Mario De Bruyn1,2, Paolo Cuello-Penaloza1, Melissa Cendejas3, Ive Hermans3, Jiayue He1, Siddarth H Krishna1, David M Lynn1, James A Dumesic1, George W Huber1, Bert M Weckhuysen2. 1. Department of Chemical and Biological Engineering, University of Wisconsin-Madison, 1415 Engineering Drive, Madison, Wisconsin 53706, United States. 2. Faculty of Science, Debye Institute for Nanomaterials Science, Utrecht University, Universiteitsweg 99, 3584 CG Utrecht, The Netherlands. 3. Department of Chemistry, University of Wisconsin-Madison, 1101 University Avenue, Madison, Wisconsin 53706, United States.
Abstract
We report on the synthesis and characterization of novel mesoporous chiral polyboronates obtained by condensation of (R,S)/(S,S)-hexane-1,2,5,6-tetrol (HT) with simple aromatic diboronic acids (e.g., 1,3-benzenediboronic acid) (BDB). HT is a cellulose-derived building block comprising two 1,2-diol structures linked by a flexible ethane bridge. It typically consists of two diastereomers one of which [(S,R)-HT] can be made chirally pure. Boronic acids are abundantly available due to their importance in Suzuki-Miyaura coupling reactions. They are generally considered nontoxic and easy to synthesize. Reactive dissolution of generally sparingly soluble HT with BDB, in only a small amount of solvent, yields the mesoporous HT/polyboronate materials by spontaneous precipitation from the reaction mixture. The 3D nature of HT/polyboronate materials results from the entanglement of individual 1D polymeric chains. The obtained BET surface areas (SAs) and pore volumes (PVs) depend strongly on HT's diastereomeric excess and the meta/para orientation of the boronic acids on the phenyl ring. This suggests a strong influence of the curvature(s) of the 1D polymeric chains on the final materials' properties. Maximum SA and PV values are respectively 90 m2 g-1 and 0.44 mL g-1. Variably sized mesopores, spanning mainly the 5-50 nm range, are evidenced. The obtained pore volumes rival the ones of some covalent organic frameworks (COFs), yet they are obtained in a less expensive and more benign fashion. Moreover, currently no COFs have been reported with pore diameters in excess of 5 nm. In addition, chiral boron-based COFs have presently not been reported. Scanning electron microscopy reveals the presence of micrometer-sized particles, consisting of aggregates of plates, forming channels and cell-like structures. X-ray diffraction shows the crystalline nature of the material, which depends on the nature of the aromatic diboronic acids and, in the specific case of 1,4-benzenediboronic acid, also on the applied diastereomeric excess in HT.
We report on the synthesis and characterization of novel mesoporous chiral polyboronates obtained by condensation of (R,S)/(S,S)-hexane-1,2,5,6-tetrol (HT) with simple aromatic diboronic acids (e.g., 1,3-benzenediboronic acid) (BDB). HT is a cellulose-derived building block comprising two 1,2-diol structures linked by a flexible ethane bridge. It typically consists of two diastereomers one of which [(S,R)-HT] can be made chirally pure. Boronic acids are abundantly available due to their importance in Suzuki-Miyaura coupling reactions. They are generally considered nontoxic and easy to synthesize. Reactive dissolution of generally sparingly soluble HT with BDB, in only a small amount of solvent, yields the mesoporous HT/polyboronate materials by spontaneous precipitation from the reaction mixture. The 3D nature of HT/polyboronate materials results from the entanglement of individual 1D polymeric chains. The obtained BET surface areas (SAs) and pore volumes (PVs) depend strongly on HT's diastereomeric excess and the meta/para orientation of the boronic acids on the phenyl ring. This suggests a strong influence of the curvature(s) of the 1D polymeric chains on the final materials' properties. Maximum SA and PV values are respectively 90 m2 g-1 and 0.44 mL g-1. Variably sized mesopores, spanning mainly the 5-50 nm range, are evidenced. The obtained pore volumes rival the ones of some covalent organic frameworks (COFs), yet they are obtained in a less expensive and more benign fashion. Moreover, currently no COFs have been reported with pore diameters in excess of 5 nm. In addition, chiral boron-based COFs have presently not been reported. Scanning electron microscopy reveals the presence of micrometer-sized particles, consisting of aggregates of plates, forming channels and cell-like structures. X-ray diffraction shows the crystalline nature of the material, which depends on the nature of the aromatic diboronic acids and, in the specific case of 1,4-benzenediboronic acid, also on the applied diastereomeric excess in HT.
