Binding interactions between cucurbit[ n ]uril hosts and tritopic, dicationic guests containing a central ferrocenyl and two terminal aminocyclohexyl sites

Article abstract of DOI:10.1021/om400717z

The binding interactions between two cucurbit[n]uril hosts (n = 7, 8, CB7 and CB8) and the three new guests 1,1′-bis(cyclohexylammoniomethyl) ferrocene (1H₂²⁺), 1,1′- bis(cyclohexylmethylammoniomethyl)ferrocene (2H₂2⁺

Full text of DOI:10.1021/om400717z

Article
pubs.acs.org/Organometallics
Binding Interactions between Cucurbit[n]uril Hosts and Tritopic,
Dicationic Guests Containing a Central Ferrocenyl and Two Terminal
Aminocyclohexyl Sites
Wei Li and Angel E. Kaifer*
Center for Supramolecular Science and Department of Chemistry, University of Miami, Coral Gables, Florida 33124-0431, United
States
S
* Supporting Information
ABSTRACT: The binding interactions between two cucurbit-
[n]uril hosts (n = 7, 8, CB7 and CB8) and the three new
guests 1,1′-bis(cyclohexylammoniomethyl)ferrocene (1H22+),
1,1′-bis(cyclohexylmethylammoniomethyl)ferrocene (2H22+),
and 1,1′-bis(cyclohexyldimethylammoniomethyl)ferrocene
1
(32+) were investigated in aqueous solution using H NMR
spectroscopy, voltammetry, and electrospray (ESI) mass
2+
spectrometry. The experimental data reveal that guests 1H2
2+
and 2H2 behave similarly with both hosts, leading to the
formation of pseudorotaxane complexes in which the host is located centrally around the ferrocenyl residue. A similar complex
forms also between 32+ and CB7, but its formation is slower (k = 2.5 × 10−4 s−1) and can be monitored by NMR spectroscopy.
Finally, 32+ and CB8 give rise to a ternary 2:1 complex, in which CB8 receptors are bound to the terminal cyclohexyl groups,
when the concentration of the host exceeds that of the guest.
them monitored by NMR spectroscopy or electrochemical
techniques.10
INTRODUCTION
■
In aqueous solution the cucurbit[7]uril host (CB7) forms
inclusion complexes of very high stability with suitable guests,
such as ferrocenyl1−3 and adamantyl2,4 derivatives. In particular,
t h e c o m p l e x b e t w e e n C B 7 a n d 1 , 1 ′ - b i s -
(trimethylammoniomethyl)ferrocene exhibits an equilibrium
association constant (K) of ca. 1015 M−1 in aqueous solution,3 a
value which is similar to that measured between avidin and
biotin.5 The fact that this guest is dicationic and the two
positively charged ammonium nitrogens interact with the two
rims of carbonyl oxygens lining the host cavity portals was
initially thought to be an important factor contributing to the
binding. However, thermodynamic studies3 and more recent
work6 have shown that the high stability of the complex largely
derives from hydrophobic forces and from the small entropic
penalty associated with complex formation.
We have investigated in detail highly stable complexes
formed between CB7 and various monocationic and neutral
ferrocenyl derivatives.7,8 In a related study, we investigated the
binding interactions between cucurbit[8]uril (CB8) and
monocationic, ditopic guests containing ferrocenyl and
2,2,6,6-tetramethylpiperidine-1-oxyl (Tempo) binding sites.9
In this case, the degree of methylation of the central
ammonium nitrogen has a pronounced effect on the binding
preference of CB8 between the two sites on the guest. This was
an unexpected result, which is still not well understood. More
recently, we have reported on the complexation between CB7
and a monocationic guest containing adamantyl and ferrocenyl
sites, in which the two possible microscopic complexes can be
readily detected and the kinetics of interconversion between
As another component of our work on highly stable
cucurbituril complexes, we decided to prepare a new series of
tritopic, dicationic guests containing a central ferrocenyl site
and two terminal aminocyclohexyl sites (see Figure 1 for
structures) and investigate their binding interactions with the
CB7 and CB8 hosts. The three guests in this work differ in the
level of methylation of the amine nitrogens separating the
binding sites. Therefore, guest 1 does not bear any methyl
groups on the amine nitrogens, while in guest 2 there is one
methyl group directly attached to each nitrogen atom. Finally,
guest 32+ is fully methylated and, thus, each of the nitrogen
atoms is quaternized and positively charged. To increase the
similarities between all guests, the first two compounds were
2+
used in their protonated, dicationic forms (1H2 and 2H22+).
All guests possess a central ferrocenyl site flanked by two
positive charges. This ferrocenyl binding site is similar to that in
1,1′-bis(trimethylammoniomethyl)ferrocene,3 affording an ex-
cellent thermodynamic well for the CB7 host in the middle of
these guest structures. On the other hand, the terminal
aminocyclohexyl groups can also become effective binding
sites for the hosts.
