Study of the Coordination of Quinuclidine to a Chiral Zinc
Phthalocyanine Dimer
Nelson Giménez-Agullóa, Gemma Aragaya, José Ramón Galán-Mascarós*a,b and
Pablo Ballester*a,b
We dedicate this work to Prof. Tomás Torres, Universidad Autónoma de Madrid on the occasion of his 65th
anniversary.
a

Institute of Chemical Research of Catalonia (ICIQ), The Barcelona Institute of Science and Technology, Av. Països

Catalans 16, 43007 Tarragona, Spain; e-mail: pballester@iciq.es b Institution for Research and Advanced Studies
(ICREA), Passeig Lluis Companys, 23, 08010 Barcelona, Spain
Received date (to be automatically inserted after your manuscript is submitted)
Accepted date (to be automatically inserted after your manuscript is accepted)
ABSTRACT: We attempted the calculation of an accurate equilibrium constant for the dimerization
process of enantiomerically pure Zn-1 using UV-vis dilution experiments. At millimolar concentration
Zn-1 is involved in a chemical exchange process between its monomeric and dimeric state that is slow on
the 1H NMR timescale. We performed variable-temperature 1H NMR experiments in CDCl3 solution to
determine the dimerization constant value at different temperatures and performed a van’t Hoff plot to
derive the thermodynamic parameters of the process. The calculated thermodynamic data revealed that
the dimerization process is entropy-driven and enthalpically opposed. We also probed the coordination of
quinuclidine, 1-azabicyclo[2.2.2]octane, 2, to the Zn-phthalocyanine using UV-vis and 1H NMR titrations
in CDCl3 solution. At micromolar concentration the Zn-1 exclusively exists in solution as a monomer and
forms a simple 1:1, 21-Zn, complex with quinuclidine having a stability constant of K(21-Zn) = 106 M1

. On the other hand, the 1H NMR titrations carried out at 298 K and at millimolar concentration showed

that Zn-1 was present in solution as the dimer and formed 2:1, 2(1-Zn)2, and 2:2, 22(1-Zn)2, complexes
by coordination to 2. In addition, the 1:1 complex, 2(1-Zn), showed a reduced dimerization constant
compared to the uncoordinated parent monomer 1-Zn. At high quinuclidine concentration, the 1:1
complex, 21-Zn, derived from the coordinated dimer dissociation was also detected. The 1H NMR
spectra of the titrations displayed separate signals for some hydrogen atoms of the Zn-phthalocyanine in
each one of the four species. Remarkably, the chemical exchange processes involving free and bound
quinuclidine in the monomeric and dimeric complexes showed different kinetics on the NMR timescale.
KEYWORDS: phthalocyanines, dimerization, coordination, quinuclidine, zinc
*Correspondence to: Prof. P. Ballester, Institution for Research and Advanced Studies (ICREA) and Institute of Chemical Research
of Catalonia (ICIQ), The Barcelona Institute of Science and Technology, Av. Països Catalans 16, 43007, Tarragona (Spain), Fax
(+34)977-920-211, e-mail: pballester@iciq.es

INTRODUCTION
Phthalocyanines (Pc) and their metallated derivatives have attracted general attention due to their interesting
physical, electronic and optical properties. Of particular interest to this work is their tendency to aggregate by
stablishing intermolecular π-π stacking interactions. This ability provides Pcs with the capacity to self-assemble into
complex supramolecular systems featuring new properties that are not present in the individual entities. The molecular
aggregation of Pcs usually leads to one-dimensional nanostructures such as, nanoribons [1] or nanofibers [2] with
interest in the field of functional materials. In recent years, the exceptional properties featured by the Pcs and their
aggregates have found application in multiple fields that cover from gas sensors [3], molecular magnets [4],
photovoltaic cells [5], photosensitizers for photodynamic therapy [6, 7], to non-linear optical materials [8].
Up to date, a large number of different substituted Pcs have been synthesized and their aggregation properties
reported [9]. However, the number of reports dealing with chiral Pcs is scarce [10,11]. The unique optical and magnetic
properties of chiral porphyrinoids make them interesting candidates for several applications including chiral recognition
[12], chiral memory [13], and asymmetric catalysis [14]. Different approaches have been used for their synthesis
leading to three diverse groups of optically active Pcs: 1) Pcs substituted with alkyl chains that contain a stereogenic
carbon [15], 2) Pcs substituted with optically active components [16] and 3) Pcs displaying supramolecular asymmetry
[17]. The most widely synthetic approach used consists on the introduction of chiral residues into the macrocycle’s
periphery. Chiral phthalocyanine Zn-1 (Figure 1) was reported in 2012 by Jiang and coworkers [18]. It constitutes a
nice example of building an enantiomerically pure Pc through peripheral decoration with chiral binaphtyl units. It was
described that Zn-1 formed dimeric aggregates held together through multiple cooperative π-π interactions. The
existence of Zn-1 as discrete dimeric entities (i.e. [Zn-1]2) featuring a face-to-face supramolecular structure was
demonstrated in the solid state by X-ray diffraction of single crystals and in solution using 1H NMR spectroscopy
techniques.
Metal coordination bonds have been extensively used to direct the self-assembly of large supramolecular
architectures. In this context, transition-metal porphyrins have been amply exploited as acceptor building blocks since
the central metal atom possesses at least one axial site available for coordination [19, 20]. In turn, the use of
metallophthalocyanines as acceptor ligands is very attractive not only due to their analogous structure with porphyrin
complexes but also due to their higher chemical and thermal stability. Some works regarding the coordination of amino
ligands to several Zn(II) phthalocyanines have been published in literature [21, 22, 23]. However, in these reports the
determination of accurate binding constants for the corresponding complexes was not addressed in detail.
Herein we report the thermodynamic characterization of the aggregation process experienced by the chiral Zn-1 in
CHCl3 solution. We show that the aggregation process of racemic Zn-1 does not lead to exclusive formation of a dimer.
We also describe detailed studies of the coordination processes of enantiopure and racemic Zn-1 with quinuclidine, a
nitrogenated ligand known to axially coordinate Zn-metallated porphyrinoids. These studies were performed using two
very different monomer concentrations, μM and mM, in order to dissect the coordination processes of the monomeric
and dimeric states of Zn-1. When coordinated to quinuclidine the thermodynamic stability of the 22[Zn-1]2 dimer is
reduced compared to that of the non-coordinated counterpart. The 22(R,R)- or (S,S)-[Zn-1]2 homodimers and the meso
(R,S)-heterodimer are isoenergetic.
Insert Figure 1

