Fluorinated porphyrin tweezer: A powerful reporter of absolute configuration for erythro and threo diols, amino alcohols, and diamines

Article abstract of DOI:10.1021/ja0752639

A general and sensitive nonempirical protocol to determine the absolute configurations of erythro and threo diols, amino alcohols, and diamines is reported. Binding of diols to the porphyrin tweezer system is greatly enhanced by increasing the Lewis acidi

Full text of DOI:10.1021/ja0752639

Published on Web 01/23/2008
Fluorinated Porphyrin Tweezer: A Powerful Reporter of
Absolute Configuration for erythro and threo Diols, Amino
Alcohols, and Diamines
Xiaoyong Li, Marina Tanasova, Chrysoula Vasileiou, and Babak Borhan*
Department of Chemistry, Michigan State UniVersity, East Lansing, Michigan 48824
Received July 14, 2007; E-mail: babak@chemistry.msu.edu
Abstract: A general and sensitive nonempirical protocol to determine the absolute configurations of erythro
and threo diols, amino alcohols, and diamines is reported. Binding of diols to the porphyrin tweezer system
is greatly enhanced by increasing the Lewis acidity of the metalloporphyrin. Supramolecular complexes
formed between the porphyrin tweezer host and chiral substrates exhibited exciton-coupled bisignate CD
spectra with predictable signs based on the substituents on the chiral center. The working model suggests
that the observed helicity of the porphyrin tweezer is dictated via steric differentiation experienced by the
porphyrin ring bound to each chiral center. A variety of erythro and threo substrates were investigated to
verify this chiroptical method. Their absolute configurations were unequivocally determined, and thus a
general mnemonic is provided for the assignment of chirality.
Introduction
the absolute stereochemical determination of the chiral molecule
based on the observed Exciton Coupled Circular Dichroism
Vicinal substituted heterofunctionalized chiral molecules such
as erythro and threo diols, amino alcohols, and diamines are
widely present in biologically active natural and synthetic
products such as alkaloids, polyketides, and carbohydrates. They
also play a crucial role in asymmetric catalysis, functioning as
fundamental building blocks or chiral auxiliaries and ligands.
The biological or catalytic activity of this class of compounds
is often governed by their configurations. Consequently, de-
velopment of methods to determine their absolute stereochem-
istry has been an active area of research.
(ECCD).1,2 Nonetheless, the need for derivatization is not ideal.
The absolute stereochemistry of erythro systems, however,
cannot be reliably assigned via existing strategies and remains
a challenging and largely unresolved issue.
Briefly, the ECCD method relies on the coupling of the
electric transition dipole moments of two or more chromophores
held in space in a chiral fashion.1 The sign of the resultant
ECCD couplet reflects the helicity of the interacting chro-
mophores and consequently the chirality of the derivatized
system (see Figure 1, dashed box). The challenge in utilizing
ECCD for absolute stereochemical determinations is not only
to orient two or more chromophores in a chiral fashion dictated
strictly by the chiral center but also to arrive at a system robust
enough that it provides consistent results with structurally
different compounds. Figure 1 illustrates the dibenzoate method
for threo and erythro diols. With threo diols, the expected high
population (low energy) rotomer leads to an observable and
distinct ECCD spectrum. For the threo diol depicted in Figure
1, the result is a positive ECCD irrespective of R1 and R2. The
expected high population rotomer for derivatized erythro diols,
however, is ECCD silent. The expected minor populations lead
to opposing ECCD spectra and thus cannot be used as a reliable
indicator of absolute stereochemistry.
For the most part, the absolute stereochemical determination
of threo substituted systems can be achieved with the imple-
mentation of the dibenzoate methodology.1-3 This is ac-
complished via derivatization of the functional groups with
benzoates (or other similar chromophores),4-15 which enables
(1) Harada, N.; Nakanishi, K. Circular Dichroic Spectroscopy: Exciton
Coupling in Organic Stereochemistry; University Science Books: Mill
Valley, CA, 1983.
(2) Berova, N.; Nakanishi, K.; Woody, R. W. Circular Dichroism: Principles
and Applications, 2nd ed.; Wiley-VCH: New York, 2000.
(3) Kawamura, A.; Berova, N.; Nakanishi, K.; Voigt, B.; Adam, G. Tetrahedron
1997, 53, 11961-11970.
(4) Wiesler, W. T.; Nakanishi, K. J. Am. Chem. Soc. 1989, 111, 9205-9213.
(5) Uzawa, H.; Nishida, Y.; Ohrui, H.; Meguro, H. J. Org. Chem. 1990, 55,
116-122.
(6) Harada, N.; Saito, A.; Ono, H.; Gawronski, J.; Gawronska, K.; Sugioka,
T.; Uda, H.; Kuriki, T. J. Am. Chem. Soc. 1991, 113, 3842-3850.
(7) Zhou, P.; Zhao, N.; Rele, D. N.; Berova, N.; Nakanishi, K. J. Am. Chem.
Soc. 1993, 115, 9313-9314.
Other approaches have been developed to transform acyclic
substrates into cyclic conformationally defined derivatives
through generation of cottonogenic derivatives with metal
(8) Zhou, P.; Berova, N.; Wiesler, W. T.; Nakanishi, K. Tetrahedron 1993,
49, 9343-9352.
