Bimolecular catalysis and turnover from a macromolecular host system

Article abstract of DOI:10.1021/jo400532r

The synthesis of a globular macromolecule and its application as a bimolecular catalyst are reported. The macromolecular structure supports (at least) two zinc-metalated porphyrin units, each capable of binding a single reactant. The proximity of the two

Full text of DOI:10.1021/jo400532r

Article
pubs.acs.org/joc
Bimolecular Catalysis and Turnover from a Macromolecular Host
System
Adam Ellis, David Gooch, and Lance J Twyman*
Department of Chemistry, The University of Sheffield, Brook Hill, Sheffield S3 7HF, U.K.
ABSTRACT: The synthesis of a globular macromolecule and
its application as a bimolecular catalyst are reported. The
macromolecular structure supports (at least) two zinc-
metalated porphyrin units, each capable of binding a single
reactant. The proximity of the two bound reactants results in
an increased local concentration, leading to a maximum 300-
fold increase in the reaction rate. In contrast to other synthetic
catalysts, where bidentate products inhibit further reactions,
this macromolecular system allows the product to be displaced
by the reactants leading to turnover and catalysis. We believe
that this is due to the dynamics of the macromolecular host system, which maintains enough flexibility to adopt a favorable/
reactive geometry, which allows the reactants to get close and react while possessing sufficient rigidity/poor geometry to reduce
and disrupt any cooperative/inhibitive bidentate binding.
structure of polymers can be tailored allowing control of the
steric and electronic environment within their internal spaces
and around any specific binding sites.10,11 Initially, we
considered using linear polymers, but this idea was quickly
rejected. Although linear polymers possess the desired globular
shape when in solution, their structures and conformations are
dynamic. This results in the constant movement of any binding
sites. Consequently, the relative positions of the binding sites
would be impossible to control and the chance of any two
being close enough to enable bound guests to react would be
very low. In addition, it is unlikely that there would be enough
“strain” in these systems to displace any bidentate product.
Dendrimers12 are structurally more rigid and have fixed
globular shape (dependent on generation)13 and are capable
of being exploited in a number of interesting and exotic
catalytic systems.14 As such, it is theoretically possible to
position and orient binding sites in a controlled manner.
However, constructing such a molecule represents a consid-
erable synthetic challenge, which does not represent an
improvement over the cavity based systems described above.
Therefore, we decided to use a polymer system whose structure
was intermediate between the overly flexible linear polymers
and the relatively rigid dendrimers. These polymers are known
as hyperbranched polymers (HBPs).15 Unlike dendrimers,
HBPs can be synthesized in one step using a branching
monomer. Due to the random nature of the polymerization,
not every branching point is reacted and imperfections occur
throughout the structure. These imperfections reduce the
degree of branching,16 which restricts internal packing. As a
result the ensuing structures have a relatively fixed globular
structure, but remain relatively open and flexible inside.
INTRODUCTION
■
The ability to design and implement hosts capable of binding
two substrates and catalyzing a bimolecular reaction represents
a considerable challenge.1 Although there have been a number
of synthetic and biomimetic attempts over the years, they are
far from perfect and demonstrate various problems limiting
their use. For instance, antibody and ribosome approaches can
exhibit poor catalytic activity and are generally expensive and
time-consuming to produce.2,3 In an attempt to overcome these
problems, synthetic hosts have been developed, and these fall
into two general areas. The first involves the use of cavities
designed to encapsulate two reactive species, resulting in a
faster reaction by virtue of an increase in local concen-
trations.4−6 A second approach uses molecules that possess
specific binding sites designed to bind both reactants in close
proximity, again allowing them to react faster. Examples include
the macrocyclic hosts designed by Sanders,7 which could
accelerate the reaction of two bound reagents to form a
bidentate product. However, the binding and reactive processes
are soon inhibited by the bidentate nature of the products,
which bind much more strongly than either set of monodentate
starting materials (when catalytic quantities of hosts were
used). In principle, this problem could be overcome by using a
smart host system capable of reversibly changing its structure
and then displacing any bound dimeric product (i.e., when
heated up or light is applied).8 Unfortunately, even at relatively
low product concentrations, the bidentate product would still
bind to the host (and inhibit catalysis) when it returned to its
reactive geometry.
To overcome this problem, we proposed the use of a
polymer system whose structure retained a certain amount of
strain and limited flexibility. An additional advantage in using
macromolecules is that their globular structures can be tailored
to mimic those of proteins and enzymes.9 That is, the bulk
Received: March 13, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jo400532r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Therefore, even at relatively high molecular weights, it is less
likely that any binding sites will become prohibitively hindered
(compared to dendrimers). As such we decided to focus our
attention on the use of functionalized hyperbranched polymers.
Our next decision regarded the choice of binding group/
interaction. We have previously reported the facile synthesis of
a porphyrin-cored hyperbranched polymer and its ability to
catalyze a (monomeric) oxidation reaction.17 In this case, the
size of the polymer did not hinder or prevent substrates binding
to the porphyrin. In fact, subsequent binding studies revealed
that as the molecular weight of the HBP increased, then so did
substrate binding affinity.18 Encouraged by these results we
decided to incorporate a number of zinc-metalated porphyrin
binding sites within our HBP (at least two binding sites). These
porphyrin units would then be able to bind reactive groups
attached to pyridine ligands through Zn−N coordination.
interaction, we focused our attention on the synthesis of the
porphyrin functionalized HBP capable of coordinating the
pyridine units of substrates 1 and 2.