RESULTS AND DISCUSSION
Synthesis of Zn-1: Chiral phthalocyanines (R)- and (S)-Zn-1 (Figure 1) and the enantiomerically pure (R)- and (S)dinitriles precursors were synthesized following reported procedures [18, 24]. The circular dichroism (CD) spectra of
the prepared dinitriles, not reported in the original paper, showed, as expected, bisignated CD spectra due to the exciton
coupling between the chromophores (naphthyls) present in the molecule.
The enantiopure dinitriles were refluxed in separated hexanol solutions in the presence of Zn(OAc) 2 and DBU to
yield the (R)-Zn-1 and (S)-Zn-1 [25]. Enantiopure Zn-1 was isolated in acceptable yields (39-45 %) after silica column
chromatography purification of the reaction crude using CH2Cl2:THF (100:2) as eluent mixture. As expected, both
enantiomers presented identical 1H NMR and UV-vis spectra that were in agreement with the ones previously reported
by Jiang and co-workers [18].
Thermodynamic characterization of the dimerization process of Zn-1:
The 1H NMR spectrum of (R)-Zn-1 displayed two singlets resonating at 9.30 and 7.83 ppm assigned to the aromatic
protons (Ha, see Figure 1 for proton assignment) of the isoindole units and two sets of six signals appearing in the range
of 8.72 to 6.79 ppm corresponding to the aromatic protons of the peripheral binaphtyl moieties. This observation was in
agreement with the described π-π dimerization experienced by (R)-Zn-1 at 1 mM concentration, that assigns a C4
symmetry of the monomer in the dimeric assembly [(R)-Zn-1]2 [26]. The 1H NMR spectrum recorded for a more diluted
sample of (R)-Zn-1 in CDCl3 solution (5 × 10-4 M) resulted in the emergence of a new set of proton signals that was
attributed to the monomeric species, (R)-Zn-1, featuring D4 symmetry. In the (R)-Zn-1 monomer the protons in the
isoindole units appear as a singlet at 9.43 ppm. Moreover, the protons corresponding to the peripheral binaphthyl
substituents produce a single set of six signals between 8.17-7.36 ppm owing to the symmetry increase for the
monomer. A dimerization constant value of KD = 7.3 × 103 M-1 in CDCl3 was previously reported by Jiang et al. This
constant value was calculated using the integral areas of the proton signals assigned to the monomer and the dimer.
Considering that the accurate assessment of the dimerization constant for enantiopure Zn-1 was crucial for the
calculation of its speciation (monomer/dimer) in solution at any concentration, we decided to investigate the
dimerization equilibria of Zn-1 in chloroform using dilution UV-vis experiments.
The aggregation of zinc-phthalocyanines is known to occur in solution even at the low concentrations used to register
their UV-vis spectra [18]. The UV-vis spectrum of enantiopure Zn-1 in CHCl3 solution at a 10-6 M concentration
displayed the typical absorption bands of a metallated phthalocyanine (i.e. Soret band with maximum at 350 nm, and an
intense Q-band centered at 680 nm together with two weak shoulders with maxima at 616 and 647 nm). The absorption
band corresponding to the binaphtyl moieties was centered at 280 nm. UV-vis dilution studies of enantiopure Zn-1
dissolved in chloroform in a concentration range of 10-5 to 10-6 M returned absorption spectra that did not show
significant changes in the epsilon scale. This result evidenced that in the used range of concentrations enantiopure Zn-1
existed in solution as a unique species, most likely the monomer considering the previously reported KD value.
Conversely, the UV-vis spectra acquired in a range of more concentrated solutions of enantiopure Zn-1 (1 × 10-5 M to
1.24 × 10-4 M) showed a non-linear relationship between concentration and molar extinction coefficient values. Upon
increasing the concentration of enantiopure Zn-1, the UV-vis spectrum (epsilon scale) displayed a small increment in
the intensity of the bands centered at 617 and 647 nm, while the band at 680 nm decreased in intensity (Figure 2). This
observation suggested the formation of aggregates in solution induced by π-π stacking interactions [18]. We performed
the mathematical analysis of the UV-vis dilution data using a simple theoretical dimerization model. The fit of the
experimental data to the model was good and returned a value of KD = 5.17 × 103 M-1 for the dimerization constant
(Figure S1).