Mesoporous materials have attracted significant
attention as their
wide pores (2–50 nm) allow for improved diffusion and accessibility,
favoring applications as diverse as, controlled drug release, chromatography,
adsorbents, electrodes, solar cells, and heterogeneous catalysis.[1,2] Exemplary are (non)siliceous mesoporous oxides, periodic mesoporous
organic silicas (PMOs), mesoporous carbons, hyperconjugated porous
polymers, and some metal–organic frameworks.[3−7] Often the introduction of mesopores in a material
is linked to the use of (supra)molecular templates such as surfactants,
but also other techniques are known like dealumination/desiliciation,
nanoassemblies, and local associations of helices, among others.[1,8]Boronate ester polymers (BEP) and their tendency to form hierarchical
supramolecular structures have been widely reported in the literature.[9] However, few micro/mesoporous materials have
been reported based on 1D polymeric boronate esters containing backbones
not involving heteroatoms (e.g., nitrogen, sulfur). Exemplary are
the flowerlike micro/mesoporous microparticles made by the condensation
of 1,4-benzenediboronic acid (1,4-BDB) with pentaerythritol (PE).
These materials have a specific surface area of 184 m2 g–1 (BET), yet no data on pore volume and pore size has
been published.[10,11] Notably, the functionalization
of these materials with Au or Pd nanoparticles allowed for (chemoselective)
catalytic reduction reactions.[10,11] Mixing diboronic acid
with a chiral indacene-type bis(1,2-diol) gave trimeric microporous
cages with surface areas (SAs) of 491 and 582 m2 g–1.[12] Hierarchical structures
using dative B–N bonds have also been reported.[9,13] A relative recent addition are the covalent 2D/3D organic frameworks
(COFs). These are micro/mesoporous materials where rigid building
blocks, often with a specific geometry and/or symmetry, are interconnected
through covalent bonds forming 2D/3D rigid crystalline porous structures.[14] Most typically, the boron-based COFs employ
catechol building blocks. To date no COF has been reported with a
pore diameter larger than 5 nm. In addition, only a few chiral COFs
are known, typically using chiral pyrrolidine and tartaric acid derivatives
as the linking or pendant groups.[15] Presently,
no chiral boron-based COFs have been reported.Very recently
a novel COF class was introduced where helical polymers
are interwoven at regular intervals by means of copper complexes.[16] Demetalation of this structure (named COF-505)
was shown to retain its morphology, at the expense of a certain decrease
in crystallinity.[16] Most importantly, owing
to the large flexibility of the 1D threads in demetalated COF-505,
a 10-fold increase of its elasticity was observed.[17] Besides weaving, entanglement in porous frameworks can
also be achieved by the interpenetration of 2D/3D frameworks or interlocking
of rings.[17]In spite of the high
surface areas and pore volumes, and the related
plethora of potential and promising applications, the commercial use
of mesoporous materials is still quite a challenge due to their chemical
instability, expensive synthesis approaches, and nonstraightforward
processability.[18] The cost of the extended
linkers, which is especially relevant for COFs and MOFs, is another
major concern with, for instance, only two commercialized MOFs employing
linkers with more than one phenyl ring.[19]Recently, we have reported on the sustainable synthesis of
hexane-1,2,5,6-tetrol
(HT) in which two 1,2-diol moieties are joined by a flexible ethane
bridge.[20] HT is derived from cellulose
in 3–4 steps many of which involve heterogeneous catalysts.
In successive order these comprise (Figure ) the following list:
Figure 1
Overview figure synthesis of (S,S)/(S,R)-hexane-1,2,5,6-tetrol
from
(waste) cellulosic feedstocks.