Special Issue: Ferrocene - Beauty and Function
Received: July 21, 2013
Published: September 23, 2013
© 2013 American Chemical Society
6091
dx.doi.org/10.1021/om400717z | Organometallics 2013, 32, 6091−6097

Organometallics
Article
2+
obtained with guest 1H2 in the presence of various
concentrations of CB7. An important observation from this
series of spectra is that the ferrocenyl protons (labeled 1−3 in
the figure) all shift upfield upon addition of CB7. This fact is a
clear indication that the ferrocenyl group is engulfed inside the
cavity of the host.1,7 In contrast, the cyclohexyl protons (labeled
4−8) shift downfield in the presence of CB7, in good
agreement with the expected behavior of protons that stay
outside the host cavity but remain in the vicinity of the cavity
portals. Therefore, the NMR spectroscopic data are consistent
with the formation of an inclusion complex in which the host
occupies the central ferrocenyl binding site, while the
cyclohexyl side arms stay outside the cavity. Furthermore, the
NMR spectroscopic data suggest that the formation of this
complex is quantitative at the millimolar concentrations used in
the NMR experiments, since the changes brought about by the
presence of CB7 level off at 1.0 equiv of host, with excess
beyond this point leading to no further spectral modifications
(Figure 2d,e). Of course, this is consistent with the substantial
stability anticipated for these complexes.
The anodic voltammetric behavior of the guests is high-
lighted by the reversible, one-electron oxidation of the
ferrocenyl unit.12 Indeed, Figure 3 shows the reversible
Figure 1. Structures of the hosts and guests used in this work.
RESULTS AND DISCUSSION
■
Synthesis. The preparation of the hosts CB7 and CB8 was
done according to reported procedures.11 Guest 1 was prepared
by reaction of 1,1′-ferrocenedicarboxaldehyde with cyclohexyl-
amine in methanolic solution, followed by reduction with
NaBH4. Guest 2 was similarly synthesized from the same
ferrocene starting material and N-methylcyclohexylamine, but
reduction was carried out with NaBH3CN. Finally, dicationic
32+ was prepared by methylation of 2 with dimethyl sulfate in
1
dichloromethane solution. All guests were characterized by H
and 13C NMR spectroscopy, as well as high-resolution
electrospray (ESI) mass spectrometry (see the Experimental
Section for details). Compounds 1 and 2 were used in mildly
acidic solutions (pH ∼4.5) to ensure that they would be
present as their diprotonated, dicationic forms.
Figure 3. Cyclic voltammetric behavior on glassy carbon (0.071 cm2)
of 1.0 mM 1H22+ in 0.1 M NaCl pH 4.5 in the absence (black) and in
the presence of 0.5 equiv (blue) and 1.0 equiv (red) of CB7. Scan rate:
0.1 V s−1.
Binding with CB7. The binding interactions between the
guests 1H22+, 2H22+, and 32+ and the CB7 host were
1
2+
investigated using H NMR spectroscopy, cyclic voltammetry,
oxidation process for guest 1H2 centered at a half-wave
and ESI mass spectrometry. Figure 2 shows the NMR data
potential (E1/2) of 0.56 V. Addition of 1.0 equiv of CB7 shifts
the observed E1/2 value to 0.75 V, which is entirely consistent
with the inclusion of the ferrocenyl group inside the host cavity.
We have observed, upon encapsulation with CB7, pronounced
anodic E1/2 shifts with a variety of cationic ferrocene
derivatives.1,7,8 These anodic potential shifts reflect the
thermodynamic hindrance to oxidation associated with the
poorer solvation of the oxidized, positively charged ferrocenium
form inside the relatively hydrophobic cavity of CB7. In the
presence of 0.5 equiv of CB7, the voltammetric behavior shows
two sets of waves, corresponding to the free and bound guests.
This “two-wave” behavior has been described in detail by our
group and is fully expected when the host−guest complex
exhibits considerable thermodynamic stability.13
1
Clearly, both H NMR spectroscopic and voltammetric data
led us to conclude that CB7 can easily reach the central binding
site in the 1H22+ guest, leading to the formation of a symmetric
complex, CB7·1H22+, in which the ferrocenyl group is
encapsulated and ion−dipole interactions develop between
each of the positively charged ammonium groups and the
carbonyl oxygens lining the portals of CB7. The resulting CB7·
1
Figure 2. Partial H NMR spectra (500 MHz, 0.1 M NaCl−D2O, pD
2+
4.5) of 1.0 mM 1H2 (a) in the absence and in the presence of (b)
0.25, (c) 0.5, (d) 1.0, and (e) 1.25 equiv of CB7.