(9) Cai, G. L.; Bozhkova, N.; Odingo, J.; Berova, N.; Nakanishi, K. J. Am.
Chem. Soc. 1993, 115, 7192-7198.
(10) Matile, S.; Berova, N.; Nakanishi, K.; Novkova, S.; Philipova, I.; Blagoev,
B. J. Am. Chem. Soc. 1995, 117, 7021-7022.
(13) Rele, D.; Zhao, N.; Nakanishi, K.; Berova, N. Tetrahedron 1996, 52, 2759-
2776.
(11) Zhao, N.; Berova, N.; Nakanishi, K.; Rohmer, M.; Mougenot, P.; Jurgens,
U. J. Tetrahedron 1996, 52, 2777-2788.
(14) Jiang, H.; Huang, X. F.; Nakanishi, K.; Berova, N. Tetrahedron Lett. 1999,
40, 7645-7649.
(12) Matile, S.; Berova, N.; Nakanishi, K.; Fleischhauer, J.; Woody, R. W. J.
Am. Chem. Soc. 1996, 118, 5198-5206.
(15) Weckerle, B.; Schreier, P.; Humpf, H. U. J. Org. Chem. 2001, 66, 8160-
8164.
9
10.1021/ja0752639 CCC: $40.75 © 2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 1885-1893
1885

A R T I C L E S
Li et al.
Figure 1. Excitonic coupling of two chromophores leads to an observable
ECCD, the sign of which reflects the helicity of the coupling chromophores
(dashed box). The major rotomer of derivatized threo diols leads to an
observable ECCD spectrum. The major rotomer of derivatized erythro diols
is ECCD silent.
complexes16-24 or conversion into dioxolanes25-28 or 4-biphe-
nylborates.29,30 These methods, however, suffer from fairly weak
CD signals due to weak effective transition moments of the
cottonogenic species, and few address the absolute stereochem-
ical determination of erythro compounds. A practical and
universally expedient method to unequivocally determine the
chirality of erythro diols, amino alcohols, and diamines is a
challenging task.
Porphyrin tweezer systems (Figure 2a) have been successfully
used to determine the stereochemistry of chiral amines,31-33
alcohols,32,34 and carboxylic acids.35-37 The principle advantage
of the porphyrin tweezer system resides with the noncovalent
Figure 2. (a) Structure of Zn-TPP tweezer 1. (b) Binding of a chiral
diamine to 1 leads to a helical disposition of the porphyrin rings dictated
by the sterics at the chiral center. The resultant ECCD is used to assign the
absolute stereochemistry of the chiral center. (c) Structure and synthesis of
Zn-TPFP tweezer 3.
(16) Dillon, J.; Nakanishi, K. J. Am. Chem. Soc. 1975, 97, 5409-5417.
(17) Dillon, J.; Nakanishi, K. J. Am. Chem. Soc. 1975, 97, 5417-5422.
(18) Snatzke, G.; Wagner, U.; Wolff, H. P. Tetrahedron 1981, 37, 349-361.
(19) Frelek, J.; Snatzke, G. Fresenius Z. Anal. Chem. 1983, 316, 261-264.
(20) Frelek, J.; Ikekawa, N.; Takatsuto, S.; Snatzke, G. Chirality 1997, 9, 578-
582.
binding of the chiral guest, which precludes the need for
chemical derivatizations. Prior work, however, has focused on
the absolute stereochemical determination of single chiral
centers. As depicted in Figure 2b, the binding of the Zn-
porphyrin tweezer to a diamine with a single chiral center leads
to an induced helical arrangement of the bound porphyrins that
yields a predictable ECCD spectrum.31 Since the tweezer
strategy is based on the steric interaction between the substit-
uents at the chiral center and one of the two porphyrins, it was
envisaged that the second porphyrin could participate in steric
differentiation with chiral guests that have two chiral centers
such as 1,2-bisfunctionalized systems. In other words, we
hypothesized that a system with two chiral centers could be
considered as two systems with a single chiral center each. In
this manner, erythro diols, amino alcohols, and diamines could
yield reproducible ECCD spectra based on the steric differentia-
tion at each chiral center. We predict that the porphyrin would
bind the hydroxyl group anti to the largest substituent on the
chiral center and stereodifferentiate between the two remaining
groups attached to the same chiral carbon. Upon steric dif-
ferentiation at each chiral center, the two porphyrins would adopt
a specific helicity. It should be noted that assuming a 1,2-
(21) Frelek, J.; Geiger, M.; Voelter, W. Curr. Org. Chem. 1999, 3, 117-146.
(22) Di Bari, L.; Pescitelli, G.; Pratelli, C.; Pini, D.; Salvadori, P. J. Org. Chem.
2001, 66, 4819-4825.
(23) Scott, A. I.; Wrixon, A. D. J. Chem. Soc., Chem. Commun. 1969, 1184-
1186.
(24) Di Bari, L.; Lelli, M.; Pintacuda, G.; Salvadori, P. Chirality 2002, 14, 265-
273.
(25) Rosini, C.; Scamuzzi, S.; Pisani-Focati, M.; Salvadori, P. J. Org. Chem.
1995, 60, 8289-8293.
(26) Superchi, S.; Casarini, D.; Laurita, A.; Bavoso, A.; Rosini, C. Angew. Chem.,
Int. Ed. 2001, 40, 451-454.
