Influence of the solvent and metal center on supramolecular chirality induction with bisporphyrin tweezer receptors. Strong metal modulation of effective molarity values

Article

pubs.acs.org/IC

Influence of the Solvent and Metal Center on Supramolecular

Chirality Induction with Bisporphyrin Tweezer Receptors.

Strong Metal Modulation of Effective Molarity Values

Inmaculada C. Pintre,^†Simon Pierrefixe,^†Alex Hamilton,^†Virginia Valderrey,^†Carles Bo,*^,†

and Pablo Ballester*^,†,‡

^†Institute of Chemical Research of Catalonia (ICIQ), Avgda. Països Catalans 16, 43007 Tarragona, Spain

^‡Catalan Institution for Research and Advanced Studies (ICREA), Passeig Lluís Companys, 23, 08018, Barcelona, Spain

S

* Supporting Information

ABSTRACT: We describe the synthesis of a bisporphyrin

tweezer receptor 1·H₄and its metalation with Zn(II) and

Rh(III) cations. We report the thermodynamic character-

ization of the supramolecular chirality induction process that

takes place when the metalated bisporphyrin receptors co-

ordinate to enantiopure 1,2-diaminocyclohexane in two

different solvents, toluene and dichloromethane. We also

performed a thorough study of several simpler systems that

were used as models for the thermodynamic characterization

of the more complex bisporphyrin systems. The initial com-

plexation of the chiral diamine with the bisporphyrins

produces a 1:1 sandwich complex that opens up to yield a

simple 1:2 complex in the presence of excess diamine. The CD

spectra associated with the 1:1 and 1:2 complexes of both metalloporphyrins, 1·Zn₂and 1·Rh₂, display bisignate Cotton effects

when the chirogenesis process is studied in toluene solutions. On the contrary, in dichloromethane solutions, only 1·Zn₂yields

CD-active 1:1 and 1:2 complexes, while the 1:2 complex of 1·Rh₂is CD-silent. In both solvents, porphyrin 1·Zn₂features a

stoichiometrically controlled chirality inversion process, which is the sign of the Cotton effect of the 1:1 complex is opposite to

that of the 1:2 complex. In contrast, porphyrin 1·Rh₂affords 1:1 and 1:2 complexes in toluene solutions with the same sign for

their CD couplets. Interestingly, in both solvents, the signs of the CD couplets associated with the 1:1 sandwich complexes of

1·Zn₂and 1·Rh₂are opposite. The amplitudes of the CD couplets are higher for 1·Zn₂than for 1·Rh₂. This observation is in

agreement with 1·Rh₂having a smaller extinction coefficient than 1·Zn₂. We performed DFT-based calculations and assigned

molecular structures to the 1:1 and 1:2 complexes that explain the observed signs for their CD couplets. Unexpectedly, the

quantification of the thermodynamic stability of the two metallobisporphyrin/diamine 1:1 sandwich complexes revealed the

existence of interplay between effective molarity values (EM) and the strength of the intermolecular interaction (K_m; N···Zn or

N···Rh) used in their assembly. The EM for the N···Rh(III) intramolecular interaction is 3 orders of magnitude smaller than that

for the N···Zn(II) interaction, both of which are embedded in the same scaffold of the 1·M₂bisporphyrin receptor.

Evidence of exciton interaction between porphyrin chromo-

phores can also be derived from spectral shifts or splitting in

the Soret band of UV−vis spectra of composite systems

compared to monoporphyrin. In the absorption spectrum,

the magnitudes of the shifts and splitting are correlated with

the relative orientation and distance between the coupled

electric transition dipole moments associated with the Soret

band. Employing a simplified model of single effective Soret

transition moment per porphyrin unit,⁶the parallel

orientation of coupled dipoles results in a blue shift of the

Soret band since only excitation in the higher level energy

level exciton state is allowed.

INTRODUCTION

■

Porphyrins feature a red shifted and very intense absorbance

band (ε ≈ 4−5 × 10⁵M⁻¹cm⁻¹), known as the Soret or B

band centered at ca. 420 nm, which makes them excellent

chromophores for structural studies of exciton-coupled circular

dichroism (CD).¹⁻³If porphyrin chromophores experiencing

exciton coupling are held in a chiral environment, that is,

covalently connected to a chiral scaffold, the phenomenon can

be clearly detected in the corresponding CD spectrum. A bisig-

nate CD curve (so-called exciton couplet) is observed, with two

bands of opposite sign and similar intensity centered around

the λ_maxof the UV−vis absorption corresponding to the transition

dipoles. The sign of the couplet is determined by the absolute

sense of the skewness (twist) of the interacting electric transition

moments associated with the chromophores’ transitions.^4,5

Received: November 22, 2011

Published: March 28, 2012

4620

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

The supramolecular application of the exciton-chirality rule

with achiral bisporphyrin receptors (host) is based on the

transfer of chirality from the chiral substrate (guest) to the

resulting host−guest complex (chirogenesis); see Figure 1.

(Rh and Zn) studied in this work have been shown to con-

stitute a minimal platform to investigate these assumptions.

When chiral recognition (stereodifferentiation) is operative

during the formation of the 1:1 “tweezer” complex, the chirality

of the guest is transferred in the formation of one or a few

diastereoisomeric complexes (stereospecific conformations)

that are energetically more favorable than others. The dia-

stereoisomeric complexes display chiral spatial arrangements

with fixed relative orientations (twist) of the porphyrin units of

the bound “tweezer”. The close spatial proximity of the por-

phyrin cores encountered in the 1:1 chiral “sandwich”

complexes induces exciton coupling between the porphyrins’

electronic transitions.

In general, the CD and UV spectra experimentally observed

represent a weighted average of corresponding spectra of the

individual diastereoisomeric complexes in solution. Normally,

the sign of the couplet inverts upon changing the absolute

configuration of the chiral guest, and the observation of small

CD signals are explained by partial cancellation of intense CD

bands for diastereoisomeric complexes present in solution and

having opposite couplet signs.

An important advantage of the exciton-chirality rule is its

nonempirical nature. There are strong theoretical grounds,

which also apply to bisporphyrin supramolecular complexes, to

relate the observed sign of the CD couplet to the relative spatial

orientation exhibited by the chromophores in the supra-

molecular chiral structure. Nowadays, ab initio calculations

constitute an invaluable methodology to predict CD spectra

from given supramolecular structures and become fundamental

in the rationalization of the supramolecular chirogenesis

processes.^11,12

Figure 1. (a) Schematic representation of a bisporphyrin “tweezer”

receptor. A simplified model with a single effective electric transition

moment (B_5,15, double arrow in bold) is adopted for each porphyrin

unit. The weight of the transverse B_1,10component is reduced by fast

rotation. (b) The exciton-chirality rule relates the sense of the torsion

angle between the two coupled electric dipole moments of the

bisporphyrin units of a 1:1 supramolecular chiral complex to the sign

of the CD couplet. (c) Diagram of the origin of UV−vis spectral shifts

due to exciton coupling.

Achiral bisporphyrin tweezer^13,14receptors are powerful

tools for the absolute configurational analysis of organic

compounds using CD spectroscopy.¹⁵⁻¹⁷Alternatively, related

CD studies of bisporphyrin “tweezer” receptors binding chiral

guests of known absolute configuration with the assistance of

theoretical calculations are used to comprehensively investigate

the various controlling factors (electronic and structural) that

influence the chirogenic process and how they are affected by

external and internal stimuli.¹⁸

In 2008, we reported the binding of bisporphyrin tweezer

1a·Zn₂with enantiopure 1,2-diaminocyclohexane 2 in dichloro-

methane solution (Scheme 1).¹⁹The system expressed the

properties of both chirality induction and stoichiometrically

controlled inversion. A related finding had already been

reported by Borovkov et al. with a simpler ethane bridged

bisporphyrin receptor.²⁰We undertook the work reported here

in order to investigate in more detail the influence of the metal

center and solvent on the chirality induction and inversion

processes displayed by the complexation of 1·M₂tweezers with

enantiopure diamine 2. We also wanted to combine the

experimental results with electronic structural methods in the

assignment of molecular structures to the 1·M₂⊃R,R-2 (1:1)

and 1·M₂@(R,R-2)₂complexes formed during the binding

process. This endeavor constitutes a necessary step to achieve a

rationalization of the observed chirality transfer and inversion

on the basis of electronic and structural factors.

In most of the examples described in the literature, the chiral

guests contain two interaction sites. That is, the ligands are

bidentate and capable of undergoing simultaneous axial

coordination with the two metalloporphyrin binding units of

the receptors. In other words, the achiral bisporphyrin

receptors function as a molecular “tweezer” grabbing the

guest in an open cavity defined by two porphyrin units bridged

by a spacer.

The resulting ditopic 1:1 complexes are usually highly

thermodynamically stable and constitute ideal systems for the

calculation of effective molarity values (EM). EM values are

used to quantify the ease of the intramolecular interaction.⁷

The physical expression of chelate cooperativity is reflected in

an increase in binding affinity for the ditopic 1:1 complex

compared to the statistically corrected value of a monotopic 1:1

counterpart (K_1:1ditopic> 4K_1:1monotopic). The EM value is not in

itself a measure of chelate cooperativity, but the different ways

proposed for the assessment of such cooperativity do require

EM calculation.⁸⁻¹⁰In general, the EM value is assumed to be

an independent property of the supramolecular scaffold of the

tweezer receptor. Thus, in a “tweezer”-like complex based on

the same scaffold, one might expect that an increase in the

strength of the intramolecular interaction (coordination bond,

N···M) should produce a boost in the yield of the cyclic ditopic

complex (EM × K_m). The two metallobisporphyrin systems

Among the different metals that can be inserted into the

porphyrin core, we were interested in investigating iodo-

Rh(III) tweezer receptors. To the best of our knowledge, there

are no reports of supramolecular chirality induction with

Rh(III) porphyrins. Halogenorhodium(III) porphyrins are

diamagnetic, and they generally bind a single amine on the

4621

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

the condensation reaction used to prepare porphyrin 4 is

tetraalkylphenylporphyrin 5·H₂. 5-Benzyloxy-2,6-pyridine di-

carbonyl chloride 6 was prepared in four steps following a

previously published procedure.^33,34Coupling of aminopor-

phyrin 4 was accomplished by cannulating a solution of diacid

chloride 6 in methylene chloride into a solution containing

2.4 equiv of the porphyrin and excess triethylamine. Column

chromatography of the crude products allowed the recovery of

unreacted porphyrin and the isolation of bisporphyrin 1·H₄as a

purple solid in 61% yield. Metalation with Zn(II) was

accomplished in almost quantitative yield by using a solution

of zinc acetate in CH₂Cl₂/MeOH (3:1), yielding pure Zn-

monoporphyrin 5·Zn and Zn-bisporphyrin 1·Zn₂after column

chromatography purification with basic alumina. Metalation

of the free base porphyrins 5·H₂and 1·H₄with a

chlorodicarbonylrhodium(I) dimer, followed by in situ oxi-

dation with iodine and purification by column chromatography

with neutral alumina, afforded the diamagnetic iodorhodium-

(III) porphyrins 5·Rh and 1·Rh₂in approximately 85% yield.