6092
dx.doi.org/10.1021/om400717z | Organometallics 2013, 32, 6091−6097

Organometallics
Article
2+
1
1H2 complex has a pseudorotaxane structure, in which the
(Figure S5, Supporting Information). We used the H NMR
spectroscopic data to investigate the kinetics of the conversion
between the two microscopic forms of the CB7·32+ complex
(Figure 5). The process was found to be first order, and the rate
constant was determined to be 2.5 × 10−4 s−1 (t1/2 = 46 min) at
25 °C.
CB7 “wheel” is centered between the two positive charges of
the “thread” compound, the 1H2 guest, with the cyclohexyl
end groups lacking the necessary bulk to keep the host trapped
and prevent dissociation (see Figure 2, top). The formation and
stoichiometry of this complex was also confirmed by ESI mass
spectrometric data (see the Supporting Information, Figure
S1).
2+
2+
The binding interactions between the guest 2H2 and CB7
follow exactly the same pattern, as evidenced by the NMR
spectroscopic, voltammetric, and mass spectrometric data
(Figures S2 and S3, Supporting Information) being similar to
those obtained for guest 1H22+. Therefore, we conclude that
replacement of a proton by a methyl group on each of the
ammonium nitrogens of the guest has no significant effect on
the formation of pseudorotaxane complexes and, thus, the
2+
2+
structures of the CB7·1H2 and CB7·2H2 complexes are
basically the same.
Figure 5. Concentration (as its natural log) of the external CB7·32+
complex as a function of time in 0.1 M NaCl at 25 °C.
Further methylation of the guest does change the observed
interactions with the CB7 host. For instance, the voltammetric
behavior of guest 32+ in the presence of 1.0 equiv of CB7
initially shows a pronounced cathodic shift in the observed E1/2
value for the reversible oxidation of the central ferrocenyl
residue, which shifts from 0.67 V in the free guest to 0.47 V in
the CB7 complex (Figure 4). The negative shift in the half-wave
The 1:1 stoichiometry of the CB7·32+ complex was also
verified by ESI mass spectrometric data (Figure S4, Supporting
Information). Ultimately, all three guests form the same
pseudorotaxane, symmetric CB7 inclusion complex, with the
host engulfing the central ferrocenyl residue. Remarkably, the
level of methylation of the ammonium nitrogens affects the
2+
kinetics of formation of the complex. With guests 1H2 and
2H22+ the kinetics of formation of the CB7 complex is too fast
to be monitored by conventional NMR spectroscopic
techniques and the system reaches equilibrium before any
spectra can be recorded. However, with the fully methylated
guest 32+, we can monitor the kinetics of formation of the
pseudorotaxane using either NMR spectroscopy or voltam-
metric measurements (data not shown). We must conclude
that full methylation of the two ammonium nitrogens on the
guest slows down the threading process by which the CB7 host
reaches the central binding site.
Several reasons can be proposed to explain the effects of
ammonium methylation on the threading rate of CB7 along the
32+ guest. The first one is based on steric reasons, as two methyl
groups attached to each ammonium nitrogen would create an
obstacle when the host slides over, trying to reach the
thermodynamic valley offered by the ferrocenyl binding site.
The presence of protons attached to the ammonium nitrogen
may facilitate its passage through the hydrophobic CB7 cavity
through a deprotonation−protonation mechanism.14 Alterna-
tively, one could also argue that methylation may increase the
binding affinity of the host for the terminal cyclohexyl residues,
resulting in a higher energetic barrier that must be overcome
for the host to reach the central binding site. As a quick way to
assess the second possibility, we explored the relative stability of
the external complexes by adding excess (over 1.0 equiv) CB7.
If the external complexes are sufficiently stable, in the presence
of 2.0 equiv of host, we may observe the formation of 2:1
complexes in preference to the already observed 1:1
pseudorotaxane. It is unlikely that a 2:1 complex will form
with a CB7 host on the central ferrocenyl site, as binding of two
CB7 hosts in close proximity has not been observed because of
the resulting electrostatic repulsions between the rims of
carbonyl oxygens lining the cavity entrances. For the same
reasons, formation of a 3:1 complex has also not been observed
with these guests. Therefore, the attempt to form 2:1
Figure 4. Cyclic voltammetric behavior on glassy carbon (0.071 cm2)
of 1.0 mM 32+ in 0.1 M NaCl pH 4.5 in the absence (black) and in the
presence of 1.0 equiv of CB7 immediately after addition (blue) and at
equilibrium (red) after ca. 4 h. Scan rate: 0.1 V s−1.
potential strongly suggests that the ferrocenyl group is outside
the CB7 cavity but in the vicinity of one of the host cavity
portals. This finding is consistent with the inclusion of one of
the terminal cyclohexyl groups inside the host cavity. However,
this is not the most thermodynamically stable complex, because
the voltammetric behavior of this solution changes as a function
of time and, after ∼4 h, we observe a single set of waves
centered at 0.81 V, which corresponds to the ferrocenyl-
included, symmetric pseudorotaxane.