The exact polymer was a variant of a hyperbranched
polyester based on the branching monomer, 3,5-diacetox-
ybenzoic acid.19 We have previously used this method to
construct a variety of functionalized HBPs, including a polymer
with a porphyrin center20 and another with an activated/
reactive core group.21 We have also shown that copolymers can
be constructed if simple carboxylic acid containing como-
nomers are added.22 This method allows a number of
functionalized units to be incorporated throughout the
structure of a HBP. Our target molecule can therefore be
obtained if the branching monomer 3,5-diacetoxybenzoic acid
is copolymerized with a monofunctionalized acid porphyrin.
The carboxylic acid porphyrin 4 was synthesized using a two
step-procedure, Scheme 2. The first step involved the formation
of a dipyrromethane 5 using standard conditions. This was then
condensed under statistical conditions with 0.5 equiv each of
benzaldeyde and 4-carboxybenzaldehyde. The monofunction-
alized porphyrin 4 was obtained after purification by column
chromatography. 1H NMR confirmed the structure and
monofuntionalized nature of the porphyrin. The doublets
from the carboxylic acid containing aromatic resonated at 8.45
and 8.35 ppm, and the integration ratio confirmed that only
one carboxylic acid was present. In addition, mass spectroscopy
indicated a molecular ion at 659 Da (MH+).
RESULTS AND DISCUSSION
■
To test the (bimolecular) catalytic activity of the proposed
HBPs, a relatively simple esterification was selected; specifically,
the reaction between an alcohol and an activated ester.
Esterifications proceed through a tetrahedral intermediate that
possess a very different structure to the trigonal planar
products. This change in geometry, coupled with some host
flexibility, should encourage a situation that favors product
release, turnover and catalysis. The particular reaction used in
our study is shown in Scheme 1. Both reactants contain a
The synthetic procedure for the construction of the
porphyrin containing HBP is shown in Scheme 3 and involved
reacting the 3,5-diacetoxybenzoic acid and the porphyrin 4 in a
1:20 ratio. The polymer obtained after initial precipitation was
purified by preparative size-exclusion chromatography (SEC)
Scheme 1. Bimolecular Reaction Used To Test the Catalytic
Host System. Reaction Followed by Monitoring the Relative
Intensities of Resonance Signals for Protons Ha and Hb (1H
NMR)
1
using a biobead column. H NMR spectra of the purified
polymer 6 indicated the presence of shielded protons around
minus 2.00 ppm. This is a characteristic resonance for the NH
protons at the center of an aromatic macrocycle, thus
confirming the presence of the porphyrin units. Other peaks
corresponding to the porphyrin’s pyrrole group could also be
seen at 8.91 ppm. Although the remaining aromatic peaks from
the tetraphenyl units are hidden beneath the aromatic peaks of
the polymer, integration of the aromatic region confirms that
additional protons were present (when compared to the
polymers acetate peak). The remainder of the spectrum was
consistent with that obtained for similar HBPs based on the
same monomer.23 The dark brownish/red polymer 6 had a
molecular weight of just over 6000 Da (Mn) as determined by
pyridine, which is required for porphyrin binding. Substrate 1 is
a simple alcohol, whereas substrate 2 is an activated ester of
nicotinic acid. These react giving the bis-pyridine product 3 and
4-nitrophenol. Having chosen our model system and binding
Scheme 2. Two-Step Procedure for the Synthesis of the Monoacid Porphyrin
B
dx.doi.org/10.1021/jo400532r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
a
Scheme 3. Synthesis of Free Base Porphyrin Cored HBP 6 and Zinc Metalloporphyrin Cored HBP 7
a
The purpose of heating to 180 °C at 5 mmHg was to remove the acetic acid byproduct and drive the equilibrium toward product.
analytical SEC. Furthermore, SEC detection using RI and UV
(set to the porphyrins λmax) produced overlapping traces;
confirming that the porphyrin was present across the polymers
complete molecular weight range. Spectroscopic analysis
determined that on average approximately two porphyrins
were incorporated per polymer molecule (exact calculation
indicated 2.4 porphyrins per polymer based on a molecular
weight of 6000). This was calculated using known amounts of
polymer 6 and comparing its porphyrin UV absorption to that
obtained from a Beer−Lambert analysis of the simple
C
dx.doi.org/10.1021/jo400532r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
porphyrin 4. Although this technique assumes that all polymers
in the poly disperse mix are the same, it does represent the
minimum number of porphyrins present in the polymer. As
such, any errors related to molecular weight are in our favor
with respect to porphyrin incorporation. This is a direct
consequence of the calibration characteristics of SEC, which are
known to underestimate the Mn values for globular branched
polymers.24 As a result, any porphyrin calculation based on
polymer molecular weight (and therefore concentration) will
always underestimate the number of porphyrins per polymer.
To help explain, let us consider an equal mass of two porphyrin
polymers with differing Mn values, the first with an Mn of 1000
Da and a second of Mn 2000 Da. At any given porphyrin
absorption (and therefore concentration) each of the two
polymer solutions would have exactly the same number of
porphyrins. However, as there are exactly half the numbers of
polymer molecules in the larger polymer fraction (for a given
mass), then the same number of porphyrins would be spread
over fewer molecules. That is, the larger polymers possess more
porphyrins. Therefore, if we acknowledge that molecular weight
measurements are underestimated via SEC, then we can
conclude that our polymers are larger than indicated via SEC
and as a consequence, porphyrin loadings are higher than the
two per polymer previously calculated.
