Porphyrin synthesis in surfactant solution: Multicomponent assembly in micelles

Article abstract of DOI:10.1021/jo9600161

A synthesis of meso-substituted porphyrins in anionic sodium dodecyl sulfate micelles has been developed. Polar, functionalized aromatic aldehydes condense reversibly with pyrrole in the micellar phase. Oxidation of the porphyrinogen then provides functionalized porphyrins in yields of 10-48%. Hydrophobic aldehydes condense irreversibly to give low yields at practical substrate concentrations. Synthesis in D₂O solution results in per-β-deuterated porphyrins. A two-phase model is used to rationalize the dependence of porphyrin yield on reactant and surfactant concentration. Micelles are viewed as potential wells which promote porphyrinogen assembly by binding products more tightly than reactants.

Full text of DOI:10.1021/jo9600161

J . Org. Chem. 1996, 61, 3623-3634
3623
P or p h yr in Syn th esis in Su r fa cta n t Solu tion : Mu lticom p on en t
Assem bly in Micelles
Richard P. Bonar-Law
Cambridge Centre for Molecular Recognition, University Chemical Laboratory, Lensfield Road,
Cambridge CB2 1EW, U.K.
Received J anuary 2, 1996X
A synthesis of meso-substituted porphyrins in anionic sodium dodecyl sulfate micelles has been
developed. Polar, functionalized aromatic aldehydes condense reversibly with pyrrole in the micellar
phase. Oxidation of the porphyrinogen then provides functionalized porphyrins in yields of 10-
48%. Hydrophobic aldehydes condense irreversibly to give low yields at practical substrate
concentrations. Synthesis in D2O solution results in per-â-deuterated porphyrins. A two-phase
model is used to rationalize the dependence of porphyrin yield on reactant and surfactant
concentration. Micelles are viewed as potential wells which promote porphyrinogen assembly by
binding products more tightly than reactants.
In tr od u ction
This paper explores the preparation of symmetrical
meso-substituted porphyrins in micelles (Scheme 1). The
initial stage of porphyrin synthesis is the acid-catalyzed
condensation of four aldehydes and four pyrroles to a
mixture of linear intermediates which then cyclize to
porphyrinogen, a process which can be reversible.6,7 The
porphyrinogen is subsequently irreversibly oxidized to a
porphyrin. The idea was that micelles would promote
porphyrin synthesis in aqueous solution by (1) collecting
and concentrating the reactants and (2) biasing the
condensation equilibria in favor of porphyrinogen by
binding successive intermediates increasingly tightly; as
the aldehyde-pyrrole chain grows, water molecules are
eliminated and the hydrophobicity of the chain should
increase, pulling it further into the nonpolar interior of
the micelle. Previous examples of reversible reactions
in micelles include formation of imines,8 protonation
equilibria,1,9 axial ligation of iron porphyrins,10 and more
recently host-guest type chemistry.11,12 While the physi-
cal properties of micelle-solubilized porphyrins have been
widely studied,9,10,12,13 the use of surfactant aggregates
to direct the course of porphyrin synthesis, or any other
multiple assembly process, does not appear to have been
investigated.
Micelles are dynamic clusters of surfactant molecules,
able to collect and concentrate species from bulk aqueous
solution.1 Polar solutes tend to reside in the aqueous
outer regions of the micelle near the head groups,
whereas more hydrophobic molecules prefer the hydro-
carbon-like interior.2 It is this ability to localize and
orient substrates according to polarity which is most
promising from a synthetic point of view, and micellar
solubilization has been exploited to speed up and some-
times change the product distribution of several types of
reaction,3 including some unusual photochemical trans-
formations.4 Most of the micelle-mediated reactions
examined to date are kinetically controlled and irrevers-
ible. However, reversible chemistry is also of preparative
interest, particularly for one-step assembly of large
structures in which many bonds have to be formed, since
wrong-way reactions can be corrected under equilibrating
conditions.5
X Abstract published in Advance ACS Abstracts, May 1, 1996.
(1) For leading references to micelle chemistry, see: (a) Myers, D.
Surfactant Science and Technology; VCH: Weinheim, 1992. (b) Fen-
dler, J . H. Membrane Mimetic Chemistry; Wiley: New York, 1982. (c)
Solution Chemistry of Surfactants; Mittal, K. L., Ed.; Plenum: New
York, 1979. (d) Micellization, Solubilization and Microemulsions;
Mittal, K. L., Ed.; Plenum: New York, 1976. (e) Fendler, J . H.; Fendler,
E. J . Catalysis in Micellar and Macromolecular Systems; Academic:
New York, 1975. (f) Menger, F. M. Acc. Chem. Res. 1979, 12, 111.
(2) (a) Mukerjee, P., in ref 1c, Vol. 1, p 153. (b) Mukerjee, P.;
Cardinal, J . R.; Desai, N. R., in ref 1d, Vol. 1, p 241. (c) Almgren, M.;
Grieser, F.; Thomas, J . K. J . Am. Chem. Soc. 1979, 101, 279.
(3) (a) Fringuelli, F.; Pani, G.; Piermatti, O.; Pizzo, F. Tetrahedron
1994, 50, 11499. (b) Nivalkar, K. R.; Mudaliar, C. D.; Mashraqui, S.
H. J . Chem. Res., Synop. 1992, 98. (c) J aeger, D. A.; Wang, J .
Tetrahedron Lett. 1992, 33, 6415. (d) Fringuelli, F.; Germani, R.; Pizzo,
F.; Santinelli, F.; Savelli, G. J . Org. Chem. 1992, 57, 1198. (e) Bassetti,
M.; Cerichelli, G.; Floris, B. Gazz. Chim. Ital. 1991, 121, 527. (f) Singh,
V. K.; Raju, B. N. S.; Deota, P. T. Synth. Commun. 1988, 18, 567. (g)
Bassetti, M.; Cerichelli, G.; Floris, B. Gazz. Chim. Ital. 1986, 116, 583.
(h) Braun, R.; Schuster, F.; Sauer, J . Tetrahedron Lett. 1986, 27, 1285.
(i) Singh, A. K.; Raghuraman, T. S. Tetrahedron Lett. 1985, 26, 4125.
(j) Bianchi, M. T.; Cerichelli, G.; Mancini, G.; Marinelli, F. Tetrahedron
Lett. 1984, 25, 5205. (k) Matsui, Y.; Orchin, M. J . Organomet. Chem.
1983, 244, 369. (l) Breslow, R.; Maitra, U.; Rideout, D. Tetrahedron
Lett. 1983, 24, 1901. (m) Onyiriuka, S. O.; Suckling, C. J .; Wilson, A.
A. J . Chem. Soc., Perkin Trans. 2 1983, 1103. (n) Nikles, J . A.; Sukenik,
C. N. Tetrahedron Lett. 1982, 23, 4211. (o) Sutter, J . K.; Sukenik, C.
N. J . Org. Chem. 1982, 47, 4174. (p) Link, C. M.; J ansen, D. K.;
Sukenik, C. N. J . Am. Chem. Soc. 1980, 102, 7798. (q) J aeger, D. A.;
Robertson, R. E. J . Org. Chem. 1977, 42, 3298. (r) Menger, F. M.; Rhee,
J . U.; Rhee, H. K. J . Org. Chem. 1975, 40, 3803.
Resu lts
Recon n a issa n ce. Several surfactants were screened
for synthesis of tetraphenylporphyrin (TPP). In a typical
(5) Lindsey, J . S. New J . Chem. 1991, 15, 153.
(6) (a) Lindsey, J . S.; MacCrum, K. A.; Tyhonas, J . S.; Chuang, Y.-
Y. J . Org. Chem. 1994, 59, 579. (b) Lindsey, J . S.; Wagner, R. W. J .
Org. Chem. 1989, 54, 828. (c) Lindsey, J . S.; Schreiman, I. C.; Hsu, H.
C.; Kearney, P. C.; Marguerettaz, A. M. J . Org. Chem. 1987, 52, 827.
(7) Mauzerall, D. J . Am. Chem. Soc. 1960, 82, 2601. Mauzerall, D.
In The Porphyrins; Dolphin, D., Ed.; Academic Press: New York, 1978;
Vol. II, Part B, p 91.
(8) (a) Zoltewicz, J . A.; Bloom, L. B. J . Phys. Chem. 1992, 96, 2007.
(b) Martinek, K.; Yatsimirski, A. K.; Osipov, A. P.; Berezin, I. V.
Tetrahedron 1973, 29, 963.
(9) Barber, D. C.; Woodhouse, T. E.; Whitten, D. G. J . Phys. Chem.
1992, 96, 5106 and references therein.
(10) Mazumdar, S. J . Chem. Soc., Dalton Trans. 1991, 2091 and
references therein.
(11) (a) Nowick, J . S.; Cao, T.; Noronha, G. J . Am. Chem. Soc. 1994,
116, 3285. (b) Nowick, J . S.; Chen, J . S.; Noronha, G. J . Am. Chem.
Soc. 1993, 115, 7636.
(12) Bonar-Law, R. P. J . Am. Chem. Soc. 1995, 117, 12397.
(13) For leading references to porphyrins in organized media, see:
Fuhrhop, J . H.; Konig, J . Membranes and Molecular Assemblies: The
Synkinetic Approach; The Royal Society of Chemistry: Cambridge,
1994.
(4) For a review, see: Ramamurthy, V. Tetrahedron 1986, 42, 5753.
S0022-3263(96)00016-3 CCC: $12.00 © 1996 American Chemical Society