Conventional/microwave pyrolysis of
(waste) cellulosic polysaccharides.[21−23]Hydrogenation of levoglucosenone to
levoglucosanol over dihydrolevoglucosenone with heterogeneous
Pd catalysts (one or two steps).[24,25] It is noteworthy
that presently a joint venture between the Australian Circa Company
and the Norwegian multinational Norskeskog is running a fully operational
50 tonnes/year prototype plant in Tasmania.[26]Combined one pot
hydrogenation/hydrolysis
of levoglucosanol (one step).[20]Overview figure synthesis of (S,S)/(S,R)-hexane-1,2,5,6-tetrol
from
(waste) cellulosic feedstocks.HT is typically obtained as a mixture of two diastereomers
notably
(S,R)-HT (mesomer) and (S,S)-HT. Boronic acids are abundantly available
due to their importance in Suzuki–Miyaura coupling reactions.[27,28] They are generally considered nontoxic and are easy to synthesize.[27,29]Here we report on the synthesis of a novel porous organic
polymer
obtained by reactive dissolution of sparingly soluble HT with simple
benzenediboronic acids (BDB) and concomitant precipitation of
the mesoporous HT polyboronate. The reaction is conducted at RT and
requires no special reaction conditions such as vacuum, inert atmosphere,
or controlled water removal. It is shown that the properties of the
resulting mesoporouspolymers are largely determined by the chirality
of HT and the meta/para orientation of the boronic acids on the phenyl
ring. Thereby HT’s chirality is likely inducing coiling of
the polymer chain, allowing for interweaving of the polymeric chains.
These materials are crystalline and in spite of their lower surface
areas (30–90 m2 g–1) display mesopore
volumes up to 0.441 mL g–1. Scheme A provides a visual of the potential polymeric
chains indicative of their stereocenters and points of potential molecular
rotation. Hereafter these materials are denoted as “HT/polyboronates”
or more specifically using the HT/1,x-BDB/solvent
formalism.
Scheme 1
(A) Schematic Visualization of Two Different HT (hexane-1,2,5,6-tetrol)-BDB
(1,4- or 1,3-Benzenediboronic Acid) Polymeric Chains with Indication
of the Stereocenters and Points of Potential Rotation of the Molecular
Chains, (B) Illustrative Variability of the Growing Polymeric Chain
When Incorporating Chirally Pure (R,S)-HT [Shown as (R-S)], and (C) Unique Chiral Polymer When Constructing
the Polymeric Chain from (S,S)-HT
[Shown as (S–S)]
Results and Discussion
Table lists the
obtained specific surface areas (SAs) and pore volumes (PVs) as a
function of the type of diboronic acid used, the applied solvent,
and HT’s diastereomeric excess. It is found that the meta orientation
of the diboronic acids on the phenyl ring (1,3-BDB) gives higher SA
and PV values compared to the para orientation of the diboronic acids
(1,4-BDB) (Table ,
entries a/b). This suggests for a distinct influence of the curvature
of the polymeric chain. Interestingly, the use of 4,4′-biphenyldiboronic
acid (4,4′-BPDB) led to quasi similar results as obtained with
1,3-BDB (Table , entry
c). This can be rationalized by the nonplanarity of biphenyl units
and the consequent absence of p–p stacking of the aromatic
rings. The use of acetone or CH2Cl2 as the reaction
solvents gives higher SAs and PVs than when THF or AcCN are used (Table entries d–g).
Additionally, the SA and PV values benefit also from higher reaction
volumes (Table , entries
d, h, and i).