The intended interaction between ligand and porphyrin
involves a metal to pyridine coordination, which first requires
insertion of a zinc atom to the porphyrin center. This was
achieved by gently warming polymer 6 in a solution of excess
zinc acetate in dichloromethane. After purification, metalated
polymer 7 was isolated in quantitative yield. SEC and UV
analysis confirmed that the metalation process had been
successful and that the reaction conditions had not damaged or
changed the polymeric structure. That is, the number of Q
bands in the UV spectra had been reduced from four to two
and the data obtained from SEC was unchanged after
metalation. Having successfully synthesized a globular polymer
possessing (at least) two internal porphyrin units we then set
about ascertaining its binding affinity to reactants 1, 2, and
product 3 (Table 1).25 Specifically, we wanted to determine
association constants were the same order of magnitude, it was
likely that the difference in binding would not be high enough
to inhibit reasonable turnover. For comparison, in earlier work
by Sanders the difference in reactant and product binding was
1000 fold.6 In this case turnover was completely inhibited and
stoichiometric quantities of host were required. Therefore, our
binding analysis clearly demonstrate that our system is capable
of binding either reactant well, but its conformation is not
predisposed to recognize or bind (strongly) the product. As
such there was a strong probability that a catalytic bimolecular
reaction could take place.
The catalytic experiments were carried out with polymer 7
using an amount that corresponded to a porphyrin concen-
tration of 5 mol % with respect to the reactants 1 and 2.
Specifically, reactions were carried out such that the final
concentrations were 0.2 M in reactants 1 and 2 and 0.01 M in
porphyrin. As UV analysis had already indicated that on average
2.4 porphyrins were present in each polymer, a polymer
concentration of 0.042 M was used for the catalytic experiments
(i.e., equivalent to a porphyrin concentration of 0.01 M or
5%).26 The reaction was followed using 1H NMR and
monitoring the signals from the product’s benzyl protons Hb,
which resonate at 5.45 ppm (compared with 4.75 for Ha in the
starting material), Scheme 1. The yield was calculated by
comparing the ratio of these signal intensities with respect to
initial concentration. The yield was then plotted against time
and the graph obtained is shown in Figure 1. The uncatalyzed
Table 1. Binding Affinities for Substrates and Products to
Zinc Metalloporphyrin Cored HBP 7
compd
binding affinity to 7 Ka/M−1
substrate 1
substrate 2
product 3
1.55 × 10−4
1.60 × 10−4
5.80 × 10−4
Figure 1. Plots showing yield of ester 3 over time in the absence of
catalyst and in the presence of HBP 7. A control reaction using Zn-
tetraphenylporphyrin is also shown.
whether or not the product would bind with sufficient
cooperatively to inhibit the reaction. This would indicate
whether or not our theory regarding polymer mobility and lack
of product inhibition, was possible. Binding constants (Ka)
were assessed by UV titration measurements, and changes in
porphyrin λmax absorption with respect to increased pyridine
concentration were fitted to a 1:1 binding analysis. The
measurements and analysis were repeated four times. The
product 3 bound with a Ka of 5.80 × 104 M−1, which is around
4 times higher than the starting materials, which possessed Ka
values of 1.55 × 104 M−1 and 1.60 × 104 M−1 for 1 and 2,
respectively. Although the product bound more tightly than the
reactants, it is clearly not binding with an efficient cooperative
effect. Therefore, the binding data confirms that the product is
not binding as an efficient bidentate ligand. Furthermore, as the
control reaction was carried out using just the reactants 1 and 2
and the same reaction conditions as described (i.e., without any
porphyrin or polymeric catalyst).
The results show that the uncatalyzed reaction was very slow,
with only 10% of product being detected after nearly 100 h.
However, when the porphyrin-containing polymer was added
the reaction proceeded to almost 80% after the same 100-h
period (the concentration of porphyrin was identical in both
cases).26 The initial rates were 1.08 × 10−5 M−1 and 1.92 ×
10−8 M−1 for the catalyzed and uncatalyzed reactions,
respectively, which corresponds to around a 60-fold increase
in rate. Therefore, the rate increase of the polymer-catalyzed
reaction can be attributed directly to the polymer structure,
which provides a framework capable of binding/supporting
D
dx.doi.org/10.1021/jo400532r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
specific binding groups/ligands in close proximity (as shown
schematically in Figure 2). Although relatively flexible, the
some rate enhancement. Specifically, when an equivalent
amount of Zn−tetraphenylporphyrin was added to the reaction
between 1 and 2, an initial rate of 2.94 × 10−8 M−1 was
observed. This corresponds to a 2-fold increase in rate when
compared to the unanalyzed reaction. This small increase is
likely due to a simple Lewis acid catalysis where coordination to
the carbonyl can increase the reactivity of the carbonyl carbon
on reactant 1, and/or stabilize the charged tetrahedral
intermediate (leading to a lower Ea and faster reaction).27
Alternatively, coordination to the pyridine unit of 1 can
enhance the electrophillicity of the carbonyl carbon (via
delocalized electron withdrawal). Following on from this,
another control experiment was carried out using similar
substrates to 1 and 2, but exchanging the pyridine moiety for
benzene on one of the substrates. Removal of either pyridine
caused a significant drop in the rate of the reaction, with both
reactions only reaching 10−15% completion after 100 h (giving
an initial rate between 1.00 × 10−8 M−1 and 1.50 × 10−8 M−1,
which corresponds to a less than 2-fold increase in rate). This is
roughly the same as the uncatalyzed reaction and confirms that
catalysis requires both ligands to bind to the host polymer 7. A
final control was performed using the catalytic host polymer 7,
the pyridine substrates (1 and 2) and an excess of pyridine. In
this case the reaction was significantly less efficient, only
reaching about 20% completion after 100 h. This reduction is
due to the fact that the excess pyridine competes for the
porphyrin binding sites and effectively inhibits substrate
binding, which slows the reaction. The fact that the reaction
reaches 20% is due to the basic nature of the pyridine, which
can help deprotonate the tetrahedral intermediate, resulting in a
slightly faster reaction.