3624 J . Org. Chem., Vol. 61, No. 11, 1996
Sch em e 1 P or p h yr in Syn th esis in Su r fa cta n t Micelles (Q ) qu in on e oxid a n t)
Bonar-Law
trial reaction, 1 equiv of aqueous HCl was added to a
solution of benzaldehyde and pyrrole (1:1) in aqueous
surfactant, following the concentration of porphyrinogen
as a function of time. Porphyrinogen concentrations were
estimated by oxidizing samples of reaction mixture with
2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) and
measuring by UV-vis absorbance the amount of TPP
present.6c Anionic surfactants gave the most promising
results, as might be expected for an acid-catalyzed
reaction involving positively charged intermediates.1b
Sodium dodecyl sulfate (SDS) was chosen because it is
easily removed during workup. Pure SDS gave the same
yields as an inexpensive, commercially available mixture
containing 72% SDS (the balance being C14 and C16
homologues), so this commercial mixture was used for
all subsequent reactions. Without surfactant, reaction
mixtures were heterogeneous and only traces of por-
phyrin could be detected. The rate of reaction, as judged
by the time to reach maximum porphyrinogen concentra-
tion, was approximately proportional to the acid concen-
tration, and the maximum yield of TPP was independent
of the type or amount of strong acid catalyst employed.
(2) Some aldehydes gave much lower yields than would
be predicted from their partition coefficients. These
include electron-rich species with two or more ortho or
para electron-donating groups such as aldehydes 22 and
24 and 2,4,6-trihydroxybenzaldehyde (30), some ortho-
substituted aldehydes such as 2-acetamidobenzaldehyde
(32) and 2-carboxybenzaldehyde (7) (but not its methyl
ester 6), and aldehydes with powerful electron-withdraw-
ing substituents. In these cases brightly colored solutions
were produced on addition of acid, and oxidized aliquots
displayed intense bands at 450-500 nm, suggesting that
the formation of conjugated species such as dipyrrins is
competitive with porphyrinogen assembly.6b
Small ((10%) deviations from 1:1 aldehyde/pyrrole
stoichiometry had little effect on yields. A 10-30%
excess of pyrrole increased yields slightly for benzalde-
hyde and 4-hydroxybenzaldehyde and a 10-30% excess
of aldehyde decreased yields. Aliphatic aldehydes gave
UV yields of 5-10% under the standard conditions
(hexanal, butanal, acetaldehyde dimethyl acetal). 2-(Hy-
droxymethyl)pyrrole condensed in 0.5 M SDS to give a
2% yield of porphine, the parent heterocycle.
Scop e of th e Rea ction . Results for a variety of
aldehydes are summarized in Table 1. All reactions were
run under standard conditions: 10 mM aldehyde and 10
mM pyrrole in 0.5 M SDS. These conditions were chosen
on the basis of trial experiments to give reasonable yields
for polar aldehydes. Aqueous HCl was used to initiate
condensations, adding enough acid so that the maximum
porphyrinogen concentration was attained on a conve-
nient time scale. Reaction mixtures developed various
colors, ranging from light orange to deep burgundy, but
remained optically clear throughout. The UV yields in
Table 1 show two trends:
Tim e Cou r se of Rea ction s. Some qualitative fea-
tures of the evolution of porphyrinogen concentration are
worth recording. Porphyrinogen from hydrophilic alde-
hydes built up to a maximum after addition of acid and
then began to decline again, resulting in low yields if
reaction mixtures were left for long periods before oxida-
tion. In contrast, the porphyrinogen concentration from
hydrophobic aldehydes built up and then remained
constant. Some aldehydes such as amides 33 and 35
showed intermediate behaviors, porphyrinogen building
up rapidly at first and then slowly increasing to a
maximum before eventually decreasing.
(1) There is a broad correlation between porphyrin
yield and the water-to-micelle partition coefficient (P) of
starting aldehydes, Figure 1. In general the more
hydrophilic the aldehyde, the higher the yield. In the
series of amide-substituted benzaldehydes, para-amides
35 and 36 gave higher yields than trichloroacetyl deriva-
tive 37 and meta-acetamide 33 gave a higher yield than
the more lipophilic meta-carbamate 34. Hydroxyl-
substituted benzaldehydes behave similarly, free phenols
giving higher yields than their methyl ethers and acid
19 giving a higher yield than its methyl ester 20. If the
aldehyde is too water soluble, porphyrinogen formation
apparently becomes unfavorable, with low yields for
diacid 26 and glucoside 12.
P r ep a r a tive Sca le Syn th esis. The higher yielding
porphyrins were isolated from 1 mmol scale reactions
under the standard conditions, using the acid concentra-
tions and optimum reaction times from small scale trials.
A stoichiometric amount of 2,3,5,6-tetrachloro-1,4-benzo-
quinone (TCQ), a milder but slower oxidant than DDQ,6c
gave the best results for porphyrinogen oxidation. Oxi-
dations proceeded smoothly if TCQ, which is only slightly
soluble in 0.5 M SDS, was added as a concentrated
solution in THF so that the final reaction mixture
contained ∼5% v/v of THF. Other oxidants such as
hydrogen peroxide or ceric ammonium nitrate were also
effective for TPP, but gave lower yields for more func-
tionalized porphyrins. Trials of a one-step method in

Multicomponent Assembly in Micelles
J . Org. Chem., Vol. 61, No. 11, 1996 3625
Ta ble 1. Rea ction Con d ition s a n d P or p h yr in Yield s (%)
meso-phenyl
substituent or aldehyde
UV yield
(DDQ)
isolated yield
(TCQ)
cmpd
log Pa
[HCl]b
reaction timec
lit. yieldd
50
1
2
3
4
5
6
7
8
9
H (benzaldehyde)
2-Me
2.0
0.3
0.3
0.3
0.3
0.3
0.1
0.1
0.1
0.3
0.1
0.05
0.1
0.3
0.2
0.3
0.3
0.1
0.1
0.3
0.3
0.1
0.1
0.1
0.1
0.1
0.3
0.2
0.1
0.1
0.1
0.1
0.3
0.3
0.3
0.3
0.3
0.3
0.1
0.1
0.1
0.1
2 h
15 h
1 h
15 min
15 min
2 h
18
14
19
2
10
9
0
10
7
(TPP)
4-Me
2.4
3.3
3.2
15
P 1
2,4,6-Me
1-naphthaldehyde
2-CO2Me
2-CO2H
3-CO2Me
2.4
2.7
2.0
2 h
2 h
25 min
15 min
2 h
1 h
1.5 h
1 h
3,5-CO2Me
2-OH
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
26
15
5
20
P 2
2-OMe
2-O-â-D-glucoside
1.4
1.8
2.3
3-OH
3-OMe
31
10
11
27
42
25
40
22
25
1
10
1
43
5
29
38
32
0
15
3
37
15
36
35
13
1-4
30
30
30
26
9e
P 3
P 4
9.7
3-O(CH2)11CH3
3-OCH2CO2H
4-OH
>5
2 h
1.7
2.3
30 min
15 min
45 min
45 min
15 min
15 min
15 min
15 min
55 min
2 h
38
18
P 5
P 6
P 7
11
57
4-OMe
4-OCH2CO2H
4-OCH2CO2Me
2,3-OH
34f
2.3
2,4-OH
2,5-OH
2,6-OH
3,4-OH
1.6
48
P 8
10
3,4-OCH2CO2H
3,5-OH
3-OMe, 4-OH
3,4,5-OH
2,4,6-OH
2,4,6-OMe
2-NHAc
3-NHAc
3-NHCO2C(CH3)3
4-NHAc
4-NHCOCF3
4-NHCOCCl3
2-NO2 or 3-NO2 or 4-NO2
2-furaldehyde
3-furaldehyde
2-thiophenecarboxaldehyde
1.5
1.9
1.5
1.8
2.6
2.1
1.9
3.1
2.0
2.3
3.0
1 h
1 h
55 min
34
21
12
P 9
P 10
P 11
7.5
2 h
45 min
8 h
15 min
5 h
2 h
2 h
1 h
5 min
20 min
1 h
23
P 12
25
40
P 13
P 14
10
1.3
1.7
10
24
P 15
P 16
1
9
a
b
Water-micelle partition coefficient P, measured by capillary electrophoresis. Concentration of acid catalyst (mol/L). c Time for
d
maximum porphyrinogen concentration as measured by DDQ oxidation of reaction samples. Best recorded yield for direct synthesis
from aldehyde and pyrrole, references for individual porphyrins given in Experimental Section. e DDQ used as oxidant. f Isolated as the
tetramethyl ester.
which solid TCQ was added before the acid produced
porphyrin directly, but gave lower yields (75-90% of the
two-step yields for P 5 and P 8).
porphyrins. The deuterated versions of P 5 and P 8 were
prepared with >97% deuterium incorporation. This
represents a simple and economical method of preparing
per-â-deuterated porphyrins, compounds of some spec-
troscopic interest.15
The best workup procedure was to neutralize reaction
mixtures with aqueous KOH and then add enough of an
aqueous potassium salt to convert the surfactant into its
insoluble potassium salt. Conventional extraction with
EtOAc was then possible without producing emulsions.
The isolated yields in Table 1 reflect the ease with which
pure material could be obtained from a single chromato-
graphic step followed by crystallization. Recoveries were
generally good (the isolated yields are occasionally higher
than UV yields since TCQ rather than DDQ was used
as oxidant), but low in some cases due to borderline
stability on silica gel or in solution (P 11 and P 15) or due
to low solubility in organic solvents (P 12 and P 13). Ester
groups hydrolyze slowly under the reaction conditions,
but ester-functionalized porphyrins were still isolable in
reasonable yields if the reaction mixtures were not left
for long periods.
Rea cta n t a n d SDS Con cen tr a tion s. To investigate
partitioning effects in more detail, benzaldehyde deriva-
tives of varying hydrophobicity were condensed with 1
equiv of pyrrole at concentrations between 1 and 100 mM
in 0.5 M SDS (Figure 2). The yield of P 5 derived from
4-hydroxybenzaldehyde goes through a well-defined maxi-
mum at ∼15 mM reactants, whereas the yield of hydro-
phobic porphyrin P 4 is low at 15 mM and increases with
dilution. The curves for the other porphyrins in Figure
2 lie in between these two. Thus hydrophobic substrates
give low yields under the standard conditions (Figure 1)
because the optimum (bulk) concentration is lower for
these than for more water soluble species.
The actual reactant concentration will be higher in
micelles than in the bulk solution and can be estimated
Pyrrole is deuterated rapidly in acidic D2O,14 so syn-
thesis in D2O solutions of SDS results in per-â-deuterated
(14) Bean, G. P.; Wilkinson, T. J . J . Chem. Soc., Perkin Trans. 2
1978, 72.
(15) Gross, Z.; Kaustov, L. Tetrahedron Lett. 1995, 36, 3735 and
references therein.

3626 J . Org. Chem., Vol. 61, No. 11, 1996
Bonar-Law
F igu r e 3. Porphyrin yield versus SDS concentration (mol/L)
at a fixed ratio R ) [SDS]/[aldehyde] ) 50.
component of reaction mixtures (log P ) 1.2), has a
concentration of ∼55 mM, and lipophilic aldehydes with
log P > 2 all have nominal concentrations of ∼80 mM.
A two-phase analysis of this sort is supported by the
dependence of porphyrin yield on SDS concentration at
a fixed ratio R ) [SDS]/[aldehyde] ) 50 (Figure 3). Solute
concentrations calculated from eq 1 are almost constant
above ∼0.2 M SDS, so yields should vary little, as
observed. Below ∼0.2 M SDS, the mole fraction of
aldehyde (and pyrrole) in the micellar phase begins to
decrease, so yields should decrease. The decrease is rapid
for P 5 since Sm also moves away from its optimum value
(Figure 2). The decrease is less marked for TPP since
yields increase with decreasing Sm for more hydrophobic
aldehydes.
F igu r e 1. Porphyrin yield versus aldehyde partition coef-
ficient (10 mM aldehyde, 10 mM pyrrole in 0.5 M SDS).
“Amides” are amide- or carbamate-substituted benzaldehydes,
“phenols” are hydroxy-substituted benzaldehydes, and “others”
refer to alkyl or ester-substituted benzaldehydes or other
heterocyclic aldehydes in Table 1.
Rever sibility of th e Rea ction . Th er m od yn a m ic
ver su s Kin etic Con tr ol. The shapes of the curves in
Figure 2 and the qualitative kinetic observations sug-
gested that porphyrinogen assembly might be irreversible
for hydrophobic aldehydes. Two types of experiment
were performed to examine reversibility:
(1) Slow Ad d ition . Concentrated solutions of 3-(do-
decyloxy)benzaldehyde (15) or 3-hydroxybenzaldehyde
(13) and pyrrole (1:1) in 0.5 M SDS were added at various
rates to a larger volume of acidified 0.5 M SDS solution
so that the final reactant concentration was 10 mM
(Figure 4). Slow addition of lipophilic aldehyde 15 gave
higher yields of porphyrin P 4, indicative of an irreversible
reaction in which reactants are effectively maintained
at high dilution in the micellar phase. Slow addition of
13 had a much smaller effect on the yield of P 3,
consistent with a reversible reaction in which the yield
of porphyrin is determined by the final concentration of
reactants.
F igu r e 2. Porphyrin yield versus aldehyde concentration
(mol/L) in 0.5 M SDS ([pyrrole] ) [aldehyde]).
from the pseudophase model16 for micellar solubilization:
Sm ) S0P/(1 + â(P - 1))
(1)
where Sm is the reactant concentration in micelles, S0 is
the bulk reactant concentration, â ) vm([SDS] - cmc) is
the volume fraction of solution occupied by the micellar
phase, vm ) 0.25 L/mol is the partial molar volume of
SDS in water, and cmc ) 8 mM is the critical micelle
concentration of SDS. Under the standard conditions (10
mM reactants, 0.5 M SDS) the concentration of 4-hy-
droxybenzaldehyde in the micellar phase is estimated
from eq 1 as ∼70 mM. Pyrrole, the most hydrophilic
(2) P or p h yr in ogen Exch a n ge.6c Two aldehydes
were condensed separately with pyrrole, and when the
porphyrinogen concentrations had reached a maximum
(time ) tmax), the solutions were mixed. At a common
time 5tmax after mixing, the distribution of crossover
products was analyzed. Three pairs of aldehydes were
chosen which reacted at similar rates and gave similar
yields of porphyrin: two pairs of hydroxybenzaldehydes
(13/27 and 17/28) and one pair of hydrophobic aldehydes
(15/37). Little or no unreacted aldehyde monomer re-
mained before mixing, as judged by TLC of oxidized
samples. The two pairs of hydroxybenzaldehydes gave
between 31 and 38% exchange, and the pair of hydro-
(16) (a) Bunton, C. A., in ref 1c, Vol. 2, p 519. (b) Romstead, L. S.,
in ref 1d, Vol. 2, p 509. (c) Elworthy, P. H.; Florence, A. T.; McFarlane,
C. B. Solubilization by Surface Active Agents, and Its Application in
Chemistry and the Biological Sciences; Chapman and Hall: London,
1968.