(27) Rosini, C.; Scamuzzi, S.; Uccellobarretta, G.; Salvadori, P. J. Org. Chem.
1994, 59, 7395-7400.
(28) Donnoli, M. I.; Scafato, P.; Superchi, S.; Rosini, C. Chirality 2001, 13,
258-265.
(29) Superchi, S.; Donnoli, M. I.; Rosini, C. Org. Lett. 1999, 1, 2093-2096.
(30) Superchi, S.; Casarini, D.; Summa, C.; Rosini, C. J. Org. Chem. 2004, 69,
1685-1694.
(31) Huang, X. F.; Rickman, B. H.; Borhan, B.; Berova, N.; Nakanishi, K. J.
Am. Chem. Soc. 1998, 120, 6185-6186.
(32) Kurtan, T.; Nesnas, N.; Li, Y. Q.; Huang, X. F.; Nakanishi, K.; Berova, N.
J. Am. Chem. Soc. 2001, 123, 5962-5973.
(33) Huang, X. F.; Fujioka, N.; Pescitelli, G.; Koehn, F. E.; Williamson, R. T.;
Nakanishi, K.; Berova, N. J. Am. Chem. Soc. 2002, 124, 10320-10335.
(34) Lintuluoto, J. M.; Borovkov, V. V.; Inoue, Y. J. Am. Chem. Soc. 2002,
124, 13676-13677.
(35) Proni, G.; Pescitelli, G.; Huang, X. F.; Quraishi, N. Q.; Nakanishi, K.;
Berova, N. Chem. Commun. 2002, 1590-1591.
(36) Yang, Q. F.; Olmsted, C.; Borhan, B. Org. Lett. 2002, 4, 3423-3426.
(37) Proni, G.; Pescitelli, G.; Huang, X. F.; Nakanishi, K.; Berova, N. J. Am.
Chem. Soc. 2003, 125, 12914-12927.
9
1886 J. AM. CHEM. SOC. VOL. 130, NO. 6, 2008

Fluorinated Porphyrin Tweezer
A R T I C L E S
Table 1. Frontier Orbital Energy and Charges of Zinc Porphyrins
bisfunctionalized molecule as a system with two independent
chiral sites is an oversimplification, since the conformation of
the molecule would clearly play a large role in the overall
helicity of the complexed porphyrin tweezer. Nevertheless,
without a clear picture of the molecule’s conformation upon
complexation with the porphyrin tweezer, one can assume that
the conformation adopted would be dictated by the fact that
each chiral site is now complexed with a large porphyrin ring,
and thus the overall conformation can be hypothesized by
adopting a rotomer that minimizes interaction at each chiral
center.
A problematic issue was the weak binding of alcoholic chiral
guests with zinc porphyrin tweezers employed previously. For
this approach to work, we required a strong binding tweezer
such that a large population of the chiral guest (for example,
1,2-diols) would bind and, thus, increase the observed signal.
Also, weak binding complexes could lead to small energetic
differences between a number of complexed conformations.
Inconsistent trends in observed ECCD are possible upon binding
of different structures if different conformations yield opposite
ECCD. In fact, we have observed a mix of positive and negative
ECCD with different substrates that have the same chirality
when weak binding tweezers were utilized. The latter criteria
led to the design and synthesis of 5-(4-methylcarboxyphenyl)-
10,15,20-tri(pentafluorophenyl) porphyrin (2, Figure 2c). The
use of this electron-deficient porphyrin yields a Lewis acidic
zincated porphyrin tweezer that binds diols strongly.
Mulliken
charge
electrostatic
charge of
ELUMO
eV
ELUMO
−
eVa
EHOMO
+
+
of Zn2
Zn2
Zn-TPFP-monoester
Zn-TPP-monoester
-2.818
-2.211
3.284
3.891
0.961
0.942
1.387
1.337
a S-Lysine methyl ester was chosen as a typical chiral guest. HOMO
energy calculated using the same method was -6.102 eV.
Herein, we demonstrate the utilization of an electron deficient
zincated porphyrin tweezer as an effective host system for
binding of 1,2-bisfunctionalized diols, amino alcohols, and
diamines. Moreover, we disclose a routine and robust system
for the absolute stereochemical determination of both threo and
erythro systems without the need for chemical derivatization.
Figure 3. Electrostatic potential surface of Zn-TPFP-monoester (4) and
Zn-TPP-monoester (5) based on B3LYP/6-31G(d)//PM3 calculation. Prop-
erty range was set from -45 (red color) to +45 (blue color).
Results and Discussion
also provide evidence that the zinc atom bound to the Zn-TPFP-
monoester (4) is more electropositive, presumably due to the
strong electron-withdrawing effect of the pentafluorophenyl
groups as compared to Zn-TPP-monoester (5). Electrostatic
potential surfaces (Figure 3) provide a visual perspective of the
charge difference between the two zinc porphyrins. The Zn-
TPFP-monoester is clearly more electron deficient than the Zn-
TPP-monoester as represented by the “colder” surface.