Figure 2. Schematic showing starting materials 1 and 2 bound
productively within the macromolecular host 7.
polymer remains rigid enough to maintain an average distance/
geometry between the porphyrins, such that reactions can take
place. After reaction, the porphyrin-containing polymer has
enough free motion (strain) to break or prevent any strong
chelate or bidentate interactions. Alternatively, it could be
argued that the binding sites are not optimally arranged to bind
the product. Thus, although the geometry does not prevent the
starting materials 1 and 2 binding, the polymer is not able to
bind product 3 with any efficiency, leading to weak or
incomplete binding to one of the porphyrin binding sites.
Either way, if bidentate binding is broken or prevented, product
3 is only bound through a single pyridine/porphyrin
interaction. Therefore, any unbound reactants 1 or 2 can easily
compete for binding and displace the product 3. This effect is
greatest at the start of the reaction, where the concentration of
reactants is much higher than the product. However, as the
reaction proceeds the concentration of product steadily
increases. When the reaction has reached 50% we have an
equimolar concentration of product 3 and reactants 1 and 2.
However, as this is a bimolecular reaction, the concentration of
reactants will be twice that of the product and the catalytic host
bound reactions can still proceed. It is only after the process has
reached 66% that we have the same concentration of products
and reactants (i.e., the concentration of 1 and 2 combined equals
that of product 3). At this stage it will become much harder for
the reactants to compete with the product for binding. This
effect is substantiated by the fact that the reaction does not go
to completion, only reaching around 70% after 100 h. This
confirms that the product is beginning to inhibit catalysis at this
point (although only as a monodentate ligand).28
CONCLUSIONS
■
The rate acceleration obtained using our hyperbranched
polymer was not as high as those observed for rigid host
systems, which are predisposed to bind intermediate transition
states and products. Nevertheless, even though our polymeric
host only displayed a modest rate enhancement, turnover was
possible and the system was catalytic. The reaction yield
plateaus around 70% after 100 h. At this point we have an
eqimolar solution of reactants and products and the product
can now compete with starting materials for the binding sites,29
and catalysis slows down. In addition, as reactions can only take
place when reactants 1 and 2 bind in a “productive” manner,
some of the binding events are wasted (i.e., the host binding the
same reactant twice or the reactant binding to the wrong face of
the porhyrin). The productive binding required for a successful
reaction cannot be controlled and is a statistical phenomenon.
Therefore, the rate of reaction for reactants 1 and 2 bound in a
productive way is much higher. If we use the porphyrin to
polymer ratio that the data suggests and we apply the
“productive-pair” argument put forward by Sanders,6 then
there are only a limited number of binding events that can lead
to catalysis. That is, the catalytic reaction can only take place
when reactants 1 and 2 are bound to the correct face on each of
the two porphyrins. As such, statistical analysis indicates that
productive binding can only take place 20% of the time.
Therefore, the polymer-catalyzed reaction between 1 and 2
bound in a product manner, corresponds to a 300-fold increase
in rate.
A number of control reactions were also carried out. The first
involved the use of the nonmetalated porphyrin host. The
reaction was repeated using the free base HBP 5, in place of the
catalytic HBP 6. On this occasion no enhancement in rate was
observed. A further control reaction using 5 mol % of the zinc-
metalated tetraphenylporphyrin was also carried out to probe
any effect of porphyrin on the yield and rate of reaction (i.e., no
polymer backbone). Interestingly, this control reaction led to
E
dx.doi.org/10.1021/jo400532r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
warmed up and turned pink before turning black. The solvent was
concentrated on a rotary evaporator to leave around 100 mL of
solvent, which was then washed with saturated sodium hydrogen
carbonate solution (5 × 500 mL). The DCM and was then removed
on the rotary evaporator. The resulting black solid was recrystallized
from 1:1 water and methanol. The resulting pale yellow solid then
collected by filtration and washed using a 1:1 cold solution of water
and methanol: yield 3.41g, (33%); yellow solid; 1H NMR ppm
(CDCl3) 7.91(b, s, 2H), 7.30(m, 5H), 6.70(s, 2H), 6.20(s, 2H),
5.95(s, 2H), 5.50(s, 1H); 13C NMR ppm (CDCl3) 143.4, 132.6, 128.7,
128.5, 127.0, 117.4, 108.4, 107.3, 44.0; MS(EI) 221 Da/MH+,
(C15H14N2 = 222 Da); FT-IR (ν/cm−1) 3338, 1737, 1553, 1454,
1366; mp 101−103 °C (lit.4 mp 102.0−102.5 °C).
EXPERIMENTAL SECTION
■
General Experimental Conditions and Equipment. 1H and
13C NMR spectra were acquired at 250 and 62.5 MHz, respectively,
using a 5 mm CH probe. Chemical shifts are quoted in ppm relative to
residual CHCl3, and J values are quoted in Hz. UV/vis was performed
in wavelength mode standardized using a user baseline configuration.
Molar extension coefficients are given in M2/mol. FT-IR transmissions
maxima are given in cm−1. Electrospray ionization mass spectrometry
(ESI-MS) and matrix-assisted laser desorption (MALDI) mass
spectrometry were carried out using a time-of-flight detector.