Multicomponent Assembly in Micelles
J . Org. Chem., Vol. 61, No. 11, 1996 3627
F igu r e 5. Capillary electrophoretograms before (upper trace)
and after (lower trace) reaction between 4-hydroxybenzalde-
hyde (1 mM) and pyrrole (1 mM) in 0.1 M SDS (A ) pyrrole,
B ) aldehyde). The arrow shows the retention time of micelles.
F igu r e 4. Porphyrin yield versus total addition time of a 1:1
mixture of aldehyde and pyrrole to acidified 0.5 M SDS. For
P 4 the yields measured at 15 min and 5 h after the end of
additions were the same, but absolute and relative yields for
P 3 after 15 min (upper curve) were higher than those after 5
h (lower curve).
and while some porphyrin was formed in all solvents,
yields were less than 5%.
Literature yields are low for direct condensation of
pyrrole with aldehydes bearing unprotected functional
groups (Table 1). For completeness, the synthesis of the
phenolic porphyrins in Table 1 was also briefly investi-
gated using Lindsey’s conditions (10 mM reactants in
CH2Cl2 or CHCl3 with BF3‚Et2O or TFA catalysis).6b,c UV
yields from DDQ oxidation of reaction samples were as
follows: P 2 (<1%), P 3 and P 5 (∼22%), P 8 (∼10%), P 9
(∼5%), and P 11 (<1%). While useful amounts of P 3 and
P 5 were formed, reactions were heterogeneous, confirm-
ing the finding by Lindsey et al. that this method gives
low yields if reactants or products are sparingly soluble
in CH2Cl2 or CHCl3.6c The ortho-amide 32 also failed to
give significant amounts of porphyrin under these condi-
tions.
phobic aldehydes gave <10% exchange (see Experimental
Section for details). In further experiments a mixture
of unreacted aldehyde and pyrrole was added directly to
the porphyrinogen derived from the other aldehyde.
Under these conditions, scrambling was about twice as
fast (54% for 13/27 and 62% for 17/28) since only one
porphyrinogen has to dissociate, at least in the early
stages of the reaction. Control experiments showed that
if SDS solutions of the aldehydes were mixed before
addition of acid, near-statistical mixtures (100% ex-
change) of porphyrins resulted after oxidation at tmax
.
The slow addition and mixing experiments confirm
that porphyrinogen assembly is under kinetic control for
hydrophobic aldehydes. For the more hydrophilic alde-
hydes, assembly is clearly reversible, but the slow
exchange rates suggest that equilibrium may not always
be achieved at tmax, therefore the reactions are not
completely under thermodynamic control. The slight
increase in yield of P 3 measured after short total addition
times (Figure 4), which disappears if the reaction mix-
tures are left to stand, is also consistent with fairly slow
kinetics. In principle, irreversible condensation should
lead to quantitative yields of porphyrinogen at suf-
ficiently high dilution. In practice, yields are probably
limited by the accompanying side reactions, which would
also be irreversible. Side reactions certainly occur in
micellar synthesis as seen by the steady decline with time
of porphyrinogen generated from hydrophilic aldehydes
and the failure of certain classes of aldehyde.
Con tr ol Rea ction s in Or ga n ic Solven ts. Aromatic
solutes experience polar environments in micelles,2,17 so
the best control for micellar porphyrin synthesis is
synthesis in a polar solvent at micelle-like reactant
concentrations. Accordingly, benzaldehyde and 4-hy-
droxybenzaldehyde were condensed with pyrrole in metha-
nol, ethanol, THF, and DMF at reactant concentrations
between 50 and 500 mM, using 1 equiv of aqueous 10 M
HCl as catalyst. The higher concentration was included
to allow for the fact that micelles are not isotropic, so
concentrations from the two-phase analysis are lower
limits. Reactions were generally slow and heterogeneous,
P a r tition in g of Rea ction P r od u cts. To determine
the relative affinity of micelles for reactants and products,
the reaction between 4-hydroxybenzaldehyde and pyrrole
in SDS was examined in more detail. Micellar capillary
electrophoresis is a useful technique for measuring
water-micelle partition coefficientssthe more hydro-
phobic a solute is, the longer its retention time, with very
hydrophobic species coeluting with the micelles.18 As
shown in Figure 5, the reactants transform into species
which coelute with the micelles and thus have high
partition coefficients with log P g 5.
1
The same reaction was also followed by H NMR in a
D2O solution of SDS. Upon addition of acid, the pair of
aromatic doublets from the aldehyde was gradually
replaced with an upfield-shifted pair of multiplets and
the aldehyde singlet at 9.7 ppm was replaced by a
multiplet at 5.3 ppm. Longitudinal relaxation rates and
self-diffusion coefficients were measured for aldehyde,
products, and micelles. Products were found to relax
about twice as fast as the starting aldehyde and to diffuse
at a rate intermediate between starting aldehyde and
micelles. These findings are also consistent with the
reaction products being more strongly bound to the
micelle than the reactants.
(18) (a) Capillary Electrophoresis: theory and practice; Camilleri,
P., Ed.; CRC Press: Boca Raton, FL, 1993. (b) Terabe, S.; Otsuka, K.;
Ando, T. Anal. Chem. 1985, 57, 834.
(17) Abraham, M. H.; Chadha, H. S.; Dixon, J . P.; Rafols, C.; Treiner,
C. J . Chem. Soc., Perkin Trans. 2 1995, 887.

3628 J . Org. Chem., Vol. 61, No. 11, 1996
Bonar-Law
Sch em e 2. Mod el for Rever sible P or p h yr in ogen
Assem bly in a Tw o-P h a se System (Li ) Lin ea r
In ter m ed ia te Com p osed of i Mon om er Un its,
Distr ibu ted w ith P a r tition Coefficien t P i betw een
Micella r (Su bscr ip t m ) a n d Aqu eou s (Su bscr ip t a q)
P h a ses)
F igu r e 6. Simulation of porphyrinogen assembly in a single
phase (solid curves) and in two phases (dotted curves). Equi-
librium constants for single phase assembly: A, Km ) 2 × 104,
Kc ) 10; B, Km ) 2 × 103, Kc ) 10; C, Km ) 200, Kc ) 10; D,
Km ) 2 × 103, Kc ) 1 × 103. Equilibrium constants for assembly
in 0.5 M SDS (â ) 0.125), assuming an average reactant
partition coefficient log P1 ) 1.5: A, Km ) 5 × 103, Kc ) 15; B,
Km ) 500, Kc ) 15; C, Km ) 50, Kc ) 15; D, Km ) 500, Kc ) 1.5
× 103. The squares are UV yields of P 5 (DDQ oxidation),
representing experimental values for the 4-hydroxybenzalde-
hyde-pyrrole equilibrium.
Discu ssion
In the classical Adler-Longo porphyrin synthesis, a
solution of aldehyde and pyrrole in a high-boiling acid
solvent is refluxed in air so that condensation and
oxidation occur simultaneously.19 This simple reaction
is still the best way of making large quantities of robust
porphyrins, particularly those which crystallize cleanly
from the reaction mixture. A major improvement in the
scope of aldehyde-pyrrole condensations was the intro-
duction of a two-step method in which aldehyde and
pyrrole are allowed to condense in a nonpolar solvent
with exclusion of air, followed by oxidation of the
porphyrinogen.6,20a The procedure optimized by Lindsey
et al. in chlorinated solvents is currently the most useful
one for organic soluble porphyrins, with yields of 20-
40%, rising to ∼50% in a few instances.6,21 Recent
improvements in the synthesis of tetraarylporphyrins
include the use of an oxidizing cosolvent,20b oxidizing
Lewis acids,22a and various clays as catalysts.22b-d
Adler-Longo synthesis gives low yields of sensitive
porphyrins, reflecting the rather vigorous conditions, and
the Lindsey method gives low yields of polar porphyrins,
due to solubility problems. Micellar synthesis is thus
complementary to these procedures, giving good yields
of sensitive, polar porphyrins. The major practical
limitation is that relatively large quantities of surfactant
are required.
favorable process in the gas phase.23 Since the Adler-
Longo synthesis is performed in polar solvents, and
â-substituted pyrroles condense quite readily in polar
solvents at room temperature,24 porphyrinogen synthesis
is also likely to be energetically downhill in polar media.
The failure of control reactions in solvents with micelle-
like polarity can therefore be ascribed to kinetic barriers,
implying that one function of the surfactant phase is to
provide a kinetically viable route to porphyrinogen.
Simply keeping reaction mixtures homogeneous may be
sufficient.
To provide further insight into the role of the surfac-
tant, reversible porphyrinogen assembly was modeled
according to Scheme 2. In this model porphyrinogen is
produced by cyclization of octamer L8, which is one of a
series of linear intermediates present at equilibrium in
both phases. Details of the model are discussed in the
Experimental Section. Most of the reaction occurs inside
micelles under the standard conditions, so simulations
were initially confined to a single phase. Porphyrinogen
yield as a function of reactant concentration is shown in
Figure 6 (solid curves) for various values of the inter-
and intramolecular equilibrium constants Km and Kc.
These were initially chosen to roughly reproduce experi-
mental data for the 4-hydroxybenzaldehyde-pyrrole
equilibrium (curve B) and then varied independently. The
simulations illustrate three general points:
Role of th e Micella r P h a se. Cyclocondensation of
four aldehydes and four pyrroles is calculated to be a
(19) (a) Kim, J . B.; Adler, A. D.; Longo, F. R. In The Porphyrins;
Dolphin, D., Ed.; Academic Press: New York, 1978; Vol. 1, p 85. (b)
Adler, D.; Longo, F. R.; Shergalis, W. J . Am. Chem. Soc. 1972, 94, 3986.
(c) Dolphin, D. J . Heterocycl. Chem. 1970, 7, 275. (d) Treibs, A.;
Ha¨berle, N. Liebigs Ann. Chem. 1968, 718, 183. (e) Adler, A. D.; Sklar,
L.; Longo, F. R.; Finarelli, J . D.; Finarelli, M. G. J . Heterocycl. Chem.
1968, 5, 669. (f) Datta-Gupta, N.; Bardos, T. J . J . Heterocycl. Chem.
1966, 3, 495.
(20) (a) Rocha Gonsalves, A. M. d’A.; Pereira, M. M. J . Heterocycl.
Chem. 1985, 22, 931. (b) Rocha Gonsalves, A. M. d’A.; Vareja˜o, J . M.
T. B.; Pereira, M. M. J . Heterocycl. Chem. 1991, 28, 635.
(21) (a) Quast, H.; Dietz, T.; Witzel, A. Liebigs Ann. Chem. 1995,
1495. (b) Wagner, R. W.; Ruffing, J .; Breakwell, B. V.; Lindsey, J . S.
Tetrahedron Lett. 1991, 32, 1703.
(1) An optimum reactant concentration arises naturally
for porphyrinogen assembly, as noted by Lindsey et al.
for TPP synthesis in CH2Cl2.6c If the reactants are too
dilute, the linear species will consist largely of monomer
and the concentration of L8,m will be low. If the reactants
(22) (a) Gradillas, A.; del Campo, C.; Sinisterra, J . V.; Llama, E. F.
J . Chem. Soc., Perkin Trans. 1 1995, 2611. (b) Shinoda, T.; Izumi, Y.;
Onaka, M. J . Chem. Soc., Chem. Commun. 1995, 1805. (c) Onaka, M.;
Shinoda, T.; Izumi, Y.; Nolen, E. Tetrahedron Lett. 1993, 34, 2625. (d)
Laszlo, P.; Luchetti, J . Chem. Lett. 1993, 449.
(23) George, P. Ann. N.Y. Acad. Sci. 1973, 206, 84.
(24) Tietze, L. F.; Geissler, H. Angew. Chem., Int. Ed. Engl. 1993,
32, 1038 and references therein.