It is worthy to note that the magnitude of the calculated KD should allow the observation of monomeric and dimeric
species in the 1H NMR spectra at the working concentration range (2 mM). The observation of a single species (dimer)
in the 1H NMR spectra at this concentration indicated that the dimerization constant value was underestimated.
Probably, the low amount of dimer present in solution in the range of concentrations covered for the UV-vis dilution
experiments (i.e. less than 20%) complicated the determination of an accurate value. However, the high absorptivity
displayed by enantiopure Zn-1 did not allow us to acquire UV-vis spectra at higher concentrations. In view of these
limitations and considering the results obtained by NMR spectroscopy, we simply estimated a value of KD > 104 M-1 for
the dimerization process. Our estimate is close but larger than the value reported by Jiang et al. using 1H NMR
spectroscopy.
The reasons why the aggregation of enantiomerically pure Zn-1 stops at the dimeric state are unknown for us.
However, the analysis of the diffraction data of single crystals obtained from a CHCl 3 solution of enantiopure Zn-1
previously described by Jiang and coworkers also revealed the exclusive presence of dimeric aggregates in the solid
state [18]. Analogous results were obtained in the solid state for the enantiopure free base phthalocyanine H2-1 [24].
Surprisingly, contrary to the sharp and well-defined signals observed in 1H NMR spectra of enantiomerically pure
samples of (R)- and (S)-Zn-1 in CDCl3, an equimolar mixture of (R)- and (S)-Zn-1 in CDCl3 solution, that is racemic
Zn-1, did not produce observable proton signals when analyzed using 1H NMR spectroscopy. This result suggested that
in the racemic mixture the aggregation process does not stop at the dimeric state but continues with the formation of
higher order colloidal aggregates that are not detectable by 1H NMR spectroscopy [27].
Insert Figure 2
According to the magnitude of the estimated dimerization constant, the 1H NMR spectrum of a 2mM solution of
enantiopure Zn-1 displayed a number of proton signals in agreement with the C4 symmetry of [Zn-1]2 in a dimeric
species. We performed a Diffusion-Ordered SpectroscopY (DOSY) experiment [28] using the above solution and
determined a diffusion constant of 3.51 ± 0.06 × 10-10 m2.s-1 for the [Zn-1]2 dimer. Using the Stokes-Einstein equation
we derived a hydrodynamic radius of 11.7 Å for the [Zn-1]2 dimer, which agrees well with the dimensions of the dimer
estimated from simple molecular modelling studies (Figure S2).
The dimerization process of enantiopure Zn-1 was also investigated in solution by means of variable temperature 1H
NMR experiments (Figure 3). As stated above, at 298 K, the 1H NMR spectrum of a 2 mM solution of enantiopure Zn1 in chloroform displayed two singlets at 9.30 and 7.83 ppm corresponding to the chemically non-equivalent Ha protons
in the isoindole units of Zn-1 in the [Zn-1]2 dimer (Figure 3). Lowering the temperature to 263 K resulted in the
appearance of a new signal at 9.43 ppm that was assigned to the chemically-equivalent isoindole Ha protons of the
monomer Zn-1. Further lowering of temperature induced an increase in the signal resonating at 9.43 ppm,
corresponding to the monomer, at the expenses of those attributed to the [Zn-1]2 dimer. At 223 K, the proton signals
corresponding to the monomer were the only ones observable in the 1H NMR spectrum. These results suggested that the
monomer is favored at low temperatures while the dimer is favored on increasing the temperature. This was not an
expected result; typically dimerization is favored at low temperatures. The thermodynamic signature for the
dimerization process was obtained from a van’t Hoff plot (Figure S4). The linear fit of the experimental data was good
and allowed the calculation of the values of the thermodynamic contributions to the dimerization as ΔH = 8.25
kcal.mol-1 and ΔS = 42.1 cal.mol-1.K-1, corresponding to a ΔG = -4.28 kcal.mol-1 at 298 K. The obtained thermodynamic
parameters indicated that the dimerization process is endothermic and driven exclusively by entropy. This result may
appear counterintuitive if one thinks that dimeric species are more ordered and both vibrationally and conformationally