Dithranol or dihydroxy benzoic acid were used as the matrices in
the MALDI mass spectrometer. Analytical GPC samples were run at
room temperature on either a low or high molecular weight column
(calibrated with polystryrene samples with Mn values ranging from 500
to 100000 Da) and a 1 mL/min flow rate. All samples were run using
GPC THF and samples were prepared in the same grade THF and
spiked with toluene as a flow rate marker. Eluent product
concentration was monitored by an refractive index detector or in
conjunction with a UV/vis LC spectrophotometer. Melting points
were recorded using open-ended capillary tubes. Flash chromatog-
Step 2: 5-(4-Carboxyphenyl)-10,15,20-triphenylporphyrin 4.32 2-
(Phenyl(1H-pyrrol-2-yl)methyl)-1H-pyrrole 5 (1.50 g, 6.80 mmol),
benzaldehyde (0.70 g, 6.70 mmol), 4-carboxybenzaldehyde (1.01g 6.70
mmol), and DCM (1500 mL) were added to a 2 L round-bottom flask.
The round-bottom was put under a nitrogen atmosphere and
protected from the light using aluminum foil. To the mixture
borontrifluoride (200 uL, 3.20 mmol) was added. The reaction was
stirred at room temperature for 1 h, during this time reaction turned
pink. After 1 h, DDQ (5.00 g, 22.00 mmol) was added, and the
reaction was stirred at room temperature for a further hour. To the
mixture was added 15.00 g of silica and the solvent removed by rotary
evaporator. The residue was dry loaded onto a 30 cm diameter silica
column. The column was initially eluted with DCM, and a purple
fraction was recovered and analyzed and found to be tetraphenylpor-
phyrin. The column was then eluted with 2% methanol in DCM, and a
purple fraction was collected. Analysis confirmed that it was 5-(4-
carboxyphenyl)-10,15,20-triphenylporphyrin. On some occasions the
product was not completely pure; when this occurred the
chromatography was repeated until the sample was pure: yield 150
́
raphy was performed using flash silica (35−70 μm particles with 60 Å
pore size) and eluted under an applied pressure supplied from a
standard bellows system. Preparative GPC was carried out using
Biobeads (available from Radleys) and eluted under gravity.30
Synthesis. 4-Nitrophenyl Isonicotinate 2.31 Isonicotinoyl chloride
(5.00 g, 36.00 mmol), 4-nitrophenol (5.00 g, 36.00 mmol), and THF
(100 mL) were added to a round-bottom flask. The mixture was
brought to reflux (65 °C), and once all the starting materials were
completely dissolved triethylamine (3.60 g, 36.00 mmol) was added
via syringe and the reaction was refluxed for 24 h. The THF was
removed on the rotary evaporator and the crude mixture was dissolved
in DCM (100 mL). The DCM layer was then washed with a saturated
sodium hydrogen carbonate solution (3 × 300 mL). The DCM layer
was collected and dried with magnesium sulfate before filtering and
rotary evaporation, to reveal a very pale yellow solid: yield 7.30 g
(80%); pale yellow solid; 1H NMR ppm (CDCl3) 8.95 (d, 2H, J = 6.0
Hz), 8.40(d, 2H, J = 9.0 Hz), 8.05(d, 2H, J = 6.0 Hz), 7.45(d, 2H, J =
9.0 Hz); 13C NMR ppm (CDCl3) 162.3, 155.0, 151.0, 146.4, 135.2,
125.4, 123.2, 122.5; FT-IR (ν/cm−1) 3024, 1738, 1520, 1486; MS(EI)
245 Da/MH+, (C12H8N2O4 = 244 Da); mp 136−137.5 °C (lit.5 137−
138 °C). Anal. Calcd for C12H8N2O4: C, 59.02; H, 3.30; N, 11.47.
Found: C, 59.04; H, 3.21; N, 11.42.
1
mg (3.0%); purple solid; H NMR-ppm (CDCl3): 8.95(s, 8H, 8.45(d,
2H, J = 7.0), 8.35(d, 2H, J = 7.0), 8.25(d, 6H, J = 6.5), 7.9(m, 9H);
FT-IR (ν/cm−1) 1687, 1605, 1557, 1471; UV/vis absorbance(CH2Cl2)
λmax 419 nm (ε 181100); MS(EI) 659 Da/MH+, (C45H30N4O2 = 659
Da). Anal. Calcd for C45H30N4O2: C, 82.05; H, 4.59; N, 8.51. Found:
C, 82.03; H, 4.25; N, 8.28.
3,5 Diacetoxybenzoic Acid.34 3,5-Dihydroxybenzoic acid (77.00 g,
0.50 mol) was added to excess acetic anhydride (250 mL, 2.65 mol),
and the mixture was refluxed (150 °C) for 5 h. Over this time the
mixture turned gradually yellow. After heating, the excess acetic
anhydride and acetic acid byproducts were removed by vacuum
distillation, being careful to keep the temperature below 100 °C (to
prevent premature polymerization). After distillation, a white solid
remained. The crude product was dissolved in hot chloroform 60 °C
(200 mL) and then filtered. Petroleum ether 60−80 (70 mL) was then
added causing a white solid to precipitate. The mixture was then left
overnight to allow further precipitation. The white solid was collected
(Pyridin-4-yl)methyl Isonicotinate 3. 4-Nitrophenyl isonicotinate 2
(330 mg, 1.40 mmol), (pyridin-4-yl)methanol 1 (150 mg, 1.40 mmol),
and chloroform (100 mL) were added to a round-bottom flask fitted
with a condenser and heated to reflux (68 °C). Triethylamine (20 mL,
excess) was the added via syringe. The reaction mixture was stirred for
10 days before being cooled to room temperature. The mixture was
washed with saturated sodium hydrogen carbonate solution (5 × 300
mL). The DCM layer was collected and the solvent removed: yield
0.26 g (88%); white solid; 1H NMR ppm (CDCl3): 8.85(d, 2H, J = 6.0
Hz), 8.67(d, 2H, J = 6.0 Hz), 7.95(d, 2H, J = 6.0 Hz), 7.35(d, 2H, J =
6.0 Hz), 5.42(s, 2H); 13C NMR ppm (CDCl3):164.7, 150.8, 150.1,
144.3, 134.3, 122.9, 122.0, 65.4; MS(EI) 215 Da/MH+, (C12H10N2O2
= 214 Da); FT-IR (ν/cm−1) 1735, 1593, 1443; mp 92.5−93.5 °C.