Table 1
Survey of the Specific Surface Areas
and Specific Pore Volumes for a Range of HT (Hexane-1,2,5,6-tetrol)
Containing Polyboronate Esters
entry
component 1
component 2
solvent
spec surface area (m2 g–1) BET
spec pore volume (mL g–1)
a
HT 38% 1 < 2
1,4-BDB
THF
42
0.227
b
HT 38% 1 < 2
1,3-BDB
THF
65.8
0.408
c
HT 38% 1 < 2
4,4′-BPDB
THF
64.5
0.359
d
HT 38% 1 > 2
1,3-BDB
THF
60.5
0.315
e
HT 38% 1 > 2
1,3-BDB
CH2Cl2
74.0
0.397
f
HT 38% 1 > 2
1,3-BDB
AcCN
61.2
0.290
g
HT 38% 1 > 2
1,3-BDB
acetone
67.2
0.440
h
HT 38% 1 > 2
1,3-BDB
THF 5 mL
49.1
0.289
i
HT 38% 1 > 2
1,3-BDB
THF 45 mL
62.8
0.407
j
HT 38% 1 < 2
1,3-BDB
acetone
51.3
0.271
k
HT 98% 1 > 2
1,3-BDB
THF
80.7
0.412
l
HT 98% 1 > 2
1,4-BDB
THF
32.8
0.148
m
HT 98% 1 > 2
1,3-BDB
acetone
87.3
0.441
Unless otherwise stated, the used
solvent volume is 15 mL; 1 = (S,R)-HT and 2 = (S,S)-HT; the 38%
and 98% refer to diastereomeric excesses (d.e.).
Unless otherwise stated, the used
solvent volume is 15 mL; 1 = (S,R)-HT and 2 = (S,S)-HT; the 38%
and 98% refer to diastereomeric excesses (d.e.).Partial enrichment of (S,R)-HT
over (S,S)-HT was described in a
previous publication.[20] Specific crystallization
of the (S,R)-HT diastereomer was
achieved from hot 1,4-dioxane, allowing for the evaluation of HT’s
chirality on the properties of the HT/polyboronates. Unfortunately,
it has presently not been found possible to crystallize/purify (S,S)-HT. It is noteworthy that (S,R)-HT is a mesomer, making that it displays
no net chirality on its own but does so effectively in a polymeric
setting. Given that (S,R)-HT can
build into the polymeric chain as (S,R) or (R,S), substantial variability in the curvature
of the polymeric chain is likely present (Scheme B/C). Actual and distinct influences of the
diastereomeric (S,R)/(S,S) ratio in HT on the properties of the materials
are observable:In using 38% d.e. (S,R)-HT, higher
SA and PV are obtained then when
38% d.e. (S,S)-HT is used (Table , entries g vs j).Elevating the d.e. in
(S,R)-HT from 38 to 98% is found
to increase both
the SA and PV by around 30% for HT/1,3-BDB/THF (Table , entries k ↔ d). In the case of HT/1,3-BDB/acetone,
only an SA increase of 30% is observed with an already high PV (Table , entries m vs g).Using diastereomerically
pure (S,R)-HT magnifies the effect
of the meta/para
orientation of BDB on the spread in SA and PV (Table , entries a, b, k, and l).It is noteworthy that no cooperative stereochemical
effect occurs
as hydrolysis of the formed HT/polyboronates does not reveal a preferential
incorporation of one of the HT diastereomers. Irrespective the nature
of the HT/polyboronates, the obtained SA and PV values correlate to
a certain degree (Figure S1_A).In
terms of yield, reactions between HT with 38% d.e. (1 > 2 or
1 < 2) and 1,4-BDB typically give between 65 and 75% yield of the
solid HT/1,4-BDB polyboronate product. This is 73–85% of the
theoretical yield which is 88.6% given the loss of two waters molecules
during the condensation of HT and BDB. Reactions between HT with 38%
d.e. (1 > 2 or 1 < 2) and 1,3-BDB typically give between 55
and
65% yield of the solid HT/1,3-BDB polyboronate product. This is 62–73%
of the theoretical yield. Reactions using pure HT (i.e., > 98%
d.e.)
are with 26–30% much lower yielding, and this irrespective
of the HT diastereomer and the BDB nature.As shown in Figure A, the N2 physisorption isotherms display a type IV isotherm
and an H3 hysteresis. The latter is indicative of aggregates of platelike
particles forming slitlike pores.[30] This
is reminiscent of the pentaerythrithol/1,4-BDB (PE/1,4-BDB) material
reported by Fujiwara et al. and Matsushima et al.[10,11] with the difference that the N2 adsorption is fully reversible,
strengthening our claim on mesoporous materials. Larger variably sized
mesopores, spanning mainly the 5–50 nm ⌀ range, are evidenced (Figure B). The sharp peak at 3.7 nm is an artifact relating to forced
closure of the sorption hysteresis.[31] As
can be inferred from Figure B, the pore size distribution is solvent dependent with the
use of acetone leading to a larger mean pore diameter than when THF
is used.