more constrained than the monomer. However, we can rationalize the obtained result by considering the
solvation/desolvation processes that are involved with the dimerization equilibrium. The formation of the dimer [Zn-1]2
is associated with a substantial reduction of the Pc’s surface that is accessible to the solvent. In other words, upon
dimerization a non-defined number of ordered solvent molecules that solvated one of the monomer’s faces must be
released to the bulk solution. The dimerization process definitively produces a gain in entropy when considering the
solvent molecules. This large gain in entropy compensates the enthalpic cost of the dimerization process that is
associated with a lower enthalpy for the π-π interactions of the dimer than for the solvated monomers. When the
temperature is increased the released solvent molecules pay for the energetic cost of the dimerization process which is
enthalpically unfavorable. This is translated in the observed increase of the dimerization equilibrium constant at higher
temperatures. Entropy-driven dimerization processes in organic solvents are uncommon but have been observed for
other systems such as Cram’s velcrands [29] and Rebek’s molecular capsules [30].
Insert Figure 3
Solution studies of quinuclidine binding to Zn-1
The binding of quinuclidine 2 to enantiopure Zn-1 was studied by means of UV-vis and 1H NMR spectroscopy. The
concentration of Zn-1 used to perform the binding studies plays an important role in determining its distribution into
monomeric and dimeric species [31]. The coordination of quinuclidine to enantiopure Zn-1 was first probed by UV-vis
titration experiments using a 4 × 10-6 M chloroform solution of Zn-1. From the dimerization studies, one can derive that
under these diluted conditions Zn-1 exists in solution exclusively as a monomer (Figure S3). The addition of increasing
amounts of quinuclidine to a 4 μM solution of enantiopure Zn-1 resulted in a gradual diminution of the Soret band
centered at 355 nm and the concomitant appearance of a red-shifted band with maximum at 362 nm (Figure 4). This 7
nm red-shift is associated with the axial coordination of the quinuclidine ligand to the Zn center of Zn-1 to form a 1:1
complex, as typically observed for Zn-porphyrins [32]. Moreover, the titration spectra provided a clear isosbestic point
centered at 361 nm. This observation suggested the presence of a single equilibrium involving two colored species in
solution. Based on previously reported coordination studies using Zn-porphyrins, we assigned the two species to free
Zn-1 and the 1:1 complex 2Zn-1 having one molecule of quinuclidine axially coordinated to the zinc metal center.
The analysis of the experimental titration data using a theoretical 1:1 binding isotherm gave a good fit and returned an
association constant value of K(2Zn-1) = 1.69 × 106 M-1 for the 2Zn-1 complex (Figure 4). This value is two orders
of magnitude larger than the one measured for coordination complexes of quinuclidine with zinc-porphyrins [33]. Most
likely, the Lewis acidity of the zinc(II) metal center is increased in the Pc derivatives compared to their porphyrin
analogues provoking the strengthening of the coordination bonds.
Insert Figure 4
When the binding process of quinuclidine with enantiopure Zn-1 was studied using 1H NMR spectroscopy
complexes of larger stoichiometry than the simple 1:1 were detected. At NMR concentration ([Zn-1]  2 × 10-3 M) and
room temperature, enantiopure Zn-1 is exclusively present in solution as the dimer [Zn-1]2 (Figure 5a). The addition of
0.3 equiv of quinuclidine to a 2 mM CDCl3 solution of Zn-1 resulted in the emergence of a new singlet (Ha’) in the
downfield region of the 1H NMR spectrum (Figure 5b). This new singlet emerging at 9.13 ppm was attributed to
isoindole aromatic protons of Zn-1 in a coordination complex of unknown stoichiometry. Concomitantly, three new
signals appeared in the upfield region of the 1H NMR spectrum. These signals resonated at -2.85, -0.7, and 0 ppm and
were assigned to the H1’, H2’ and H3’ protons of bound quinuclidine. The notable upfield shift experienced by the
quinuclidine protons compared to their chemical shifts in the free ligand (Δδ = -5.7, -2.2 and -1.7 ppm for the H1, H2

and H3, respectively), pointed to the formation of an axially coordinated complex between quinuclidine and Zn-1. The
protons of the bound quinuclidine experienced a large upfield shift induced by the ring-current that was more evident
for the α-quinuclidine protons (H1’) due to its close proximity to the Pc π-system.
Insert Figure 5
Taken together, these observations suggested the initial formation of a 2:1 complex 2[Zn-1]2, in which one
quinuclidine molecule is bound to one of two faces of the [Zn-12] dimer. Addition of incremental amounts of 2 (up to
0.5 equiv) provoked a decrease in the intensity of the proton signals assigned to the initially formed 2:1 complex.
Simultaneously, a new broad signal resonating at 9.29 ppm (Ha’’) emerged and a new set of three signals for bound
quinuclidine, resonating at -3.06, -0.83 and -0.08 ppm, and assigned as H1’’, H2’’ and H3’’, appeared (Figure 5c). We
attributed this new set of signals to the formation of a 2:2 complex 22[Zn-1]2 having one quinuclidine axially
coordinated in each one of the two faces of the [Zn-12] dimer. A further increase in the quinuclidine concentration
gradually shifted the equilibrium towards the preferential formation of the 2:2 complex. Nonetheless, in the presence of
0.7 equiv of quinuclidine (Figure 5d) a third set of signals for the bound ligand emerged (-1.6, -0.1 and 0.4 ppm, H1’’’,
H2’’’ and H3’’’, respectively) in the upfield region of the spectrum together with a new broad aromatic signal resonating
at 9.43 ppm (Ha’’’). The downfield shift of this aromatic signal suggested that the newly formed complex should have a
1:1 stoichiometry (2Zn-1). We have observed an analogous downfield shift of the isoindole protons in the Zn-1
monomer. The emergent 2Zn-1 complex will result from the dissociation of the previously formed 2:2 complex,
whose concentration had increased (Figure 6). The simulated speciation profile for the quinuclidine binding to
enantiopure Zn-1 at millimolar concentration, considering dimerization constants for Zn-1 and 2Zn-1 as KD = 10 M
4