Anal. Calcd for C12H10N2O2: C, 67.28; H, 4.71; N, 13.08. Found: C,
67.52; H, 4.81; N, 12.92.
1
by filtration and dried under vacuum: yield 74.3 g (62%); H NMR
ppm (CDCl3) 7.75 (d, 2H, J = 2.0), 7.25 (t, 1H, J = 2.0), 2.30 (s, 6H);
13C NMR ppm (CDCl3): 170.3, 168.9, 151.0, 131.4, 121.1, 120.9, 20.0;
FT-IR (ν/cm−1) 2944, 1763, 1688, 1594; MS(EI) 239-MH+
(C11H10N4O6 = 238 Da); mp 158−160 °C (lit.1 mp 157−159 °C).
Anal. Calcd for C11H10O6: C, 55.47; H, 4.23. Found: C, 55.29; H, 4.19.
Multi Zn-Porphyrin Hyperbranched Polymer 7. 5-(4-Carboxy-
phenyl)-10,15,20-triphenylporphyrin 4 (200 mg, 0.30 mmol), 3,5
diacetoxybenzoic acid (1.00 g, 4.10 mmol), and diphenyl ether (5.00
g) were heated in a 50 mL round-bottom flask to 225 °C for 1 h. The
temperature was then reduced to 180 °C, and the reaction was fitted
with a one piece distillation kit and put under high vacuum for 2 h.
During this time acetic acid could be seen to distill off from the
reaction vessel. The reaction was then allowed to cool, but while still
warm the crude mixture was dissolved in 20 mL of THF and then
precipitated into 500 mL of ice-cold methanol. After filtering, a brown
precipitate was collected and then washed with cold methanol. To
remove unincorporated porphyrin the crude product were loaded to a
15 mm diameter biobead column (preparative SEC), which was eluted
with DCM (to avoid overloading the column, the chromatography was
run a number of times with a maximum loading of 200 mg crude
5-(4-Carboxyphenyl)-10,15,20-triphenylporphyrin 4.32 A two-step
procedure was used to synthesis the required monofunctionalized
porphyrin. The first step was the preparation of a dipyrromethane
intermediate, 2-(phenyl(1H-pyrrol-2-yl)methyl)-1H-pyrrole. The sec-
ond step involved a mixed condensation of this dipyrromethane and
two aldehydes.
Step 1: 2-(Phenyl(1H-pyrrol-2-yl)methyl)-1H-pyrrole 5.33 Benzal-
dehyde (5.00 g, 47.20 mmol), pyrrole (20 mL, 289 mmol), and DCM
(500 mL) were added to a round-bottom flask which was protected
from the light by aluminum foil. To the mixture was added
trifluoroacetic acid (200 mg, 2.60 mmol). The reaction was stirred
at room temperature for 30 min, during which time the reaction
F
dx.doi.org/10.1021/jo400532r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
product). The collected fractions of free base multiporphyrin polymer
6 were redisolved in DCM (25 mL) and an excess of zinc acetate
added (1.00 g). The resulting solution was then gently warmed for 30
min before being filtered (to remove excess zinc acetate) and
concentrated on a rotary evaporator. The red/brown solid was then
dissolved in the minimum amount of THF and the final product
obtained via trituation into methanol. The solid was collected and
dried under vacuum. The crude product was then purified using
(5) Chen, J.; Korner, S.; Craig, S. J.; Rudkevich, D. M.; Rebek, J., Jr.
Nature 2002, 415, 386.
̈
(6) Lee, S. J.; Cho, S. H.; Mulfort, K. L.; Tiede, D. M.; Hupp, J. T.;
Nguyen, S. B. T. J. Am. Chem. Soc. 2008, 130 (50), 16829.
(7) (a) Walter, C. J.; Anderson, H. L.; Sanders, J. K. M. Chem.
Commun. 1993, 458. (b) Walter, C. J.; Sanders, J. K. M. Angew. Chem.,
Int. Ed. Engl. 1995, 34 (2), 217. (c) Anderson, H. L.; Sanders, J. K. M.
J. Chem. Soc., Perkin Trans. 1 1995, 18, 2223. (d) Anderson, H. L.;
Anderson, S.; Sanders, J. K. M. J. Chem. Soc., Perkin Trans. 1 1995,
No. 18, 2231. (e) Reinhoudt, D. N.; Crego-Calama, M. Science 2002,
295, 2403.
(8) (a) Hartley, G. S. Nature 1937, 140, 281. (b) Zimmermann, G.;
Chow, L. Y.; Paik, U. J. J. Am. Chem. Soc. 1958, 80, 3528.
(9) (a) Twyman, L. J.; Ge, Y. Chem. Commun. 2006, 15, 1658.
(b) Chiba, F.; Hu, T. C.; Twyman, L. J.; Wagstaff, M. Chem. Commun.
2008, 36, 4351. (c) Dandliker, P. J.; Diederich, F.; Zingg, A.; Gissel-
brecht, J.-P.; Gross, M.; Louati, A.; Sanford, E. Helv. Chim. Acta 1997,
80 (6), 1773.