Scheme 1. Synthesis and Structures of the Porphyrins Used

in This Study

a

1

The H NMR spectra of the metalated monoporphyrins

5·Zn and 5·Rh when dissolved in CH₂Cl₂-d₂or toluene-d₈show

sharp and well-defined signals that are easily assigned. We

observed two pair of doublets for the protons of the meso-

phenyl substituents in 5·Rh, indicating slow rotation of the

C_meso−C_phenylbond on the ¹H NMR time scale and confirming

that the two porphyrin faces are not chemically equivalent. The

proton signals characteristic of an axially coordinated methanol

molecule in the opposite face of the axial iodo ligand could also

be detected.²³Single crystals of 5·Rh·MeOH suitable for X-ray

diffraction analysis were grown from a dichloromethane/

methanol solvent mixture. Two crystalographically inequivalent

molecules of 5·Rh·MeOH are present in the asymmetric unit of

the crystal (Figure 2). A packing plot shows that the molecules

stacked on one another confronting the iodo ligand with the

coordinated methanol and forming a tilted columnar arrange-

ment. The planes of adjacent porphyrin units in the stack are

rotated 36° with respect to each other, whereas the porphyrin

planes of alternate units are parallel.

a

The line drawing structures of the amines used in the study are also

shown at the bottom of the scheme.

axial site trans to the halogen atom featuring a high binding

affinity and very slow dissociation kinetics.²¹Halogenorhodium(III)

porphyrins can be envisaged as structural analogues of Zn-

porphyrins providing thermodynamically and kinetically more

stable supramolecular assemblies based on ligand axial

coordination chemistry with amines. As mentioned above, the

extraction of EM values from the thermodynamic stability

constants calculated for 1:1 Zn(II) and Rh(III) “tweezer”-like

complexes will constitute a real test for the general assumption

that EM values can be extrapolated among analogous

supramolecular structures independently of the strength of

the intermolecular interactions. In addition, we found that

reports on the thermodynamic properties of covalently linked

Rh(III) porphyrin dimers are relatively scarce,²²although

Rh(III) monoporphyrins have received attention for the con-

struction of mono- and multimetallic oligoporphyrin systems

through axial coordination with ligands,²³⁻²⁷as receptors for

amino acid/esters^21,28,29or nucleobases, as chemical shift chiral

reagents for amines,³⁰and even in the template synthesis of

rotaxanes.³¹

1

The H NMR spectra of the bisporphyrins 1·Zn₂and 1·Rh₂

when dissolved in CH₂Cl₂-d₂(dielectric constant, dipolar

moment; 9.1, 1.6 D), CHCl₃-d (4.8, 1.04 D), or toluene-d₈(2.4,

0.36 D) show broad and unresolved proton signals, suggesting

the existence of aggregation processes or processes of inter-

conversion between conformers with exchange rate constants

that are intermediate on the ¹H NMR chemical shift time scale

(Figure 3). On the contrary, when bisporphyrins 1·Zn₂and

1

1·Rh₂are dissolved in THF-d₈(7.5, 1.75 D), H NMR spectra

with sharp and well-resolved proton signals are observed. This

result suggests that the broadening of the proton signals

observed in the nonoxygenated solvents is probably due to

aggregation processes induced by π−π interactions between the

porphyrin rings. Because the polar properties (dielectric con-

stant and dipolar moment) of THF and CH₂Cl₂are similar, it is

likely that the oxygen atom in the THF molecule coordinates to

the metal center, disrupting these intermolecular interactions.

The aggregation behavior of the bisporphyrin 1·Rh₂was

evidenced by a significant red shift of their Soret band during

dilution experiments in TOL solution in the range of

concentrations of 10⁻⁵−10⁻⁷M (Supporting Information).

The position of the porphyrin’s Soret band remained constant

at concentrations below 0.1 μM. The mathematical analysis of

the dilution data assuming a simple dimerization process

afforded K_d= 1.6 × 10⁵M⁻¹. In addition, the aggregation

RESULTS AND DISCUSSION

■

Synthesis of Porphyrins and Structural Considera-

tions. Aminoporphyrin 4 was prepared according to methods

described in the literature.³²We note that the major product of

4622

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

A simplified analogue of bisporphyrin 1a·Zn₂, in which three

meso-phenyl substituents in each porphyrin unit and the benzyl

ether in the linker are removed, was used to compute the gas-

phase energy of three conformers. Structures of the conformers

were optimized by the method described in the computational

details (Supporting Information). The calculations show a

5 kcal mol⁻¹energy preference for the out-out conformation.

The in-out conformation is 3 kcal mol⁻¹less stable than the out-

out conformation. This result suggest that, in the gas-phase,

bisporphyrin 1a·Zn₂adopts an out-out conformation almost

exclusively.

1

The H NMR spectrum of 1·Zn₂in THF-d₈shows a single

set of proton signals, implying either that 1·Zn₂is present in

solution as a single conformer, probably in the out-out con-

formation based on previous calculation results, or that the

interconversion between conformers is fast on the NMR time

scale (Figure 4). On the other hand, the ¹H NMR spectrum of

1·Rh₂in THF-d₈displays multiple signals for the same type of

protons, that is, NH, β-pyrrole, meso-phenyl aromatics, or

benzylic CH₂. We ascribe this different behavior to a slow

rotation on the NMR time scale of the C_meso−C_amidophenylbond

in the porphyrin units of 1·Rh₂that have chemically non-

equivalent faces. This slow rotation provides additional

conformers in equilibria with any of the three basic

conformations that 1·Rh₂can adopt with respect to the relative

spatial orientation of its porphyrin units (in-in, out-out, or in-

out; see Figure 3). As mentioned above, the similar dynamic

behavior observed with monoporphyrin 5·Rh doubled the

number of the signals of meso-phenyl protons.

To substantiate the chemical integrity of 1·Rh₂, we investi-

Figure 2. (a) Solid-state structure of 5·Rh·MeOH. (b) Packing of

5·Rh·MeOH in the solid state. Calculated planes defined by the four

nitrogen atoms of the porphyrin rings are shown highlighting the

relative orientation of the units in the stack. Hydrogen atoms have

been removed for clarity. All atoms in the porphyrin structure are

shown as thermal ellipsoids.

gated its complexation with the symmetric ditopic diamine

1

DABCO (1,4-diazabicyclo(2,2,2)octane) using H NMR spec-

troscopy at millimolar concentrations. The addition of

incremental amounts of DABCO to the solution of 1·Rh₂in

THF-d₄produces the appearance of a new set of sharp proton

signals that grew at the expense of the multiple protons signal

of free 1·Rh₂(Figure 5). We allocate the new set of signals to

the sandwich 1:1 complex 1·Rh₂⊃DABCO. Diagnostic signals

for the protons of the DABCO sandwiched guest are observed

at δ = −5 ppm. When 1 equiv of DABCO is added, only the set

of new proton signals is observed. The assignment of the new

set of signals to the protons of bound bisporphyrin is simple,

indicating that the ditopic coordination of DABCO locks

bisporphyrin 1·Rh₂into the cofacial conformer and confirms

the chemical integrity of the sample. The 1:1 sandwich complex

1·Rh₂⊃DABCO is highly thermodynamically stable and, even

at a millimolar concentration, does not open up in the presence

of excess DABCO (3.6 equiv).

Figure 3. DFT-optimized structures for the conformers of structurally

simplified bisporphyrins related to 1·M₂and proposed interconversion

equilibria in which they are involved.

properties of Zn-bisporphyrins analogous to 1·Zn₂have already

been reported, and for TOL solution, dimerization constant

values on the order of 5 × 10³M⁻¹were calculated.³⁵In DCM

solution, the aggregation process is sensibly reduced (i.e.,

K_d≈ 3 × 10²M⁻¹for 1·Zn₂) and completely eliminated, for

both 1·Zn₂and 1·Rh₂, if THF is used as solvent (Supporting

Information).

We intentionally employed spacer 6 as a mean to restrict the

conformational freedom of bisporphyrins 1·M₂.³⁶The internal

hydrogen bonds present in the 2,6-dicarboxamidopyridine unit

force the spacer to predominantly adopt a syn-syn con-

formation. However, the free rotation that still exists around the

NH−C_{meta‑meso‑phenyl}in 1·M₂enables the existence of at least

three different conformers incorporating the original preorga-

nization motif.

Thermodynamic Characterization of Model Com-

plexes. Before examining the induced chirality transfer of

enantiopure 1,2-diaminocyclohexane 2 to bisporphyrins 1·Zn₂

and 1·Rh₂, we investigated the binding properties of simpler

systems that cannot assemble into intramolecular sandwich

complexes. The values of the binding interactions derived from

the latter will be used as references or standards for the relevant

binding interactions that occur in the former. The binding

models used in the fit of the titration data derived from the

model complexes are described in detail in the Supporting

Information.

Determination of Microsocopic Binding Constants for the

Interactions R-NH₂···Zn(II) and R-NH₂···Rh(III)-I. The coordi-

nation of monoamine 7 (cyclohexylamine) with monoporphyr-

in 5·Rh was probed by UV−vis titrations in toluene (TOL),

4623

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

Figure 4. ¹H NMR spectra of 1·Rh₂and 1·Zn₂in THF-d₈solution. See Scheme 1 for proton assignment. Please notice that the two ortho and meta

protons in each meso-phenyl substituent in 1·Rh₂are chemically nonequivalent (different substitution of the two faces of the porphyrin unit) and

split into four different signals as a result of the slow rotation in the NMR time scale of the C_meso−C_phenylbond.

7−10 nm is characteristic for the formation of 1:1 axial

complexes of amines with Zn porphyrins. The mathematical

analysis of the titration data using the 1:1 binding model

allowed the accurate calculation of a binding constant value K_1:1

(5·Zn@7) = 1.0

0.2 × 10⁴M⁻¹that we assigned to the

microscopic interaction of R-NH₂···Zn(II) in TOL. We had

previously also determined the value of K_1:1(5·Zn@7) in

dichloromethane (DCM) solution as 5.0 × 10⁴M⁻¹.¹⁹This

result indicates a 5-fold reduction in the value of the

microscopic binding constant determined in the aromatic

solvent TOL compared to DCM.

Calculation of the Cooperativity Factor for the Bispor-

phyrin 1·Zn₂. As mentioned above, bisporphyrins 1·Zn₂and

1·Rh₂seem to aggregate at micromolar concentrations in TOL.

Dilution studies in the concentration range of 10⁻⁶−10⁻⁷M⁻¹

showed no evidence of this behavior for 1·Zn₂in both solvents

and a reduced aggregation process operating for 1·Rh₂below

1 μM concentration in TOL solutions. We studied the binding

of monoamine 7 with bisporphyrin 1·Rh₂in TOL and DCM

and with bisporphyrin 1·Zn₂in TOL using spectrophotometric

titrations. (See Scheme S4, Supporting Information, for the

detailed binding model.) The titrations of 7 with 1·Zn₂and

1·Rh₂were nonisosbestic. Probably, the nonisosbestic behavior

reflects small differences in the binding properties of the 1:1

and 2:1 complexes. In all cases, the titration data were analyzed

by considering three colored states for the bisporphyrin (free,

1:1, and 2:1).

