Mixed cyclic trimers of porphyrins and dioxoporphyrins: Geometry vs. electronics in ligand recognition

Article abstract of DOI:10.1039/a805504a

A series of metalloporphyrin-containing cyclic hosts has been prepared, which are electronically and geometrically programmed to recognize guests in a particular configuration. Electronic control is achieved by substituting the C-H meso positions with carbonyl groups, while geometric control is possible in the choice of the length of the rigid acetylene porphyrin linkers.

Full text of DOI:10.1039/a805504a

View Article Online / Journal Homepage / Table of Contents for this issue
Mixed cyclic trimers of porphyrins and dioxoporphyrins: geometry vs.
electronics in ligand recognition
Zoe Clyde-Watson, Nick Bampos and Jeremy K. M. Sanders*
Cambridge Centre for Molecular Recognition, University Chemical L aboratory, L ensÐeld Road,
Cambridge CB2 1EW , UK
L e t t e r
A series of metalloporphyrin-containing cyclic hosts has been prepared, which are electronically and
geometrically programmed to recognize guests in a particular conÐguration. Electronic control is achieved by
substituting the CwH meso positions with carbonyl groups, while geometric control is possible in the choice of
the length of the rigid acetylene porphyrin linkers.
We describe here a new series of cyclic porphyrin oligomers
containing a mixture of monomer units with di†erent individ-
ual binding and structural properties; these oligomers address
new questions about ligand preferences and dynamics within a
host cavity. Previously we reported the convergent synthesis
of an unsymmetrical porphyrin trimer1 that has been used to
change the stereochemical outcome of a DielsÈAlder reac-
tion.2 This convergent approach allows access to a much
more diverse array of host molecules than the one-pot GlaserÈ
Hay coupling route to symmetrical trimers.3h5 We describe
here trimers containing electron-deÐcient dioxoporphyrin
units in combination with standard porphyrins. Studying
ligand orientation within these hosts allows us to assess the
relative importance of electronic and geometric factors in
intra-cavity binding.
The key features of the synthesis involve the palladium-
catalysed cross-coupling6 of a central diiodo porphyrin frag-
ment 1 with two equivalents of a mono-alkyne porphyrin frag-
ment 2 (Scheme 1). The resulting linear trimer 3a is then dep-
rotected and cyclized to give trimer 4 via an intramolecular
GlaserÈHay coupling reaction, with the aid of an s-tri(4-
into two separate dipyrromethane units, leading to the loss of
conventional NMR ring current.9 Conversions of porphyrins
to their dioxo analogues were carried out efficiently using
thallium triÑuoroacetate in CH Cl ÈTHF (3 : 1) followed by a
sodium sulÐte quench.8 The mono-dioxo linear trimer was
formed in 60% yield from the dioxo-diiodo porphyrin frag-
ment, dioxo-1, and the monoalkyne porphyrin fragment 2.
Subsequent deprotection with TBAF (86%) followed by intra-
2
2
molecular coupling with the Py T template gave trimer
3
mono-dioxo-4 É Py T complex in 32% yield.
A similar
3
sequence gave bis-dioxo-4 É Py T in an overall yield of 12%
3
from monomers.
The 1H NMR spectra of the 1 : 1 Py T complexes of trimers
3
mono-dioxo-4 and bis-dioxo-4 show sharp and well-resolved
resonances. Two-dimensional NMR experiments were
employed to characterize the hosts and to investigate the
mode of binding within the cavities. In the 1H NMR spectrum
of the 1 : 1 mono-dioxo-4 É Py T complex, the pyridyl protons
3
of the bound ligand are split in a ratio of 2 : 1, as predicted for
a symmetrical tridentate ligand binding strongly to a host
with C symmetry. Well-resolved doublets (J \ ca. 5 Hz) are
2
pyridyl)triazine (Py T) template.4,5 ModiÐcation of either the
observed at 2.17 (H ) and 6.23 (H ) ppm, corresponding to
3
a2
b2
diiodo porphyrin or the monoalkyne porphyrin by oxidation
the two pyridyl arms binding to standard porphyrins, while
the pyridyl arm binding to the dioxoporphyrin gives rise to
downÐeld doublets (J \ ca. 5 Hz) at 8.86 (H ) and 7.91 (H )
before coupling leads to the mixed trimers.
Dioxoporphyrins7 are more electron deÐcient than stan-
dard porphyrins and thus show an increased affinity for
pyridyl ligands.8 Furthermore, interruption of the conjugated
system by the oxo functionalities divides the dioxo porphyrin
a1
b1
ppm [Fig. 1(a)].