(10) (a) Suslick, K.; Cook, B.; Fox, M. Chem. Commun. 1985, 9, 580.
(b) Bhyrappa, P.; Young, J. K.; Moore, J. S.; Suslick, K. S. J. Am. Chem.
Soc. 1996, 118 (24), 5708. (c) Bhyrappa, P.; Young, J. K.; Moore, J. S.;
Suslick, K. S. J. Mol. Catal A: Chem. 1996, 113 (1−2), 109.
(d) Dandliker, P. J.; Diederich, F.; Gross, M.; Knobler, C. B.; Louati,
A. Angew. Chem. 1994, 106 (17), 1821.
(11) (a) Twyman, L. J.; Beezer, A.; Esfand, R.; Hardy, M. J.;
Mitchell., J. C. Tetrahedron Lett. 1999, 40 (9), 1743. (b) Twyman, L. J.
Tetrahedron Lett. 2000, 41 (35), 6875.
1
preparative SEC/GPC: total yield, 1.02 g, red/brown solid; H NMR
ppm (CDCl3) 8.90 (br, m, 8H), 8.60 (br, m, 4H, CTPP*), 8.35 (br, m,
6H, CTPP*), 8.25 (br, m, 9H, CTPP*), 8.05−7.55 (br, m, 2H,
polymer*), 7.55−7.10 (br, m, 1H, polymer*), 2.25(br, s, 3H, CH3
polymer*); UV/vis absorbance nm (CH2Cl2) λmax 418; FT-IR (ν/
cm−1) 3019, 1746 (COOH), 1595, 1444; GPC Mn 6000, PD 2.36.
*The exact ratio of porphyrin to repeat units is dependent on
molecular weight. As a result, integration ratios are reported relative to
either porphyrin or polymer.
Multi Zn−Porphyrin Hyperbranched Polymer Catal-
ysis Procedure. 4-Nitrophenyl isonicotinate 2 (33.0 mg, 0.14
mmol), (pyridin-4-yl)methanol 1 (15 mg, 0.14 mmol), and
multi Zn−porphyrin HBP 7 (18.5 mg, 5 mol % with respect to
porphyrin) were dissolved in 0.7 mL of CDCl3. The reaction
1
was then monitored via HNMR at various time intervals for a
period of 100 h. The reaction was followed by monitoring the
signal intensity of the product’s emerging benzyl protons (5.45
ppm) and those of the starting material (4.75 ppm). The yield
was calculated by comparing the ratio of these signal intensities
with respect the substrates initial concentration.
(12) (a) Tomalia, D. A.; Baker, H.; Dewald, J.; Hall, M.; Kallos, G.;
Martin, S.; Roeck, J.; Ryder, J.; Smith, P. Polym. J. (Tokyo) 1985, 17
(1), 117. (b) Tomalia, D. A.; Baker, H.; Dewald, J.; Hall, M.; Kallos,
G.; Martin, S.; Roeck, J.; Ryder, J.; Smith, P. Macromolecules 1986, 19
General Procedure for UV Titrations. A stock solution of
porphyrin containing material in DCM was prepared with a
concentration of 1 × 10−6 M with respect to porphyrin.
Pyridine solutions were made up to 1 × 10−2 M with respect to
pyridine using the stock porphyrin solution described above (so
as to maintain a constant porphyrin concentration). The
porphyrin solution (1 mL) was measured into a quartz cuvette,
and a UV/vis wavelength scan between 350 nm and 800 nm
was performed. To the cuvette were added aliquots of ligand
solution, between 10 and 20 μL, and the absorption of the
Soret band at 430 nm (bound peak) was monitored by
measuring the peak intensity. Absorptions were plotted against
pyridine concentrations and binding constants calculated using
fitting software (GraphPad). Titrations on each material were
performed a minimum of four times to ensure consistent
results.
́
(9), 2466. (c) Frechet, J. M. J.; Gitsov, I.; Grubbs, R. B.; Hawer, C. J.;
Henmi, M.; Leduc, M.; Sanford, E.; Yui, K. Polym. Mater. Sci. Eng.
1995, 73, 271. (d) Newkome, G. R.; Yao, Z.; Baker, G. R.; Gupta, V.
K. J. Org. Chem. 1985, 50 (11), 2003. (f) Gittins, P. J.; Twyman, L. J.
Supramol. Chem. 2003, 15 (1), 5.
(13) (a) De Gennes, P. G.; Hervet, H. J. Phys., Lett. 1983, 44 (9),
351. (b) Burnett, J.; King, A. S. H.; Twyman, L. J. React. Funct. Polym.
2006, 66 (1), 187.
(14) (a) Twyman, L. J.; King, A. S. H. J. Chem. Res., Synop. 2002,
No. 2, 43. (b) Twyman, L. J.; King, A. S. H.; Martin, I. K. Chem. Soc.
Rev. 2002, 31 (2), 69.
(15) (a) Kim, Y. H. J. Polym. Sci., Part A: Polym. Chem. 1998, 36 (11),
1685. (b) Jikei, M.; Kakimoto, M. Prog. Polym. Sci. 2001, 26 (8), 1233.
(c) Gao, C.; Yan, D. Prog. Polym. Sci. 2004, 29 (3), 183. (d) Uhrich, K.
E.; Hawker, C. J.; Frec
25 (18), 4583. (e) Wooley, K. L.; Hawker, C. J.; Lee, R.; Frec
J. Polym. J. (Tokyo) 1994, 26 (2), 187. (f) Yates, C. R.; Hayes, W. Eur.
Polym. J. 2004, 40 (7), 1257. (g) Feast, W. J.; Stainton, N. M. J. Mater.