Multicomponent Assembly in Micelles
J . Org. Chem., Vol. 61, No. 11, 1996 3629
are too concentrated, [L8,m] will again be low, as the
equilibrium overshoots to polymer.
porphyrins in Table 1 are difficult to make by direct
aldehyde-pyrrole condensation under conventional con-
ditions. SDS was used exclusively, but other surfactants
with different local microenvironments and partitioning
properties would be expected to have their own particular
selectivities.
The variation of porphyrin yield with substrate parti-
tion coefficient, reactant concentration, and surfactant
concentration was rationalized within a two-phase scheme
and supported by numerical simulations. It was con-
cluded that micelles act as potential wells, binding
products more tightly than reactants, and may also
catalyze condensation reactions. Multiple assembly within
micelles is an interesting strategy which may be extend-
ible to other reactions and other phases.27
(2) For a fixed value of Kc, the optimum reactant
concentration is inversely proportional to Km, as il-
lustrated by the shift of curve C to curve A when log Km
is increased by a factor of 2. An increase in Km may
account for the higher yields from hydrophobic aldehydes
at low reactant concentrations (Figure 2), although an
equilibrium treatment eventually becomes inappropriate
when condensation is irreversible.
(3) The maximum yield for a fixed Km depends only on
Kc. Curve D shows that if Kc is large enough, the entire
equilibrium is pulled over to porphyrinogen, as expected.
In reality the aqueous phase is also present, tending
to extract the more polar species out of micelles. For
assembly to be favorable under these conditions, Km
should remain approximately constant for each step, as
in homogeneous solution, and simulations with this
assumption (dotted curves in Figure 6) are qualitatively
similar to the single-phase results. If Km is constant,
then Scheme 2 requires that the binding energies of
successive intermediates to the micelle increase in an
additive fashion. In fact this is a reasonable expectation
since group additivity schemes are well established for
partitioning processes.25 Experimental support comes
from the finding that the products of the 4-hydroxy-
benzaldehyde-pyrrole equilibrium do indeed bind to
micelles much more strongly than the reactants. In
addition Martinek et al. found that the product from
condensation of benzaldehyde and aniline in SDS solution
is bound about twice as well as the reactants.8b
Exp er im en ta l Section
Gen er a l. Pyrrole was distilled and stored under argon.
One liter stock solutions of 0.5 M SDS were prepared by
stirring 150 g of “95% lauryl sulfate” (Aldrich, SDS content
72%) in MilliQ water under a flow of argon. SDS solutions
were sealed with rubber septa and used within a few months
of preparation. Reactions were run under argon, transferring
surfactant solutions with plastic syringes. Silica gel 60 (Merck
9385, 230-400 mesh) was employed for flash chromatography.
FAB mass spectra were obtained using a m-nitrobenzyl alcohol
matrix. NMR spectra are referenced to CHCl3 (7.25 ppm) or
acetone (2.05 ppm). Extinction coefficients and λmax values
were measured in “solvent A” which refers to a mixture of
CH2Cl2 and MeOH (4:1) containing 1% v/v pyridine. Capillary
electrophoresis was performed on an Applied Biosystems 270-
HT instrument.
Micelles can thus be said to fulfill the basic thermo-
dynamic requirements for efficient assembly, acting as
potential wells for porphyrinogen. An inevitable conse-
quence of the progressive solubilization mechanism is
that if species sink too far into the potential well, the
rate of dissociation back to reactants will start to
decrease, and this provides a qualitative explanation of
why hydrophobic aldehydes condense under kinetic con-
trol. The very slow rate of monomer exchange between
hydrophobic porphyrinogens clearly shows that porphy-
rinogens can become trapped in micelles; a similar
observation was made for clay-mediated porphyrin syn-
thesis, in which products are compartmentalized within
the pores of the mineral.22c
Sm a ll Sca le Tr ia ls (R ) [SDS]/[a ld eh yd e] ) 50). Alde-
hyde (0.1 mmol) was stirred with aqueous SDS (10 mL, 0.5
M) until a clear solution was obtained, and then pyrrole (7
µL, 0.1 mmol) was added. Hydrochloric acid (25-300 µl, 10
M, plastic syringe and needle) was injected to initiate reaction,
and samples (0.5 mL) were withdrawn at intervals. The
samples were quenched by addition to 100 µL portions of a 10
mg/mL solution of DDQ in THF in 5 mL volumetric flasks
(stock solutions of DDQ in THF were prepared in dry THF,
shielded from light, and used within 1 day). After allowing
the solution to stand for ∼30 min, the flasks were filled up to
the mark with THF, and 50 or 100 µL of this solution was
added to 2.0 mL of solvent A in a cuvette. The difference in
absorbance was measured between the top of the Soret band
and the inflection between the Soret and the peak due to DDQ
reduction products at ∼350 nm. An operational extinction
coefficient (ꢀexp) measured between 350 nm and the top of the
Soret for pure porphyrin in solvent A was used to calculate
porphyrin concentrations (generally ꢀexp ∼0.95ꢀ where ꢀ is the
true extinction coefficient). This method underestimates the
concentration of porphyrin slightly, particularly for low-
yielding reactions due to band overlap. In those cases where
the porphyrin was not isolated on a preparative scale, the ꢀexp
value measured for a similarly substituted porphyrin was used
to estimate the yield. For alkyl- and ester-substituted tetra-
arylporphyrins, the extinction coefficient for TPP was used (log
ꢀexp ) 5.69). For tetraalkylporphyrins an extinction coefficient
of log ꢀexp ) 5.3028 was assumed.
Su m m a r y a n d Con clu d in g Rem a r k s
Porphyrins have been prepared from a range of alde-
hydes in anionic micelles. Porphyrinogen assembly was
found to be reversible for polar aromatic aldehydes but
irreversible for hydrophobic ones. Moderate to good
yields of sensitive functionalized porphyrins, and also
their â-deuterated versions, could be obtained directly
from the corresponding aldehydes without the need for
protecting groups. The reactions are simple to perform,
and the conditions are mild,26 although large quantities
of surfactant are required. The present work is unusual
among micelle-mediated reactions in that it is prepara-
tively uniquesmany of the more highly functionalized
R ep r esen t a t ive P r oced u r e for P r ep a r a t ive Sca le
Syn t h esis: m eso-Tet r a k is(4-h yd r oxyp h en yl)p or p h yr in
(P 5).22a Hydrochloric acid (1.0 mL, 10 M) was added to a
rapidly stirring solution of 4-hydroxybenzaldehyde (122 mg,
1.0 mmol) and pyrrole (70 µL, 1.0 mmol) in aqueous SDS (100
(25) (a) Comprehensive Medicinal Chemistry; Pergamon Press: New
York, 1990; Vol. 4, p 295. (b) Smith, G. A.; Christian, S. D.; Tucker, E.
E.; Scamehorn, J . F. Langmuir 1987, 3, 598.
(26) The ease with which porphyrinogens can be assembled in a
micelles hints at a role for such supramolecular aggregates in the
prebiotic synthesis of porphyrins. For prebiotic type mineral-catalyzed
porphyrin synthesis, see: Cady, S. S.; Pinnavaia, T. J . Inorg. Chem.
1978, 17, 1501.
(27) For example, recently developed water soluble dendrimers
behave in some respects as unimolecular micelles and may function
in the same way: (a) Hawker, C. J .; Wooley, K. L.; Fre´chet, J . M. J . J .
Chem. Soc., Perkin Trans. 1 1993, 1287. (b) Newkome, G. R.; Moore-
field, C. N.; Baker, G. R.; Saunders, M. J .; Grossman, S. H. Angew.
Chem., Int. Ed. Engl. 1991, 30, 1178.
(28) Albers, V. M.; Knorr, H. V. J . Chem. Phys. 1936, 4, 422.