Similar data were obtained in 1H NMR spectroscopic
experiments, indicating the initial formation of a complex in
which one of the terminal cyclohexyl groups is included by CB7
followed by the gradual development of a set of peaks
corresponding to the included ferrocenyl protons. In other
words, the spectra clearly reveal the time evolution of the
supramolecular system from an external complex (included
cyclohexyl group) to an internal complex (pseudorotaxane), in
which CB7 occupies the central binding site in the 32+ guest
6093
dx.doi.org/10.1021/om400717z | Organometallics 2013, 32, 6091−6097

Organometallics
Article
complexes relies on the possibility that each of the terminal
cyclohexyl groups will be included by a host, while the central
ferrocenyl site remains unbound. Our attempts to observe any
such 2:1 species were not successful. Once we observed the
formation of the 1:1 pseudorotaxane complex, we failed to
observe any convincing evidence for the formation of 2:1 CB7
complexes. This matter will be revisited later in the paper.
Binding with CB8. Since the rate of formation of the CB7
complexes appears to be affected by the size of the groups
attached to the ammonium nitrogens on the guests, we decided
to investigate the binding interactions with the next higher
analogue, the cucurbit[8]uril (CB8) host. Our experimental
approach relied on the same techniques. However, the lower
aqueous solubility of CB8 forced us to carry out experiments
with lower concentrations of guests in solution.
The 1H NMR spectroscopic data for mixtures of guest 1H2
2+
and CB8 are shown in Figure 6. Clearly, the addition of CB8
1
Figure 7. Partial H NMR spectra (500 MHz, 0.1 M NaCl-D2O pD
4.5) of 0.1 mM 32+ (a) in the absence and in the presence of (b) 0.44,
(c) 0.88, (d) 1.0, (e) 1.8, (f) 2.6, and (g) 3.5 equiv of CB8.
of CB8 results in a clear upfield shift of the cyclohexyl protons,
a finding which is consistent with the formation of a 2:1
complex in which each of the terminal cyclohexyl residues is
bound by a CB8 host, leading to the formation of a dumbbell-
type complex.
The differences in the structures and stoichiometries of the
2+
CB8 complexes formed with guests 1H2 and 2H22+, on one
side, and guest 32+, on the other side, are again triggered by the
full methylation of the ammonium nitrogens on the guest. To
confirm the conclusions derived from NMR spectroscopic data,
we carried out square wave voltammetric (SWV) experiments.
SWV was used instead of cyclic voltammetry because of its
higher sensitivity, which is required in these experiments
because of the low solubility of CB8. The reversible one-
electron oxidation of free guest 1H22+ is observed at a half-wave
potential of 0.56 V vs Ag/AgCl, while upon addition of 1.0
equiv of CB8 the E1/2 value shifts to 0.71 V. Addition of a
second equivalent of CB8 does not result in any further
potential changes. As mentioned before, the anodic shift of the
E1/2 value correlates well with the inclusion of the ferrocenyl
group inside the cavity of the host. Therefore, the NMR
spectroscopic and voltammetric data are fully consistent with
1
Figure 6. Partial H NMR spectra (500 MHz, 0.1 M NaCl−D2O pD
2+
4.5) of 0.1 mM 1H2 (a) in the absence and in the presence of (b)
0.5, (c) 1.0, (d) 1.5, and (e) 2.0 equiv of CB8.
leads to an upfield shift for the protons (labeled 1−3) on the
ferrocenyl residue. As in the case of CB7, this finding indicates
the formation of a symmetric complex with the host engulfing
the central ferrocenyl residue. Notice that addition of CB8
beyond 1.0 equiv does not lead to any changes in the structure
of the complex. In other words, the formation of a 2:1 complex
is not favored in this case and the CB8·1H2 pseudorotaxane-
type complex prevails, even under conditions of excess CB8 in
the solution. Entirely similar NMR spectroscopic data were
2+
2+
the formation of a 1:1 pseudorotaxane CB8·1H2 complex.
2+
The SWV data for the guest 2H2 are also very similar, with
E1/2 values of 0.62 and 0.72 V for the free and CB8-bound
complex, respectively.
2+
obtained for mixtures of guest 2H2 and CB8 (Figure S6,
Supporting Information), revealing that the corresponding
As expected from the NMR spectroscopic results, the
voltammetric data obtained with guest 32+ are different. The
free guest undergoes reversible oxidation at a half-wave
potential of 0.68 V; upon addition of 2.0 equiv of CB8, the
E1/2 value shifts cathodically to 0.32 V! This considerable
potential shift, equivalent to −0.36 V, is consistent with the
formation of a dumbbell complex in which the ferrocenyl group
is not included but suffers the electrostatic effects exerted by
two CB8 hosts, which engulf the cyclohexyl residues at the ends
of the guest. The formation of this 2:1 complex is, of course,
consistent with the NMR spectroscopic data.
2+
CB8·2H2 complex has a similar structure.