The ROESY spectrum of this complex shows nOe contacts
between the H protons of the Py T and the H aromatic
a
3
2
protons of the host that are directed into the cavity. The
asymmetry of the complex results in three distinct subsets of
* E-mail: jkms=cam.ac.uk
New J. Chem., 1998, Pages 1135È1138
1135

View Article Online
I
I
1
2
i
2 x
H
TMS
R
R
3a R = TMS
3b R = H
ii, iii
N
N
N
N
N
N
4
Scheme 1 (i) Cat. Pd (dba) , Ph As, Et N, THF, 35 ¡C; (ii) TBAF;
2
3
3
3
(iii) Py T, CuCl, TMEDA, CH Cl
3
2 2
aromatic protons, which are assigned with the aid of the
COSY spectrum. Exchange peaks are not observed in the
ROESY spectrum between resonances of the bound ligand,
indicating that once bound inside the cavity, the tridentate
ligand is tightly anchored on the exchange (T ) timescale. The
1
corresponding spectra for the bis-dioxo-4 É Py T complex are
3
similar in appearance, but with the pyridyl protons split in a
reversed 1 : 2 ratio [Fig. 1(b)].
Addition of one equivalent of the bidentate 4@-phenyl-
4,2@ : 6@,4A-terpyridyl (Py Py)10 ligand to mono-dioxo-4 leads
2
to more than one possible binding orientation inside the
cavity. Due to the chelate e†ect, the ligand will preferentially
bind to the interior face of the host, but the two pyridyl arms
may bind either to the dioxoporphyrin and one of the two
standard porphyrins, or to both porphyrin units. As dioxo-
porphyrin monomers bind pyridyl ligands more strongly than
standard porphyrins,11 one would expect the ligand to orien-
tate itself asymmetrically within the host, between the dioxo-
porphyrin and one of the other porphyrins across an acetylene
linker, rather than between two porphyrins across the buta-
diyne linker [Fig. 2(a)].
Fig. 1 Bound chemical shifts for Py T within trimer complexes
3
porphyrin across the shorter acetylene link, or instead may
bind across the butadiyne link to both dioxounits, driven by
the higher intrinsic binding to the dioxoporphyrins. Model-
1H NMR spectroscopy conÐrms this expectation, showing
signiÐcant upÐeld shifts for the phenyl protons pointing
towards a porphyrin unit (4.42, 4.96 and 6.00 ppm), the
ling studies complemented by the binding behaviour of Py T-
3
like ligands in the 1,1,1 and 1,1,2 trimers suggest that geo-
porphyrin-bound pyridyl protons (1.95 and 5.81 ppm for H
metrically, Py Py will prefer to bind across the shorter acety-
a2
2
and H , respectively) and the pyridyl protons bound to a
lene linker.1h4 However, due to the relative di†erence in
b2
dioxoporphyrin (8.12 and 7.29 ppm for H
and H ,
binding affinities, there will be an electronic preference to bind
across the butadiyne link to both dioxo units. Alternatively, it
is feasible that the ligand exchanges between both conÐgu-
a1
b1
respectively). Exchange peaks are observed in the ROESY
spectrum between the H and H protons, the H and H
a1
a2
b1
b2
protons and also between the two r protons of the ligand,
indicating ligand rotation within the cavity; two-point binding
inside a host with three possible binding sites is not sufficient
to anchor the ligand in a single orientation. As the two
porphyrin units in the host are equivalent, ligand rotation
rations. NMR evidence though, shows the Py Py ligand
2
adopting a single orientation within the cavity [Fig. 2(b)],
with resonances at 6.18 and 2.11 ppm for the pyridyl binding
to a standard porphyrin unit; the second pyridyl, binding to a
dioxo porphyrin, gives rise to signals at 8.66 and 7.85 ppm. If
the ligand was binding to both dioxo units, the phenyl group
would point into the shielding region of the porphyrin and
upÐeld shifts would be expected for these protons, as observed
about the C dioxo axis is also possible, with one pyridyl arm
2
binding to the dioxo unit while the other Ñips between the two
remaining porphyrin units. This motion though, is too rapid
to be detected on the chemical shift timescale, and only the
average chemical shifts are recorded.
for Py Py binding in mono-dioxo-4. As no other signals are
2
observed in the region between 4.5 and 7.0 ppm, it seems that
Addition of one equivalent of Py Py to a chloroform solu-
geometrical factors are dominant in determining the ligand
orientation for this complex.