-1

and KD = 0.6 × 10 M (vide infra), respectively, and a microscopic binding constant value of K = 1.69 × 10 M for the
3

-1

6

-1

interaction of quinuclidine with enantiopure Zn-1 in any of its stoichiometric states, was in complete agreement with
the results obtained in the above 1H NMR titration (Figure S5).
Insert Figure 6
The addition of more than 1 equiv of quinuclidine to the solution, provoked the selective disappearance of the
quinuclidine signals assigned to the 1:1 complex but not to those of the parent 2:2 complex (Figure 5f). In contrast, the
proton signals corresponding to the isoindole units in the 1:1 and the 2:2 complex did not experience noticeable
changes. These observations together with the lack of well-defined proton signals in the 1H NMR spectra of mixture for
free quinuclidine added in excess, suggested the existence of a chemical exchange process that was intermediate on the
chemical shift timescale between the 1:1, 2Zn-1, complex and free quinuclidine. The differences in the exchange
kinetics observed for free quinuclidine and bound quinuclidine in the 1:1, 2Zn-1, (intermediate) or the 2:2, 22[Zn-1]2,
(slow) complexes suggested that they were involved in exchange mechanisms with different energy barriers (Figure
S6). We hypothesize that in the 2:2 complex the exchange of bound quinuclidine implies the complete cleavage of the
Zn(II)···N coordination bond prior to the exchange with the free ligand (i.e. “dissociative-like” ligand exchange
mechanism). Conversely, in a 1:1 complex the face opposite to the bound ligand is available for the entrance of the
exchange partner. Most likely, the approach of the incoming quinuclidine occurs synchronously to the departure of the
one leaving. Thus, the ligand exchange occurs through an “associative-like” mechanism. The energy barrier for the
“dissociative-like” exchange (2:2 complex) should be higher compared to the “associative-like” one (1:1) because the
former requires the complete cleavage of the coordination bond prior to the exchange. This translates into the slow
exchange process on the 1H NMR timescale observed for the free/bound quinuclidine in the 2:2, 22[Zn-1]2, complex.
Likewise, the lower energy barrier that might be associated with the ligand exchange in the 1:1, 2Zn-1, complex

would explain the intermediate rate on the 1H NMR timescale observed for exchange process of the free/bound
quinuclidine in this complex.
In order to reduce the exchange rate of quinuclidine in the 1:1, 2Zn-1, complex we performed variable temperature
1

H NMR experiments using a 2 mM CDCl3 solution of Zn-1 in the presence of a large excess of quinuclidine (Figure

S7). Unfortunately, we were not able to observe separate proton signals for free and bound quinuclidine in the solution
containing the 1:1, 2Zn-1, complex even at 213 K. The only significant change we observed in the v.t. 1H NMR
spectra was the decrease in intensity of the proton signals corresponding to the 2:2, 22[Zn-1]2, complex and the
concomitant increase of the ones corresponding to the 1:1 complex. This observation indicated that at lower
temperatures the 1:1 ↔ 2:2 equilibrium was shifted towards the formation of the 1:1, 2Zn-1, complex whereas at
higher temperatures the π-aggregation was favored. This result was in complete agreement with the previous findings
made for the v.t. study of the dimerization process of Zn-1.
We observed that at room temperature the 1H NMR spectrum of a CDCl3 solution containing 2 mM enantiopure Zn1 and 3 equiv of 2 showed proton signals for dimeric and monomeric coordinated species (2:2 and 1:1 complexes,
respectively). This observation indicated a reduction in the dimerization affinity for the coordinated monomer compared
to the non-coordinated counterpart. Using the values of the integral areas of the most downfield shifted proton signals
of the 2Zn-1 and [2Zn-1]2 complexes we calculated a dimerization constant value of KD ([2Zn-1]2)= 0.6 × 103 M-1 at
298 K for the enantiopure 22[Zn-1]2 complex. This value is indeed lower than our estimate for the dimerization
constant of free Zn-1 (KD > 104 M-1, vide supra) and Jiang’s reported value (KD = 7.3 × 103 M-1). The strength of π-π
interactions responsible for the dimer formation is reduced upon coordination of the axial ligand. Probably, the
complexation of quinuclidine to the Zn(II) metal center of the Zn-1 Pc distorted its planarity resulting in weaker π-π
interactions between the two coordinated monomers in the dimer [34]. We also determined thermodynamic quantities of
the dimerization process for the enantiopure and axially coordinated 2Zn-1 complex as ΔH = 7.90 kcal.mol-1 and ΔS =
39.41 cal.mol-1.K-1 using a van’t Hoff plot. Not surprisingly, these values are very similar to the ones determined for the
non-coordinated Zn-1 counterpart (Figure S8).
As expected, separated CDCl3 solutions of enantiomerically pure complexes (R) and (S)-[Zn-1] containing an excess
of quinuclidine ligand (3 equiv) displayed identical 1H NMR spectra that showed the diagnostic proton signals for the
chiral monomer and the corresponding chiral dimer. Remarkably, the 1H NMR spectrum of a solution containing
racemic Zn-1 and 3 equiv of 2 displayed an additional set of proton signals. This was especially visible in the upfield
region of the spectrum (Figure 7b) and we assigned the new set of signals to quinuclidine protons in the meso-dimer
22[(R,S)-Zn-1]2. Simple integration of the signals corresponding to the homo (22[(R,R)-Zn-1]2 , 22[(R,R)-Zn-1]2) and
hetero dimeric complexes (22[(R,S)-Zn-1]2) allowed the calculation of an equilibrium constant value for the selfsorting process of the chiral homodimers into the meso-heterodimer as Kss = 4. This value coincides with the statistical
estimate indicating that all dimers are isoenergetic and the self-sorting process of racemic 2Zn-1 complexes into
dimers does not show signs of chiral discrimination (self-recognition or self-discrimination).
Insert Figure 7