On the one hand, the fit of the titration data of 1·Zn₂with 7

allowed the calculation of reliable stability constant values for

the complexes 1:1 K_1:1(1·Zn₂@7) and 1:2 K_1:2(1·Zn₂@7₂), as

well as their corresponding UV-spectra. The microscopic con-

stant value derived for the R-NH₂···Zn(II) interaction using the

equation K_m2= 2K_1:1↔1:2is very close to the value we obtained

in the titration of monoporphyrin 5·Zn with minomaine 7

(1.0 0.2 × 10⁴M⁻¹). However, the microscopic constant for

1

Figure 5. Downfield and upfield selected regions of the H NMR

spectra of 1·Rh₂in THF-d₈solution with increasing amounts of

DABCO. The inset shows the energy-minimized molecular structure

of the sandwich complex 1·Rh₂⊃DABCO. The complex is stable in the

presence of more than 1 equiv of DABCO. The small signals

resonating in the region of 8.6−8.8 ppm after the addition of 1.3 equiv

of DABCO are impurities. See the Supporting Information for a

detailed proton assignment.

showing a shift of the Soret band from 415 to 432 nm (see the

Supporting Information). We assigned the 17 nm shift to the

formation of the 1:1 axially coordinated complex 5·Rh@7.

Analysis of the binding isotherm using a 1:1 binding model

(Scheme S3a, Supporting Information) demonstrated that the

binding constant is too high to be measured accurately at this

concentration ([5·Rh] = 5.8 × 10⁶M]). Thus, we simply esti-

mated the binding constant value of K_1:1(5·Rh@7) as 1 × 10⁸M⁻¹

and assigned a similar value to the reference microscopic con-

stant K_mfor the interaction R-NH₂···Rh(III)-I in toluene.³⁷The

formation of 1:1 complexes between Rh(III) porphyrins and

amines having high thermodynamic stability and even surviving

column chromatography was previously reported by Sanders²²

and Aoyama.²⁹

the same interaction when derived from the relationship K_m1

=

K_1:1/2 is slightly smaller. This result indicates a decrease in the

binding affinity of the first interaction of 1·Zn₂with 7 com-

pared to the strength calculated for the model system (7 with

monoporphyrin 5·Zn). This small anticooperative behavior,

possibly produced by an interaction with the adjacent porp-

hyrin unit in 1·Zn₂, was quantified by a cooperativity factor

α_P= 0.5 calculated from the ratio of stepwise binding constants

K_1:1(1·Zn₂@7)/K_1:1↔1:2(1·Zn₂@7₂) = 4α_P(Scheme S4,

Similarly, we investigated the complexation of cyclohexyl-

amine 7 with 5·Zn in TOL. During the titration, 5·Zn experi-

enced a shift of the Soret band from 424 to 431 nm. A shift of

4624

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

Supporting Information). We derived a similar value for α_Pin a

closely related Zn bisporphyrin in the complexation with

quinuclidine.³⁶On statistical grounds, the expected value of α_P

for truly independent binding sites is 1.0; on the basis of our

definition of α_P, smaller numbers correspond to the

anticooperative behavior of the two porphyrin units.

On the other hand, the mathematical analyses of the titration

data of bisporphyrin 1·Rh₂with 7, in either TOL or DCM,

were carried out in a completely different manner. We fixed the

stability constants of the complexes to their statistically

estimated values using the relationships K_1:1= 2α_PK_mand

2

K_1:2= α_PK_m(K_m= 1 × 10⁸M⁻¹in TOL and 5 × 10⁸M⁻¹in

DCM). As determined for Zn-porphyrins 5·Zn and 1·Zn₂, we

assumed here that, for bisporphyrin 1·Rh₂, the K_mvalue also

displays a 5-fold increase when moving from TOL to DCM.

We also assigned a cooperativity factor of 0.5 to 1·Rh₂based on

the results obtained with 1·Zn₂. The UV−vis spectra of the

three species were the only variables considered during the

fitting procedure. For the two solvents, we obtained a very

good fit of the titration data to the theoretical binding

isotherms. In addition, the UV−vis spectra calculated for the

three species, in which 1·Rh₂is involved, displayed the spectral

features expected for their stoichiometry and geometry (see the

Supporting Information). These observations give some

indication of the quality and correctness of the data analysis,

as well as the values statistically estimated as lower limits for the

stability constants of the complexes, which, in turn, were

derived from our estimate of the microscopic binding constant

for the R-NH₂···Rh(III)-I.

Figure 6. Downfield and upfield regions of ¹H NMR spectra acquired

during the titration of 5·Rh with R,R-2 in CH₂Cl₂-d₂. Top: proton

signals for the β-pyrrole and ortho and meta meso-phenyl protons in

free and 1:1 and 2:1 complexes are indicated in the downfield

expansion. Bottom: upfield expansion showing separate signals for the

protons of the R,R-2 diamine in the 1:1 (assigned) and the 2:1

complexes (not assigned). See Scheme 1 for proton assignment.

Whereas the protons of the bound R,R-2 amine in the C₂symmetric

2:1 complex appear as 8 different signals, each one containing two

chemically, but not magnetically, equivalent protons, the same protons

in the nonsymmetric 1:1 complex should give rise up to 16 different

signals.

Calculation of the Cooperativity Factor for the R,R-2

Diamine. Next, we calculated the cooperativity factor for the

ditopic chiral diamine R,R-2 in both solvents, DCM and TOL.

We used titrations performed at millimolar concentrations with

1

monoporphyrins 5·Zn and 5·Rh using H NMR spectroscopy.

In DCM at 298 K, the system 5·Rh and R,R-2 displayed slow

exchange on the NMR time scale. Proton signals for free 5·Rh,

the 2:1 (porphyrin/amine) complex, and the 1:1 complex could

be assigned (Figure 6). In the presence of less than 0.5 equiv

of R,R-2, only free 5·Rh and (5·Rh)₂⊃R,R-2 were observed

in slow exchange. When 0.5 equiv of R,R-2 is added,

(5·Rh)₂⊃R,R-2 is the exclusive species in solution. The addition

of more than 0.5 equiv of R,R-2 provoked the observation of

separate proton signals for (5·Rh)₂⊃R,R-2 (2:1) and 5·Rh@

R,R-2 (1:1) complexes. As the concentration of R,R-2

increased, the equilibrium gradually shifted in favor of the 1:1

complex. The signal of the β-pyrrole protons in the 2:1 sand-

wich complex was shifted upfield relative to the corresponding

signal in free 5·Rh or bound in the 1:1 complex. This is due to

the proximity of the two porphyrin rings in the 2:1 sandwich

complex, causing a large ring-current-induced shift. The two

ortho and meta protons of the meso-phenyl substituents in free

5·Rh and in its complexes are chemically nonequivalent and

split in two doublets due to nonequivalent faces of the

porphyrin unit together with slow rotation on the NMR time

scale of the C_meso−C_p‑phenylbond (vide supra).

complex. The observation of separate proton signals for the two

complexes of different stoichiometry indicates that they are

involved in slow chemical exchange on the ¹H NMR time scale.

When 1 equiv of diamine is added, the 5·Rh@R,R-2 (1:1)

complex is the only species present in solution. The higher

symmetry (C₂) of the (5·Rh)₂⊃R,R-2 (2:1) complex compared

with that of 5·Rh@R,R-2 (1:1) becames apparent from the

analysis of the number of proton signals of the bound diamine

in the two respective states.

The association constants for the 1:1, 5·Rh@R,R-2, and 2:1,

(5·Rh)₂⊃R,R-2, complexes are too high to be measured

accurately at NMR concentration. However, the value of the

cooperativity factor α_Lcan be calculated using the following

relationship, K_2:1↔1:1= 4/α_L(Scheme S3b, Supporting

Information), where K_2:1↔1:1is the equilibrium constant of

the destruction of the 2:1 sandwich complex to produce the 1:1

complex. The value of K_2:1↔1:1was calculated from the

integration of the resonances of each of the two complexes

after 0.5 equiv of diamine was added, and a value of α_L= 0.05

was determined. This value indicates that the two nitrogen

atoms of the diamine R,R-2 experience a severe anticooperative

behavior upon coordinating to two porphyrin units.

In the upfield region of the ¹H NMR spectra, the behavior of

the signals of the protons for the bound diamine R,R-2 is in

agreement with the observations made in the aromatic region

of the spectra for the β-pyrrole protons. Until 0.5 equiv of

diamine is added, the observed proton signals for the bound

diamine correspond to the (5·Rh)₂⊃R,R-2 (2:1) complex.

Upon addition of increasing amounts of R,R-2, a new set of proton

signals start to appear, corresponding to the 5·Rh@R,R-2 (1:1)

For the systems 5·Rh with R,R-2 in TOL and 5·Zn with

R,R-2 in both TOL and DCM solutions, we observed fast chemical

exchange on the NMR time scale between free and bound

species. The reduction in kinetic stability detected for the Zn

4625

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

Table 1. Calculated and Estimated Macroscopic Binding Constants Derived from the Model Systems

b

porphyrin

ligand

solvent

K_1:1[10⁵M⁻¹

]

K_1:2(1·M₂) K_2:1(5·M) [10⁹M⁻²

]

K_1:1↔1:2(1·M₂) K_1:1↔2:1(5·M) [10⁵M⁻¹

]

c

5·Rh

5·Zn

1·Zn₂

1·Rh₂

5·Rh

5·Zn

7

TOL

DCM

TOL

DCM

TOL

DCM

TOL

DCM

TOL

1000

7

0.1 0.02

a

7

0.5 0.1

a

7

0.14 0.06

0.093 0.008

0.06 0.002

0.4 0.1

a

7

1.2 0.5

4.5 0.5

c

7

1000

5 × 10⁶

c

7

5000

125 × 10⁶

c

R,R-2

2000

1 × 10⁵0.2 × 10⁵

5

1

c

0.2

0.005

0.079

0.0025

0.0079

c

DCM

1.0

a

b

c

Values taken from refs 19 and 36. K_1:1↔1:2= K_1:2/K_1:1for 1·M₂and K_1:1↔2:1= K_2:1/K_1:1for 5·M. Statistically estimated; see text for details.

Table 2. Microscopic Binding Constants for the M···N Interaction and Cooperativity Factors for the Ditopic Ligand R,R-2 and

Bisporphyrins 1·M₂

d

porphyrin

ligand

solvent

K_m1[10⁵M⁻¹

]

K_m2[10⁵M⁻¹

]

α_P

α_L

a

5·Rh

5·Zn

1·Zn₂

5·Rh

5·Zn

7

TOL

DCM

TOL

DCM

TOL

DCM

TOL

1000

a

7

0.1

a

7

0.5

b

c

7

0.07 0.003

0.13 0.06

0.5 0.2

0.8 0.3

b

c

7

0.6 0.2

0.7 0.3

e

f

R,R-2

0.01

0.05

0.03

e

5·Zn

R,R-2

c

DCM

d

a

b

e

f

K_m1= K_1:1. K_m1= K_1:1/2. K_m2= 2K_11↔21. α_P= K_1:1/4K_1:1↔1:2. α_L= 4/K_2:1↔1:1, where K_2:1↔1:1= K_1:1/K_1:1↔2:1. Calculated from integration of the

¹H NMR signals.

system correlates well with the diminution of thermodynamic

stability. Specifically, the microscopic binding constant

calculated for the R-NH₂···Zn(II) is 4 orders of magnitude

smaller than our estimate for the R-NH₂···Rh(III) interaction.

