Generation of a solid Bronsted acid site in a chiral framework

Article abstract of DOI:10.1039/b718443c

Protonation of chiral porous materials introduces a Bronsted acid centre, the structure of which is unique to the heterogeneous phase requiring pore wall confinement for stable isolation. The Royal Society of Chemistry.

Full text of DOI:10.1039/b718443c

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Generation of a solid Brønsted acid site in a chiral frameworkw
Michael J. Ingleson, Jorge Perez Barrio, John Bacsa, Calum Dickinson, Hyunsoo Park and
Matthew J. Rosseinsky*
Received (in Cambridge, UK) 29th November 2007, Accepted 15th January 2008
First published as an Advance Article on the web 14th February 2008
DOI: 10.1039/b718443c
Protonation of chiral porous materials introduces a Brønsted
acid centre, the structure of which is unique to the heterogeneous
phase requiring pore wall confinement for stable isolation.
thermal stability.9 A suspension of 1 in anhydrous Et2O upon
treatment with HCl in ether led to the isolation of the
protonated phase Ni(L-asp)bipy0.5(HCl)0.9(MeOH)0.5 2.
,
EDX analysis confirmed this assignment with a Ni : Cl ratio
consistently close to 1 : 1,w whilst the degree of solvation was
elucidated by a combination of micro- and thermogravimetric
(TG) analysis.w Powder X-ray diffraction (pXRD) confirmed
the crystallinity of 2, with only minor differences observed on
comparison to the parent material 1, confirming comparable
framework connectivities.w Anhydrous conditions are essen-
tial for the isolation of 2; an analogous reaction performed
with HCl in the aqueous phase led to partial dissolution of the
nickel salt and no resultant framework material. Despite
various approaches the pXRD continually revealed two minor
impurity phases; one phase is unreacted 1, the other (which
can be synthesised as the major product by use of 5 eq. of HCl
in anhydrous ether for 120 h) is Ni(bipy)Cl2,19 consistent with
an ‘over-protonation’ of the framework by excess acid. Opti-
mised conditions of 1.2 equivalents of HCl (per Ni) in anhy-
drous ether and 18 h reaction time minimise both impurities.
Infrared (IR) spectroscopy is particularly informative in iden-
tifying the site of protonation. In non-protonated 1, the
Zeolitic materials are important industrial heterogeneous cat-
alysts, however their application in asymmetric transforma-
tions has been limited by the scarcity of chiral zeolite phases.1,2
The requirement for new chiral porous catalytic materials may
be fulfilled by metal organic frameworks (MOFs), with an
increasing number of enantiopure materials reported,3–6 in-
cluding examples both porous and robust.7–12 Excluding
several notable exceptions,13–15 realising both chirality and
enantioselective catalytic activity with MOFs remains a for-
midable challenge. The synthetic conditions for MOF forma-
tion are in conflict with the requirements for catalytically
active sites due to coordinative saturation at all metal centres
(eliminating metal Lewis acidity), inaccessibility of the Lewis
basic sites (involved in bonding to the metal) and absence of
Brønsted acid sites (candidate low pKa moieties are deproto-
nated by the unavoidable presence of Lewis basic ligands). An
alternative approach to the synthesis of catalytically active
MOFs, which reduces the probability for catalyst deactivation,
is to introduce the active site after the construction of the
framework; as exemplified by the work of Lin et al.16 We were
interested in using this methodology to introduce a Brønsted
acid site into a homochiral MOF that would be of sufficiently
low pKa to be an effective catalyst. To date there has been only
one reported chiral, Brønsted acidic MOF that is catalytically
active (albeit with enantiomeric excesses (ee) o 5%).17 Here
we report a fully reversible, post-synthetic, acid functionalisa-
tion of a homochiral, porous MOF family. This introduces
Brønsted acidic centres that are heterogeneous asymmetric
catalysts for organic transformations. The catalysts presented
are novel as they are not heterogenised homogeneous cata-
lysts;18 in fact, homogeneous phase analogues are inaccessible.
Stable catalyst isolation is dependent on active group confine-
ment in the MOF scaffold wall.
ꢀ
nCO2 symmetric and asymmetric stretches are in the range
1650 to 1300 cmꢀ1 as expected,20,21 on protonation the
analogous region now includes a stretch at 1720 cmꢀ1 (Fig.