EXPERIMENTAL
General

1

H NMR spectra were recorded on a Bruker Avance 400 (400 MHz for 1H NMR spectroscopy), or a Bruker Avance

500 (500 MHz for 1H NMR spectroscopy) ultrashield spectrometer. The deuterated solvents (Aldrich) used are
indicated and chemical shifts are given in ppm. For CDCl3, the peaks were referenced relative to the solvent residual
peak δH = 7.26 ppm. High-resolution mass spectra were obtained using a Bruker Autoflex MALDI-TOF mass
spectrometer. Thin-layer chromatography (TLC) and flash column chromatography were performed with DC-aufolien
Kieselgel 60 F254 (Merck) and Scharlab 60 silica gel, respectively. UV-vis measurements were carried out using a
Shimadzu UV-2410PC spectrophotometer equipped with a photomultiplier detector, double-beam optics, and D2 and
W light sources and using a 1 mm path cell for the dimerization studies.
Materials
All commercial solvents and chemicals were of reagent grade quality and were used without further purification. Dry
DMF was acquired form a Solvent Purification System MBraun SPS-800. Deuterated chloroform and chloroform were
deacidified by passing through a short column of aluminum oxide 90 active, neutral (Merck).
Synthesis
Preparation of (R)- and (S)-phthalocyanato zinc (Zn-1)
A mixture of enantiomerically pure (R) or (S)-benzo[b]dinaphtho[2,1-e:1’,2’-g][1,4]dioxocine-5,6-dicarbonitrile
(200 mg, 0.489 mmol), Zn(OAc) 2·2H2O (22.4 mg, 0.122 mmol) and DBU (0.05 mL) were refluxed at 160 ºC in 1hexanol (2 mL) for 4 h. After reaction time the mixture was cooled at room temperature, the volatiles were removed
under reduced pressure and the crude was purified by silica gel column chromatography using CH 2Cl2:THF (100:2) as
eluent mixture. The solvent of the collected fractions was evaporated in vacuum to yield 52 mg of the (R)-enantiomer as
a green powder (25%). 1H NMR (400 MHz; CDCl3; Me4Si): δH, ppm: 6.84 (1H, d, J = 8.46 Hz), 6.96 (1H, t, J = 7.58
Hz), 7.11 (1H, d, J = 8.46 Hz), 7.20 (1H, t, J = 7.58), 7.32 (1H, t, J = 7.58 Hz), 7.44 (1H, t, J = 7.58 Hz), 7.50 (1H, d, J
= 8.46 Hz), 7.69 (1H, d, J = 8.46 Hz), 7.83 (1H, s), 7.87 (1H, d, J = 8.46 Hz), 7.97 (1H, d, J = 8.46), 8.46 (1H, d, 8.46
Hz), 8.66 (1H, d, 8.46 Hz), 9.30 (1H, s). HRMS (MALDI+): 1704.3527 m/z (calcd. for [M]+ 1704.3513). 1H NMR data
are in complete agreement with those reported in the literature [18].
Titrations and data analysis
1

H NMR and UV-vis titrations were performed by addition of incremental amounts of a solution containing the

ligand to a solution of the Zn-1 in either a 5 mm NMR tube or a 1 cm path quartz cuvette by using microliter syringes.
After each addition the sample mixture was hand shacked and the spectrum recorded. In both types of titration
experiments, Zn-1 was present in the ligand solution at the same concentration than that in the NMR tube or cuvette to
avoid dilution effects. Deacidified chloroform and deacidified deuterated chloroform were used as solvent for the UVvis and 1H NMR titration, respectively. UV-vis spectrophotometric titrations were analyzed by fitting the whole series
of spectra at 1 nm intervals by using the software ReactLab Equilibria 1.1 from Jplus Consulting Pty Ltd (8 Windsor
Road, East Fremantle, WA 6158, Australia).