2
tion of bis-dioxo-4 leads to a similar competitive binding
process. The ligand may now bind to one dioxo unit and one
The increased affinity of Py T for the mixed dioxo trimers
3
1136
New J. Chem., 1998, Pages 1135È1138

View Article Online
Experimental
Chemical shifts are quoted relative to residual solvent (7.25
ppm for 1H and 77.0 ppm for 13C) and coupling constants are
given in Hz. All spectra were acquired at room temperature.
Distilled solvents were used throughout. The Ðnal puriÐcation
step in the preparation of the various porphyrins was crys-
tallization by layered addition of hexane to a dichloromethane
solution of the compound, followed by drying in vacuo.
Dioxo-diiodo porphyrin monomer (dioxo-1)
Thallium(III) triÑuoroacetate (376 mg, 0.69 mmol) in dry
CH Cl (40 ml) was added dropwise to a solution of the
2
2
diiodo porphyrin monomer (200 mg, 0.17 mmol) in CH Cl È
2
2
THF (60 ml : 20 ml) at 0 ¡C over a period of 15 min. The
mixture was stirred for a further 90 min before allowing to
warm up to room temperature and then washing with
aqueous sodium sulÐte solution (0.25 M, 2 ] 150 ml) followed
by water (2 ] 150 ml). The organic phase was dried (MgSO ),
Ðltered and evaporated to give a dark brown solid. Column
4
chromatography on silica, eluting with 1 : 1 hexaneÈethyl
acetate, followed by recrystallization from CH Cl Èhexane
2
2
yielded 139 mg of pure dioxo-diiodo porphyrin dioxo-1 (68%).
1H NMR (250 MHz, CDCl ): d \ 1.17 (s, 12H, Me), 2.45 (t,
3
J \ 7.7, 8H, CH ), 2.86 (t, J \ 7.7, 8H, CH ), 3.59 (s, 12H,
2
2
MeO), 7.18 (d, J \ 4.9, 2H, aromatic H), 7.55 (s, 2H, aromatic
H), 7.82 (t, J \ 4.9, 2H, aromatic H); 13C NMR (62.5 MHz,
CDCl ): d \ 12.67 (4 Me), 21.00 (4 CH ), 34.33 (4 CH ), 51.75
3
2
2
(4 MeO), 94.39 (2 aromatic CwI), 128.79, 130.63, 138.16,
138.49, (8 aromatic CwH), 136.18, 140.99, 141.51, 143.95,
151.41, 154.98 (20 quaternary aromatic and pyrrolic carbons),
174.46 (4 CxO ester), 182.13 (2 CxO dioxo); FAB-MS: calcd
for
C
H
I O N Zn 1208.2, found 1207 [M]`, 1229
[M ] Na]`; j (CH Cl )/nm: 280.4, 343.8, 471.3, 566.4.
52 48 2 10
4
max
2 2
TMS-protected linear trimer (mono-dioxo-3a)
Dioxo-diiodo porphyrin dioxo-1 (40 mg, 0.033 mmol) and
mono-TMS porphyrin 2 (73 mg, 0.070 mmol, 2.1 equiv) in
tetrahydrofuranÈtriethylamine (20 ml : 5 ml) were degassed (2
freezeÈthaw cycles) and heated to 35 ¡C under argon. Tri-
phenylarsine (24 mg, 0.079 mmol) and tris(dibenzylidene
acetone)dipalladium(0) (9 mg, 9.85 mmol) were added and the
mixture Ñushed through with argon for 5 min. The mixture
was heated at 35 ¡C for 90 min before removal of the solvent
by rotary evaporation to give a red solid. This was chromato-
graphed on silica, eluting initially with hexaneÈethyl acetateÈ
chloroform (2 : 1 : 1 v/v) to remove any monomers, followed
by neat chloroform to remove the product that appeared as a
red-orange band. Recrystallization from dichloromethaneÈ
hexane yielded 60 mg of pure product (60%). 1H NMR (250
Fig. 2 Bound chemical shifts for Py Py within trimer complexes
2
was investigated more quantitatively by a series of 1H NMR
competitive binding experiments. Addition of one equivalent
of mono-dioxo-4 to
a solution of the standard 1,1,2-
trimerÈPy T complex results in signiÐcant ligand transfer to
3
the more electron deÐcient cavity (K
/K
B
monovdioxov4 1,1,2vtrimer
9) with an estimated dissociation rate of B1 ] 10~4 s~1. A
dissociation rate of similar magnitude is observed for the
MHz, CDCl ): d \ 0.26 (s, 18H, TMS Me), 1.19 (s, 12H, dioxo
3
transfer of Py T from mono-dioxo-4 to bis-dioxo-4 and
Me), 2.37 (m, 8H, dioxo CH ), 2.46, 2.49 (s, 24H, porphyrin
3
2
K
/K
is B9. Surprisingly, the transfer of
Me), 2.82 (m, 8H, dioxo CH ), 3.00, 3.01 (m, 16H, porphyrin
CH ), 3.65 (s, 12H, dioxo MeO), 3.52, 3.58 (s, 24H, porphyrin
bisvdioxov4 monovdioxov4
2
Py T from the 1,1,2-trimer to bis-dioxo-4 is so much faster
that a reliable estimate of the dissociation rate is no longer
3
2
MeO), 4.22 (m, 16H, porphyrin CH ), 7.11È8.26 (m, 24H, aryl
2
H), 10.07 (s, 4H, meso H); 13C NMR (62.5 MHz, CDCl ):
3
possible. This striking discrepancy in ligand dissociation rates
suggests an associative mechanism involving intermediates in
which the ligand is bound simultaneously to both the original
and competing host molecules:12 the rate of transfer depends
on the di†erence between the number of dioxo units in the two
competing hosts.
