoride anions to stabilize the large
cavities and the D4R cages of the LTA structure, respectively.4
The presence of fluoride anions in the synthesis media not only
favors the crystallization of the high-silica LTA structure due to
the stabilization of the D4R cages but also helps to control the
charge distribution within the crystalline structure by
preferential location of fluoride anions within the D4Rs,
allowing the preparation of the high-silica and even pure silica
LTA polymorphs.
We have shown above that the use of self-assembled DDBQ
dimers, whose size is slightly higher than the size of MTPQ
dimers, is able to direct the crystallization of SAPO-42 with low
Si content and highly distributed Si species under fluoride-free
conditions. However, the introduction of fluoride anions in the
synthesis gel of the optimized SAPO-42 material could produce
significant alterations in its physicochemical properties, such as
crystal size, silicon distribution, and acid strength, among
others.
Si Distribution and Surface Acidity. The 29Si MAS NMR
spectrum of the LTA-5 material exhibits a main signal at −111
ppm, which is characteristic of Si-rich domains in SAPO
materials (see LTA-5 in Figure 9). This result suggests that,
under these synthesis conditions, the paired-MTPQ OSDA is
not able to preferentially direct the Si atoms as isolated Si
species in SAPO-42. If this is so, other anionic species must be
introduced within the zeolitic framework or some structural
defects must be created in the zeolite to properly balance the
two positive charges of the occluded OSDA molecules per
cavity. The 19F MAS NMR spectrum (Figure 12) shows a main
signal at −95 ppm, which has been assigned to F− within
D4Rs.6a The quantification of the signal assigned to F− in the
D4R (1.6 wt % of F−) indicates that the negative charge
Thus, the synthesis of the SAPO-42 zeotype in a fluoride
medium using DDBQ molecules as OSDA was first attempted.
For this purpose, similar synthesis conditions to that described
above for the optimum SAPO-42 under fluoride-free conditions
were selected in this part of the work but in fluoride media: [P/
Al = 0.8, Si/(Al+P) = 0.11, MTPQ/(Al+P) = 0.5, HF/(Al+P)
= 0.5, H2O/(Al+P) = 10] at 175 °C for 5 days. Unfortunately,
an unstable lamellar material was obtained under these
synthesis conditions.
2987
DOI: 10.1021/acs.chemmater.5b00337
Chem. Mater. 2015, 27, 2981−2989

Chemistry of Materials
Article
(note that the LTA structure is composed of 24 T atoms per
unit cell).
The PXRD pattern of the resultant material indicates the
formation of the SAPO-42 material (see LTA-6 in Figure 4a).
With respect to the crystal size of the LTA-6 sample, Figure 5e
shows a mixture of crystals with different shapes (tetragonal
and cubic) and sizes (ranging from 2 to 5 μm). The chemical
composition of different crystals has been measured by EDX,
and similar chemical compositions were found that correspond
well with that of the bulk sample (see bulk chemical analysis of
LTA-6 in Table 1). Thus, despite the presence of different
crystal morphologies, the chemical distribution within the
whole sample is homogeneous. As occurred above for the LTA-
5, sample LTA-6 also shows the presence of two aromatic
organic molecules per cavity forming a supramolecular self-
assembled bulky dimer by π−π interactions of the aromatic
rings (see chemical analyses in Table 1 and photoluminescence
study in Figure 11d).
Si Distribution and Surface Acidity. The LTA-6 sample has
been studied by 19F MAS NMR spectroscopy to determine the
coordination and the amount of the fluoride anions in the as-
prepared solid. The 19F MAS NMR spectrum of the LTA-6
sample has a single signal centered at −95 ppm (see LTA-6 in
Figure 12), indicating that all the fluoride anions present in the
silicoaluminophosphate LTA-6 are occluded inside D4Rs. The
amount of fluoride anions is 0.8 wt % of F−, giving almost 1 F−/
u.c. Thus, if we know that there are two positive charges
provided by the OSDA per unit cell (see Table 1) and there is
almost one of these positive charges properly balanced by one
fluoride anion, it should be another positive charge that would
be able to selectively place the Si species in isolated form.