S3w) fully consistent with a protonated COOH nCQO
stretch.29 The 1 : 1 Ni : Cl EDX ratio confirms addition of
one equivalent of H+ (per aspartate) to the entire framework,
precluding the presence of the COOH moiety in the IR as a
minor impurity. TG analysis for 2 revealed a two stage mass
loss before framework decomposition, an initial rapid solvent
loss followed by gradual loss of HCl (from 130 to 200 1C); the
combined mass loss of 14.0% approaches that predicted from
the analysed composition (15.1%). Microanalysis post TG
and pXRD confirms the regeneration of 1 by HCl loss. HCl
removal can also be achieved chemically by action of the
small, non-nucleophilic base Et2NH to regenerate 1. Desol-
vated 2, though non-porous to CO2 (at ꢀ78 1C and 25 1C) and
H2 (ꢀ196 1C), does adsorb MeOH at 25 1C, albeit at a slow
rate.w The chloride counterion, with a van der Waals radius of
1.75 A,26 is of sufficient diameter to hinder the transport of
even the smallest molecules through the narrow (3.2 A by 4.3
A) pores.9 The selective porosity observed for 2 is attributed to
the superior hydrogen bonding properties of MeOH, enabling
it to overcome the frameworkꢁ ꢁ ꢁanion interactions and access
the interior pores. In combination the analytical data clearly
defines 2 as a new material produced by HCl introduction into
the open framework structure of 1. The detailed outcome
The amino acid-based open framework Ni(L-asp)bipy0.5, 1
(L-asp = L-aspartate, bipy = 4,40-dipyridyl), is an ideal
starting material for protonation studies, as it possesses per-
manent porosity, homochirality and excellent chemical and
Department of Chemistry, University of Liverpool, Liverpool, UK L69
7ZD. E-mail: m.j.rosseinsky@liv.ac.uk; Fax: +44 151 794 3598;
Tel: +44 151 794 3504
w Electronic supplementary information (ESI) available: Full experi-
mental information, data and crystallographic details for all com-
pounds. CCDC 662349–662350. See DOI: 10.1039/b718443c
ꢂc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 1287–1289 | 1287

View Article Online
corresponds to full proton transfer from HCl to a framework
aspartate carboxylate group (either a or b) forming a COOH
moiety that by its confinement within the pore wall is still tethered
to the metal centre. The chloride conjugate base is located within
the channels, demonstrated by a dramatic decrease in the porosity
of 2 (in comparison to 1). Framework 2 is a rare example of a
post-synthesis chemically modified MOF23 that can be shuttled,
reversibly, between protonated and non-protonated states with
no framework degradation. It was of interest to determine if the
acidic sites in 2 were of sufficient strength to be active as a
Brønsted acid catalyst. The test reaction examined was the
methanolysis of rac-propylene oxide (PO) expected to yield 2-
methoxy-1-propanol and 1-methoxy-2-propanol. Framework 2 is
active for this conversion (albeit with limited turnover numbers at
25 1C and 48 h),w with catalyst activity limited by its effectively
non-porous nature. With the low activity observed for 2, a related
material with larger pores was targeted to improve reactant
adsorption kinetics (and increase active site accessibility) with
the Clꢀ counterion no longer effectively blocking the channel.
Fig. 1 Left, COOH region of the IR spectra before and after
protonation of Cu(asp)bpe0.5. Right, Cu environment showing the
protonation position, b-carboxylate depicted protonated arbitrarily.