In summary, the synthetic strategy of the 1,1,2 system
allows easy access to a wide range of unsymmetrical cyclic
hosts in which the electronic and geometric properties may be
Ðne-tuned to inÑuence and probe intra-cavity binding. Else-
where we describe the ability of these trimers to accelerate a
hetero DielsÈAlder reaction.13
d \ 0.03 (TMS, 6 Me), 12.65, 14.19 (12 Me), 20.81, 21.76, 29.75,
34.23, 36.93 (24 CH ), 51.72 (12 MeO), 89.03, 90.94, 94.66 (6
2
alkyne C), 97.17 (4 meso CwH), 105.07 (2 alkyne CwTMS),
127.68, 127.93, 128.51, 129.00, 129.33, 131.99, 132.00, 133.15,
135.99 (24 aromatic CwH), 118.45, 118.63, 122.28, 122.66,
124.22, 136.21, 138.77, 138.99, 139.76, 141.35, 143.39, 143.67,
144.18, 145.81, 147.52, 152.24 (66 quaternary aromatic and
pyrrolic carbons), 173.57, 174.31 (12 CxO ester), 182.43 (2
CxO dioxo); FAB-MS: calcd for C
H
N
O
Zn Si
170 166 12 26
3
2
3045.5572, found 3046 [M]`, 1523 [M]2`; j (CH Cl )/nm:
max
2 2
345.7, 411.5, 469.8, 539.1, 571.6.
New J. Chem., 1998, Pages 1135È1138
1137

View Article Online
Mono-dioxo-4 Æ Py T complex
References
3
Py T14 (4.8 mg, 0.0154 mmol) was added to a solution of the
1 A. VidalÈFerran, N. Bampos and J. K. M. Sanders, Inorg. Chem.,
1997, 36, 6117.
2 Z. Clyde-Watson, A. Vidal-Ferran, L. J. Twyman, C. J. Walter,
D. W. J. McCallien, S. Fanni, N. Bampos, R. S. Wylie and J. K. M.
Sanders, New J. Chem., 1998, 22, 493.
3 S. Anderson, H. L. Anderson and J. K. M. Sanders, J. Chem. Soc.,
Perkin T rans. 1, 1995, 2247.
4 A. Vidal-Ferran, Z. Clyde-Watson, N. Bampos and J. K. M.
Sanders, J. Org. Chem., 1997, 62, 240.
5 V. Marvaud, A. Vidal-Ferran, S. J. Webb and J. K. M. Sanders,
J. Chem. Soc., Dalton T rans., 1997, 985.
6 K. Sonogashira, Y. Tohda and N. Hagihara, T etrahedron L ett.,
1975, 4467.
7 G. H. Barnett, B. Evans and K. M. Smith, T etrahedron, 1975, 31,
2711.
3
deprotected linear trimer mono-dioxo-3b (34 mg, 0.012 mmol)
in dry CH Cl (55 ml) and the mixture stirred at room tem-
2
2
perature for 5 min under dry air before the addition of freshly
prepared copper(I) chloride15 (80 mg, 0.81 mmol, 70 equiv)
and TMEDA (125 ll, 0.81 mmol, 70 equiv). Stirring was con-
tinued in the dark for 15 h after which the mixture was
washed with water (6 ] 50 ml), dried (MgSO ), Ðltered and
4
the solvent removed under reduced pressure to give a red-
brown solid. Chromatography on silica, eluting with 2 : 1 : 1
CHCl ÈEtOAcÈhexane allowed separation of the product
3
from baseline material. Recrystallization from CH Cl Èhexane
2
2
yielded 12 mg of mono-dioxo-4 É Py T complex (32%). 1H
3
NMR (500 MHz, CDCl ): d \ 1.11 (s, 12H, dioxo Me), 2.17
3
8 D. W. J. McCallien and J. K. M. Sanders, J. Am. Chem. Soc., 1995,
117, 6611.