3630 J . Org. Chem., Vol. 61, No. 11, 1996
Bonar-Law
mL, 0.5 M) under a slow flow of argon. The initial straw-
colored solution quickly turned orange and then orange-brown
but remained clear. A solution of TCQ (220 mg, 0.89 mmol)
in THF (5 mL, warmed briefly to dissolve the TCQ) was added
after 30 min, and the mixture was stirred overnight open to
the air. The dark suspension was poured into a separatory
funnel containing a mixture of EtOAc (100 mL), aqueous
potassium hydroxide (5 mL, 2 M), potassium phosphate pH 7
buffer (10 mL, 1 M), and aqueous potassium chloride (10 mL,
3 M), rinsing out the reaction flask with water (50 mL). Phase
separation after shaking was rapid to give a dark organic layer.
The aqueous layer was extracted with a further portion of
EtOAc (50 mL), and the combined organic extracts were
washed with water (2 × 100 mL; a little brine was added after
the initial water shake to induce rapid phase separation) and
dried (Na2SO4). The desiccant was filtered off through a coarse
glass frit, which also removed some insoluble black material,
and the filtrate rotary evaporated with the addition of ∼5 g of
flash silica gel to give a friable brown-black solid. This was
tipped onto a prepacked (5% THF in CH2Cl2) silica gel column
(2.2 × 25 cm), and the column was eluted with 5-12% THF
in CH2Cl2, collecting the main purple band. The crude
porphyrin was dissolved in warm THF (∼5 mL), filtered
through a plug of cotton wool, and then diluted with hot
cyclohexane (10-15 mL) until crystallization started. After
allowing the solution to stand overnight, the crystals that
formed were filtered, washed with cyclohexane, and dried at
180 °C (0.1 mmHg) for 12 h to give porphyrin P 5 (65 mg, 38%)
as fine purple needles: 1H NMR (d6-acetone, 250 MHz) δ -2.71
(br s, 2H), 7.30 (d, J ) 8.5 Hz, 8H), 8.07 (d, J ) 8.5 Hz, 8H),
8.89 (br s, 4H, OH), 8.93 (s, 8H); λmax (solvent A) 421 (log ꢀ )
5.66), 518, 556, 591, 650 nm; HRFABMS m/ z 679.2327 (MH+),
m eso-Tet r a k is(2-h yd r oxyp h en yl)p or p h yr in
(P 2).30
Chromatography: 2-10% THF in CH2Cl2, collecting four red
bands. Part of the slowest moving R,R,R,R-atropisomer band
was contaminated with brown material and was discarded.
The combined red fractions were crystallized from THF/
cyclohexane and then recrystallized from MeOH/cyclohexane.
The mixture of crystals and amorphous purple powder was
dried at 150 °C (0.1 mmHg) for 12 h to give P 2 (35 mg, 20%):
1H NMR (CDCl3, 250 MHz) δ -2.75 (br s, 2H), 5.0 (br s, 4H,
OH), 7.32 (m, 8H), 7.72 (m, 4H), 7.97 (m, 4H), 8.9 (m, 4H);
λmax (solvent A) 417 (log ꢀ ) 5.65), 511, 545, 586, 642 nm;
FABMS m/ z 679.4 (MH+).
m eso-Tet r a k is(3-h yd r oxyp h en yl)p or p h yr in
(P 3).19d
Chromatography: 3-12% THF in CH2Cl2. The crude porphy-
rin was washed with methanol, then dissolved in hot reagent-
grade chloroform, and left to evaporate to a small volume over
several days. The small purple-black prisms were dried at 150
°C (0.1 mmHg) for 12 h to give P 3 (44 mg, 26%): 1H NMR
(d6-acetone, 250 MHz) δ -2.77 (br s, 2H), 7.33 (d, J ) 7.3 Hz,
4H), 7.63 (t, J ) 7.3 Hz, 4H), 7.73 (d, J ) 7.3 Hz, 4H), 7.74 (s,
4H), 8.83 (br s, 4H), 8.95 (s, 8H); λmax (solvent A) 417 (log ꢀ )
5.67), 514, 548, 587, 643 nm; FABMS m/ z 679.4 (MH+).
m eso-Tetr a k is(3-(d od ecyloxy)p h en yl)p or p h yr in (P 4).
Extraction: 2 × 100 mL and then 50 mL of EtOAc (phase
separation was best seen with the help of a hand-held UV lamp
since both aqueous and organic layers were very dark).
Chromatography: 25-35% CH2Cl2 in hexane. The crude
porphyrin was dissolved in warm MeOH with the minimum
volume of CH2Cl2. On standing, a slowly solidifying shiny
purple oil was deposited. This was washed with MeOH and
dried at 150 °C (0.1 mmHg) for 12 h to give porphyrin P 4 (33
mg, 9%): 1H NMR (CDCl3, 250 MHz) δ -2.80 (br s, 2H), 0.84
(br t, 12 H), 1.12-1.55 (m, 72H), 1.86 (m, 8H), 4.13 (br t, 8H),
7.31(d, J ) 7.3 Hz, 4H), 7.61 (t, J ) 7.3 Hz, 4H), 7.77 (s, 4H),
7.78 (d, J ) 7.3 Hz, 4H), 8.88 (s, 8H); λmax (solvent A) 418 (log
ꢀ ) 5.73), 514, 547, 587, 643 nm; HRFABMS m/ z 1351.9851
(MH+), C92H127N4O4 requires 1351.9857. Anal. Calcd for
C
44H31N4O4 requires 679.2341.
Other porphyrins were prepared on the same scale from the
appropriate aldehydes using the acid concentrations and
reaction times listed in Table 1. The solvents for extraction
(if different solvents or solvent volumes were used), chroma-
tography, and crystallization are summarized for each por-
phyrin, along with any other differences in procedure. Unless
indicated otherwise, reaction mixtures were neutralized with
aqueous KOH, adding enough of an aqueous potassium salt
before extraction to convert the surfactant into its potassium
salt. This material is only slightly soluble in water and most
organic solvents and remains largely suspended in the aqueous
layer (magnesium or calcium salts were also effective). EtOAc
was found to be the best solvent for extractions because it is
a good solvent for polar porphyrins and has little tendency to
form emulsions. Chlorinated solvents could also be employed,
but generally give slower phase separations.
C
92H126N4O4: C, 81.73; H 9.39; N, 4.14. Found: C, 81.72; H,
9.45; N, 4.02.
m eso-Tetr a k is(4-m eth oxyp h en yl)p or p h yr in (P 6).20b,22a
Chromatography: 5% THF in CH2Cl2. The crude porphyrin
was refluxed briefly in methanol, and the crystals were washed
well with methanol and dried at 200 °C (0.1 mmHg) for 12 h
to give porphyrin P 6 (34 mg, 18%): 1H NMR (CDCl3, 200 MHz)
δ -2.75 (br s, 2H), 4.09 (s, 12H), 7.28 (d, J ) 8.5 Hz, 8H), 8.11
(d, J ) 8.5 Hz, 8H), 8.65 (s, 8H); λmax (solvent A) 420 (log ꢀ )
5.66), 516, 554, 591, 649 nm; FABMS m/ z 635.7 (MH+).
m eso-Tetr a k is(4-(2-(m eth oxyca r bon yl)eth oxy)p h en yl)-
p or p h yr in (P 7). Aqueous potassium hydroxide (15 mL, 2 M),
aqueous potassium chloride (15 mL, 3 M), and water (100 mL)
were added after TCQ oxidation, and the precipitated potas-
sium dodecyl sulfate was filtered from the basic reaction
mixture using a coarse glass frit and washed with water. The
filtrate was acidified with aqueous HCl until the porphyrin
precipitated (pH ∼ 5), filtered through a 3 × 3 cm plug of flash
silica gel, and washed with water. The porphyrin-containing
silica gel was partially dried under aspirator pressure, and
then porphyrin was eluted with 10% MeOH in THF and the
filtrate evaporated. The crude porphyrin tetraacid was stirred
overnight in a mixture of 10% w/w HCl in MeOH (50 mL) and
CH2Cl2 (30 mL). The volatiles were evaporated, and a solution
of the crude porphyrin tetraester in CH2Cl2 was washed with
pH 7 buffer, concentrated to a small volume, and applied
directly to a silica gel column, eluting with 1-2% THF in
CH2Cl2. The product was washed with a little THF and then
MeOH and dried at 100 °C (0.1 mmHg) for 12 h to give P 7 (82
mg, 34%) as a purple powder: 1H NMR (CDCl3, 250 MHz) δ
-2.81 (br s, 2H), 3.94 (s, 12H), 4.93 (s, 8H), 7.28 (d, J ) 8.5
Hz, 8H), 8.12 (d, J ) 8.5 Hz, 8H), 8.84 (s, 8H); λmax (solvent A)
419 (log ꢀ ) 5.72), 516, 552, 590, 647 nm; FABMS m/ z 967.9
(MH+).
The progress of DDQ or TCQ oxidations was conveniently
followed by injecting 5 µL of reaction mixture directly into
solvent A in a cuvette. Porphyrin concentrations were usually
monitored by UV at various stages during purification to check
for material recovery. Porphyrins bearing polar functionality
frequently crystallize with solvent or water of crystallization,29
and prolonged drying at high temperatures was sometimes
necessary to produce material of constant weight. All isolated
porphyrins were pure by 1H NMR with no residual organic
solvent.
m eso-Tet r a k is(4-m et h ylp h en yl)p or p h yr in
(P 1).22a
Extraction: 100 mL and then 2 × 50 mL of CH2Cl2 (slight
emulsion formation). Chromatography: 50% hexane in CH2Cl2.
The crude porphyrin was refluxed briefly in methanol, and the
crystals were washed well with methanol and dried at 200 °C
(0.1 mmHg) for 12 h to give P 1 (25 mg, 15%): 1H NMR (CDCl3,
250 MHz) δ -2.79 (br s, 2H), 2.70 (s, 12H), 7.54 (d, J ) 8.5
Hz, 8H), 8.09 (d, J ) 8.5 Hz, 8H), 8.84 (s, 8H); λmax (solvent A)
417 (log ꢀ ) 5.69), 515, 551, 589, 645 nm; FABMS m/ z 671.4
(MH+).
(29) (a) Byrn, M. P.; Curtis, C. J .; Hsiou, Y.; Khan, S. I.; Sawin, P.
A.; Tendick, S. K.; Terzis, A.; Strouse, C. E. J . Am. Chem. Soc. 1993,
115, 9480. (b) Krupitsky, H.; Stein, Z.; Goldberg, I. Z. Kristallogr. 1995,
210, 665.
(30) Sta¨ubli, B.; Fretz, H.; Piantini, U.; Woggon, W.-D. Helv. Chim.
Acta 1987, 70, 1173.