Figure 2
N2 adsorption isotherms (A) and pore size distributions
(B) for HT (hexane-1,2,5,6-tetrol)-1,3-BDB (1,3-benzenediboronic
acid) made with four different solvents.
N2 adsorption isotherms (A) and pore size distributions
(B) for HT (hexane-1,2,5,6-tetrol)-1,3-BDB (1,3-benzenediboronic
acid) made with four different solvents.The here reported maximal PV of 0.44 mL g–1 is
comparable to the one reported for the [MeOAc]50-H2P (0.42 mL g–1) COF yet obtained in a less expensive and more benign
fashion than the COF.[32] The marked difference
in SA between the HT/polyboronates and the latter COF material (90
vs 754 m2 g–1) relates to the much larger
pore diameters of the materials reported here (⌀ 5–50
nm) vis-à-vis the one for the above stated COF (⌀ 1.8
nm).SEM analysis of 98% d.e. (R,S)-HT/1,4-BDB/THF (Figure A series), 98% d.e. (R,S)-HT/1,3-BDB/THF (Figure B series), and 98% d.e. (R,S)-HT/1,3-BDB/acetone (Figure C series) reveals the following:
Figure 3
Scanning electron microscopy (SEM) images of 98% d.e. (R,S)-HT/1,4-BDB/THF (parts A1, A2, A3),
98% d.e. (R,S)-HT/1,3-BDB/THF (parts
B1, B2, B3), and 98% d.e. (R,S)-HT/1,3-BDB/acetone
(parts C1 C2 C3).
Micrometer-sized particles (Figure A1/B1/C1).Aggregates
of plates for 98% d.e. (R,S)-HT/1,4-BDB/THF
(Figure A2/A3).A more pronounced pore structure when 1,3-BDB
is used
over 1,4-BDB (Figure B1/B2). Here the plate aggregates form channels and cell-like structures.
Interestingly, some of the observed plate-like structures seem to
have delaminated which may have contributed positively to the observed
enhanced porosity (Figure B3).Using acetone as the reaction
solvent yields equally
a highly porous structure yet with a substantially different fine-structure
(on the micrometer level) than observed when THF is used (Figure B2/B3 vs Figure C2/C3).Scanning electron microscopy (SEM) images of 98% d.e. (R,S)-HT/1,4-BDB/THF (parts A1, A2, A3),
98% d.e. (R,S)-HT/1,3-BDB/THF (parts
B1, B2, B3), and 98% d.e. (R,S)-HT/1,3-BDB/acetone
(parts C1 C2 C3).Overall the dimensions
of the irregular voids between the platelets,
as channels or cell-like structures, are in the tens of nanometers
and thus consistent with the N2 physisorption results and
the appearance of an H3 hysteresis in the N2 physisorption
experiments. In spite of the high melting temperatures of the HT/polyboronates
(∼600 K), TEM imaging was found impossible due to melting of
the materials under the electron beam. XRD revealed a degree of crystallinity
in all HT/polyboronates with the nature of the BDB (length, substitution
pattern) affecting markedly the appearance of the different XRD patterns
(Figure ). Given the
characteristic H3 hysteresis loop in the N2 physisorption
experiments and the absence of order in the SEM images (Figure ), we assume that the observed
crystallinity reflects some ordering in the pore wall structure. The
nature of the used solvent has no effect on the crystallinity (Figure S2). The diastereoselectivity of HT affects
the crystallinity only in the case of the 1,4-BDB linker (Figure ).
Figure 4
X-ray diffractograms
(XRD) of (A) HT/1,4-BDB, (B) HT/1,3-BDB, and
(C) HT/4,4′-BPDB; no background subtraction.
Figure 5
X-ray diffractograms (XRD) of HT/1,3-BDB (A 38% HT d.e.;
B 98%
HT d.e.) and HT/1,4-BDB (C 38% HT d.e.; D 98% HT d.e.); PHT = chirally
pure HT.