(d, J \ 5.6, 4H, H , bound Py T), 2.25 (t, J \ 7.0, 8H, dioxo
a2
3
9 E. D. Becker, R. B. Bradley and C. J. Watson, J. Am. Chem. Soc.,
1961, 83, 3743.
CH ), 2.50 (s, 24H, porphyrin Me), 2.74 (t, J \ 7.0, 8H, dioxo
2
2
CH ), 3.02 (m, 16H, porphyrin CH ), 3.41 (s, 24H, porphyrin
10 F. Krohnke, Synthesis, 1976, 1.
2
MeO), 3.51 (s, 12H, dioxo MeO), 4.19 (m, 16H, porphyrin
11 Due to the presence of a bound water molecule there is no shift in
the Soret band of dioxo porphyrin monomers on binding
pyridine-based ligands. Detailed studies of the Q-band region of
the visible spectrum did not yield simple 1 : 1 titration curves and
failed to reproduce very large earlier estimates8 for the magnitude
of these binding constants. However, the competition experiments
described here do conÐrm that dioxo porphyrins bind more
strongly than conventional porphyrins to pyridine ligands.
12 This is precedented in other porphyrin oligomers: C. A. Hunter,
M. N. Meah and J. K. M. Sanders, J. Am. Chem. Soc., 1990, 112,
5773.
CH ), 6.23 (d, J \ 5.6, 4H, H , bound Py T), 7.19 (d, J \ 7.8,
2
b2
3
2H, H4), 7.38 (s, 2H, H2@), 7.50 (t, J \ 7.8, 2H, H5), 7.79È7.81
(m, 6H, H6, H5@, H5A), 7.91 (d, J \ 4.5, 2H, H , bound Py T),
b1
3
8.03 (m, 4H, H4@, H4A), 8.14, 8.20 (2 d, J \ 6.8, 4H, H6@, H6A),
8.31, 8.37 (2 s, 4H, H2@, H2@), 8.86 (d, J \ 4.5, 2H, H , bound
a1
Py T), 9.93 (s, 4H, meso H); 13C NMR (62.5 MHz, APT,
3
CDCl ): d \ 12.36, 15.36 (12 Me), 20.68, 33.85, 36.85 (24 CH ),
3
2
50.71, 51.17 (12 MeO), 82.43, 88.70, 90.90, 93.68 (8 alkyne C),
96.90 (4 meso CwH), 127.70, 127.87, 129.25, 129.82, 130.94,
131.08, 131.72, 133.01, 133.54, 133.88, 135.62, 137.65 (24 aro-
matic CwH), 111.54, 117.53, 123.88, 129.73, 130.95, 131.19,
131.49, 131.57, 132.78, 133.32, 134.04, 135.39, 136.73, 136.88,
137.66, 138.06, 139.82, 141.01, 141.29, 146.96, 154.24 (66 quat-
ernary aromatic and pyrrolic carbons), 173.35, 173.44 (12
CxO ester), 181.10 (2 CxO dioxo), 120.19, 123.13, 143.79,
13 M. Marty, Z. Clyde-Watson, L. J. Twyman, M. Nakash and
J. K. M. Sanders, Chem. Commun., 1998, 2265.
14 H. Biedermann and K. Wichmann, Naturforsch., B, 1974, 29, 360.
15 R. N. Keller and H. D. Wyco†, in Inorganic Syntheses, ed. W. C.
Fernelius, McGraw-Hill, New York, 1946, vol. 2, p. 1.
148.38 (12 aromatic CwH, Py T), 120.05, 130.67, 168.33,
174.28 (6 quaternary carbons, Py T); FAB-MS: calcd for
3
3
C
j
H
N
O
Zn É C
H
N
3211.5, found 3213 [M]`;
164 148 12 26
3
18 12
6
(CH Cl )/nm: 289.5, 339.4, 418.4, 468.5, 550.6.
max
2 2
Acknowledgements
We thank the EPSRC for Ðnancial support and the EPSRC
Mass Spectrometry Service (Swansea) for FAB mass spectra.
Received in Montpellier, France, 13th July 1998;
L etter 8/05504A
1138
New J. Chem., 1998, Pages 1135È1138