In contrast to these findings, the binding interactions
between 32+ and CB8 were different, as evidenced by the
1
corresponding H NMR spectroscopic data (Figure 7). In this
case the addition of up to 1.0 equiv of CB8 leads to the
formation of the pseudorotaxane complex, as revealed by the
upfield shift of the ferrocenyl protons. However, addition of
excess CB8 (more than 1.0 equiv) leads to the return of the
ferrocenyl proton signals to chemical shifts close to those
observed with the free guest. Furthermore, addition of 2.0 equiv
6094
dx.doi.org/10.1021/om400717z | Organometallics 2013, 32, 6091−6097

Organometallics
Article
ESI mass spectrometry was also used to investigate the CB8
the final symmetric complex, probably because the host sliding
process over the ammonium group is associated with a higher
energy barrier. The unavailability of a deprotonation−
protonation mechanism14 to assist the sliding of the CB7
over the central binding site may also be an important factor.
However, an steric component must also play a role, because
the kinetics of formation of the analogous pseudorotaxane
complex between 32+ and CB8 is fast on the time scale of NMR
spectroscopic and voltammetric experiments.
The binding events observed between CB8 and the tritopic
guest 32+ suggest a better balance between the relative
stabilization of the larger host at the two types of binding
sites (terminal cyclohexyl or central ferrocenyl). In fact, the
data in Table 1 show that the K value for the CB8 complexation
of 4+ is only slightly lower than that for the CB8·62+ complex.
Therefore, the binding interactions between 32+ and CB8
initially lead to the formation of the symmetric complex, but
the terminal sites compete effectively when [CB8] > [32+] and
the 2:1 complex predominates under these conditions. This
behavior is not observed with guests 1H22+ and 2H22+, because
the protonated cyclohexyl ammonium site has a lower binding
affinity for CB8, as reflected by the K value in Table 1 between
CB8 and 5H+. This finding is also supported by a previous
report from our group9 on the larger binding affinity with CB8
of methylated ammonium Tempo sites compared to similar
sites with lower levels of methylation.
complexes. While we could confirm again the formation and
2+
2+
stoichiometry of the CB8·1H2 and CB8·2H2 complexes
(Figures S7 and S8, Supporting Information), we could not
observe any signals corresponding to the (CB8)2·32+ 2:1
complex. However, we clearly detected the CB8·32+ 1:1
complex (Figure S9, Supporting Information). The failed
mass spectrometric detection of the 2:1 CB8 complex of guest
32+ cannot dispel the strong combination of evidence obtained
in NMR spectroscopic and electrochemical experiments, which
consistently support the formation of this ternary complex.
Binding of Model Compounds. Each of the guest
compounds surveyed here is composed of three binding sites:
two terminal cyclohexyl sites and a central ferrocenyl site. We
decided to investigate a small series of model compounds
resembling the individual binding sites in order to gain a better
understanding of the key factors controlling the binding
differences between the fully methylated 32+ and the other
two guests. Therefore, we focused our effort on the three model
compounds shown in Table 1, which were readily accessible to
Table 1. Equilibrium Association Constants (K) and Gibbs
Free Energies (ΔG°) for the Binding Interactions between
Model Guest Compounds 4+, 5H+, and 62+ and the hosts
CB7 and CB8 in 50 mM NaAc pH 4 solution at 25 °C
In conclusion, we have shown that the most common
binding behavior between the tritopic, dicationic guests
surveyed here and the hosts CB7 and CB8 is the rapid
formation of a symmetric, pseudorotaxane 1:1 complex (see
Scheme 1). The kinetic rate of formation of this complex is
Scheme 1. Possible Equilibria between the Tritopic Guests
and the CB7 and CB8 Hosts
us. The cyclohexyl derivatives 4+ and 5H+ are intended to
represent the terminal binding sites on the tritopic guests at
either of the two extremes of full methylation (4+) and full
protonation (5H+) of the ammonium group. The model guest
62+ represents the central ferrocenyl site on the tritopic guests.
While it is tempting to try to estimate the ΔG° values of the
various complexes between CB7 or CB8 and the tritopic guests
on the basis of the thermodynamic parameters shown in Table
1, this exercise is likely to be marred by large errors. However,
there are some trends in the data that are very useful to explain
our experimental results with the tritopic guests. One important
factor is that binding between 62+ and either host leads to more
stable complexes than binding between either host and the
other two cyclohexyl derivatives. Therefore, the general
tendency to form symmetric, pseudorotaxane-type complexes
with the ferrocenyl site engulfed by the host is supported by the
data in Table 1. The slower formation of the pseudorotaxane
complex, in the case of 32+ and CB7, can be rationalized by the
more stable complex formed by CB7 with 4+ in comparison to
5H+. In other words, methylation of the ammonium nitrogen
increases the stability of the complexes in which CB7 interacts
with the cyclohexyl terminus of the tritopic guest. The larger
relative stabilization of the host slows down the formation of
considerably slow between CB7 and 32+, probably because of
the substantial stabilization of the 1:1 external complex. Also,
the interactions between the same guest and CB8 are
particularly interesting, as the predominant complex formed
depends on the relative concentrations of both partners. When
the host concentration is 1.0 equiv or less, the predominant
supramolecular species is the symmetric pseudorotaxane.