Multicomponent Assembly in Micelles
J . Org. Chem., Vol. 61, No. 11, 1996 3631
m eso-Tetr a k is(3,4-d ih yd r oxyp h en yl)p or p h yr in (P 8).31
Extraction: 200 mL and then 150 mL of EtOAc. Chro-
matography: 30% THF in CH2Cl2 containing 0.01% v/v AcOH.
The crude porphyrin was crystallized from THF/cyclohexane
and dried at 100 °C (0.1 mmHg) for 24 h to give P 8 (89 mg,
48%) as a purple-black microcrystalline powder: 1H NMR (d6-
acetone, 250 MHz) δ -2.73 (br s, 2H), 7.27 (d, J ) 8.3 Hz,
4H), 7.57 (dd, J ) 8.3, 2.2 Hz, 4H), 7.73 (d, J ) 2.2 Hz, 4H),
8.39 (br s, 8H), 8.97 (s, 8H); λmax (solvent A) 423 (log ꢀ ) 5.51),
518, 556, 591, 648 nm; HRFABMS m/ z 743.2138 (MH+),
C44H31N4O8 requires 743.2142.
m eso-Tet r a k is(3,5-d ih yd r oxyp h en yl)p or p h yr in (P 9).
Extraction: 200 mL and then 100 mL of EtOAc. Chro-
matography: 30% THF in CH2Cl2 containing 0.01% v/v AcOH.
The crude porphyrin was washed twice with CH2Cl2, and the
purple flakes were recrystallized from THF/cyclohexane and
dried at 150 °C (0.1 mmHg) for 24 h to give P 9 (65 mg, 34%)
as a purple powder: 1H NMR (d6-acetone, 250 MHz) δ -2.81
(br s, 2H) 6.85 (t, J ) 2.1 Hz, 4H), 7.25 (d, J ) 2.1 Hz, 8H),
8.71 (br s, 8H), 9.03 (s, 8H); λmax (solvent A) 419 (log ꢀ ) 5.47),
514, 548, 587, 642 nm; HRFABMS m/ z 743.2200 (MH+),
C44H31N4O8 requires 743.2142.
m eso-Tet r a k is(4-h yd r oxy-3-m et h oxyp h en yl)p or p h y-
r in (P 10).19d Extraction: 2 × 100 mL of EtOAc. Chro-
matography: 1-5% THF in CH2Cl2. The crude porphyrin was
crystallized from THF/MeOH, washed thoroughly with MeOH,
and dried at 180 °C (0.1 mmHg) for 12 h to give P 10 (42 mg,
21%) as a purple powder: 1H NMR (d6-acetone, 250 MHz) δ
-2.72 (br s, 2H), 4.02 (s, 12H), 7.28 (d, J ) 8.3 Hz, 4H), 7.68
(br d, J ) 8.3 Hz, 4H), 7.85 (s, 4H), 8.14 (br s, 4H), 8.96 (s,
8H); λmax (solvent A) 423 (log ꢀ ) 5.61), 518, 555, 590, 650 nm;
FABMS m/ z 799.8 (MH+).
mmHg) for 12 h to give P 13 (52 mg, 25%) as purple crystals:
1H NMR (d6-acetone, 250 MHz) δ -2.73 (br s, 2H), 2.26 (s,
12H), 8.10, 8.18 (ABq, J ) 8.8 Hz, 16H), 8.92 (s, 8H), 9.64 (br
s, 4H); λmax (solvent A) 421 (log ꢀ ) 5.67), 517, 554, 591, 648
nm; FABMS m/ z 843.7 (MH+).
m eso-Tet r a k is(4-(t r iflu or oa cet a m id o)p h en yl)p or p h -
yr in (P 14). The starting aldehyde was not completely soluble
in 0.5 M SDS, but a clear solution was formed ∼30 min after
the addition of acid. Extraction: 150 mL and then 50 mL of
EtOAc. Chromatography: 3-7.5% THF in CH2Cl2, discarding
contaminated tailings. The crude product was crystallized
twice from THF/cyclohexane and dried at 180 °C 0.1 mm Hg
for 12 h to give P 14 (107 mg, 40%) as a purple powder: 1H
NMR (d6-acetone, 400 MHz) δ -2.75 (br s, 2H), 8.23, 8.31 (ABq,
J ) 8.5 Hz, 16H), 8.94 (s, 8H), 10.74 (br s, 4H); λmax (solvent
A) 419 (log ꢀ ) 5.71), 515, 552, 590, 646 nm; HRFABMS m/ z
1059.2247 (MH+), C52H31N8O4F12 requires 1059.2276.
m eso-Tetr a k is(fu r a n -2-yl)p or p h yr in (P 15).19d Extrac-
tion: 150 mL and then 100 mL of EtOAc. Chromatography:
1-5% THF in CH2Cl2. The crude porphyrin, which contained
some coeluted SDS, was slurried with MeOH, filtered, and
washed well with MeOH. Crystallization from THF/cyclohex-
ane gave P 15 (15 mg, 10%) as small purple crystals after being
dried at rt 0.1 mmHg for 2 days. 1H NMR (d6-acetone, 400
MHz) δ -2.70 (br s, 2H), 7.50 (dd, J ) 3.3, 2 Hz, 4H), 7.48 (d,
J ) 3.3 Hz, 4H), 8.32 (d, J ) 2 Hz, 4H), 9.23 (s, 8H); λmax
(solvent A) 430 (log ꢀ ) 5.37), 525, 572, 670 nm; HRFABMS
m/ z 575.1702 (MH+), C36H23N4O4 requires 575.1719.
This porphyrin was previously prepared in trace amounts
by Triebs and Haberle.19d P 15 decomposes slowly in solution,
coating glassware with a yellow film, but appears to be quite
stable when crystalline.
m eso-T e t r a k is (3,4,5-t r ih y d r o x y p h e n y l)p o r p h y r in
(P 11).32 The starting aldehyde required ∼45 min of stirring
to dissolve completely. Extraction: 2 × 100 mL and then 50
mL of EtOAc. Chromatography: 0.2% v/v AcOH in EtOAc
(41% porphyrin recovery as measured by UV). The crude
porphyrin was crystallized under argon from THF/cyclohexane
and dried at 100 °C (0.1 mmHg) for 12 h to give P 11 (24 mg,
12%) as a black powder: 1H NMR (d6-DMSO, 250 MHz) δ
-2.96 (br s, 2H), 7.11 (s, 4H), 8.60 (br s, 4H), 8.93 (s, 8H),
9.31 (br s, 8H); λmax (solvent A) 421 (log ꢀ ) 5.32), 517, 555,
591, 649 nm; HRFABMS m/ z 807.1955 (MH+), C44H31N4O12
requires 807.1938.
The absorbance of this porphyrin decreases slowly on being
allowed to stand in solvent A. P 11, previously prepared by
demethylation of its permethyl ether, showed the same
changes in UV-vis spectra on addition of base and acid as
reported by Milgrom et al.32
m eso-Tet r a k is(3-a cet a m id op h en yl)p or p h yr in (P 12).
The starting aldehyde required 1 h of stirring to dissolve
completely. DDQ (200 mg) in THF (5 mL) was used as the
oxidant. Extraction: 200, 150, and then 100 mL of EtOAc.
Chromatography: 30% THF in CH2Cl2, discarding some green-
ish forerun and contaminated tailings. The crude porphyrin
was crystallized from MeOH/trichloroethylene, washed well
with trichloroethylene, and dried at 180 °C (0.1 mmHg) for
12 h to give P 12 (50 mg, 23%) as purple crystals: 1H NMR
(d6-acetone, 250 MHz) δ -2.76 (br s, 2H), 2.16 (s, 12H), 7.75
(br t, 4H), 7.95 (br d, 4H), 8.17 (br d, 4H), 8.56 (br s, 4H), 8.94
(s, 8H), 9.54 (br s, 4H); λmax (solvent A) 418 (log ꢀ ) 5.65), 513,
548, 587, 643 nm; HRFABMS m/ z 843.3401 (MH+), C52H43N8O4
requires 843.3407.
m eso-Tetr a k is(4-a ceta m id op h en yl)p or p h yr in (P 13).19d
Extraction: 150, 100, and then 50 mL of EtOAc. Chro-
matography: 1-10% MeOH in 1:1 THF/CH2Cl2, discarding
fast running yellow and brown bands and contaminated
tailings. The crude product was dissolved in a small volume
of hot pyridine and hot trichloroethylene added until crystals
began to separate. This process was repeated, and the product
washed was well with methanol and dried at 180 °C (0.1
m eso-Tet r a k is(t h ien -2-yl)p or p h yr in (P 16).19d Extrac-
tion: 150 mL and then 50 mL of EtOAc. Chromatography:
0.01% pyridine in CH2Cl2. The crude porphyrin was slurried
with MeOH, filtered, washed well with MeOH, and dried at
100 °C 0.1 mm Hg for 12 h to give P 16 (39 mg, 24%) as purple
crystals: 1H NMR (CDCl3, 400 MHz) δ -2.66 (br s, 2H), 7.50
(dd, J ) 5.3, 3.4 Hz, 4H), 7.85 (dd, J ) 5.3, 1.3 Hz, 4H), 7.91
(dd, J ) 3.4, 1.3 Hz, 4H), 9.03 (s, 8H); λmax (solvent A) 430 (log
ꢀ ) 5.56), 525, 572, 670 nm; HRFABMS m/ z 639.0785 (MH+),
C36H23N4S4 requires 639.0806.
P or p h in e.33 Hydrochloric acid (0.25 mL, sp gr 1.16) was
added to a stirred solution of 2-(hydroxymethyl)pyrrole (97 mg,
1.0 mmol) in aqueous SDS (100 mL, 0.5 M). The solution
rapidly turned bright orange. DDQ (200 mg) in THF (5 mL)
was added after 10 min, and after a further 30 min the reaction
mixture was worked up in the usual manner. Extraction: 3
× 100 mL of EtOAc (the aqueous layer was filtered after the
first extraction to remove finely divided black material).
Volatiles were evaporated, and a solution of the crude material
in CH2Cl2 was eluted through a short column of silica gel. The
crude porphin thus obtained (∼2 mg, 2% UV yield assuming
logꢀ ) 5.2628) was not purified further.
P er -â-d eu t er a t ed m eso-Tet r a k is(4-h yd r oxyp h en yl)-
p or p h yr in (d 8-P 5). This reaction was run with
a low
surfactant/substrate ratio (R ) 13) to maximize the amount
of porphyrin produced per mL of D2O rather than the chemical
yield. A mixture of 4-hydroxybenzaldehyde (50 mg, 0.41
mmol), pyrrole (28.5 µL, 0.41 mmol), and SDS (1.5 g, 5.2 mmol)
was stirred in D2O (5 g) under argon until a clear solution
was obtained. Deuteriotrifluoroacetic acid (32 µL, 0.41 mmol)
was added, followed after 15 min by a suspension of TCQ (100
mg, 0.39 mmol) in dry THF (1 mL). After overnight stirring,
workup as for P 5 and crystallization from THF/cyclohexane
gave d 8-P 5 (8.5 mg, 12%). 1H NMR (20% d6-acetone in CDCl3)
integration of the residual â-pyrrole resonance indicated 97.5%
deuterium incorporation: FABMS m/ z 687.5 (MH+).
P er -â-deu ter ated meso-Tetr akis(3,4-dih ydr oxyph en yl)-
p or p h yr in (d 8-P 8). The same procedure as for d 8-P 5 was
followed starting from 3,4-dihydroxybenzaldehyde (50 mg, 0.36
(31) Bogatskii, A. V.; Zhilina, Z. I.; Stepanov, D. E. Zh. Org. Khim.
1982, 18, 2309.
(32) Alberry, W. J .; Bartlett, P. N.; J ones, C. C.; Milgrom, L. R. J .
Chem. Res., Synop. 1985, 364.
(33) Longo, F. R.; Thorne, E. J .; Adler, A. D.; Dym, S. J . Heterocycl.
Chem. 1975, 12, 1309.