The reduced kinetic stability observed for the rhodium

complexes in TOL also reflects a decline in binding affinity

in this solvent compared with DCM. This observation supports

the supposition that the ratio between the microscopic constant

values in both solvents that we derived from Zn-porphyrin

titrations can be extrapolated to the Rh(III) analogues.

experiencing fast chemical exchange in the NMR time scale.

The increase in the concentration of R,R-2 induces downfield

shifts of the signals because the free and 1:1 complex states of

the ligand become more populated. The chemical shift changes

for the meso and β-pyrrolic protons are in agreement with

those discussed for the system 5·Rh with R,R-2 in DCM but are

rather involved in fast chemical exchange (Supporting

Information). The determined α_Lvalues for the diamine R,R-

2 in the systems experiencing fast exchange are on the same

order of magnitude and also coincide with the α_Lvalue

calculated for the system 5·Rh with R,R-2 in DCM featuring

slow chemical exchange. In conclusion, the two binding sites

(nitrogen atoms) in diamine R,R-2 show a markedly stronger

anticooperative behavior than the two porphyrin units of the

bisporphyrins 1·M₂. The cooperativity factor values calculated

for the ditopic ligand R,R-2 and bisporphyrins 1·M₂also

evidence a reduced solvent modulation effect. Tables 1 and 2

summarize the results derived from the model systems.

Molecular modeling studies assigned several binding geo-

metries to the 2:1 sandwich complexes (5·M)₂⊃R,R-2. For

example, Figure 7 depicts two CAChe-minimized structures for

(5·Zn)₂⊃R,R-2; in both of them, the diamine R,R-2 places the

amino groups in an equatorial orientation. However, the axially

coordinated porphyrin units adopt different spatial arrange-

ments in trying to avoid steric crowding. Most likely, in

solution, dynamic equilibria between different geometries of the

complexes are at work.

1

From the previous H NMR titration experiment of 5·Rh in

DCM, we learned that, up to the addition of 0.5 equiv of R,R-2,

the (5·Rh)₂⊃R,R-2 sandwich complex is the species predom-

inantly formed. When more than 0.5 equiv of R,R-2 is added,

the 2:1 complex, (5·Rh)₂⊃R,R-2, is destroyed to yield the 1:1

complex, 5·M@R,R-2. For the systems in fast exchange, we

determined the equilibrium constant for the destruction of the

2:1 complex, K_2:1↔1:1, from the ratio of the macroscopic stabi-

lity constants of the two complexes, K_2:1↔1:1= K_2:1/K_1:1↔2:1. In

turn, the stability constants of the complexes were determined

by fitting the changes in chemical shifts recorded in NMR

titrations to the corresponding binding model (1:1 and 2:1

complex formation). To reduce the number of variables during

the fitting procedure, we fixed K_1:1to the statistically estimated

1

value K_1:1= 2K_m. At a low guest-to-host molar ratio, the H

NMR spectra of the titrations of the monoporphyrin 5·Zn

showed the existence of four highly upfield-shifted proton

signals (Supporting Information). The relative integral ratio of

these signals was 1:1:1:1, and their chemical shift values must

correspond to the weighted average chemical shifts of some of

the protons of R,R-2 in the three different stoichiometric states

Thermodynamic Equilibria Involved in the Binding of

Ditopic Amine Ligand with Bismetalloporphyrins. At

micromolar concentrations, the interaction of ditopic amines

with bisporphyrins produces sandwich motifs encountered in

4626

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

1,2-diaminocyclohexane 2 with Zn bisporphyrin 1a·Zn₂in

DCM solution.¹⁹UV−vis titrations of 1·Zn₂with R,R-2

revealed the initial formation of a thermodynamically stable

1:1 sandwich complex, 1·Zn₂⊃R,R-2, at low molar guest-to-host

ratios. In the presence of excess diamine, the sandwich complex

opened up to form a simple 2:1 complex, 1·Zn₂@(R,R-2)₂. The

sandwich complex, 1·Zn₂⊃R,R-2, displayed only a 2 nm red

shift of the Soret band with respect to the maximum of free

1·Zn₂. This is characteristic of sandwich complex formation in

which the two porphyrin units of the tweezer are simulta-

neously bound to a ditopic ligand and experience strong exciton

coupling between their transitions.³⁸Interestingly, the CD

titration revealed that the sandwich complex, 1·Zn₂⊃R,R-2, was

endowed with a very strong (A = 783 M⁻¹cm⁻¹) negative

exciton couplet. With respect to the higher stoichiometry 1:2

complex, 1·Zn₂@(R,R-2)₂, the red shift of the Soret band was

close to 10 nm, typical for open 1:2 bisporphyrin/amine. Quite

unexpectedly, the mathematical analysis of the titration CD

spectra assigned a weak (A = 139 M⁻¹cm⁻¹) positive bisignate

signal to the 1·Zn₂@(R,R-2)₂complex, for which we had

anticipated a silent CD. A direct visual analysis of the raw data

obtained in the CD titration already made clear the existence of

CD sign inversion when moving form low to high guest-to-host

ratios.

Figure 7. CAChe-minimized structures of different binding geometries

for the 2:1 complex (5·Zn)₂⊃R,R-2: (a) The R,R-2 ligand is

sandwiched between two porphyrin units located at the edges of the

ditopic guest. (b) The R,R-2 ligand is sandwiched between two

porphyrin units located on the top and bottom faces of the guest. 5·Zn

is shown as a stick representation. Hydrogen atoms and pentyl chains

are removed for clarity. R,R-2 is displayed as a CPK model.

intermolecular “ladders” or intramolecular “tweezer” complexes

(Scheme 2). The intermolecular ladder-like complexes are

Scheme 2. Schematic Representation of the Possible

Equilibria Involved in the Binding of a Bisporphryin with a

a

Ditopic Diamine

Our previous studies with model systems evidenced a subtle

influence of the solvent (DCM and TOL) in the microscopic

binding constant value, K_m, for the RNH₂···Zn(II) interaction

(Table 1). The K_mvalue is 5-fold higher in DCM than in TOL.

It is probable that TOL better solvates the π-system of the

porphyrin ring. We were interested in finding out how the

reduction in binding affinity would be expressed in the

chirogenesis process. We described above how we determined

the α_pand α_Lvalues for 1·Zn₂and R,R-2, respectively, using

model systems, and we noticed that these values were almost

not affected by changing the solvent. Consequently, they

should not be responsible for alterations in the chirogenesis

process upon a change in solvent.

First, we probed the interaction of 1·Zn₂with R,R-2 using

UV−vis titrations in TOL solution. The bisporphyrin

concentration was maintained constant at ca. 1.25 × 10⁻⁶M.

The initial addition of R,R-2 led to a shift of the Soret

absorption band of 2 nm (422−424 nm) and a concomitant

reduction of the half-bandwidth. This small 2 nm red shift is

diagnostic of 1:1 ditopic bisporphyrin/diamine sandwich

complexes experiencing exciton coupling.³⁹Upon addition of

more R,R-2, a new absorption appeared at 428 nm

corresponding to the 1:2 open complex. As observed for the

reference compound 5·Zn, a Soret band shift of 10 nm is

expected for axially coordinated complexes of Zn-porphyrins.

The reduced red shift (6 nm compared with 10 nm) of the

Soret band observed by the end of the titration could be

indicative of a residual exciton in the 1:2 complex (Figure 8).

The half-bandwidth of the final Soret band is noticeably smaller

than that of the initial band and comparable with the one

obtained after the initial additions. A reduction in the half-

bandwidth suggests that the bisporphyrin adopts a preferred

conformation and exhibits a reduction in conformational

freedom. We did not observe any isosbestic point during the

titration. This finding is in agreement with the existence of

simultaneous multiple equilibria in the binding process. The

titration data were analyzed using multivariate factor analysis

and considering a binding model with three colorimetric

stoichiometric states for 1·Zn₂(free, 1:1 and 1:2 complexes);

a

EM is the effective molarity for the intramolecular interaction

required for cyclization in the sandwich complex. Overall stability

constants (K_1:2and K_2:2) and stepwise equilibrium constants (K_1:1

K_1:1↔1:2, K_2:2↔1:2) are indicated. The relevant equilibrium constants for

this study are related to K_m, EM, α_L, α_P, and statistical correction

factors.

,

stable in the presence of a moderate excess of diamine, whereas

the intramolecular sandwiches usually resist higher concen-

trations of diamine. Increasing the concentration of the amine

finally breaks down the sandwich complexes, yielding 1:2

(porphyrin/amine) complexes. We demonstrated previously¹⁹

that Zn-bisporphyrins, 1·Zn₂, form exclusively 1:1 intra-

molecular sandwich complexes with ditopic amines that are

shape- and size-complementary, that is, DABCO and R,R-2

(R,R-1,2-diaminocyclohexane). For this reason, the formation

of “ladder”-like complexes will not be considered in the binding

model used to analyze the titration data of “tweezer”

bisporphyrins 1·M₂with enantiopure 1,2-diaminocyclohexane

described in the following sections.

Supramolecular Chirality Induction in the Coordina-

tion of Bisporphyrin Tweezer 1·Zn₂with Enantiopure

Ligand 2. We reported that the properties of chirality

induction and inversion were both expressed and stoichio-

metrically controlled in the complexation of enantiopure

4627

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

Figure 8. (a) UV−vis and (d) CD spectra obtained during the titration of 1·Zn₂and R,R-2 in toluene solution at 298 K. The concentration of 1·Zn₂

was maintained constant throughout the titration (1.25 × 10⁻⁶M). The curves shown correspond to the addition of 0, 1.8, 7.4, 23.4, and 367 equiv

of R,R-2. Calculated spectra for the 1:1 (continuous line) and 2:1 complexes (dashed line): (b) UV−vis (Soret band) and (e) CD. Fit of the titration

data to the theoretical binding isotherm at selected wavelengths: (c) UV−vis and (f) CD.

see Scheme 2. The UV−vis spectra of free 1·Zn₂was fixed, and

the fitting procedure returned the following values of K_1:1

formation and produces a 1:2 complex 1·Zn₂@(R,R-2)₂

possessing a positive couplet with an unexpectedly high

amplitude. The analysis of the CD titration data fixing the

values of the stability constants of the complexes to those

calculated from the UV−vis titration allowed the calculation of

the CD spectra for the two species (Figure 8e). The reasonable

good fit observed between the experimental titration CD data

and the theoretical binding isotherms supports the quality of

the data analysis and of the predicted CD spectra.

In trying to assign a molecular structure to the 1:1 sandwich

complex and rationalize the observed CD couplet by means of

electronic and structural parameters, the low-energy geometries

of the diastereoisomeric sandwich 1:1 complexes 1a·Zn₂⊃R,R-2

were investigated by DFT-based calculations. In Figure 7, we

showed that, in a sandwich complex involving diamine R,R-2

and two monoporphyrins, it is possible for the diamine to be

oriented in at least two different arrangements. Similarly, the

1:1 complex 1a·Zn₂⊃R,R-2 could arrange the diamine in two

possible geometries (Figure 9). Both structures of the 1:1

complex lead to calculated UV/vis and CD spectra in

agreement with the experiments. However, the complex in

which the amine adopts a parallel orientation with respect to

the porphyrin rings is significantly higher in energy, approxi-

mately 10 kcal mol⁻¹, than the one in which the diamine is

perpendicular to the porphyrin planes. Consequently, all fur-

ther calculations consider the diamine placed in a perpendicular

orientation.