6095
dx.doi.org/10.1021/om400717z | Organometallics 2013, 32, 6091−6097

Organometallics
Article
1
characterized as the sulfate salt (36 mg, yield 43%). H NMR (D2O,
500 MHz): δ 1.10−1.12 (m, 2H, CH2), 1.23−1.26 (m, 4H, CH2),
1.46−1.48 (m, 4H, CH2), 1.60−1.62 (d, 2H, CH2), 1.88−1.91 (d, 4H,
CH2), 2.12−2.14 (d, 4H, CH2), 2.81 (s, 12H, CH3), 3.15 (t, 2H, CH),
4.37 (s, 4H, CH2), 4.46 (s, 4H, C5H4), 4.55 (s, 4H, C5H4) ppm. 13C
NMR (D2O, 500 MHz): δ 24.22 (2C, CH2), 24.78 (C, CH2), 25.62
(4C, CH2), 46.60 (2C, CH), 62.24 (2C, CH2), 71.18 (4C, CH3),
71.73 (4C, C5H4), 73.06 (4C, C5H4), 73.64 (2C, C5H4) ppm. ESI-MS:
found m/z 233.1503 [M]2+, C28H46FeN22+, calcd 233.1533.
However, as the concentration of host exceeds 1.0 equiv, a 2:1
complex, in which the two terminal cyclohexyl groups are
bound by CB8 (Scheme 1), starts to appear and competes
effectively with the pseudorotaxane. Once the CB8 concen-
tration is 2.0 equiv or larger, this ternary “dumbbell” complex
becomes the predominant supramolecular species. As shown in
Scheme 1, we did not observe either the ternary complex, in
which the two hosts occupy adjacent binding sites, or the
quaternary complex, in which all three binding sites are
occupied by host molecules. Presumably, these supramolecular
species are destabilized by the electrostatic repulsions between
nearby carbonyl oxygens on the host cavity portals.
Synthesis of Cyclohexyltrimethylammonium (4+). N,N-Dimethyl-
cyclohexylamine (500 mg, 589 μL, 3.98 mmol) was dissolved in 50 mL
of CH2Cl2, followed by addition of MeI (837 mg, 367 μL, 5.89 mmol).
The mixture was stirred under N2 for 24 h at room temperature. A
white precipitate was formed and washed with acetone. The white
product was collected as the iodide salt (856 mg, yield: 80%). ESI-MS:
found 142.1593 [M]+, C9H20N+, calcd 142.1596.
EXPERIMENTAL SECTION
■
Materials. Cyclohexylamine, N,N-dimethylcyclohexylamine, 1,1′-
ferrocenecarboxaldehyde, N-methylcyclohexylamine, MeI, NaBH4,
NaBH3CN, and all other chemicals and solvents were commercially
available. All amines were passed through basic alumina before use.
The purity of the CB7 and CB8 hosts, prepared as described in the
literature,11 was assayed by the method previously reported by our
group.15
Synthesis of 1,1′-Bis(trimethylammoniomethyl)ferrocene (62+).
1,1′-Ferrocenedicarboxaldehyde (200 mg, 0.83 mmol) was dissolved in
an ethanolic solution of methylamine (244 mg, 7.86 mmol in 30 mL)
under a N2 atmosphere. The red solution was refluxed for 1 h at 45 °C.
After the mixture was cooled to 0 °C, NaBH4 (125 mg, 0.3.30 mmol)
was added in one portion. The mixture was stirred for 12 h at 45 °C.