3632 J . Org. Chem., Vol. 61, No. 11, 1996
Bonar-Law
Ta ble 2. Rela tive P er cen t Yield s of P or p h yr in s fr om P or p h yr in ogen Mixin g Exp er im en ts
A ) 17, B ) 28
[A + B]a
A ) 13, B ) 27
[A + B]
statistical
distribution
λmax (nm)
[A] + [B]b
λmax (nm)
[A] + [B]
A4
6.25
25
37.5
25
417.0
417.6
418.2
418.8
419.2
8c
28
37
23
4
37
13
9
5
36
422.9
422.5
422.0
421.4
421.0
7
19
41
27
6
34
11
14
8
A3B
A2B2
AB3
B4
6.25
33
a
b
SDS solutions of A and B mixed before addition of acid. DDQ oxidation at tmax
.
SDS solutions of A and B condensed separately with
pyrrole and then mixed at tmax. DDQ oxidation at 5tmax
.
c Relative yields reproducible to within a few percent.
mmol) to give d 8-P 8 (10.5 mg, 15%). 1H NMR (d6-acetone)
indicated >98% deuterium incorporation: FABMS m/ z 751.3
(MH+).
aldehyde (13) was performed in exactly the same fashion. After
the additions had been completed, yields from 15 stayed
constant, but yields from 13 declined slowly. Yields measured
5 h after completion of addition were ∼65% of those after 15
min.
Con tr ol Rea ction s in Or ga n ic Solven ts. Aqueous HCl
(20 µL, 10 M) was added to a solution of benzaldehyde or
4-hydroxybenzaldehyde (0.2 mmol) and pyrrole (0.2 mmol) in
a solvent (0.4, 2, or 4 mL) and the mixture stirred under argon.
Aliquots of reaction mixture were withdrawn periodically and
oxidized with DDQ, measuring the porphyrin content by UV
absorbance as described above. Solvents used were MeOH,
EtOH, THF, and DMF. Trifluoroacetic acid (1 equiv) was also
employed as catalyst for the reaction of benzaldehyde and
pyrrole in MeOH, THF, and DMF at 10 mM reactants. Yields
in all cases were less than 5%.
P or p h yr in ogen Exch a n ge Exp er im en ts. The procedure
is illustrated for the pair of aldehydes 3-hydroxybenzaldehyde
(13) and 3,5-dihydroxybenzaldehyde (27). Three solutions
were prepared. Solution A contained aldehyde 13 (0.1 mmol)
and pyrrole (0.1 mmol) in aqueous SDS (10 mL, 0.5 M).
Solution B was the same except with aldehyde 27. Solution
C was prepared by mixing 3.33 mL of A and 3.33 mL of B in
a separate flask so that A, B, and C were all of a volume of
6.66 mL. Reaction was initiated by addition of aqueous HCl
(200 µL, 10 M) to A, B, and C. After 1 h, DDQ (15 mg) in
THF (0.5 mL) was added to C, and solutions A and B were
mixed. Samples (2 mL) of the A/B mixture were then
withdrawn at various times and added to DDQ (4 mg) in THF
(0.5 mL). After the solution was allowed to stand for 1 h,
sufficient potassium phosphate buffer (pH 7, 1 M) was added
to the oxidized samples to precipitate surfactant and the crude
porphyrins were isolated by extraction with EtOAc. Prepara-
UV yields and conditions for synthesis under Lindsey’s
conditions (10 mM reactants in anhydrous chlorinated solvent)6c
are summarized in the format porphyrin {aldehyde, solvent,
catalyst, maximum yield, time}: P 2 {2-hydroxybenzaldehyde,
CH2Cl2 or CHCl3 (with or without EtOH), 20 mM TFA or 3.3
mM BF3‚Et2O, <1%}, P 3 {3-hydroxybenzaldehyde, CH2Cl2, 20
mM TFA, 22%, 2 h}, P 5 {4-hydroxybenzaldehyde, CH2Cl2, 20
mM TFA, 22%, 1 h}, P 8 {3,4-dihydroxybenzaldehyde, CH2Cl2,
33 mM TFA, 10%, 3 h}, P 9 {3,5-dihydroxybenzaldehyde,
CH2Cl2, 33 mM TFA, 5%, 6 h}, P 11 {3,4,5-trihydroxybenz-
aldehyde, CH2Cl2, 20 mM TFA, <1%}, P 32 {2-acetamidobenz-
aldehyde, CH2Cl2, 3.3 mM BF3‚Et2O, 3%, 8 h}. BF3‚Et2O (8
mM) in place of TFA gave much lower yields for P 3 and P 5.
1H NMR Stu d y of th e Rea ction betw een 4-Hyd r oxy-
ben za ld eh yd e a n d P yr r ole. 4-Hydroxybenzaldehyde (12.2
mg, 0.1 mmol) and pyrrole (7 µL, 0.1 mmol) were stirred under
argon in 0.825 M SDS in D2O (1.0 mL) until a clear solution
was obtained. A portion of this solution (25 µL) was injected
into 0.14 M SDS in D2O (0.575 mL) in an argon-filled NMR
tube, and deuteriotrifluoroacetic acid (5 µL, 65 µmol) was
added to initiate reaction. The pyrrole resonances dissap-
peared immediately, and the aromatic resonances from the
starting aldehyde {9.69 (s, 1H, CHO), 7.83 (d, 2H), and 7.0 (d,
2H) ppm} began to reduce in intensity with the appearance
of another set of resonances {7.1 (m, 2H), 6.8 (m, 2H), and 5.3
(m, 1H) ppm}. After 15 min, the reaction had reached ∼50%
conversion and was stopped at this stage by addition of 60 µL
of buffered aqueous NaOH (0.725 M Na2HPO4/NaH2PO4 pH
6.5 buffer which had been made 1.05 M in NaOH). The final
mixture of species was of approximate composition: 1.9 mM
aldehyde, 1.9 mM pyrrole, 1.9 mM product(s), 160 mM SDS,
96 mM sodium trifluoroacetate, and 65 mM buffer salts. The
intensities of the reactant and product peaks did not change
further during subsequent measurements. Longitudinal re-
laxation rates (R1) were measured by inversion recovery in
standard fashion. R1 values for aldehyde, product(s), and the
tive TLC (15% MeOH in CH2Cl2) of a portion of the 5 h (5tmax
)
sample from the A/B mixture gave P 3, P 9, and three mixed
porphyrins of intermediate polarity. A portion of the reaction
mixture from C was also chromatographed. The relative yields
given in Table 2 were calculated from absorbance measure-
ments, assuming that extinction coefficients for mixed por-
phyrins were weighted linear combinations of extinction
coefficients for P 3 and P 9. The A2B2 fraction was assumed to
be an unresolved mixture of the two possible isomeric porphy-
rins. The remaining portions of the 5 h sample and the C
reaction mixture were eluted separately through short columns
of silica gel to remove base line material and analyzed by
FABMS. The relative intensities of the five MH+ peaks (679
(A4 ) P 3), 695 (A3B), 711 (A2B2), 727 (AB3), and 743 (B4
)
P 9)) were in qualitative agreement with chromatographically
isolated porphyrin yields.
In similar experiments, solid aldehyde 13 or 27 (0.1 mmol)
along with an equivalent of pyrrole was added separately to
solutions of porphyrinogen derived from the other aldehyde 1
h after the reaction had been initiated, and the products were
analyzed as above. Adding aldehyde 13 to the porphyrinogen
formed from 27 gave 54% exchange after 5tmax, and the
opposite order of addition gave 63% exchange after the same
time.
For aldehydes 17 and 28 (Table 2), the reaction was initiated
with aqueous HCl (100 µL, 10 M), solutions were mixed after
0.5 h, and mixtures were chromatographed with 7% MeOH in
CH2Cl2. For aldehydes 15 and 37, the reaction was initiated
with aqueous HCl (200 µl, 10 M), and the solutions were mixed
after 1 h. For this pair of aldehydes only three fractions were
isolated by chromatography (CH2Cl2): P 4, a mixed fraction of
all four resolvable crossover products, and meso-tetrakis(4-
(trichloroacetamido)phenyl)porphyrin (FABMS m/ z 1256.6
(MH+)).
Slow Ad d ition Exp er im en ts. A mixture of 3-(dodecyl-
oxy)benzaldehyde (15) (1.0 mmol) and pyrrole (1.0 mmol) was
stirred in aqueous SDS (10 mL, 0.5 M) until a clear solution
was obtained. Portions (1.0 mL) of this solution were added
by syringe pump at different rates to separate flasks contain-
ing aqueous SDS (9.0 mL, 0.5 M) and aqueous HCl (0.3 mL,
10 M). Samples were withdrawn at 15 min, 1 h, and 5 h after
completion of addition, and the porphyrin concentration was
assayed in the usual manner. Slow addition of 3-hydroxybenz-
main SDS peak at 1.3 ppm were 0.4, 0.8, and 1.5 s-1
,
respectively. Self-diffusion coefficients (D) were measured
with a pulsed-gradient spin-echo (PGSE) sequence34a which
minimized transverse relaxation effects.34b The diffusion
coefficient (corrected for R1 weighting) for the product(s) was
0.6 relative to aldehyde (i.e. Daldehyde ) 1) and for SDS was 0.4
relative to aldehyde. Similar results were obtained from
experiments in which R1 rates and relative D values were
(34) (a) Stilbs, P. J . Colloid Interface Sci. 1983, 94, 463. (b) Tanner,
J . E. J . Chem. Phys. 1970, 52, 2523.