To confirm this point, the coordination of the silicon species
in the LTA-6 sample has been characterized by 29Si MAS NMR
spectroscopy. As can be seen in Figure 9, the presence of a
main peak centered at −93 ppm reveals the preferential
formation of isolated Si species. Interestingly, ICP analysis
indicates that the amount of Si/TO2 is 0.05, corresponding to
∼1.2 Si atoms per u.c. (note that there are 24 atoms per u.c.),
which is the expected amount of potential Si species to be
properly balanced by the OSDA.
Figure 11. Photoluminescence spectra (λex = 284 nm) of a diluted
aqueous solution of MTPQ at 5.10−4 M (a), concentrated aqueous
solution of MTPQ at 3 M (b), as-prepared LTA-5 material (c), and as-
prepared LTA-6 material (d).
Figure 12. 19F MAS NMR of as-prepared LTA-5 and LTA-6 samples.
The crystalline nature of the LTA-6 sample is preserved after
being calcined in the air at 580 °C (see PXRD pattern of LTA-
6_calc in Figure 4b). N2 adsorption reveals high volume and
micropore area (see Table 2).
introduced by the fluoride anions per unit cell is almost 2.
Therefore, the two positive charges provided by the self-
assembled paired-MTPQ molecules per unit cell fit well with
the two negative charges introduced by fluoride anions. This
high interaction between the OSDA molecules and the fluoride
anions precludes the preferential location of silicon as isolated
species.
3.3. Synthesis of SAPO-42 Material Using MTPQ
Organic Molecule as OSDA with Stoichiometric Amount
of Fluoride in the Synthesis Gel. In order to improve Si
isolation, a synthesis has been designed where the amount of
fluoride introduced in the synthesis is that required to only fill
one D4R per unit cell (D4R/u.c.). Then, if we assume that two
positive charges per unit cell have been introduced by the self-
assembled dimeric OSDA and one of those positive charges will
be balanced by one fluoride anion entrapped within a D4R, the
other positive charge should be able to locate, at least, one Si
atom in isolated form in the unit cell of SAPO-42. Working
under the same synthesis conditions as for the LTA-5 sample,
but decreasing the theoretical HF/(Al+P) ratio to 0.05, would
correspond to the presence of one fluoride anion per unit cell
Finally, the acid properties of the calcined LTA-6 have been
evaluated by NH3-TPD. As seen in Figure 10, the LTA-6
sample shows three overlapped desorption peaks centered at
200, 250, and 340 °C, indicating the presence of weak, medium,
and strong acid sites, respectively. The amount of the different
acid sites can be estimated integrating the area under the
ammonia desorption peaks. Following this methodology, the
concentration of each acid site could be estimated, obtaining
0.04, 0.27, and 0.09 mmol NH3/g for weak, mild, and strong
acidities (see Table 2). The amount of desorbed ammonia for
the LTA-4 sample is significantly higher than for the LTA-6
sample, probably due to the higher content of isolated Si atoms
in LTA-4 (see Table 1 and Figure 9).
The catalytic activity and hydrothermal stability of the small
pore SAPO-42 materials synthesized along the present work
with different acid strength will be evaluated for different
industrially relevant chemical processes, such as methanol-to-
olefins (MTO) or selective catalytic reduction (SCR) of
NOx.19
2988
DOI: 10.1021/acs.chemmater.5b00337
Chem. Mater. 2015, 27, 2981−2989

Chemistry of Materials
4. CONCLUSIONS
Article
(15) Brun, A. M.; Harriman, A.; Tsuboi, Y.; Okada, T.; Mataga, N. J.
Chem. Soc., Faraday Trans. 1995, 91, 4047.
The synthesis of the silicoaluminophosphate form of the LTA,
SAPO-42, with low silicon content, a very large amount of
isolated Si sites, and high thermal stability, has been
accomplished by using bulky OSDAs formed by the self-
assembling of two aromatic organic molecules through π−π
interactions in fluoride as well as in fluoride free synthesis
conditions. Two different quinolinium derived aromatic
molecules, MTPQ and DDBQ, are able to direct the
crystallization of the SAPO-42 material. The rational
combination of self-assembled aromatic molecules and the
presence or not of fluoride anions, as well as the amount of
fluoride content, allows synthesizing SAPO-42 materials with
different crystal sizes, morphologies, silicon distributions, and
acid strength, depending on the synthesis conditions.