from a minor impurity due to the stoichiometric, 1 : 1 inclusion
of H+ per metal centre (by EDX) combined with the micro-
analysis.w The similarity in the pXRD patterns of 4 and 5w
confirms closely related framework connectivities, precluding
any cleavage of aspartate to metal bonding as observed in the
homogeneous phase.25–27 There is an unavoidable minor im-
purity in the synthesis of 5, that increases with any over-
equivalence of HCl used; we tentatively formulate this impurity
as Cu(bpe)Cl2 based on comparison to the Ni system, though it
does not correspond to the reported literature phase.28 The
impurity can be strictly kept to a minimum (o5%) through
the use of an exact stoichiometric equivalence of HCl and a
reaction time of 1 h. The porosity of 5 was confirmed by MeOH
uptake, whilst its porosity to propylene oxide (catalysis starting
material) was confirmed by TG and micro-analysis on a des-
olvated sample of 5 soaked in propylene oxide overnight.w 5
however is non porous to CO2 (ꢀ78 1C) frustrating attempts to
probe the acid strength of the protonated site via IR studies.20
The nCQO stretch of adsorbed acetone was also uninformative,
obscured in the carboxylate region of 5.29
A new family of materials closely related to 1, Cu(asp)L0.5
-
(guests) (L = bipy, 3, L = bpe, 4) can be readily accessed
phase pure by standard solvothermal methods.w The structure
(Fig. 2 shows the bpe congener 4) of both is similar to that
previously reported for 1, correspondingly consisting of neu-
tral chiral M(asp) layers bridged by the trans disposed bipyr-
idyl ligands. The major noteworthy distinction between the Ni
and Cu congeners is the Jahn–Teller distortion around the d9
octahedral Cu centre elongating the ‘axial’ Cu–O bonds. The
bond lengthening observed around the metal in 3 and 4
reduces the thermal stability (framework decomposition tem-
perature B190 1C in 3 and 4, compared with 4300 1C for 1).w
Chiral gas chromatography (GC) confirmed the enantiopurity
of the starting amino acid is delivered to the products 3 and 4
(495% ee) by this route when either D- or L-aspartate is used
as the starting material. Frameworks 3 and 4 are robust to
guest loss and are truly porous, 4 for example reversibly
adsorbs 6.5 wt% of methanol at 25 1C.w PLATON24 revealed
22.1% solvent accessible volume for 3 and 25.2% for 4
(compared to 23.1% for 1).9 Framework 4 is highly significant
to this study as it is an analogue of 1 but with larger pores. The
channel topology of 4 is defined by pore windows of 4.1 ꢃ
4.3 A, and elongated pores of 8.6 ꢃ 3.2 A, arising from the
increased interlayer separation of the bpe pillared framework
(compared to a continuous channel, 4.3 by 3.2 A in 1).
Protonation studies using HCl in Et2O under anhydrous
conditions revealed that 4 is considerably more sensitive to
acid stoichiometry and reaction time than the Ni analogue, 1,
consistent with a less robust phase. Careful control of reaction
time and stoichiometry led to the successful isolation of the
protonated material Cu(L-asp)bpe0.5(HCl)(H2O), 5 (H2O is ad-
sorbed from atmospheric moisture, confirmed by micro- and TG
analysis; observed 5.3% mass loss, predicted for 1 H2O 5.2%).w
The ratio of Cu : Cl obtained by EDX is constantly close to 1 : 1,
consistent with this formulation. Again IR spectroscopy is the
most useful diagnostic tool for identifying the position of
protonation, with a signal at 1720 cmꢀ1 observed in the proto-
nated framework 5 that is not present in the parent material 4
supporting the formation of a COOH moiety fully analogous to
the Ni system (Fig. 1). Again, the COOH group cannot arise
Catalysis studies performed using 5 for the methanolysis of
PO demonstrated a significant increase in turnover frequency
(compared to 2),w as anticipated on increasing the pore
dimensions. Catalysis utilising 5 is unambiguously heteroge-
neous (the filtered supernatant is inactive, precluding catalysis
by leeched HCl); catalysis is occurring in the pores, not just at
Fig. 2 Left, Cu(asp) 2D layer in 4 dashed red lines emphasise
Jahn–Teller elongated Cu–O bonds. Right, cross sectional view down
the a axis of 4 (solvent removed for clarity) displaying the 1D porous
channels between chiral Cu(asp) layers. Red = O, dark blue = N,
cyan = Cu, dark grey = C, light grey = H.
ꢂc
This journal is The Royal Society of Chemistry 2008
1288 | Chem. Commun., 2008, 1287–1289

View Article Online
Table 1 Catalysis results for methanolysis of cis-2,3-epoxybutane
metal centre, or protonation (and dissociation) of the NH2 site.25–27
It is only the confinement of the aspartate moiety in the extended
MOF structure that allows for the protonation of one of the two
carboxylate groups per aspartate ligand whilst anchoring the
resulting Brønsted acidic –COOH group to the metal centre. This
is essential as the strong acidic nature of 5 (required for catalysis)
arises only from the continual binding of the COOH to Cu,
increasing proton acid strength via the stabilisation of the conjugate
base through stronger coordination to the metal.