CONCLUSIONS
The dimerization process of enantiopure Zn-1 was studied using UV-vis dilution experiments and variable
temperature 1H NMR experiments. Unexpectedly, we found that the dimerization process was favored at high
temperatures. The calculated thermodynamic contributions of the dimerization process revealed that it is entropy-driven

and enthalpically opposed. We rationalized these results considering the solvation/desolvation processes that are
involved in the dimerization equilibria.
The coordination of quinuclidine 2 to enantiopure Zn-1 was studied at two different concentrations using UV-vis and
1

H NMR titrations. At micromolar concentration, enantiopure Zn-1 is exclusively present in solution as a monomer and

formed simple 1:1, 2Zn-1, complexes. The calculated association constant for the 2Zn-1 complex is larger than 106
M-1. This value is two orders of magnitude higher than the ones typically reported for the coordination of 2 with Znporphyrins. At millimolar concentration, enantiopure Zn-1 is present in solution mainly as a dimer. The incremental
addition of quinuclidine to mM solutions of enantiopure Zn-1, produced quinuclidine:Zn-1 complexes of larger
stoichiometry than the simple 1:1. Using 1H NMR spectroscopy, it was possible to demonstrate the formation of 1:2
(2[Zn-1]2) and 2:2 (22[Zn-1]2) complexes in addition to the known 1:1 (2Zn-1) complex. We measured a reduction
in the dimerization constant value of the coordinated enantiopure monomer 2Zn-1 compared to the parent
uncoordinated metallo-phthalocyanine Zn-1. Most likely, the axial coordination of quinuclidine to the Zn(II) metal
centers bows the planarity of the Pc weakening the π-π interactions responsible for dimer 22[Zn-1]2 formation. The
addition of an excess of quinuclidine to a millimolar solution of enantiopure Zn-1 induced the selective disappearance
of the quinuclidine signals assigned to the 1:1 complex. The protons of the quinuclidine in the 2:2 complex remained
unaltered. This observation prompted us to propose a different exchange mechanism between free and bound
quinuclidine in the two complexes. The dissimilar kinetics observed on the NMR chemical shift timescale for the two
exchange processes are related to different energy barriers for them (dissociative exchange and associative exchange).
Finally, we demonstrated that the dimerization of racemic Zn-1 (equimolar mixture of (R)- and (S)-Zn-1) produced
large oligomers. Remarkably, when the dimerization study of racemic Zn-1 was performed in a solution containing an
excess of quinuclidine, we observed the assembly of both the chiral homodimers and the meso-heterodimer. The ratio of
the dimers coincided with the expected statistical estimate assuming isoenergetic species, indicating a lack of chiral
discrimination in the self-sorting process of racemic (R,S)-2Zn-1 into dimers

ACKNOWLEDGEMENTS
The authors thank the European Union (ERC Stg grant 279313, CHEMCOMP), Gobierno de España MINECO
(projects CTQ2014-56295-R, CTQ2014-52974-REDC and Severo Ochoa Excellence Accreditation 2014-2018 SEV2013-0319), FEDER funds (project CTQ2014-56295-R) and the ICIQ Foundation for funding.

REFERENCES

Tong WY, Djurisic AB, Xie MH, Ng ACM, Cheung KY, Chan WK, Leung YH, Lin HW and Gwo S. J. Phys. Chem.
B 2006; 110: 17406-17413.
1

2

Duzhko V and Singer KD. J. Phys. Chem. C. 2007; 111: 27-31.

3

Guillaud G, Simon J and Germain JP. Coord. Chem. Rev. 1998; 178: 1433-1484.

4

Ishikawa N, Sugita M, Ishikawa T, Koshihara S and Kaizu Y. J. Am. Chem. Soc. 2003; 125: 8694-8695.

Loi MA, Denk P, Hoppe H, Neugebauer H, Winder C, Meissner D, Brabec C, Sariciftci NS, Gouloumis A, Vazquez P
and Torres T. J. Mater. Chem. 2003; 13: 700-704.
5

Hofman JW, van Zeeland F, Turker S, Talsma H, Lambrechts SAG, Sakharov DV, Hennink WE and van Nostrum CF.
J. Med. Chem. 2007; 50: 1485-1494.
6

7

Kolarova H, Nevrelova P, Bajgar R, Hrova D, Kejlova K and Strnad M. Toxicol. in Vitro 2007; 21: 249-253.

8

de la Torre G, Vazquez P, Agullo-Lopez F and Torres T. J. Mater. Chem. 1998; 8: 1671-1683.

9

Nemykin VN and Lukyanets EA. Arkivoc 2010: 136-208.

10

Kobayashi N. Coord. Chem. Rev. 2001; 219: 99-123.

11

Kimura M, Kuroda T, Ohta K, Hanabusa K, Shirai H and Kobayashi N. Langmuir 2003; 19: 4825-4830.

12

Hembury GA, Borovkov VV and Inoue Y. Chem. Rev. 2008; 108: 1-73.

13

Furusho Y, Kimura T, Mizuno Y and Aida T. J. Am. Chem. Soc. 1997; 119: 5267-5268.

14

Lu HJ and Zhang XP. Chem. Soc. Rev. 2011; 40: 1899-1909.