X-ray diffractograms
(XRD) of (A) HT/1,4-BDB, (B) HT/1,3-BDB, and
(C) HT/4,4′-BPDB; no background subtraction.X-ray diffractograms (XRD) of HT/1,3-BDB (A 38% HT d.e.;
B 98%
HT d.e.) and HT/1,4-BDB(C 38% HT d.e.; D 98% HT d.e.); PHT = chirally
pure HT.The (S,R)-HT/1,4-BDB and (S,R)-HT/1,3-BDB polyboronates were characterized
by 13C CP-MAS and 11B spin echo MAS NMR, and
the results are shown in Figure . Analytical simulations of the 11B spectra
confirmed the existence of only one type of boron species, and this
is irrespective of the nature of the BDB unit. (Figure B/D, red curve). FT-IR spectroscopy further
confirmed the presence of boronate esters with characteristic vibrational
peaks in the 1320–1300 cm–1 (B–O stretch)
and 690–650 cm–1 (characteristic boronate
ester peak[33]) spectral ranges (Figure S3_A).[11,34] In addition,
FT-IR spectroscopy revealed the clear absence of OH groups (3500–3000
cm–1), characteristic for the residual presence
of HT or incomplete boronate ester formation in the synthesized HT/polyboronates
(Figures S3_B/C).
Figure 6
13C CP-MAS
NMR of (A) (S,R)-HT/1,4-BDB and
(C) (S,R)-HT/1,3-BDB; 11B spin echo MAS NMR of (B) (S,R)-HT/1,4-BDB and (D) (S,R)-HT/1,3-BDB;
* = spinning sideband.
13C CP-MAS
NMR of (A) (S,R)-HT/1,4-BDB and
(C) (S,R)-HT/1,3-BDB; 11B spin echo MAS NMR of (B) (S,R)-HT/1,4-BDB and (D) (S,R)-HT/1,3-BDB;
* = spinning sideband.
Conclusions
A range of new crystalline (chiral) mesoporous
materials were obtained
by the facile and benign condensation of (chiral), biobased hexane-1,2,5,6-tetrol
(HT) and simple aromatic diboronic acids. The so formed mesoporousHT/polyboronates are obtained by spontaneous precipitation from the
reaction mixture, not requiring an extensive workup procedure. HT
is easily accessible from cellulose and chiral pure (R,S)-HT is straightforward to crystallize. The applied
simple diboronic acids are abundantly available due to their commercial
use in Miyaura–Suzuki reactions, easy to synthesize, and generally
considered nontoxic.[35−37] The low cost of these materials is ultimately considered
to benefit potential commercial applications. The reported materials
show an appreciable pore volume of up to 0.44 mL g–1, rivalling the pore volumes of some reported COF materials yet obtained
in a less expensive way. Importantly, presently no COF has been reported
with pore diameters in excess of 5 nm. The (R,S)-HT (98% d.e.)/polyboronate represents a first example
of a chiral mesoporous polyboronate. The high pore volume and the
larger pore sizes of 5–50 nm offer great potential to the incorporation
of enzymes or as supports to immobilize catalysts to convert bulkier
substrates. The meta/para orientation of the boronic acids on the
phenyl ring was shown to influence the obtainable surface areas and
pore volumes. Uniquely, a distinct influence of HT’s chirality
(its d.e. value) on the materials’ properties was evidenced.
Future work will focus on testing these new materials toward a wide
range of different applications, thereby hopefully demonstrating its
potential as a valuable material.
Authors: Robin J White; Peter S Shuttleworth; Vitaliy L Budarin; Mario De Bruyn; Anna Fischer; James H Clark Journal: ChemSusChem Date: 2016-01-19 Impact factor: 8.928
Authors: James Sherwood; Mario De bruyn; Andri Constantinou; Laurianne Moity; C Rob McElroy; Thomas J Farmer; Tony Duncan; Warwick Raverty; Andrew J Hunt; James H Clark Journal: Chem Commun (Camb) Date: 2014-09-04 Impact factor: 6.222