=

6.7 × 10⁴M⁻¹and K_1:2= 4.5 × 10⁹M⁻². The K_1:1value deter-

mined for the 1:1 complex and its calculated spectrum are in

support of the existence, to a certain extent, of a sandwich

complex geometry with a ditopic interaction between the

diamine and the bisporphyrin tweezers. Taken together, the

results obtained are completely in agreement with the initial

formation of a 1:1 complex displaying a sandwich-like geo-

metry, which is destroyed by the addition of excess diamine,

affording a 1:2 complex in which, surprisingly, residual exciton

coupling and conformational bias are still present. Having

determined the stability constant value of the 1:1 sandwich

complex, we became interested in calculating its effective

2

molarity (EM) using the relationship K_1:1= 2α_Pα_LK_mEM. The

use of this equation implicitly assumes the formation of a single

closed state for the 1:1 complex.⁴⁰We determined a value of

EM = 0.01 M for the 1:1 complex, indicating that the con-

formation adopted by the bisporphyrin is not highly strained.

Similar values of EM (0.03 M) were calculated using DABCO

as the ligand for a related bisporphyrin tweezer.³⁶

When the interaction of 1·Zn₂and R,R-2 was monitored

using electronic CD spectroscopy, we observed that the initial

addition of the amine led to the appearance and stepwise en-

hancement of a negative bisignate Cotton effect centered close

to the maximum of the porphyrin Soret band. Nevertheless, the

increase in concentration of R,R-2 produced a rapid decrease in

the initial ellipticity values of the sample and the generation

of a more intense, but positive, CD couplet. We rationalize

these results through the formation of an initial 1:1 complex

1·Zn₂⊃R,R-2 with reduced thermodynamic stability and as-

sociated with a negative couplet. The reduction in the strength

of the intramolecular interaction Zn···N upon moving from

DCM to TOL (5-fold reduction in K_m) produced a strong

impact in the thermodynamic stability of the sandwich. The

destruction of the 1:1 complex occurs simultaneously to its

The optimized structures and calculated UV−vis and CD

spectra obtained from the DFT calculations of the two energy

minima diastereoisomeric complexes, with the diamine oriented

in a perpendicular arrangement, are shown in Figure 10. There

is little energetic preference for either of the two different chiral

diastereoisomeric complexes. In diastereoisomer 1a·Zn₂⊃R,R-

2-(-), the relative orientation of the “privileged” or effective

electric transition moments,⁶B_5,15, which corresponds to the

direction of the covalent linkage between the porphyrin and the

4628

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

favorable energy (relative energy 0.6 kcal/mol) than the related

diastereoisomer with a clockwise orientation of dipoles

1a·Zn₂⊃R,R-2-(+). We assume that the effective transition

picture is justified in bisporhyrin 1·Zn₂due to fast and

unhampered rotation of the porphyrin units around the 5-15

axis.⁴¹The calculated energy difference translates in a 73%

percentage preference (K = e^ΔG/(−RT)= 2.75) for the

1a·Zn₂⊃R,R-2-(-) in solution considering the exclusive formation

of the two diastereoisomers under stoichiometric control (1:1).

We are aware that this energy is in the range of experimental

error in the calculations; nevertheless, according to the exciton-

chirality method, the observed relative orientation between

electronic transitions in the energy-minimized structure of

1a·Zn₂⊃R,R-2-(-) nicely predicts the negative chirality observed

experimentally. Furthermore, using the structure of 1·Zn⊃R,R-

2-(-) as input, the experimental electronic absorption and CD

spectra are nicely reproduced by TD-DFT calculations at both

the BP86/TZP and the SAOP/TZP levels. Structures for the

1·Zn⊃S,S-2 complexes were also calculated. The lower energy

structure for the 1a·Zn₂⊃S,S-2-(+) diastereoisomer is very close

to that of the enantiomer of 1a·Zn₂⊃R,R-2-(-) and shows an

energy preference of 0.6 kcal/mol with respect to the

1a·Zn₂⊃S,S-2-(-) diastereoisomer of minimum energy. This

result is in complete agreement with the fact that the

experimental spectral and binding parameters obtained for

S,S-2 are identical to those of R,R-2 within experimental error,

except for the sign of the couplet, which is opposite.

Figure 9. DFT-optimized structures of two different binding

geometries for the 1:1 sandwich complex. (a) Parallel orientation of

the diamine with respect to the porphyrin planes. (b) Perpendicular

orientation of the diamine with respect to the porphyrin units. The

parallel orientation of the diamine provides a 1:1 complex that is

thermodynamically 10 kcal mol⁻¹less stable than the perpendicular

one. Bisporphyrin 1a·Zn₂is shown as a stick representation and

diamine 2 as a CPK model.

We were also interested in assigning a molecular structure to

the 1:2 complex produced by the destruction of the 1:1 com-

plex in the presence of excess diamine. This high stoichiometry

complex displayed chirality inversion with respect to its 1:1

precursor and an unexpectedly high amplitude of its CD spectra

spacer (C-5/C-15), is anticlockwise with a dihedral angle of 34°

between them. This diastereoisomer features a slightly more

Figure 10. Side, front, and top views of the two energy-minimized diastereoisomeric 1:1 sandwich complexes 1a·Zn₂⊃S,S-2-(-) and 1a·Zn₂⊃S,S-2-

(+). The plane defined by the atoms C-5, C5′, and C15′ is shown in all structures. The effective transition dipole moments of the porphyrin units are

represented by green double arrows. The value of the dihedral angle between the two planes (C-15, C-5, C5′, C-15′) that contain the effective dipole

moments and the predicted sign for the expected bisignate Cotton effect based on the exciton rule are indicated. Calculated UV−vis and CD spectra

of the complexes are depicted.

4629

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

Figure 11. Side, front, and top views of the two energy-minimized diastereoisomeric 2:1 sandwich complexes 1a·Zn₂@(R,R-2)₂-(-) and 1·Zn@(R,R-

2)₂-(+). The effective transition dipole moments of the porphyrin units are represented by green double arrows. The predicted sign for the expected

bisignate Cotton effect based on the exciton rule is indicated. Calculated UV−vis and CD spectra of the complexes are depicted.

especially in TOL solution. In addition, the UV−vis spectra of

the 1:2 complex in DCM and more noticeably in TOL solution

showed signs of the existence of residual exciton coupling

between the porphyrin chromophores. That is, the maximum

of the Soret band of the 1:2 complex was slightly blue shifted

(418 nm) with respect to the absorption maximum (420 nm)

provided by a model system consisting of monoamine 7 axially

coordinated to monoporphyrin 5·Zn. Molecular modeling

studies suggested that the observed residual exciton coupling

between the porphyrin chromophores in the 1:2 1a·Zn₂@(R,R-2)₂

complex could be the result of the establishment of intra-

molecular interactions between the two axially coordinated

ditopic amines 2. Different chiral monoamines we assayed, that

is, S-8, did not give rise to 1:1 or 1:2 CD-active complexes;

thus, the ditopic nature of the ligand seems to be fundamental

to yield CD activity. Figure 12 shows the DFT-optimized structures

adjacent bound R,R-2 ligands. In general, amines are poor

hydrogen-bond donor groups; we propose here that the

coordination of the nitrogen atom to the Zn metal center in-

creases the acidity and hydrogen-bonding capabilities of its hydro-

gen atoms. The energy inclination for the 1a·Zn₂@(R,R-2)₂-(+)

diastereoisomer is in complete agreement with the observed

inversion of chirality. The CD spectra calculated for both

diastereoisomeric complexes are depicted in Figure 11. The

energetic stabilization of the in-in conformer is expected to

increase with stronger intramolecular hydrogen-bonding inter-

actions. We speculate that the higher amplitude observed in the

CD of the 1:2 complex registered in TOL solution compared to

DCM is a direct consequence of the formation of stronger

intramolecular hydrogen bonds in the less-polar solvent

(TOL).

The 1:2 complex could also adopt the out-out conformation

(Figure 12). Optimization of this conformer leads to a C₂

symmetric structure, showing the formation of hydrogen

bonding between the NHs of the free amine and the carbonyl

oxygen atoms of the amide groups of the spacer. Because of the

unfavorable characteristics of amines as hydrogen-bond donors,

the formed hydrogen bonds should be very weak. Nevertheless,

the calculations assign a lower energy to the out-out conformer

of the 1a·Zn₂@(R,R-2)₂complex compared with that of the in-in

analogues. Although the formation in solution of the out-out

conformer cannot be ruled out, its structure does not support

the blue shift observed in the Soret band of the UV−vis spec-

trum of the 1:2 complex. Additionally, the predicted CD

spectrum for the out-out conformer does not agree with that

observed experimentally. A plausible scenario is the formation

of the 2:1 complex as a mixture of conformers involved in fast

chemical equilibria, and the in-in conformer with a positive

twist between the effective transition dipole moments of the

porphyrin units being the major contributor to the experi-

mentally observed CD spectrum.

Figure 12. DFT-optimized structure for the out-out conformer of the

1:1 complex 1a·Zn₂@(R,R-2)₂.

for two diastereoisomeric complexes 1a·Zn₂@(R,R-2)₂with the

sense of the twist between the effective transition dipole

moments of the porphyrin rings reversed. The calculated

energy difference between 1a·Zn₂@(R,R-2)₂-(+) and

1a·Zn₂@(R,R-2)₂-(-) is 2.6 kcal mol⁻¹in favor of the former.

Probably, the 1:2 complex adopts the in-in conformation as a

consequence of the formation of intramolecular hydrogen

bonds between the NHs and the nitrogen atoms of the two

4630

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

Supramolecular Chirality Induction in the Coordina-

tion of Bisporphyrin Tweezer 1·Rh₂with Enantiopure

Ligand 2. We commented above that the value of the

microscopic binding constant for the interaction R-NH₂···Rh-

(III)-I is too high to be measured accurately even using

absorption spectroscopy. This is also the case for the stability

constants of the 1:1 and 1:2 complexes if they were produced in

separated phases of the chirogenesis process of 1·Rh₂induced

by coordination with 2. We performed the spectroscopic

titrations of bisporphyrin 1·Rh₂in TOL and DCM solutions

with diamine R,R-2 using the same methodology described for

1·Zn₂. We observed that the maximum of the absorption Soret

band of free 1·Rh₂in TOL solution is centered at λ = 414 nm.