[Cat]a
Temperature/1C
Yield (%)b
eec
TOFd
5 (L-asp)
6 (D-asp)
5 (L-asp)
6 (D-asp)
H2SO4
a
25
25
0
0
25
59
65
30
32
+10
ꢀ6
4.8
4.7
2.6
2.7
—
+17
ꢀ13
+2
In conclusion, we report the rational, post-synthesis mod-
ification of a porous, homochiral MOF leading to a functional
Brønsted acidic material that is active as a uniquely hetero-
geneous (inaccessible in the solution phase) asymmetric cata-
lyst. Current work is targeting the functionalisation of chiral
materials with significantly larger pore windows, allowing for
increased catalytic reaction scope.
100
Catalysts are activated overnight in vacuo before suspension in 5 ml
b
dry MeOH under argon. Combined R,R and S,S produced after 48 h
c
from 10 mg of 5 (or 6) by GC. ees were assessed using chiral GC and
are fully reproducible; negative signs signify the alternate enantiomer
is in excess. TOF units, molꢀ1 dayꢀ1
d
.
Notes and references
the surface (the observed turnover from the attempted metha-
nolysis of a significantly bulkier epoxide (2,3-epoxypropyl)-
benzene was effectively zero).w The protonated material 5 is
the active catalyst with both the impurity phase, ‘Cu(bpe)Cl2’
and the parent material 4 inactive (in separate control experi-
ments). Framework 5 is returned essentially intact (by pXRD)
after catalysis, with IR spectroscopy confirming the persistence
of the COOH functionality. Furthermore, pore confined HCl is
not the catalyst with significantly different observed selectivities
(for homogeneous HCl compared to 5).w Racemic starting
materials (PO and trans-2,3-epoxybutane) produced effectively
racemic products, with no evidence for any enantioselective
reagent sorption.9 A racemisation step in the catalysis is pre-
cluded by a test reaction using chirally pure S-PO, producing
single enantiomer products.w
1 M. E. Davis and R. F. Lobo, Chem. Mater., 1992, 4, 756.
2 J. M. Newsham, M. M. J. Treacy, W. T. Koestier and C. B. de-
Gruyter, Proc. R. Soc. London, Ser. A, 1988, 420, 375.
3 S. Thushari, J. A. K. Cha, H. H.-Y. Sung, S. S.-Y. Chui, Andy L.-F.
Leung, Y.-F. Yen and I. D. Williams, Chem. Commun., 2005, 5515.
4 J. Weng, M. Hong, Q. Shi, R. Cao and A. S. C. Chan, Eur. J.
Inorg. Chem., 2002, 2553.
5 B.-Y. Lou, R.-H. Wang, D.-Q. Yuan, B.-L. Wu, F.-L. Jiang and
M.-C. Hong, Inorg. Chem. Commun., 2005, 8, 971.
6 E. V. Anokhina and A. J. Jacobson, J. Am. Chem. Soc., 2004, 126, 3044.
7 D. Bradshaw, T. J. Prior, E. J. Cussen, J. B. Claridge and M. J.
Rosseinsky, J. Am. Chem. Soc., 2004, 126, 6106.
8 N. G. Pschirer, D. M. Ciurtin, M. D. Smith, U. H. F. Bunz and H.-
C. zur-Loye, Angew. Chem., Int. Ed., 2002, 41, 583.
9 R. Vaidhyanathan, D. Bradshaw, J.-N. Rebilly, J. P. Barrio, J. A. Gould,
N. G. Berry and M. J. Rosseinsky, Angew. Chem., Int. Ed., 2006, 45, 6495.
10 E. V. Anokhina, Y. B. Go, Y. Lee, T. Vogt and A. J. Jacobson, J.
Am. Chem. Soc., 2006, 128, 9957.
11 D. N. Dybtsev, M. P. Yutkin, E. V. Peresypkina, A. V. Virovets, C.
Serre, G. Ferey and V. P. Fedin, Inorg. Chem., 2007, 46, 6843.
12 D. N. Dybtsev, A. L. Nuzhdin, H. Chun, K. P. Bryliakov, E. P.
Talsi, V. P. Fedin and K. Kim, Angew. Chem., Int. Ed., 2006, 45, 916.