Vannostrum CF, Bosman AW, Gelinck GH, Picken SJ, Schouten PG, Warman JM, Schouten AJ and Nolte RJM. J.
Chem. Soc., Chem. Commun. 1993: 1120-1122.
15

16

Kobayashi N, Kobayashi Y and Osa T. J. Am. Chem. Soc. 1993; 115: 10994-10995.

17

Kobayashi N, Narita F, Ishii K and Muranaka A. Chem-Eur. J. 2009; 15: 10173-10181.

18

Wang K, Qi DD, Wang HL, Cao W, Li WJ and Jiang JZ. Chem-Eur. J. 2012; 18: 15948-15952.

19

Hunter CA, Meah MN and Sanders JKM. J. Am. Chem. Soc. 1990; 112: 5773-5780.

20

Taylor PN and Anderson HL. J. Am. Chem. Soc. 1999; 121: 11538-11545.

21

Liu CHJ and Lu WC. J. Chin. Inst. Chem. Eng. 2007; 38: 483-488.

22

Przybyl B and Janczak J. Dyes Pigm. 2014; 100: 247-254.

23

D'Souza F and Ito O. Chem. Commun. 2009: 4913-4928.

24

Wang K, Wang HL, Mack J, Li WJ, Kobayashi N and Jiang JZ. Acta Chim. Sinica. 2012; 70: 1791-1797.

25

The absolute configuration of all the BINOL units is the same, all-(R)-Zn-1 or all-(S)-Zn-1.

26

The chemical exchange of the monomers within the dimeric species is slow on the 1H NMR timescale.

Dynamic Light Scattering (DLS) experiments on 0.1 mM CHCl3 solutions of enantiomerically pure (S)-Zn-1 and of
an equimolar mixture of (R)- and (S)-Zn-1 gave an average radii of 1.71 nm and 337.4 nm, respectively. These results
demonstrated that in the racemic mixture of (R)- and (S)-Zn-1 aggregates of bigger dimensions are formed compared to
the enantiomerically pure solutions.
27

28

Johnson CS. Prog. Nucl. Magn. Reson. Spectrosc. 1999; 34: 203-256.

29

Cram DJ, Choi HJ, Bryant JA and Knobler CB. J. Am. Chem. Soc. 1992; 114: 7748-7765.

30

Kang JM and Rebek J. Nature 1996; 382: 239-241.

31

Chi XL, Guerin AJ, Haycock RA, Hunter CA and Sarson LD. J. Chem. Soc., Chem. Commun. 1995: 2567-2569.

32

Hunter CA, Sanders JKM and Stone AJ. Chem. Phys. 1989; 133: 395-404.

Ballester P, Costa A, Castilla AM, Deya PM, Frontera A, Gomila RM and Hunter CA. Chem-Eur. J. 2005; 11: 21962206.
33

34

Andryushkevich SO, Birin KP, Gorbunova YG and Tsivadze AY. Prot. Met. Phys. Chem. Surf., 2011; 47: 494-502.

FIGURE CAPTIONS
Figure 1. Dimerization equilibrium of the chiral the chiral zinc phthalocyanine Zn-1. The structure of Zn-1 monomer is shown as line-drawing
structure and the dimer [Zn-1]2 as MM3 energy minimized structure (protons are omitted for clarity).
Figure 2. UV-Vis spectra of Zn-1 at different concentrations (1 × 10-5 M to 1.24 × 10-4 M) in chloroform recorded during the dilution experiment.
Figure 3. Left: Selected region of the variable temperature 1H NMR spectra of a 2 mM chloroform solution of Zn-1 showing the proton signal
corresponding to the isoindole unit of the monomer (M) and one of the two isoindole protons displayed by the dimer (D). Right: MM3 energy
minimized structures of the monomeric and dimeric species of Zn-1.
Figure 4. MM3 energy minimized structures representing the quinuclidine coordination to the Zn-1 monomer (top) and UV-vis titration spectra
(Soret region) of Zn-1 with increasing amounts of quinuclidine in chloroform at 298K (from 0 to 7 eq) (bottom). The concentration of the zincphthalocyanine was maintained constant throughout the titration (4 × 10-6 M). Inset: Inset: Fit of the experimental data at 355 and 365 nm to the
calculated theoretical 1:1 binding isotherm.
Figure 5. Selected regions of the 1H NMR spectrum recorded during a titration of Zn-1 with increasing amounts of quinuclidine 2 displaying the
progressive formation of the 2:1, 2:2 and 1:1 complexes. ([Zn-1] = 2·10-3 M, CDCl3, 400 MHz). Primed letter corresponded to 1:2 complex 2•[Zn1]2, doubly primed to the 2:2 complex 22•[Zn-1]2, and triply primed to the 1:1 complex 2•[Zn-1].
Figure 6. Binding model used to rationalize the binding of quinuclidine 2 to the [Zn-1]2. MM3 energy minimized structures are shown for the species
involved.
Figure 7. a) Selected regions of the 1H NMR spectrum of (R)-Zn-1 with 3 equiv of quinuclidine and b) Selected regions of the 1H NMR spectrum
of the equimolar mixture of (R) and (S)-Zn-1 with 3 equiv of quinuclidine.

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7