The initial addition of R,R-2, induces a decrease in the intensity

of the Soret absorption and the appearance of a new absorption

band at 425 nm. We assign this 11 nm shift of the Soret band to

the formation of a 1:1 complex, 1·Rh₂⊃R,R-2, in which the

porphyrin binding sites are excitonically coupled. During the

initial phase of the titration, a sharp isosbestic point was

detected, suggesting that the formation of the 1:1 complex is

the main process. Upon addition of more R,R-2, the titration

entered a nonisosbestic phase. This is a result of the simul-

taneous existence of two binding events: the formation of the

1:1 complex and its destruction to afford the 1:2 complex. In

the last stages, the titration entered another isosbestic phase

since the only process taking place is the formation of the

1:2 complex from the 1:1. During the last phase, the band at

425 nm shifted to 430 nm. As observed for the reference system

5·Rh/7 in TOL, a Soret absorption at 435 nm is typical for a

1:1 complex of an axially coordinated monoamine to a Rh(III)

porphyrin. It is probable that the 5 nm blue shift observed for

the final Soret band is due to the existence of residual exciton

coupling between the porphyrin units of the 1:2 complex,

similar to the case explained above for bisporphyrin 1·Zn₂. The

complexation process of R,R-2 with 1·Rh₂was also monitored

using NMR spectroscopy in TOL solution. The incremental

addition of up to 1 equiv of amine produces the appearance of

of R,R-2 sandwiched between two porphyrin units, and the

similarity of spectra hints to an analogy of the sandwich

complexes’ structures. In the case of 5·Rh, the addition of more

than 0.5 equiv of R,R-2 induced the destruction of the 2:1 inter-

molecular sandwich complex (5·Rh)₂⊃R,R-2 and produced the

appearance of a new set of proton signals corresponding to the

1:1 complex 5·Rh@R,R-2. In contrast, for bisporphyrin 1·Rh₂,

the sandwich complex formed with R,R-2, 1·Rh₂⊃R,R-2 has a

1:1 stoichiometry and is intramolecular (see the Supporting

Information for additional characterization). In the presence of

more than 1 equiv of R,R-2, the 1:1 sandwich complex opens

up to form a 1:2 complex 1·Rh₂@(R,R-2)₂. The high number of

signals observed for the protons of the coordinated diamine in

the 1·Rh₂@(R,R-2)₂complex is probably due to both the slow

rotation on the NMR time scale of the C_meso−C_amidophenylbonds

and the desymmetrization of the bound ligand.

First, we attempted the mathematical analysis of the UV−vis

titration data of 1·Rh₂with R,R-2 in TOL solution using the

Specfit software.⁵³We fixed the stability constants of the two

complexes, 1:1 and 1:2, to their statistically estimated values

2

using the following equations, K_1:1= 2α_Pα_LK_mEM and K_1:2

=

2

4α_PK_m. We applied the values of K_mand cooperativity factors

(α_Pα_L) derived from the study of the model systems. We had

used a similar strategy for the successful fitting of the titration

data of 1·Rh₂with 7 in the study of model systems. In the case

at hand, we also considered that the EM value calculated for the

systems 1·Zn₂/R,R-2 was applicable to the 1·Rh₂analogue.

Surprisingly to us, this procedure led to unacceptable fits of the

experimental UV−vis titration data to the theoretical binding

isotherms.

Next, the fit of the UV−vis titration data in TOL was assayed

by fixing the spectrum of free bisporphyrin 1·Rh₂and K_1:2to

the statistically estimated lower value, as indicated above. With

this procedure, the spectral signatures calculated for the UV−

vis spectra of the two complexes, 1:1 and 1:2, were in full

agreement with their expected binding geometries. The

calculated absorption spectra displayed absorption maxima at

425 and 430 nm, for the 1·Rh₂⊃R,R-2 and 1·Rh₂@(R,R-2)₂

species, respectively. Moreover, the good fit obtained for the

titration data to the theoretical binding isotherms at different

wavelengths provided an indication of the quality of this data

analysis. The optimized value of K_1:1returned from the fitting

was 5.0 1.0 × 10⁹M⁻¹, close to 3 orders of magnitude smaller

than the value statistically estimated as the lower limit (1.5 ×

10¹²M⁻¹).

1

proton signals in the upfield region (0 to −8 ppm) of the H

NMR spectra (Figure 13). The number of signals and their

The absolute value of K_1:1is clearly too high to be deter-

mined accurately from the fit of the titration data corres-

ponding to its formation (initial phase of the titration). How-

ever, since the formation and destruction of the 1:1 complex

occurs simultaneously in the middle phase of the titration, the

mathematical analysis of the titration data described above

allowed the precise determination of the ratio K_1:2/K_1:1

independently of the K_mvalue used, as long as it is considered

to be equal or higher than the lower estimate 10⁸M⁻¹.⁴²The

equilibrium constant for the destruction process of the 1:1

complex to afford the 2:1 corresponds to the ratio of stability

constants of the complexes K_1:1↔1:2= K_1:2/K_1:1. In short, fixing

K_1:2to a statistical estimate, the fitting procedure allows for a

satisfactory evaluation of the magnitude of K_1:1in relationship

to the estimated K_1:2. Sound values for the EM operative in the

1:1 complex can subsequently be calculated from K_1:1or,

1

Figure 13. Upfield region of the H NMR spectra registered during

the titration of bisporphyrin 1·Rh₂with R,R-2 in TOL solution. The

fact that the 1:1 sandwich complex opens up in the presence of more

than 1 equiv of diamine is indicative of a low EM value (vide infra).

chemical shifts are almost coincident with the ones observed for

the titration of 5·Rh with R,R-2 up to the addition of 0.5 equiv

of diamine (Figure 6). These signals correspond to the protons

alternatively, K_1:1↔1:2

.

4631

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

Figure 14. (a) UV−vis and (d) CD spectra obtained during the titration of 1·Rh₂and R,R-2 in toluene solution at 298 K. The curves correspond to

the addition of 0, 0.4, 1.1, 2.2, and 28.1 equiv of R,R-2. Calculated spectra for the 1:1 (continuous line) and 1:2 complexes (dashed line): (b) UV−vis

(Soret band) and (e) CD. Fit of the titration data to the theoretical binding isotherm at selected wavelengths: (c) UV−vis and (f) CD.

Table 3. Calculated Stability Constants Values, K_1:1and K_1:2, for the R,R-2/1·Zn₂System. Calculated Stability Constants K_1:1

and Statistically Estimated K_1:2Values for R,R-2/1·Rh₂. Measured Spectral UV−vis Characteristics of Free Bisporphyrins and

Calculated Values for the Two Complexes in Which They Are Involved Are Represented

free

1:1 complex

1:2 complex

a

solvent,

porphyrin

λ_max, fwhm , ε [nm, nm,

10⁵M⁻¹cm⁻¹

]

K_1:1[M⁻¹

]

10⁵M⁻¹cm⁻¹

]

K_1:2[M⁻²

]

10⁵M⁻¹cm⁻¹

]

TOL, 1·Zn₂

DCM, 1·Zn₂

TOL, 1·Rh₂

DCM, 1·Rh₂

423, 14, 8.4

421, 15, 9.8

414, 31, 2.5

418, 35, 3.4

6.7 1.3 × 10⁴

1.0 0.2 × 10⁶

5.0 1.0 × 10⁹

2.9 0.6 × 10¹¹

424, 9, 9.5

423, 11, 9.4

424, 30, 2.5

4.5 0.9 × 10⁹

426, 12, 8.8

429, 15, 8.3

429, 28, 2.8

431, 27, 4.6

7.9 1.6 × 10⁹

b

1.0 × 10¹⁶

b

424, 29, 3.8

4.0 × 10¹⁷

a

b

2

fwhm: full width at half-maximum. Statistical estimates K_1:2= 2α_PK_m

.

We then probed the supramolecular chirogenesis of the 1·Rh₂/

bisporphyrin, we obtained a good fit of the UV titration data to

the theoretical binding isotherm, which gives some support to

the quality of the data analysis. The returned value of K_1:1was

2.9 × 10¹¹M⁻¹, again 3 orders of magnitude smaller than the

statistical estimate (1.1 × 10¹⁴M⁻¹). The calculated absorption

spectra for 1·Rh₂⊃R,R-2 and 1·Rh₂@(R,R-2)₂returned from

the fit displayed absorption maxima at 425 and 435 nm,

respectively. The 10 nm blue shift of the Soret band assigned to

the 1·Rh₂⊃R,R-2 with respect to 1·Rh₂@(R,R-2)₂supported

the existence of exciton coupling in the 1:1 complex (Figure

14). On the contrary, the coincidence in absorption maxima for

1·Rh₂@(R,R-2)₂and the model system 7@5·Rh indicates the

lack of exciton in the 1:2 aggregate. The multivariate factor

analysis of the CD titration data in DCM, using the same

methodology described for the analysis in TOL solution,

assigned a positive couplet to the 1:1 complex 1·Rh₂⊃R,R-2

and a silent CD to 1·Rh₂@(R,R-2)₂. The amplitudes of the

positive couplets assigned to the 1·Rh₂⊃R,R-2, 1:1 complex, in

TOL and DCM, are of the same order (1 × 10²). A silent CD

for the 2:1 complex, 1·Rh₂@(R,R-2)₂, in DCM agrees with the

lack of exciton coupling we assigned to the high stoichiometry

complex in this solvent. This result is also in full agreement

with our hypothesis stating that the amount of the 1:2 complex

that adopts the in-in conformer is higher in TOL than in DCM.

The intramolecular hydrogen bonding that stabilizes the in-in

conformer is more favorable in the less-polar TOL solvent.

Table 3 lists the stability constant values K_1:1and K_1:2calcu-

lated for the system R,R-2/1·Zn₂from the fit of the UV−vis

R,R-2 system also in TOL solution using CD spectroscopy.

The incremental addition of R,R-2 to a TOL solution of 1·Rh_2,

maintained at a constant concentration, initially produced the

appearance and stepwise increase of a positive CD couplet,

followed by a decrease in its amplitude. The multivariate analy-

sis of the CD data titration fixing K_1:2to the statistically

calculated value and K₁₁to the calculated value from the UV

titration provided a good fit and returned the spectral para-

meters for the observed chirality induction processes. Both

1·Rh₂⊃R,R-2 and 1·Rh₂@(R,R-2)₂complexes were assigned

with positive couplets. As could be expected, the amplitude of

the 1:1 complex 1·Rh₂⊃R,R-2 is higher (6-fold) than that of the

2:1 complex 1·Rh₂@(R,R-2)₂. This result shows that the bis-

porphyrin 1·Rh₂does not express stoichiometrically controlled

chirality inversion. It is worth noting that the sign of the couplet

associated with the 1:1 complex 1·Rh₂⊃R,R-2 (positive) is op-

posite with respect to that of the couplet (negative) displayed

by the analogous 1·Zn₂⊃R,R-2 complex. The changes observed

in the CD spectra are associated with the initial formation of

the 1:1 complex, followed by its destruction in the presence of

more than 1 equiv of R,R-2, to afford the 1:2 complex.

Similar titrations were performed with the same 1·Rh₂/R,R-2

system in DCM solution. The value of K_1:2was again statis-

tically calculated with the same equation described above, but

taking into account a 5-fold increase in the K_mvalue when

moving from TOL to DCM and α_p= 0.8 (see the model

systems). Fixing K_1:2= 3.7 × 10¹⁷M⁻²and the spectrum of free

4632

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

Table 4. Determined CD Spectral Data (CD Sign, Peak Position/nm, A Total Amplitude/M⁻¹cm⁻¹) of the Complexes (1:1 and

1:2) of Bisporphyrins 1·Zn₂and 1·Rh₂upon Coordination with R,R-2

1:1 complex

1:2 complex

a

b

c

a

b

c

solvent

porphyrin

FC

SC

A [M⁻¹cm⁻¹

]

FC

SC

A [M⁻¹cm⁻¹

]

TOL

DCM

TOL

DCM

1·Zn₂

1·Rh₂

b

−439

−438

441

425

423

757

620

79

441

−427

−426

420

150

−420

−418

441

d

−423

17

d

438

102

a

c

d

FC: first Cotton effect. SC: second Cotton effect. A: total amplitude; A = |Δε₁− Δε₂|. As the CD signal was too small to be significant, a silent

CD was assigned in the fitting.

titration data to a theoretical binding model considering three

colored stoichiometric states of the porphyrin (free, 1:1 and 1:2

complexes). Calculated K_1:1and estimated K_1:2values for the

system R,R-2/1·Rh₂are also listed in the table. Table 4

summarizes the CD spectral parameters of the two complexes

returned from the data fitting of the CD titrations fixing the

stability constant values too those obtained from the UV−vis

titrations.