Multicomponent Assembly in Micelles
J . Org. Chem., Vol. 61, No. 11, 1996 3633
measured before reaction and after complete disappearance
of starting material. Less acid was added in these experiments
so that condensation proceeded more slowly and the por-
phyrinogen concentration did not decrease significantly during
the final measurements.
The percentage yield of cyclic n-mer (100R/[L0]) was obtained
by solving for z with various values of [L0], K, Kc, P, and â.
For assembly in a single phase, P ) 1 and eq 2 reduces to a
special case of ring-chain equilibrium in homogeneous
solution.35b Given the simplicity of the model, the numerical
values of Km and Kc necessarily reflect the various approxima-
tions made, and so it is the relative valves rather than the
absolute values of these parameters which are of significance.
More elaborate schemes can be constructed at the expense of
extra parameters, but the qualitative conclusion that there is
an optimum reactant concentration for reversible assembly in
a two-phase system provided binding energies that are ap-
proximately additive remains unchanged.
P a r tition Coefficien ts. The partition coefficients in Table
1 were measured in 100 mM SDS containing 50 mM pH 7
buffer by micellar capillary electrophoresis18 using an uncoated
75 µm, 70 cm long silica gel capillary. The column tempera-
ture was 27 °C voltage 25 kV, current ∼50 µA, and detection
by UV at 200 or 250 nm. One to five microliters of ligand
solutions (2-10 mg/mL in the mobile phase or saturated
solutions for the less soluble aldehydes) were added succes-
sively to 200 µL of the mobile phase in the source vial, running
a chromatogram after each addition to identify the peaks.
(Dodecyloxy)benzene or TPP was used as markers to measure
the micelle retention time (tm). After the ligands and the
micelle marker had been added, the solvent front time (t0) was
obtained by adding enough methanol to the mixture (1-2%
v/v) to produce a detectable inflection in the base line.
Provided the amount of methanol added was small, the
retention times of ligands and marker were not significantly
affected. P values were calculated from P ) (tr/t0 - 1)/{(1 -
tr/tm)([SDS] - cmc)vm}, where tr is the solute retention time.
The reaction between 4-hydroxybenzaldehyde and pyrrole
was monitored by micellar capillary electrophoresis in the
following manner. A portion (200 µL) of a solution of the
reactants in 100 mM SDS (1 mM in both aldehyde and pyrrole)
was placed in the source vial, and a chromatogram was run
using 100 mM SDS containing 100 mM pH 7 buffer as the
mobile phase (detection at 200 nm). Aqueous HCl (20 µL, 1.0
M) was then added to the source vial, and chromatograms were
run after various times. After the starting materials had
disappeared, (decyloxy)benzene was added to mark the mi-
celles (arrow in Figure 5). Adding the marker before the
addition of acid gave the same results. Transient broad peaks
appeared with retention times between reactants and prod-
uct(s) during the course of reaction, but were not well resolved.
The reaction was also run on a larger scale, analyzing
aliquots by electrophoresis and UV absorbance in parallel with
the same results as the above in situ method.
Ald eh yd es. Commercially available aldehydes were used
as received. 4-Acetamidobenzaldehyde and 4-hydroxybenz-
aldehyde were recrystallized from water.
Methyl 2-formylbenzoate (6)37 was prepared from 2-formyl-
benzoic acid by alkylation with methyl iodide. Methyl 3-formyl-
benzoate (8)38 was prepared by chromium trioxide-acetic
anhydride oxidation39 of 3-methylbenzoic acid and treatment
of the resulting 3-bis(acetoxymethyl)benzoic acid with HCl in
anhydrous MeOH followed by aqueous workup.
2-(Hydroxymethyl)pyrrole was obtained as a yellow oil by
reduction of 2-formylpyrrole following the literature method,40
but using 1 equiv of NaBH4. The crude material was pure by
1H NMR and was used directly.
5-F or m ylben zen e-1,3-d ica r boxylic a cid d im eth yl ester
(9) was prepared in the same manner as 8 from 5-methyl-
benzene-1,3-dicarboxylic acid.41 For 9: mp 86-88 °C; 1H NMR
(CDCl3, 250 MHz) δ 3.98 (s, 6H), 8.71 (d, J ) 1.5 Hz, 2H),
8.91 (t, J ) 1.5 Hz, 1H), 10.12 (s, 1H); EI m/ z 222.1 (M+, 50%),
191 (100), 163 (10).
3-(Dod ecyloxy)ben za ld eh yd e (15).42 A mixture of 3-hy-
droxybenzaldehyde (1.22 g, 10 mmol), 1-bromododecane (2.24
g, 9 mmol), and K2CO3 (1.5 g) in DMF (5 mL) was stirred under
argon at 80 °C. After 30 min the mixture was cooled and
partitioned between EtOAc and water. The organic layer was
washed successively with 1 M aqueous NaOH, water, and pH
7 buffer and then dried and evaporated to a yellow oil. This
was eluted through a short plug of silica gel to give 15 (2.3 g,
80%) as a colorless slowly crystallizing oil: 1H NMR (CDCl3,
200 MHz) δ 0.87 (br t, 3H), 1.2-1.5 (m, 18H), 1.80 (m, 2H),
4.0 (t, J ) 6.5 Hz, 2H), 7.16 (m, 1H), 7.36 (m, 1H), 7.42 (m,
2H), 9.96 (s, 1H). Anal. Calcd for C19H30O2: C, 78.57; H,
10.41. Found: C, 78.67; H, 10.19.
Sim u la tion of P or p h yr in ogen Assem bly. The following
assumptions were made to simplify numerical treatment of
Scheme 2: (1) aldehyde and pyrrole can be treated as
equivalent monomer units, (2) a single equilibrium constant
can be used to describe stepwise chain growth,35a (3) water
activity is constant and the same in both phases, and (4)
The acids 16,43 19,43 and 2644 were obtained by addition of
a slight excess of aqueous 2 M KOH to solutions of their methyl
esters (prepared in near-quantitative yields by alkylation of
the corresponding phenols with a slight excess of methyl
bromoacetate using the procedure given above for 15) in
MeOH/THF and crystallization from hot water as needles.
3,4-Bis(2-ca r boxyeth oxy)ben za ld eh yd e (26):44 mp 244-
246 °C; 1H NMR (d6-DMSO, 250 MHz) δ 4.68 (s, 2H), 4.71 (s,
2H), 7.04 (d, J ) 8.4 Hz, 1H), 7.28 (d, J ) 1.5 Hz, 1H), 7.51
(dd, J ) 8.4, 1.5 Hz, 1H), 9.80 (s, 1H); EI m/ z (254, M+).
2-Aceta m id oben za ld eh yd e (32)45 was prepared in an
analogous fashion to 36 by acetylation of 2-aminobenzyl alcohol
with acetic anhydride (1 h, rt). The intermediate N-acetylated
benzyl alcohol was crystallized from THF/cyclohexane before
oxidation. Crude 32 was purified by crystallization from
cyclohexane (31% overall).
equilibrium constants in water and micelles are equal (Km
)
Kaq ) K). The last assumption requires that the linear i-mer
Li has the partition coefficient Pi, where P ) [L1,m]/[L1,aq] is
the partition coefficient of monomer L1. As discussed in the
main text, micellar binding energies (∆G ) -RT ln Pi ) -iRT
ln P) are in fact likely to be roughly additive. Assumption 3
is harder to justify, since there is presumably a gradient of
water concentration from the surface to the center of the
micelle. However, the reversible reactions involve polar
reactants that are likely to prefer the wet outer regions of
micelles, which may be quite extensive,1f,36 so constant water
activity seems reasonable. With assumptions 1-4, mass
balance on Scheme 2 gives
3-Aceta m id oben za ld eh yd e (33)46 was prepared from
[L0] ) (z/K){(1 - â)/(1 - z/P)2 + â/(1 - z)2} + R;
R ) n(Kc/K)zn{â + (1 - â)/Pn} (2)
(37) Brown, C.; Sargent, M. V. J . Chem. Soc. C 1969, 1818.
(38) Simonis, H. Chem. Ber. 1912, 45, 1584.
(39) Irreverre, F.; Kny, H.; Asen, S.; Thompson, J . F.; Morris, C. J .
J . Biol. Chem. 1961, 236, 1093.
(40) Silverstein, R. M.; Ryskiewicz, E. E.; Chaikin, S. W. J . Am.
Chem. Soc. 1954, 76, 4485.
(41) Armitage, B. J .; Kenner, G. W.; Robinson, M. J . T. Tetrahedron
1964, 20, 723.
(42) Ulian, F.; Vio, L. Farmaco, Ed. Sci. 1969, 24, 528.
(43) Elkan, T. Chem. Ber. 1886, 19, 3041.
(44) Rodriguez-Tabor, A.; Aran V.; Vazquez, D. Eur. J . Biochem.
1984, 139, 287.
where [L0] is the bulk concentration of monomer units, z )
K[L1,m] and is the degree of polymerization of the chain fraction
in the micellar phase, â is the phase ratio defined in the main
text, and n is the number of monomer units in the ring in
equilibrium with linear polymer (n ) 8 for a porphyrinogen).
(35) (a) Flory, P. J . Principles of Polymer Chemistry; Cornell
University Press: Ithaca, New York, 1953; p 317. (b) J acobson, H.;
Stockmayer, W. H. J . Chem. Phys. 1950, 18, 1600.
(36) Menger, F. M.; Mounier, C. E. J . Am. Chem. Soc. 1993, 115,
12222.
(45) Gassman, P. G.; Drewes, H. R. J . Am. Chem. Soc. 1978, 100,
7600.
(46) Davey, W.; Gwilt, J . R. J . Chem. Soc. 1957, 1008.

3634 J . Org. Chem., Vol. 61, No. 11, 1996
Bonar-Law
3-aminobenzyl alcohol in the same way as 32. Crude 33 was
purified by crystallization from toluene (37% overall).
mL) and THF (4 mL), and the brown suspension was stirred
rapidly at rt. After 2.5 h, the reaction mixture was diluted
with EtOAc and filtered into a separating funnel. The tarry
residue was extracted several times with EtOAc, and the
combined organic layers were washed with saturated aqueous
NaHCO3 and brine and evaporated. Crystallization from
MeOH/water (decolorizing charcoal) gave 36 as white crystals
(600 mg, 68%): mp 185 (softens)-200 °C; 1H NMR (d6-acetone,
250 MHz) δ 7.98 (narrow ABq, 4H), 10.01 (s, 1H), 10.58 (br s,
1H). Anal. Calcd for C9H6NO2F3: C, 49.78; H, 2.79; N, 6.45.
Found: C, 49.86; H, 2.86; N, 6.31.
3-((ter t-Bu toxyca r bon yl)a m in o)ben za ld eh yd e (34) was
obtained from 3-aminobenzyl alcohol following the literature
procedure for 4-((tert-butoxycarbonyl)amino)benzaldehyde47
and crystallized as needles from hexane, mp 93 °C. For 34:
1H NMR (CDCl3, 250 MHz) δ 1.52 (s, 9H), 6.65 (br s, 1H), 7.44
(t, J ) 7.7 Hz, 1H), 7.54 (d, J ) 7.7 Hz, 1H), 7.63 (d, J ) 7.7
Hz, 1H), 7.91 (s, 1H), 9.97 (s, 1H). Anal. Calcd for C12H15
-
NO3: C, 65.14; H, 6.82; N, 6.33. Found: C, 64.83; H, 6.71; N,
6.09.
4-(Tr iflu or oa ceta m id o)ben za ld eh yd e (36).48,49 Trifluo-
roacetic anhydride (0.75 mL, 5.1 mmol) was added dropwise
to a stirred solution of 4-aminobenzyl alcohol (0.5 g, 4.07 mmol)
in THF (3 mL) at ice-bath temperature. After 15 min, aqueous
NaOH (2.0 mL, 2.5 M) was added followed by saturated
aqueous NaHCO3 (2.0 mL) and MeOH (5 mL), and the
suspension was stirred at rt for 15 min to selectively hydrolyze
O-trifluoroacetates (TLC, 1:1 hexane/EtOAc). The reaction
mixture was partitioned between EtOAc and brine, and the
organic layer was dried and evaporated. Pyridinium chloro-
chromate (1.3 g, 6.0 mmol) was added to a solution of the crude
4-(trifluoroacetamido)benzyl alcohol in a mixture of CH2Cl2 (4
4-(Tr ich lor oa ceta m id o)ben za ld eh yd e (37)50 was pre-
pared in an analogous fashion to 36 by acetylation of 4-amino-
benzyl alcohol with trichloroacetic anhydride and crystallized
as yellow needles from MeOH/water (71%), mp 121 (softens)-
129 °C. For 37: 1H NMR (d6-acetone, 200 MHz) δ 7.98 (s, 4H),
10.01 (s, 1H), 10.26 (br s, 1H). Anal. Calcd for C9H6NO2Cl3:
C, 40.56; H, 2.27; N, 5.26. Found: C, 40.36; H, 2.34; N, 5.07.
Ack n ow led gm en t. Thanks to Christ’s College Cam-
bridge for financial support, Prof. J eremy Sanders for
use of facilities, Katherine Stott for help with NMR
experiments, and Prof. Alan Fersht for use of the
capillary electrophoresis instrument.
(47) Rai, R.; Katzenellenbogen, J . A. J . Med. Chem. 1992, 35, 4150.
(48) Teshima, T.; Griffin, J . C.; Powers, J . C. J . Biol. Chem. 1982,
257, 5085.
J O9600161
(49) Kwiatkowski, M.; Bandoli, G. J . Chem. Soc., Dalton Trans.
1992, 3, 379.
(50) Nyquist, R. A. Spectrochim. Acta 1963, 19, 1595.