(16) Xu, L.; Du, A.; Wei, Y.; Wang, Y.; Yu, Z.; He, Y.; Zhang, X.; Liu,
Z. Microporous Mesoporous Mater. 2008, 115, 332.
(17) Yu, T.; Wang, J.; Shen, M.; Li, W. Catal. Sci. Technol. 2013, 3,
3234.
(18) Koller, H.; Wolker, A.; Eckert, H.; Panz, C.; Behrens, P. Angew.
̈
Chem., Int. Ed. 1997, 36, 2823.
(19) Moliner, M.; Martínez, C.; Corma, A. Chem. Mater. 2014, 26,
246.
AUTHOR INFORMATION
Corresponding Authors
*E-mail: mmoliner@itq.upv.es.
*E-mail: acorma@itq.upv.es.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Financial support by the Spanish Government-MINECO
through “Severo Ochoa” (SEV 2012-0267), Consolider Ingenio
2010-Multicat, MAT2012-37160, and Intramural-201480I015 is
acknowledged.
REFERENCES
■
(1) International Zeolite Association Web page: http://www.iza-
online.org/.
(2) Barrer, R. M.; Denny, P. J. J. Chem. Soc. 1961, 971.
(3) Sherman, J. D. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 3471.
(4) Corma, A.; Rey, F.; Rius, J.; Sabater, M. J.; Valencia, S. Nature
2004, 431, 287.
(5) (a) Moliner, M.; Rey, F.; Corma, A. Angew. Chem., Int. Ed. 2013,
52, 13880. (b) Moliner, M. Topics Catal. 2015, DOI: 10.1007/s11244-
015-0390-0. (c) Davis, M. E. Chem. Mater. 2014, 26, 239.
(6) (a) Sierra, L.; Deroche, C.; Gies, H.; Guth, J. L. Microporous
Mater. 1994, 3, 29. (b) Schreyeck, L.; Dagosto, F.; Stumbe, J.; Caullet,
P.; Mougenel, J. C. J. Chem. Soc., Chem. Commun. 1997, 13, 1241.
(c) Schreyeck, L.; Stumbe, J.; Caullet, P.; Mougenel, J. C.; Marler, B.
Microporous Mesoporous Mater. 1998, 22, 87. (d) Wright, P. A.; Morris,
R. E.; Wheatley, P. S. Dalton Trans. 2007, 5359. (e) Huang, A.; Caro, J.
Microporous Mesoporous Mater. 2010, 129, 90.
(7) Lok, B. M.; Messina, C. A.; Patton, R. L.; Gajek, R. T.; Cannan,
T. R.; Flanigen, E. M. U.S. Patent 4440871, 1984.
(8) (a) Lok, B. M.; Messina, C. A.; Patton, R. L.; Gajek, R. T.;
Cannan, T. R.; Flanigen, E. M. J. Am. Chem. Soc. 1984, 106, 6092.
(b) Yu, J.; Xu, R. Acc. Chem. Res. 2010, 43, 1195. (c) Li, J.; Yu, J.; Xu,
R. Proc. R. Soc. A 2012, 468, 1955.
(9) Man, P. P.; Briend, M.; Peltre, M. J.; Lamy, A.; Beaunier, P.;
Barthomeuf, D. Zeolites 1991, 11, 563.
(10) Kuhl, G. H.; Schmitt, K. D. Zeolites 1990, 10, 2.
(11) Moscoso, J. G. U.S. Patent 2014/0170062, 2014.
(12) Schmidt, J. E.; Zones, S. I.; Xie, D.; Davis, M. E. Microporous
Mesoporous Mater. 2014, 200, 132.
(13) (a) Martinez-Franco, R.; Moliner, M.; Franch, C.; Kustov, A.;
Corma, A. Appl. Catal., B 2012, 127, 273. (b) Martinez-Franco, R.;
Moliner, M.; Concepcion, P.; Thogersen, J. R.; Corma, A. J. Catal.
2014, 314, 73. (c) Martinez-Franco, R.; Moliner, M.; Corma, A. J.
Catal. 2014, 319, 36.
(14) Martinez-Franco, R.; Cantín, A.; Moliner, M.; Corma, A. Chem.
Mater. 2014, 26, 4346.
2989
DOI: 10.1021/acs.chemmater.5b00337
Chem. Mater. 2015, 27, 2981−2989