Synthesis of Tweezer 3 and Binding Affinity Measure-
ments. Porphyrin 2 (Figure 2c) was synthesized using a
modified literature procedure.10 The zincated porphyrin tweezer
3 was prepared using protocols described previously (Figure
2c, λmax ) 416 nm, ꢀ ) 1 120 000 cm-1 M-1 in CH2Cl2; λmax
) 412 nm, ꢀ ) 890 000 cm-1 M-1 in hexane).31 Calculations,
1H NMR binding experiments, and UV-vis titrations cor-
roborate the enhanced Lewis acidity and binding affinity of the
new fluorinated porphyrin.
1
The enhanced binding could also be observed through H
NMR analysis. Binding of ethanol to zincated porphyrin 4 (1:1
ratio) led to a 0.5 ppm upfield shielding of the methylene
protons. Ethanol bound to zincated tetraphenyl porphyrin 5
upfield shifted the methylene protons by only 0.1 ppm, clearly
indicating a stronger binding affinity of porphyrin 4 with ethanol.
Computational modeling was utilized to investigate the
change in Lewis acidity of the zinc atom in the fluorinated
porphyrin as compared to its nonfluorinated counterpart. Zinc
5-(4-methylcarboxyphenyl)-10,15,20-tri(pentafluorophenyl) por-
phyrin (Zn-TPFP-monoester, Table 1, 4) and zinc 5-(4-meth-
ylcarboxyphenyl)-10,15,20-triphenyl porphyrin (Zn-TPP-mo-
noester, Table 1, 5) were used as model compounds. Geometry
optimization was performed by the semiempirical method (PM3)
followed by DFT single-point energy calculation at the B3LYP/
6-31G(d) level of theory.
As shown in Table 1, the fluorinated porphyrin has a lower
LUMO energy. This indicates an increased Lewis acidity as
well as binding affinity toward nucleophilic guest molecules
since the decreased LUMO energy would lead to a smaller
energy gap between the LUMO of the electrophilic host and
the HOMO of the nucleophilic guest. The calculated charges
Quantitative measures of amine and alcohol binding with both
Zn-TPP and Zn-TPFP were obtained via titration of the ligands
and analysis of the spectroscopic changes. Zn-TPFP-ester 4 and
Zn-TPP-ester 5 were used as simplified models of each tweezer
for initial measurements. Figure 4 illustrates the changes
observed in the UV-vis spectra upon titration of isopropanol
and isopropylamine with Zn-TPFP-ester 4. The change of
absorption at 418 nm for isopropanol and 412 nm for isopro-
pylamine as a function of concentration of ligand added yields
an exponential saturation curve. The binding constants can be
derived from fitting the latter data via a nonlinear least-squares
9
J. AM. CHEM. SOC. VOL. 130, NO. 6, 2008 1887

A R T I C L E S
Li et al.
Figure 5. Titration of Zn-TPFP-tweezer 3 with amino alcohol 12. Two
isosbestic points are apparent. The first at lower equivalence of amino
alcohol (leading to the absorption at 415 nm) signifies the 1:1 ligand/
complex formation. The second (leading to the absorption at 423 nm)
appearing at higher concentrations of amino alcohol indicates the formation
of the 2:1 ligand/tweezer complex (see dashed box).
Figure 4. Titration of Zn-TPFP ester 4 with (a) iPrNH2 (0-30 equiv,
bathochromic shift from 412 to 424 nm) and (b) iPrOH (0-12 000 equiv,
bathochromic shift from 412 to 418 nm) in hexane. Both systems exhibit
a bathochromic shift upon addition of ligand. The inset graphs are nonlinear
least-squares fits of the change in absorption vs equiv of ligand for each.
Table 2. Binding Affinity of iPrOH and iPrNH2 with Zn-TPP-Ester
and Zn-TPFP-Estera
titration of amino alcohol 12 with tweezer 3. Close inspection
of the UV-vis spectra reveals the presence of two processes
as the concentration of the ligand is increased. From prior work
it is well recognized that a ∼12 nm bathochromic shift is
expected upon coordination of a strong nucleophile such as an
amino group with the zinc porphyrin.39 It has also been
demonstrated previously that a hypsochromic shift occurs when
the two porphyrin rings are brought close in space.40 Thus, there
are two opposing effects that in combination lead to the observed
Kassoc (iPrOH)
Kassoc (iPrNH2)
1
1
porphyrin
M-
M-
Zn-TPFP-monoester (4)
Zn-TPP-monoester (5)
2170 ( 140
49 ( 2
473 000 ( 8700
11 400 ( 950
a Kassoc for each ligand/porphyrin complex was obtained via fitting the
data obtained from the change in the absorption upon titration of the
porphyrin with ligand in hexane to a nonlinear least-squares analysis routine
described in the Supporting Information.
λ
max. For example, upon complexation of 1,2-diaminoethane, a
method as has been previously reported.38 Table 2 lists the
binding constants obtained for Zn-TPFP-ester 4 and Zn-TPP-
ester 5. As expected, isopropylamine binds more strongly to
the Zn-porphyrins as compared to isopropanol. Of note,
however, is a greater than 1 order of magnitude increase in
binding affinity of the ligands (both amine and alcohol) for the
fluorinated Zn-porphyrin 4 as compared to the nonfluorinated
Zn-porphyrin 5.