Chem. 1995, 5 (3), 405.
́
het, J. M. J.; Turner, S. R. Macromolecules 1992,
het, J. M.
́
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: l.j.twyman@sheffield.ac.uk. Fax: (+)44-(0)114-22-
29436.
(16) (a) Hoelter, D.; Burgath, A.; Frey, H. Acta Polym. 1997, 48 (1/
2), 30. (b) Hawker, C. J.; Lee, R.; Frec
1991, 113 (12), 4583.
́
het, J. M. J. J. Am. Chem. Soc.
Notes
(17) Zheng, X.; Oviedo, I. R.; Twyman, L. J. Macromolecules. 2008,
41 (21), 7776.
The authors declare no competing financial interest.
(18) Bhyrappa, P.; Vaijayanthimala, G.; Suslick, K. S. J. Am. Chem.
Soc. 1999, 121 (1), 262.
ACKNOWLEDGMENTS
■
(19) (a) Voit, B. J. Polym. Sci., Part A: Polym. Chem 2000, 38 (14),
2505. (b) Voit, B. I. C. R. Chim. 2003, 6 (8−10), 821.
(20) Twyman, L. J.; Ge, Y.; Gittins, P. J. Supramol. Chem. 2006, 18
(4), 357.
(21) Gittins, P. J.; Twyman, L. J. J. Am. Chem. Soc. 2005, 127 (6),
1646.
(22) Gittins, P. J. PhD Thesis, University of Sheffield, 2004.
(23) Gittins, P. J.; Alston, J.; Ge, Y.; Twyman, L. J. Macromolecules
2004, 37 (20), 7428.
We thank the EPSRC and the University of Sheffield for
funding.
REFERENCES
■
(1) (a) Brady, P. A.; Sanders, J. K. M. Chem. Soc. Rev. 1997, 26, 327.
(b) Rowan, S. J.; Sanders, J. K. M. Curr. Opin. Chem. Biol. 1997, 1, 483.
(2) Benkovic, S. J.; Napper, A. D.; Lerner, R. A. Proc. Natl. Acad. Sci.
U.S.A. 1988, 85 (15), 5355.
(24) SEC/GPC calibrated with linear standards underestimates the
molecular weights of globular polymers; see: (a) Hawker, C. J.; Lee,
(3) Hilvert, D.; Hill, K. W.; Nared, K. D.; Auditor, M. T. M. J. Am.
Chem. Soc. 1989, 111 (26), 9261.
(4) Yoshizawa, M.; Tamura, M.; Fujita, M. Science 2006, 312, 251.
R.; Frec
́
het, J. M. J. J. Am. Chem. Soc. 1991, 113, 4583. (b) Twyman, L.
G
dx.doi.org/10.1021/jo400532r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
J.; Beezer, A. E.; Mitchel, J. C. J. Chem. Soc. Perkin Trans I 1994, 4, 407
and references cited therein.
(25) (a) Dugas, H. Bioorganic Chemistry, 3rd ed.; Springer-Verlag:
New York, 1995). (b) Hunter, C. A.; Meah, M. N.; Sanders, J. K. M. J.
Am. Chem. Soc. 1990, 112 (15), 5773.
(26) Porphyrin concentrations were kept constant and solutions
checked for a consistent UV absorbance for all catalysts (porphyrin
polymer 7 and tetraphenylporphyrin). With respect to calculating the
polymers concentration, the Mn value was used as an initial guide for
molecular weight. Where necessary solvent or sample was added to
adjust UV absorption (i.e., adjust porphyrin concentration) using
Beer−Lambert analysis.
(27) Hoydonckx, H. E.; De Vos, D. E.; Chavan, S. A.; Jacobs, P. A.
Top. Catal. 2004, 27 (1−4), 83−96.
(28) Although the reaction slows down after 60−70% completion, we
are not suggesting that the reaction is completely inhibited (but is
slowed down). This is because binding of the product is only slightly
higher than the reactants, which are still capable of displacing the
product. In addition, the background reaction continues as does the
simple porphyrin-catalyzed reaction.
(29) After the reaction had proceeded to 66% completion we had an
equal concentration of all species (0.066 M for both starting materials
and product, while the concentration of host remained 0.01 M). At
these concentrations all binding sites will be occupied (Ka ∼ 10−4 M−1
for all species). However, as the product binds with a slightly higher
affinity, it can better compete for binding sites and will therefore start
to inhibit/slow down the reaction.
(30) (a) Gittins, P. J.; Twyman, L. J. J. Am. Chem. Soc. 2005, 127 (6),
1646. (b) Gittins, P. J.; Alston, J.; Ge, Y.; Twyman, L. J. Macromolecules
2004, 37 (20), 7428.
(31) Christensen, J. B. Molecules 2001, 6, 47.
́ ́
(32) Tome, J. P. C.; Neves, M. G. P. M. S.; Tome, A. C.; Cavaleiro, J.
A. S.; Mendonc, A. F.; Pegado, I. N.; Duarteb, R.; Valdeirab, M. L.
Bioorg. Med. Chem. 2005, 13, 3878.
(33) Boyle, R. W.; Bruckner, C.; Posakony, J.; James, B. R.; Dolphin,
D. Organic Synthesis. Coll. 2004, 10, 370.
(34) Turner, S. R.; Voit, B. I.; Mourey, T. H. Macromolecules. 1993,
26, 4617.
H
dx.doi.org/10.1021/jo400532r | J. Org. Chem. XXXX, XXX, XXX−XXX