1·Zn₂. As discussed above, we deduced that the ditopic nature

of the ligand is fundamental to yield CD activity in the 1:2

complexes of 1·Zn₂.

The stability constants reported here for the 1:1 and 1:2

complexes of R,R-2 with 1·Rh₂in TOL and DCM solutions

should be considered as lower limit estimates. As far as we are

aware, there are no reports of the accurate thermodynamic

characterization of iodo-Rh(III)-bisporphyrin complexes with

amines due to their high thermodynamic stability. It is known,

however, that Rh(III)-alkyl porphyrins having a trans organic

ligand, that is, methyl in place of the halogen, show a significant

reduction of binding affinity toward amines (K_avalues on

the order of 10⁵M⁻¹).^46,22The UV−vis absorption spectra of

1·Rh₂compares well with the electronic absorption data

reported for the parent 5·Rh monoporphyrin lacking the

p-pentyl substituents.⁴⁷The spectral signatures of the 1:1 and

1:2 complexes of R,R-2 with 1·Rh₂are in line with those

observed for the analogous complexes with 1·Zn₂. The

coordination of the amine ligand induces a red shift in the

Soret band of 1·Rh₂, similar to what have been reported for the

coordination with P, S, and Se ligands.^48,49This study con-

stitutes the first report of supramolecular chirality induction

using achiral iodo-Rh(III)-bisporphyrin and ditopic enantio-

pure amine. Most likely, the decrease in the calculated A values

for the 1:1 and 1:2 complexes of 1·Rh₂compared with 1·Zn₂is

related to the 3-fold diminution of the exciton coefficient of the

former. The couplet amplitude is proportional to the square of

the extinction coefficient of the coupled chromophores.⁵

However, changes in the geometrical parameters of the

complexes (interchromophoric distance and angle) caused by

the metal substitution cannot be ruled out.

Several examples of related Zn-porphyrin tweezer receptors

binding enantiopure diamine 2 have been described in the

literature.^{19,11,20,43,44}The stability constant values determined

for the 1·Zn₂⊃R,R-2 are in complete agreement with the

magnitudes reported for related sandwich complexes (10⁵−10⁷

M⁻¹). Only in two of the previous reports,^19,20however, the

stability constants of the open 1:2 complexes are indicated and

again we observed a complete coincidence with the values

presented here (10⁹−10¹⁰M⁻²). The spectral changes (CD and

UV−vis) observed in the titration experiments of 1·Zn₂and

R,R-2 show very good correlation with those described for the

interaction of the analogous bifunctional diamine R,R-1,2-

diphenylethylenediamine (DPEA) and the ethane-bridged bis-

Zn-octaethylporphyrin.²⁰DPEA exhibited the unique property

of both chirality induction and inversion. Analogously, diamine

R,R-2 turned out to be an appropriate ditopic ligand for con-

trolling supramolecular chirality induction and inversion in

achiral 1·Zn₂. The process is also occurring via a consecutive

two-step binding mechanism involving the formation of two

complexes with 1:1 and 1:2 (bisporphyrin−amine) stoichiom-

etry.

The A values of 757 and 620 M⁻¹cm⁻¹calculated for the

1·Zn₂⊃R,R-2 in TOL and DCM solutions, respectively, are

extremely high. To the best of our knowledge, these are one of

the largest values reported for the chirality induction process

involving achiral Zn-bisporphyrin receptors and R,R-2. Ema

et al. reported an A value of approximately 500 M⁻¹cm⁻¹in

the interaction of R,R-2 and a chiral Zn-bisporphyrin.⁴²The

remarkable high values of A for the 1·Zn₂⊃R,R-2 complex are

due to optimal geometry factors, that is, intertransition distance

and angle. We have not found reported A values for the open

1:2 complexes resulting from the interaction of Zn-bisporphyr-

ins with ditopic amines, although we can deduce from the

figures shown in the reports that they should be considerable

smaller than those assigned to the sandwich 1:1 complexes. In

this regard, Borovkov et al. described A values for several 1:2

complexes of the bis-Zn-octaethylporphyrin with monotopic

chiral amines on the order of 10²M⁻¹cm⁻¹.⁴⁵This value is in

line with the calculated A for the 1·Zn₂@(R,R-2)₂complex

formed in DCM solution. In striking contrast, the A value for

the same complex in TOL solutions increases to 420 M⁻¹cm⁻¹.

It is worth to note that monotopic chiral amines, such as S-8,

did not induce the formation of a CD-active 1:2 complex with

As commented above, the calculated values for K_1:1↔1:2

=

K_1:2/K_1:1corresponding to the destruction of the 1:1 sandwich

complex 1·Rh₂⊃R,R-2 in TOL and DCM solutions are

appropriate to determine the corresponding EM values.⁴²We

used the relationship K_1:1↔1:2= 2/α_LEM, where α_L= 0.03, we

determined EM values of 3.3 1.6 × 10⁻⁵and 4.8 2.4 × 10⁻⁵

M in TOL and DCM, respectively. Interestingly, the values of

EMs determined for the N···Rh(III)-I intramolecular inter-

action in the 1·Rh₂⊃R,R-2 complex are 3 orders of magnitude

smaller that the corresponding values for the N···Zn(II)

interaction embedded in the analogous scaffolding complex

1·Zn₂⊃R,R-2. This observation suggests the existence of

interplay between the strength of K_mand EM. Since EM is

expected to decrease with restriction of conformational

flexibility and conformational strain, we hypothesize that the

much stronger N···Rh(III)-I intermolecular interaction is

responsible for the observed reduction in EM values. The

complex 1·Rh₂⊃R,R-2 is vibrationally more restricted than

1·Zn₂⊃R,R-2. A more far-reaching implication of our finding

has to do with the fact that EM values cannot be considered as

4633

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

independent properties of supramolecular scaffolds. EM values

seem to be modulated by the strength of the interaction. This is

also the basis to explain the unacceptable fit of the titration

UV−vis data we obtained in the 1·Rh₂/R,R-2 system using

statistically estimated values of K_1:1based on the EM value

determined for the 1·Zn₂/R,R-2 system. The existence of

interplay between K_mand EM has been postulated previously

by Hunter et al. using zinc porphyrin−pyridine complexes that

contain additional intramolecular H-bonding interactions.^50,51

The investigation of the structural and electronic factors that

control the observed changes in the induction of chirality for

the 1·Rh₂system compared to the Zn(II) analogue proved to

be difficult due to associated complexity in DFT calculations of

the former. Assuming that the 1:1 complexes 1·Rh₂⊃R,R-2 and

1·Zn₂⊃R,R-2 can adopt similar structures, the change in sign of

their associated couplet must be related to the change in the

sense of the twist between the “privileged” transition dipole

moments of the porphyrin units. On the basis of the chirality

exciton theory, the sense of the twist is anticlockwise in the

more stable diastereoisomer of 1·Zn₂⊃R,R-2 and should be

clockwise for the lower-energy diastereoisomer of 1·Rh₂⊃R,R-2.

A change in the relative distribution of the two chiral sandwich

1:1 diastereoisomeric complexes 1·M₂⊃R,R-2-(-) and

1·M₂⊃R,R-2-(+) as a function of the metal center, Zn or Rh,

is not unexpected because of the small difference in energy

computed for the two Zn analogues.

sensitive to solvent change. On the contrary, the chiral activity

of the 1:2 complexes is highly solvent-dependent. TOL favors

the formation of 1:2 complexes with higher amplitude values.

Most likely, this is due to a favorable cofacial-like arrangement

(conformational preference) of the porphyrin units in the 1:2

complexes formed in TOL solution. We speculate on the

possible formation of intramolecular hydrogen bonds between

the amino groups of the ligands involved in the 1:2 complex.

These interactions will be fundamental in fixing the relative

spatial orientation and distance of the chromophores attained

in the in-in conformation of the complex. Our hypothesis is

substantiated by the observation of silent CDs for the 1:1 and

1:2 complexes derived from 1·M₂and monotopic chiral amines.

Amines are not very good hydrogen-bond donors, but reason-

ably good hydrogen-bond acceptors through the basic nitrogen

atom. We assume that the coordination of the metal (Rh or Zn)

to the nitrogen atom should increase the Lewis and Bronsted

̈

acidity of the attached hydrogen atoms and their hydrogen-

bond donor capabilities. The hydrogen-bonding interactions

are expected to be maximized in TOL due to its reduced

dielectric constant, dipole moment, and hydrogen-bonding

capabilities compared to those of DCM. In fact, it has speci-

fically been shown that the interaction of DCM with polar

groups (methyl esters) renders the chirality induction

mechanism solvent-dependent.⁵²Interestingly, the values of

EM for the N···Rh(III)-I intramolecular interaction in the

1·Rh₂⊃R,R-2 complex are 3 orders of magnitude smaller than

the corresponding values for the N···Zn(II) interaction

embedded in the analogous scaffold of 1·Zn₂⊃R,R-2. This

observation suggests the existence of interplay between the

strength of K_mand EM. Contrary to current dogmas, our

results indicate that the EM value cannot be considered an

independent property of supramolecular scaffolds. EM values

are modulated by the strength of the intermolecular interaction

(K_m) used in the assembly process. In turn, for the system

under study, K_mvalues are strongly metal-dependent but show

weak solvent effects.

CONCLUSIONS AND GENERAL COMMENTS

■

The CD activity observed for the 1:1 and 1:2 complexes of

bisporphyrins 1·M₂upon binding R,R-2 is due to the existence

of energetically more favored structures (stereodifferentiation)

having an asymmetric twist between the “privileged” transition

dipole moments associated with the porphyrin units and to the

existence of exciton coupling between them. A shorter distance

and an energetically preferred cofacial arrangement of the por-

phyrin units and the interacting dipole moments are common

to the structures of the 1:1 complexes, resulting in associated

CD spectra displaying bisignate Cotton effects (couplets) of

large amplitude. Unexpectedly, some 1:2 (bisporphyrin/

diamine) complexes are also associated with CD spectra

showing a bisignate Cotton effect but having much smaller

amplitudes. This observation suggests that, in the 1:2

complexes, the distance between chromophores is larger and

their relative arrangement is probably less defined (more

conformers of similar energies are possible).

ASSOCIATED CONTENT

■

S

* Supporting Information

Computational details; xyz coordinates for all minimized

structures; experimental procedures for the synthesis of

bisporphyrins 1·M₂and related characterization data; 2D

spectra of the 1·M₂, their sandwich complexes, and R,R-2;

general procedure employed in the spectroscopic titrations;

binding equilibria for the model systems; mathematical analyses

of the obtained titration data; and the X-ray crystallographic file

(CIF) for 5·Rh·MeOH. This material is available free of charge

via the Internet at http://pubs.acs.org.