4 nm bathochromic shift is observed (as opposed to the expected
12 nm); as the separation between the porphyrin rings is
increased the λmax gradually redshifts such that binding of 1,-
10-diaminodecane results in a 9 nm bathochromic shift.40
The titration curves depicted in Figure 5 exhibit an initial
bathochromic shift of the absorption maximum from 412 to 415
nm when the first 3 equiv of the 1,2-amino alcohol 12 are added
to tweezer 3 in hexane. As the equivalence is increased,
however, a second bathochromic shift to 423 nm is evident.
We next turned our attention to the binding of bis-function-
alized substrates with the Zn-porphyrin tweezers derived from
TPP-ester and TPFP-ester. Figure 5 illustrates the binding
(39) Hunter, C. A.; Meah, M. N.; Sanders, J. K. M. J. Am. Chem. Soc. 1990,
112, 5773-5780.
(40) Huang, X.; Borhan, B.; Berova, N.; Nakanishi, K. J. Indian Chem. Soc.
1998, 75, 725-728.
(38) Shoji, Y.; Tashiro, K.; Aida, T. J. Am. Chem. Soc. 2006, 128, 10690-
10691.
9
1888 J. AM. CHEM. SOC. VOL. 130, NO. 6, 2008

Fluorinated Porphyrin Tweezer
A R T I C L E S
Figure 6. Titration of Zn-TPFP tweezer 3 with diol 16 (0-300 equiv) in
hexane. The nonlinear least-squares fit of the change in absorption vs equiv
of ligand provides the binding constant.
Figure 7. Proposed binding conformation between tweezer 3 and erythro
amino alcohol 12, yielding a positive ECCD spectrum. It is assumed that
P1 and P2 bind to their respective functional groups opposite the large
group at each chiral center. Thus P1 binds the alcohol in an anti fashion
with respect to the phenyl group and sterically differentiates between the
smaller hydrogen group and C2. This results in a counterclockwise
rotation of P1. In an analogous fashion, P2 rotates in a clockwise
manner, leading to an overall positive helicity of the tweezer and the
observed ECCD spectrum. The model illustrates the proposed binding of
tweezer 3 with amino alcohol 12 (the pentafluorophenyl groups in the
modeled picture have been omitted for clarity). A simplified mnemonic is
illustrated in the dashed box. The sign of the ECCD spectrum will depend
only on the absolute stereochemistry of the chiral center that bears the largest
group.
This behavior can be explained based on the observations
discussed above; i.e., the initial 3 nm bathochromic shift is due
to a 1:1 complexation of the 1,2-amino alcohol with tweezer 3,
which brings the porphyrins close to each other. This results in
the observed 3 nm redshifting that gives rise to the peak at 415
nm. Further increase in the concentration of the 1,2-amino
alcohol leads to the breakup of the 1:1 complex (binding of an
amine to the zincated porphyrin is much stronger than that of
an alcohol) such that each porphyrin is bound to an amino group
(1:2 complex), effectively maximizing the distance between the
porphyrin rings that leads to an 11 nm redshift. This is analogous
to a monoamine binding to the porphyrin, which would result
in the same level of redshifting. We would thus expect that
excess addition of chiral 1,2-amino alcohols to the porphyrin
tweezer would diminish the anticipated ECCD.
Based on the latter discussion, the binding constant for a 1:1
complexation of 12 with tweezer 3 was measured based on the
data obtained for up to 3 equiv of amino alcohol added.
Nonlinear least-squares analysis of the change in absorption at
412 nm as a function of guest concentration provided a binding
constant (Kassoc) of 1.31 × 107 M-1 for binding of 12 with
tweezer 3 in hexane. Similar UV titrations in methylcyclohexane
(MCH) with guest 12 rendered a binding constant of 2.71 ×
106 M-1. The better binding in hexane might lead to stronger
ECCD amplitudes.
As anticipated, the binding of 1,2-diols with tweezer 3 is
weaker due to the tempered nucleophilicity of alcohols compared
to amines. Figure 6 depicts the UV-vis titration curves of diol
16 complexed with tweezer 3 (Kassoc ) 152 000 M-1 in hexane,
36 100 M-1 in MCH). Two observations are notable; first a
large bathochromic shift is not present upon binding of the
hydroxyl group with the zincated porphyrin. This is the result
of two opposing effects, the anticipated 6 nm bathochromic shift
as a result of alcohol binding with the zincated porphyrin (see
Figure 4b for alcohol-porphyrin binding) and the hypsochromic
shift that results from bringing the two porphyrins close to each
other. Second, since addition of excess 1,2-diol does not compete
with the 1:1 complexation (previously, excess amine competed
with the hydroxyl bound porphyrin), breakup of the 1:1 complex
is not apparent. Addition of excess 1,2-diol to tweezer 3 mimics
addition of excess diamines to porphyrin tweezers, which does
not exhibit the breakup of the 1:1 complex until ∼500 equiv
are added.37
Determination of Chirality of erythro Substrates. Our
initial foray focused on the use of the fluorinated tweezer 3
with erythro diols and amino alcohols. As can be seen from
Figure 7 and Table 3, the binding of the chiral erythro guests
yielded a host-guest complex with observable ECCD couplets.
A consistent and predictable trend is observed, where the R,S
erythro compounds (the first stereochemical designation refers
to the chiral site with the largest substituent as determined by
A-values) give rise to positive ECCD, while the S,R erythro
compounds produced negative ECCD spectra. The
Cahn, Ingold, and Prelog stereochemical designation should be
used with caution since priorities do not necessarily correlate
with sterics as can be seen in 7, an example of an R,R-erythro
diol.