13 J. S. Seo, D. Whang, H. Lee, S. I. Jun, J. Oh, Y. J. Jeon and K.
Kim, Nature, 2000, 404, 982.
We next examined if any enantioselectivity is observed
during the methanolysis of the meso compound, cis-2,3-epoxy-
butane. This would arise from a different mechanism to the
selective sorption of chiral reagents. Indeed, the methanolysis of
cis-2,3-epoxybutane (Table 1) to give (R,R)- and (S,S)-3-meth-
oxybutan-2-ol catalysed by 5 at 25 1C proceeded with modest ee.
This increased when the catalysis was repeated at 0 1C (to 17%).
Importantly, methanolysis catalysed by a framework con-
structed with D-aspartate, Cu(D-asp)bpe0.5(HCl)(H2O), 6, re-
sulted in the opposite enantiomer dominating as expected.
Methanolysis catalysed by sulfuric acid30 (and HCl)w resulted
in an effectively racemic mixture (o3% ee) of products. Whilst
the ee generated is modest (a maximum of 17%) it is a rare
example of enantioselective catalysis by a porous MOF material.
The catalytic reactions chosen were enforced by the limited pore
size inherent in these materials (imposing a starting material size
restriction); we acknowledge the excellent homogeneous cata-
lysts available for these reactions, but that does not diminish
their validity as proof of Brønsted acid strength of the synthe-
tically generated acid centre in these frameworks.
14 S.-H. Cho, B. Ma, S.-B. T. Nguyen, J. T. Hupp and T. E. Albrecht-
Schmitt, Chem. Commun., 2006, 2563.
15 S. Hasegawa, S. Horike, R. Matsuda, S. Furukawa, K. Mochizuki,
Y. Kinoshita and S. Kitagawa, J. Am. Chem. Soc., 2007, 129, 2607.
16 C.-D. Wu, A. Hu and W. Lin, J. Am. Chem. Soc., 2005, 127, 8940.
17 O. R. Evans, H. L. Ngo and W. Lin, J. Am. Chem. Soc., 2001, 123, 10395.
18 P. McMorn and G. J. Hutchings, Chem. Soc. Rev., 2004, 33, 108.
19 N. Masciocchi, P. Cairati, L. Carlucci, G. Mezza, G. Ciani and A.
Sironi, J. Chem. Soc., Dalton Trans., 1996, 2739.
20 A. Vimont, J.-M. Goupil, J.-C. Lavalley, M. Daturi, S. Surble, C.
Serre, F. Millange, G. Ferey and N. Audebrand, J. Am. Chem.
Soc., 2006, 128, 3218.
21 D. H. Williams and I. Fleming, Spectroscopic Methods in Organic
Chemistry, McGraw-HillPublishing Company, Maidenhead, 1995.
22 A. Bondi, J. Phys. Chem., 1964, 68, 441.
23 Z. Wang and S. M. Cohen, J. Am. Chem. Soc., 2007, 129, 12368.
24 A. L. Spek, J. Appl. Crystallogr., 2003, 36, 7.
25 H. Schmidbaur, I. Bach, J. Riede, G. Muller, J. Helbig and G.
Hopf, Chem. Ber., 1987, 120, 867.
26 R. Barfod, L. Bendahl, A. Hammershoi, D. Kjaergaard-Jensen, A. M.
Sargeson and A. C. Willis, J. Chem. Soc., Dalton Trans., 1999, 449.
27 L.-P. Lu, M.-L. Zhu and P. Yang, Acta Crystallogr., Sect. C:
Cryst. Struct. Commun., 2004, 60, m21.
28 Z.-G. Li, J.-W. Xu, H.-Q. Jia and N.-H. Hu, Acta Crystallogr.,
Sect. C: Cryst. Struct. Commun., 2006, 62, m205.
29 A. Panov and J. J. Fripiat, Langmuir, 1998, 14, 3788.
30 G. K. Helmkamp and H. J. Lucas, J. Am. Chem. Soc., 1952, 70, 951.
The protonated frameworks 2, 5 and 6 are MOFs that have
been chemically transformed to introduce a novel, catalytically
active site, a distinct process to the heterogenisation of an existing
homogeneous catalyst. It is important to re-emphasise that these
catalysts are inaccessible in the homogeneous phase where proto-
nation leads to either dissociation of the COOH group from the
ꢂc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 1287–1289 | 1289