The ethanol was evaporated, and the residue was extracted between
water and ethyl ether (10 mL/10 mL). The aqueous solution was
further extracted by ethyl ether (3 × 10 mL). All the organic solutions
were collected and dried over Na2SO4, and then the solvent was
removed. The red oily residue was separated on a neutral Al2O3
column with ethyl ether/MeOH (1/5) as the eluent. A major yellow
band was collected and characterized as the diamine (158 mg, 0.58
Synthesis of 1,1′-Bis(cyclohexylaminomethyl)ferrocene (1). Cy-
clohexylamine (370 μL, 320 mg, 3.28 mmol) was dissolved in 20 mL
of methanol. Aqueous HCl solution was used to adjust the solution pH
value to around 5, followed by addition of 1,1′-ferrocenedicarbox-
aldehyde (132 mg, 0.54 mmol). Molecular sieves were used to absorb
water in the reaction. The orange mixture was refluxed at 45 °C for 4 h
under N2. NaBH4 (123 mg, 3.28 mmol) was then added to the
solution stirred for 24 h at 45 °C and then cooled. After filtration
through Celite, methanol was removed. The resulting orange oil was
washed and extracted by ethyl ether/water, and the organic layers were
collected for product purification. The mixture was separated through
a neutral Al2O3 column with ethyl ether/methanol (1/9). The major
product 1 was collected and dried, resulting in a yellow solid (132 mg,
yield 60%). 1H NMR (D2O, 500 MHz): δ 1.13 (t, 2H, CH2), 1.25 (m,
6H, CH2), 1.60−1.63 (d, 2H, CH2), 1.77 (s, 2H, CH2), 2.01 (s, 2H,
CH2), 2.99 (s, H, CH), 4.01 (s, 2H, CH2), 4.33 (s, 4H, C5H4), 4.40 (s,
4H, C5H4) ppm. 13C NMR (D2O, 500 MHz): δ 23.87 (1C, CH2),
24.50 (2C, CH2), 29.05 (2C, CH2), 43.58 (C, CH), 56.06 (2C, CH2),
70.55 (4C, C5H4), 70.70 (4C, C5H4), 77.7 (2C, C5H4) ppm. ESI-MS:
found m/z 205.1174 [M + 2H]2+, C24H38FeN22+, calcd 205.1166.
Synthesis of 1,1′-Bis(cyclohexylmethylaminomethyl)ferrocene
(2). N-Methylcyclohexylamine (390 μL, 335 mg, 2.94 mmol) was
dissolved in 20 mL of methanol. An aqueous HCl solution was used to
adjust the solution pH value to around 5, followed by addition of 1,1′-
ferrocenedicarboxaldehyde (119 mg, 0.49 mmol). Molecular sieves
were used to absorb water in the reaction. The orange mixture was
refluxed at 45 °C for 4 h under N2. The mild reducing agent
NaBH3CN (186 mg, 2.96 mmol) was then added to the solution,
which was stirred for 24 h at 45 °C and then cooled. After filtration
through Celite, methanol was removed. The orange oil was washed
and extracted by CH2Cl2/water, and the organic layers were collected
for product purification. The mixture was separated on a neutral Al2O3
column with CH2Cl2/methanol (1/10). The product 2 was collected
1
mmol, yield 70%). The diamine structure was confirmed by H NMR
spectroscopy and ready for the next step.
The diamine (158 mg, 0.58 mmol) was redissolved in CH2Cl2 (30
mL) under N2 at room temperature. MeI (361 μL, 824 mg, 5.8 mmol)
was added to the yellow solution. The reaction mixture was covered by
aluminum foil and stirred in the dark for 24 h. A yellow precipitate was
formed. After filtration and washing with CH2Cl2, the iodide salt of 62+
1
was collected as a yellow solid (67 mg, yield: 20%). H NMR (D2O,
500 MHz): δ 2.94 (s, 18H, CH3), 4.36 (s, 4H, CH2), 4.47 (s, 4H,
C5H4), 4.56 (s, 4H, C5H4) ppm. ESI-MS: found m/z 165.0881 [M]2+,
C18H30FeN22+, calcd 165.0874.
Methods. Cyclic voltammetric and square wave voltammetric
experiments were carried out with a BAS 100 W electrochemical
workstation. A single-compartment cell fitted with a glassy-carbon
working electrode, Pt counter electrode, and Ag/AgCl reference
electrode was used. The surface of the working electrode was polished
before each measurement on a felt surface with a slurry of 5 μm
alumina powder and water.
1H NMR spectra were recorded with a 500 MHz Bruker Avance
spectrometer equipped with a cryoprobe. DCl and D2O were
purchased from Cambridge Isotopes. The pH measurements were
done using a PHR-146 microcombination pH electrode on an
Accumet model 50 pH/ion/conductivity meter, calibrated using
standard buffers (pH 4, 7, and 10).
1
as a yellow solid (107 mg, yield 50%). H NMR (D2O, 500 MHz): δ
The equilibrium association constants between the hosts (CB7 and
CB8) and the model guests (4+, 5H+, and 62+) were determined in
competition binding experiments using cobaltocenium as the reference
guest. The competition between the two guests under conditions in
which the host concentration is not enough to bind both guests
completely can be monitored by following the cobaltocenium
absorbance at 261 nm. These experiments were all done in 50 mM
sodium acetate (pH 4.5) solution, a medium in which the equilibrium
association constants of cobaltocenium with CB7 and CB8 have been
reported by us15 as 5.7 × 109 and 1.9 × 108 M−1, respectively. Details
of similar competition binding experiments have been reported in
previous publications from our group.16,17
1.10−1.12 (m, 2H, CH2), 1.24−1.27 (m, 4H, CH2), 1.41−1.48 (m,
4H, CH2), 1.61−1.63 (d, 2H, CH2), 1.87 (m, 6H, CH2), 1.97−1.98 (d,
2H, CH2), 3.17 (t, 2H, CH), 4.08−4.25 (m, 4H, CH2), 4.40 (s, 4H,
C5H4), 4.47 (s, 4H, C5H4) ppm. 13C NMR (D2O, 500 MHz): δ 24.31
(10C, CH2), 34.79 (2C, CH), 52.79 (2C, CH2), 63.27 (2C, CH3),
71.21 (4C, C5H4), 71.74 (4C, C5H4), 75.54 (2C, C5H4) ppm. ESI-MS:
found m/z 2219.1377 [M + 2H]2+, C26H42FeN22+, calcd 219.1343.