Figure 7 illustrates a conceptual model that suggests each
chiral center orients the bound porphyrin through steric dif-
ferentiation. This is illustrated with the binding of norephedrine
12 with tweezer 3. We propose that P2 approaches the amino
group of R,S-norephedrine opposite to the largest substituent,
in this case the methyl group. In this manner, the methyl group
is anti to the complexed porphyrin and does not participate in
the steric differentiation. The two remaining groups attached
to C1, namely H and the C2, dictate the disposition of the bound
porphyrin ring; as such P2 rotates clockwise toward the smaller
hydrogen atom. Similarly, binding of P1 with the OH group
places P1 anti to the phenyl group attached to C1. Steric
differentiation of the remaining substituents (H and C2) leads
to the rotation of P1 in a counterclockwise manner to reduce
repulsion with the bulky C2 substituent. As described, P1 adopts
a clockwise (positive) helicity relative to P2, which would
predict a positive ECCD spectrum. Indeed, as illustrated in
Figure 7, the complexation of 12 with porphyrin tweezer 3 leads
to the anticipated spectrum. This is a working model that fits
the data and by no means is the only possible mode of binding
9
J. AM. CHEM. SOC. VOL. 130, NO. 6, 2008 1889

A R T I C L E S
Li et al.
Table 3. Tweezer 3 Bound erythro-1,2-Diols and Amino Alcohols
Figure 8. Increasing concentration of amino alcohol 12 leads to diminished
ECCD signals. This is presumably a result of breaking up the ECCD active
1:1 ligand/tweezer complex (see Figure 5) due to the fact that amines have
a much higher binding affinity for the zincated porphyrin. The resultant
2:1 complex is not ECCD active.
potential coordinating groups such as ketones, esters, and ethers
(Table 3, compounds 7 and 9). As expected, enantiomeric pairs
yield opposite ECCD spectra (compounds 6-8 and 12 along
with their enantiomers). Of note are the higher ECCD ampli-
tudes obtained for amine containing guests 11 and 12, which is
presumably due to their stronger binding with tweezer 3 as
compared to diols.
ECCD amplitudes of tweezer/guest complexes in hexane were
consistently larger than those measured in MCH. This is most
probably attributed to the observed stronger binding of 1,2-diols
and 1,2-amino alcohols with tweezer 3 in hexane as compared
to MCH. It seems reasonable that binding constants could be
used as a guide for choosing the appropriate solvent in such
cases. Other solvents such as dichloromethane or acetonitrile
failed to yield observable ECCD for many substrates under the
same conditions. Cooling of the solution to 0 °C produced strong
observable ECCD spectra; however, further lowering of the
temperature to -10 °C only led to a limited increase in the
amplitude.
a All substrates were >98% ee except for 10 (25% ee). b 2 µM tweezer
concentration. c Tweezer/substrate ratio, 1:100. dTweezer/substrate ratio,
e
1:40. 1 µM tweezer concentration. f Tweezer/substrate ratio, 1:5.
(see Supporting Information for alternate binding conformation
and molecular modeling studies).
Careful analysis of the results listed in Table 3 leads to a
simplified mnemonic where the helicity of the tweezer can be
determined by analyzing the chirality of one carbon center.
Based on the proposal above, in which the two porphyrins (P1
and P2) rotate counter to each other, one can consider solely
the arrangement of P1, the porphyrin bound to the chiral center
with the largest substituent, as the stereochemical determinant
group (assuming the other porphyrin P2 does not rotate and is
static). As depicted in Figure 7 (dashed box) the stereochemistry
of the chiral center with the largest substituent leads to the
predicted ECCD; namely, looking at the Newman projection, a
clockwise arrangement of S f OH f L leads to a positive
ECCD and Vice Versa. This device should be used as a
mnemonic for the given system and does not reflect the bound
conformation of the chiral molecule.
A note of caution is warranted with amino alcohols. As was
eluded to above, the large difference in binding affinity of
alcohols and amines with zincated porphyrins does lead to the
breakup of the 1:1 complex with increasing equivalence of
amino alcohols added. The resultant 2:1 complex is not ECCD
active. Thus, increasing equivalence of amino alcohols is
detrimental to obtaining a strong ECCD signal. Figure 8
demonstrates the decrease in signal as a function of increasing
concentration of amino alcohol 12, clearly suggesting the
breakup of the 1:1 complex at higher equivalences.