A sensitive reduction in the amplitude values of the couplets

corresponding to the supramolecular complexes of 1·Rh₂/R,R-2

compared to those of 1·Zn₂/R,R-2 is observed. Because the

couplet amplitude is proportional to the square of extinction

coefficients of the coupled chromophores, the reduction in

extinction coefficients of Rh(III) porphyrin compared to Zn(II)

counterparts is responsible for the observed amplitude

diminution. On the other hand, 1·Rh₂bisporphyrin produces

complexes with R,R-2 that are more kinetically and

thermodynamically stable than 1·Zn₂. The most important

effects of the modification of the metal center in the

chirogenesis process are (a) the change in sign of the bisignate

Cotton effect assigned to the 1:1 complexes and (b) the lack of

chirality inversion for the 1:2 complex of the 1·Rh₂bis-

porphyrin. The change in the solvent used to probe the chiro-

genesis process (TOL or DCM) has subtle effects in the kinetic

and thermodynamic stability of the complexes. Likewise, the

amplitudes of the couplets for the 1:1 complexes are not

AUTHOR INFORMATION

■

Corresponding Author

*E-mail: pballester@iciq.es (P.B.).

Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS

■

We thank Spanish Ministerio de Economía y Competitividad

(CTQ2011-23014 and CTQ2011-29054-C02-02), Generalitat

de Catalunya (2009SGR00462, 2009SGR00686), and ICIQ

Foundation for funding.

4634

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Inorganic Chemistry

Article

(35) Haino, T.; Fujii, T.; Fukazawa, Y. J. Org. Chem. 2006, 71, 2572−

2580.

(36) Hunter, C. A.; Sarson, L. D. Angew. Chem., Int. Ed. Engl. 1994,

REFERENCES

■

(1) Berova, N., Nakanishi, K., Woody, R. W., Eds. Circular Dichroism:

Principles and Applications; John Wiley: New York, 2000.

(2) Huang, X.; Nakanishi, K.; Berova, N. Chirality 2000, 12,

237−255.

(3) Matile, S.; Berova, N.; Nakanishi, K.; Novkova, S.; Philipova, I.;

Blagoev, B. J. Am. Chem. Soc. 1995, 117, 7021−7022.

(4) Harada, N.; Chen, S. L.; Nakanishi, K. J. Am. Chem. Soc. 1975, 97,

5345−5352.

(5) Heyn, M. P. J. Phys. Chem. 1975, 79, 2424−2426.

(6) Matile, S.; Berova, N.; Nakanishi, K.; Fleischhauer, J.; Woody,

R. W. J. Am. Chem. Soc. 1996, 118, 5198−5206.

(7) Ercolani, G. J. Am. Chem. Soc. 2003, 125, 16097−16103.

(8) Ercolani, G.; Schiaffino, L. Angew. Chem., Int. Ed. 2011, 50,

1762−1768.

(9) Hunter, C. A.; Anderson, H. L. Angew. Chem., Int. Ed. 2009, 48,

7488−7499.

(10) Whitty, A. Nat. Chem. Biol. 2008, 4, 435−439.

(11) Borovkov, V. V.; Fujii, I.; Muranaka, A.; Hembury, G. A.;

Tanaka, T.; Ceulemans, A.; Kobayashi, N.; Inoue, Y. Angew. Chem., Int.

Ed. 2004, 43, 5481−5485.

(12) Grimme, S.; Siering, C.; Torang, J.; Kruse, H.; Waldvogel, S. R.

Chem. Commun. (Cambridge, U.K.) 2010, 46, 1625−1627.

(13) Borovkov, V. V.; Bhyrappa, P.; Inoue, Y. Org. Lett. 2007, 9,

433−435.

(14) Beletskaya, I.; Tyurin, V. S.; Tsivadze, A. Y.; Guilard, R.; Stern,

C. Chem. Rev. (Washington, DC, U.S.) 2009, 109, 1659−1713.

(15) Berova, N.; Di Bari, L.; Pescitelli, G. Chem. Soc. Rev. 2007, 36,

914−931.

(16) Hembury, G. A.; Borovkov, V. V.; Inoue, Y. Chem. Rev.

(Washington, DC, U.S.) 2007, 108, 1−73.

(17) Berova, N.; Pescitelli, G.; Petrovic, A. G.; Proni, G. Chem.

Commun. (Cambridge, U.K.) 2009, 5958−5980.

(18) Borovkov, V. V.; Inoue, Y. Top. Curr. Chem. 2006, 265, 89−146.

(19) Etxebarria, J.; Vidal-Ferran, A.; Ballester, P. Chem. Commun.

(Cambridge, U.K.) 2008, 5939−5941.

(20) Borovkov, V. V.; Lintuluoto, J. M.; Inoue, Y. Org. Lett. 2002, 4,

169−171.

(21) Aoyama, Y.; Asakawa, M.; Yamagishi, A.; Toi, H.; Ogoshi, H.

J. Am. Chem. Soc. 1990, 112, 3145−3151.

(22) Kim, H. J.; Redman, J. E.; Nakash, M.; Feeder, N.; Teat, S. J.;

Sanders, J. K. M. Inorg. Chem. 1999, 38, 5178−5183.

(23) Redman, J. E.; Feeder, N.; Teat, S. J.; Sanders, J. K. M. Inorg.

Chem. 2001, 40, 2486−2499.

(24) Fukushima, K.; Funatsu, K.; Ichimura, A.; Sasaki, Y.; Suzuki, M.;

Fujihara, T.; Tsuge, K.; Imamura, T. Inorg. Chem. 2003, 42, 3187−

3193.

33, 2313−2316.

(37) The reported K_mvalue must be considered as a lower-limit

estimate. UV−vis titrations carried out in toluene solution at the

highest possible dilution (5 × 10⁻⁷M) could not provide an accurate

value of the binding constant. The shift of the Soret band reached

saturation after the addition of just 1 equiv of amine. In a 1:1 THF/

toluene mixture, a value of K_m= 6 × 10⁷M⁻¹could be determined

(Supporting Information). However, in neat THF or 1:1 THF/

toluene solutions containing an excess of amine (more than 1 equiv),

we detected the formation of a new species having two axially

coordinated amines. In the more polar solvent, the substitution of the

iodide in the 1:1 complex by a second equivalent of amine occurs and

is facilitated by irradiation with UV light.

(38) Ballester, P.; Costa, A.; Castilla, A. M.; Deya, P. M.; Frontera, A.;

Gomila, R. M.; Hunter, C. A. Chem.Eur. J. 2005, 11, 2196−2206.

(39) Hunter, C. A.; Sanders, J. K. M.; Stone, A. J. Chem. Phys. 1989,

133, 395−404.

(40) Chekmeneva, E.; Hunter, C. A.; Packer, M. J.; Turega, S. M.

J. Am. Chem. Soc. 2008, 130, 17718−17725.

(41) Pescitelli, G.; Gabriel, S.; Wang, Y. K.; Fleischhauer, J.; Woody,

R. W.; Berova, N. J. Am. Chem. Soc. 2003, 125, 7613−7628.

(42) The fit of the titration data considering K_m= 1 × 10¹⁰M⁻¹for

the interaction N···Rh(III)-I afforded calculated EM values for the 1:1

complex identical to the ones obtained using K_m= 1 × 10⁸M⁻¹.

(43) Ema, T.; Ouchi, N.; Doi, T.; Korenaga, T.; Sakai, T. Org. Lett.

2005, 7, 3985−3988.

(44) Borovkov, V. V.; Lintuluoto, J. M.; Sugiura, M.; Inoue, Y.;

Kuroda, R. J. Am. Chem. Soc. 2002, 124, 11282−11283.

(45) Borovkov, V. V.; Lintuluoto, J. M.; Sugeta, H.; Fujiki, M.;

Arakawa, R.; Inoue, Y. J. Am. Chem. Soc. 2002, 124, 2993−3006.

(46) Aoyama, Y.; Yamagishi, A.; Asagawa, M.; Toi, H.; Ogoshi, H.

J. Am. Chem. Soc. 1988, 110, 4076−4077.

(47) Grass, V.; Lexa, D.; Momenteau, M.; Saveant, J. M. J. Am. Chem.

Soc. 1997, 119, 3536−3542.

(48) Redman, J. E.; Feeder, N.; Teat, S. J.; Sanders, J. K. M. Inorg.

Chem. 2001, 40, 3217−3221.

(49) Stulz, E.; Scott, S. M.; Bond, A. D.; Otto, S.; Sanders, J. K. M.

Inorg. Chem. 2003, 42, 3086−3096.

(50) Hunter, C. A.; Misuraca, M. C.; Turega, S. M. Chem. Sci. 2012,

3, 589−601.

(51) Hunter, C. A.; Misuraca, M. C.; Turega, S. M. J. Am. Chem. Soc.

2011, 133, 582−594.

(52) Borovkov, V. V.; Hembury, G. A.; Inoue, Y. Angew. Chem., Int.

Ed. 2003, 42, 5310−5314.

(53) (a) SPECFIT, version 3.0; Spectra Software Associates.

(b) Gampp, H.; Maeder, M.; Meyer, C. J.; Zuberbuhler, A. D. Talanta

̈

(25) Redman, J. E. Ph.D. dissertation, University of Cambridge,

Cambridge, UK, 2000.

1985, 32, 95−101. (c) Gampp, H.; Maeder, M.; Meyer, C. J.;

Zuberbuhler, A. D. Talanta 1986, 33, 943−351.

̈

(26) Tsutsumi, O.; Sato, H.; Takeda, K.; Ogawa, T. Thin Solid Films

2006, 499, 219−223.

(27) Bond, A. D.; Sanders, J. K. M.; Stulz, E. New J. Chem. 2011, 35,

2691−2696.

(28) Aoyama, Y.; Nonaka, S.; Motomura, T.; Ogoshi, H. Chem. Lett.

1989, 1877−1880.

(29) Aoyama, Y.; Yamagishi, A.; Asagawa, M.; Toi, H.; Ogoshi, H.

J. Am. Chem. Soc. 1988, 110, 4076−4077.

(30) Simonato, J. P.; Pecaut, J.; Marchon, J. C. Inorg. Chim. Acta

2001, 315, 240−244.

(31) Mullen, K. M.; Gunter, M. J. J. Org. Chem. 2008, 73, 3336−

3350.

(32) Hunter, C. A.; Gardner, M.; Guerin, A. J.; Michelsen, U.; Rotger,

C. New J. Chem. 1999, 23, 309−316.

(33) Planells, M.; Pelleja, L.; Ballester, P.; Palomares, E. Energy

Environ. Sci. 2011, 4, 528−534.

(34) Forneli, A.; Planells, M.; Sarmentero, M. A.; Martinez-Ferrero,

E.; O’Regan, B. C.; Ballester, P.; Palomares, E. J. Mater. Chem. 2008,

18, 1652−1658.

4635

dx.doi.org/10.1021/ic202515v | Inorg. Chem. 2012, 51, 4620−4635

Article Doi

Influence of the solvent and metal center on supramolecular chirality induction with bisporphyrin tweezer receptors. Strong metal modulation of effective molarity values

DOI: 10.1021/ic202515v

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ic202515v

Full text of DOI:10.1021/ic202515v

Products guided by the article

R&D Labs maybe for 1370442-26-4

Relevant to this article

Hot Product