Synthesis of 1,1′-Bis(cyclohexyldimethylammoniomethyl)-
ferrocene (32+). Compound 2 (66 mg, 0.15 mmol) was dissolved in
15 mL of CH2Cl2 under a N2 atmosphere. Methyl sulfate (57 μL, 76
mg, 0.60 mmol) was added to the solution, and the mixture was stirred
for 24 h. A brown precipitate was formed. The solid was collected by
filtration and washed with CH2Cl2. The resulting light brown solid was
6096
dx.doi.org/10.1021/om400717z | Organometallics 2013, 32, 6091−6097

Organometallics
Article
ASSOCIATED CONTENT
* Supporting Information
Figures giving additional NMR and mass spectroscopic data as
mentioned in the text. This material is available free of charge
via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail for A.E.K.: akaifer@miami.edu.
■
Author Contributions
The manuscript was written through contributions of both
authors. Both authors have given approval to the final version
of the manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are grateful to the National Science Foundation for the
generous support of this research work (to A.E.K., Grant No.
CHE-0848637).
REFERENCES
■
(1) Jeon, W. S.; Moon, K.; Park, S. H.; Chun, H.; Ko, Y. H.; Lee, J.
Y.; Lee, E. S.; Samal, S.; Selvapalam, N.; Rekharsky, M. V.; Sindelar, V.;
Sobransingh, D.; Inoue, Y.; Kaifer, A. E.; Kim, K. J. Am. Chem. Soc.
2005, 127, 12984−12989.
(2) Liu, S.; Ruspic, C.; Mukhopadhyay, P.; Chakrabarti, S.; Zavalij, P.
Y.; Isaacs, L. J. Am. Chem. Soc. 2005, 127, 15959−15967.
(3) Rekharsky, M. V.; Mori, T.; Yang, C.; Ko, Y. H.; Selvapalam, N.;
Kim, H.; Sobransingh, D.; Kaifer, A. E.; Liu, S.; Isaacs, L.; Chen, W.;
Moghaddam, S.; Gilson, M. K.; Kim, K.; Inoue, Y. Proc. Natl. Acad. Sci.
U.S.A. 2007, 104, 20737−20742.
(4) Moghaddam, S.; Yang, C.; Rekharsky, M.; Ko, Y. H.; Kim, K.;
Inoue, Y.; Gilson, M. K. J. Am. Chem. Soc. 2011, 133, 3570−3581.
(5) Green, N. M. Biochem. J. 1966, 101, 774−778.
(6) Biedermann, F.; Uzunova, V. D.; Scherman, O. A.; Nau, W. M.;
De Simone, A. J. Am. Chem. Soc. 2012, 134, 15318−15323.
(7) Cui, L.; Gadde, S.; Li, W.; Kaifer, A. E. Langmuir 2009, 25,
13763−13769.
(8) Yi, S.; Li, W.; Nieto, D.; Cuadrado, I.; Kaifer, A. E. Org. Biomol.
Chem. 2013, 11, 287−293.
(9) Yi, S.; Captain, B.; Kaifer, A. E. Chem. Commun. 2011, 47, 5500−
5502.
(10) Tootoonchi, M. H.; Yi, S.; Kaifer, A. E. J. Am. Chem. Soc. 2013,
135, 10804−10809.
(11) Day, A.; Arnold, A. P.; Blanch, R. J.; Snushall, B. J. Org. Chem.
2001, 66, 8094−8100.
(12) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877−910.
(13) Miller, S. R. G., D. A.; Chen, Z.-H.; Gokel, G. W.; Echegoyen,
L.; Kaifer, A. E. Anal. Chem. 1988, 60, 2021−2024.
(14) Ling, X. X.; Samuel, E. L.; Patchell, D. L.; Masson, E. Org. Lett.
2010, 12, 2730−2733.
(15) Yi, S.; Kaifer, A. E. J. Org. Chem. 2011, 76, 10275−10278.
(16) Yi, S.; Captain, B.; Ottaviani, M. F.; Kaifer, A. E. Langmuir 2011,
27, 5624−5632.
(17) Li, W.; Kaifer, A. E. Langmuir 2012, 28, 15075−15709.
6097
dx.doi.org/10.1021/om400717z | Organometallics 2013, 32, 6091−6097