Determination of Chirality for threo Substrates. Next we
investigated threo vicinal bis-functionalized systems. Aliphatic
and aromatic R,R-threo substrates yielded positive ECCD signals
while the S,S-threo compounds produced negative ECCD spectra
(Table 4). Seven pairs of threo enantiomers all exhibited mirror
image CD couplets (see Supporting Information). Figure 9
illustrates the ECCD spectrum obtained from the complexation
of tweezer 3 with the threo diol 13. The expected positive
couplet is again assumed to derive from steric differentiation
of bound porphyrins at each chiral site. As depicted in Figure
9, each chiral center induces the same helical twist of the
As can be seen from Table 3, a number of alkyl and aryl
systems produced the expected results upon binding with
tweezer 3. In all cases the predicted sign was arrived at after
considering the A-values of the substituents on the carbinol
carbon (the structures in Table 3 are drawn such that the larger
group is on the left-hand side). The system is tolerant of other
9
1890 J. AM. CHEM. SOC. VOL. 130, NO. 6, 2008

Fluorinated Porphyrin Tweezer
A R T I C L E S
Table 4. Tweezer 3a Bound threo-1,2-Diols,b Amino Alcohols,c
and Diaminesc
Figure 9. Proposed conformation of the complex generated between the
binding of tweezer 3 (two arrows indicate the directions which the
porphyrins would rotate to minimize steric repulsion) and threo diol 13,
resulting in a positive ECCD spectrum. The depicted arrangement is arrived
at by placing P1 and P2 opposite the large group at each chiral center.
Steric differentiation at each chiral center leads to the anticipated positive
helical arrangement of the porphyrin rings (similar to the erythro case, P1
rotates in a counterclockwise manner to avoid interactions with C2 in
preference for the smaller hydrogen atom). The model illustrates the binding
of diol 13 with tweezer 3 (the pentafluorophenyl groups have been omitted
for clarity). The dashed box depicts a simplified mnemonic for the absolute
stereochemical determination of threo systems based on analysis of the chiral
center with the largest substituent.
The latter statement would suggest that C2 symmetric diols and
diamines would lead to observable ECCD spectra. This is indeed
observed (Table 4, compounds 20-23 and 26-27) making this
methodology applicable to the absolute stereochemical identi-
fication of C2 symmetric compounds, which are an important
subgroup of chiral molecules used in various enantioselective
processes. The same simplified mnemonic that was used for
erythro systems is applicable for threo compounds as well. As
illustrated in Figure 9 (dashed box), the sign of the ECCD
couplet can be predicted by analyzing the substituent on the
chiral center that bears the largest group.
As can be seen from Table 4, a variety of substituents and
functional groups produced the expected results. Also, the
presence of other potential coordinating groups such as esters
(Table 4, compounds 13, 16, 18, and 22) or acetonide (23) did
not interfere with the complexation of the substrates with
tweezer 3. As expected, threo diamines produced much higher
ECCD amplitudes than diols due to their stronger binding with
tweezer 3. Interestingly, tweezer 3 was capable of binding
diepoxide 27 and produced the expected ECCD spectrum based
on the binding mode explained above (see Supporting Informa-
tion). The absolute stereochemical determination of epoxy
alcohols via binding of the epoxide will be reported in due
course.
As with the erythro compounds, the CD for threo systems
was measured at 0 °C in order to improve the observed ECCD
amplitudes. Although the use of MCH as the solvent (tradition-
ally used as the preferred solvent for ECCD measurements with
Zn tweezers) produced ECCD spectra with large amplitudes, it
was observed that hexane in most cases would increase the
strength of the signal. For instance, the two weakest performing
threo substrates 14 (A ) +33) and 18 (A ) +20) exhibited
much larger ECCD amplitudes (+62 and +76, respectively) in
hexane. As described above, the proposed binding of porphyrin
tweezer 3 with erythro and threo systems is similar, in which
a 1 µM tweezer concentration was used for all measurements except
for 19 and 27 (2 µM). b Tweezer/substrate ratio, 1:40. c Tweezer/substrate
d
ratio, 1:5. All substrates were >95% ee except for 15 and 27 (91% ee).
e The enantiomer showed mirror image CD spectrum. f Tweezer/substrate
ratio, 1:200, methyl cyclohexane was used as solvent for all CD measure-
ments.
porphyrin tweezer (P1 and P2 twist counter to each other as
indicated by the two arrows). As was described for erythro
systems, we assume that the complexed porphyrins approach
their ligands such that the largest group on the chiral center is
anti with respect to the bound chromophores. The conformation
illustrated for threo diol 13 in Figure 9, where the large groups
(Ph on C1 and CO2Me on C2) are placed anti to each other
(similar to the conformation assumed for erythro systems), leads
to the least sterically encumbered arrangement for the bound
porphyrins. Steric differentiation at P1 (H vs C2) leads to a
counterclockwise rotation of the porphyrin toward the smaller
hydrogen atom, while a clockwise rotation of P2 (steric
differentiation of H vs C1) is anticipated. This would lead to a
positive helical arrangement of the porphyrins, which is verified
by the positive ECCD spectrum observed for diol 13 bound to
porphyrin tweezer 3. Each stereocenter of a threo system would
thus reinforce the same overall helicity of the bound tweezer.
9
J. AM. CHEM. SOC. VOL. 130, NO. 6, 2008 1891

A R T I C L E S
Li et al.
Table 5. Tweezer 1a Bound erythro and threo 1,2-Diolsb
each porphyrin binds the coordinating functional group opposite
the largest substituent on the chiral center. It should be noted
that in all cases discussed thus far the nature of the acyclic
system could lead to a number of rotomers upon com-
plexation (Figures 7 and 9 only depict the rotomer believed to
lead to the observed ECCD spectra for erythro and threo
systems, respectively). It is difficult to predict whether or not
the conformation of the complexed guest molecules retains the
lowest energy conformation of unbound molecules (such as
those depicted in Figure 1) since complexation with the large
porphyrin tweezer can lead to compensating interactions with
an overall effect of the guest system adopting a higher energy
conformation.
NMR experiments have not been useful in assigning the
conformation of the bound guest molecules since aggregation
of the Zn-TPFP porphyrin tweezer is observed at concentrations
necessary for observing NMR signals (based on our observa-
tions, the fluorinated porphyrin has a much higher tendency for
aggregation as compared to the nonfluorinated porphyrin). At
the same time, the solutions used for NMR analysis (10-3
-
10-4 M) were found to be ECCD silent. In solvents that reduced
the level of aggregations, such as CHCl3, CH2Cl2, and CH3-
CN, ECCD was absent upon binding with the chiral guests. This
is presumably due to the high Lewis acidity of the Zn-TPFP
porphyrin tweezer and its ability to interact with solvents more
polar than hydrocarbons. Computational minimizations of the
complexed system, albeit at a low level of theory due to the
large size of the complex, agree with the proposed models (see
Supporting Information for details and alternate conformations
of the bound guests leading to the expected ECCD spectra). In
summary, the conformations of the bound chiral guest molecules
that are depicted in Figures 7 and 9 to yield the observed ECCD
spectra are speculative; however, the mnemonic that is derived
from these observations consistently provides the anticipated
ECCD spectra and can be used reliably.
a 1 µM tweezer 1 concentration was used for all measurements
except for 8 and 9 (2 µM). b Tweezer/substrate ratio, 1:40. c Tweezer/
substrate ratio, 1:100. d Hexane was used as solvent for all CD measure-
ments.
All other erythro diols tested did not exhibit an ECCD signal.
As shown in Table 5, many threo diols were also ECCD silent
in the presence of Zn-TPP tweezer 1. From those that did
provide a signal, compounds 15 and 16, which led to the highest
amplitudes among the diols with Zn-TPFP tweezer 3, showed
much weaker signals when bound with Zn-TPP tweezer 1. Other
threo diols such as 19 generated signals with signs that are
opposite to the predicted model.
Comparison between TPFP Tweezer and TPP Tweezer.
Initially, this project was pursued based on the supposition that
steric differentiation of each chiral site bound to the porphyrin
tweezer will result in an induced helicity that is observable as
an ECCD signal. Although, as described above, this has come
to fruition, the initial attempts were discouraging due to the
fact that different structures with the same chirality produced
differing ECCD signs or none at all. It should be noted that
initially the Zn-TPP tweezer was utilized, and it became
apparent that weak binding of substrates potentially contributes
to a number of different, energetically close conformations of
the complex. As we have documented above, the solution to
this problem was to increase the binding of the guest molecules
with the host porphyrin tweezer by increasing the Lewis acidity
of the zincated porphyrins. Functional comparison of the
porphyrin tweezers (Zn-TPP 1 vs Zn-TPFP 3) in deducing
the absolute stereochemistry of a number of threo and erythro
diols is provided in Table 5. The measurements were taken in
hexane to ensure maximum binding. As was described above
and shown in Table 5, the predicted sign for ECCD matched
the observed spectra for all substrates tested with Zn-TPFP
tweezer 3. However, among erythro diols only 8 and 9 yielded
observable ECCD signals with Zn-TPP tweezer 1, albeit with
lower amplitudes as compared to those obtained with Zn-TPFP
tweezer 3. Moreover, diol 8 did not yield the expected ECCD.
It seems reasonable to conclude that the binding affinity of
host-guest complexes used in absolute stereochemical deter-
minations must be considered, especially if weak CD signals
and/or inconsistent signs are obtained. In this case, with the
use of the electron deficient Zn-TPFP tweezer, the binding of
the diols is sufficiently strong to produce consistent ECCD
spectra leading to a general approach for absolute stereochemical
determination of 1,2-systems.
Conclusion
We demonstrate the prompt microscale determination of
absolute configurations for threo and most importantly the often
difficult to determine erythro diols, amino alcohols, and
diamines utilizing the porphyrin tweezer methodology. Con-
servative estimates are that 1 nmol of material is required for
analysis of diamines and amino alcohols and 40 nmol are needed
for analysis of diols (along with 1 nmol of porphyrin tweezer
for each analysis). To the best of our knowledge this is the first
routine method that addresses the absolute stereochemical
determination of erythro compounds. This methodology also
requires no derivatizations, which is often used for the absolute
stereochemical determination of threo compounds. Utility of
the fluorinated tweezer 3 for binding with other functional
groups is under investigation.
9
1892 J. AM. CHEM. SOC. VOL. 130, NO. 6, 2008

Fluorinated Porphyrin Tweezer
A R T I C L E S
Acknowledgment. Generous support was provided by the
NSF CAREER (CHE-0094131) grant.
diols; ECCD data for seven enantiomers of threo substrates 13-
16, 22, 25, 26; Molecular modeling of TPFP tweezer complexed
with 12, 15, and diepoxide 27. This material is available free
of charge via the Internet at http://pubs.acs.org
Supporting Information Available: Experimental procedures
for synthesis of porphyrin tweezer 3, diols 6-10, 13-19, 21,
23, 27; Spectra data and optical purity of all the synthesized
JA0752639
9
J. AM. CHEM. SOC. VOL. 130, NO. 6, 2008 1893