Sequence-Specific Self-Assembly of Positive and Negative Monomers with Cucurbit[8]uril Linkers

Article abstract of DOI:10.1021/jacs.7b13304

The self-assembly into dynamic oligomers of Cucurbit[8]uril (CB[8]), a positive ditopic Ir(III) bis-terpyridine complex, and a negative ditopic Fe(II) bis-terpyridine complex flanked by four butyrate side chains was assessed to answer a seemingly straight

Full text of DOI:10.1021/jacs.7b13304

Subscriber access provided by THOMAS JEFFERSON UNIV
Article
Sequence-specific self-assembly of positive and
negative monomers with Cucurbit[8]uril linkers
Mersad Raeisi, Kondalarao Kotturi, Ian Del Valle, Jan Schulz, Paulina Dornblut, and Eric Masson
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b13304 • Publication Date (Web): 14 Feb 2018
Downloaded from http://pubs.acs.org on February 14, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Sequence-specific self-assembly of positive and negative
monomers with Cucurbit[8]uril linkers
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Mersad Raeisi, Kondalarao Kotturi, Ian del Valle, Jan Schulz, Paulina Dornblut and Eric Masson*
Department of Chemistry and Biochemistry, Ohio University, Athens, Ohio 45701, United States.
ABSTRACT: The self-assembly into dynamic oligomers of Cucurbit[8]uril (CB[8]), a positive ditopic Ir(III) bis-terpyridine complex, and
a negative ditopic Fe(II) bis-terpyridine complex flanked by four butyrate side chains was assessed to answer a seemingly straightforward
question: does CB[8] adopt a social self-sorting pattern by encapsulating both positive and negative units into a hetero-ternary complex? We
showed that this is indeed the case, with CB[8] linking a positive Ir unit to a neighboring negative Fe unit whenever possible. Furthermore,
the solubility of the dynamic oligomers was significantly affected by their sequence; upon addition of 0.6 – 1.2 equiv. positive Ir oligomer to its
negative Fe counterpart, the predominant assembly present in solution was a mixed oligomer with a (Fe–Ir–Ir–)n sequence. Weak interac-
tions between the negative butyrate side chains and the partially positive outer wall of CB[7] were also identified by two-dimensional nuclear
magnetic resonance techniques, and resulted in a negative pKa shift (0.10 pKa unit) for the terminal carboxylic groups.
INTRODUCTION
The recognition properties of the cavity of Cucurbit[n]uril mac-
rocycles (CB[n]) have been studied for more than 30 years, and
the major driving forces responsible for guest encapsulation in
aqueous medium, very often with extreme affinity, are now firmly
established.1-8 The ejection of water from the CB[n] cavity back to
the bulk, a process that is enthalpically and entropically favorable,
accounts for much of the affinity. Rigid guests, tight fits into the
macrocycle cavity, and the presence of positively charged substitu-
ents that interact with the carbonyl portals of CB[n] are major
affinity enhancing factors. In the latter case, binding affinities typi-
cally increase by 103 to 104- fold for each portal/positive substituent
interaction.2 Neutral guests,1 including hydrocarbons, bind to
CB[6],4 CB[7],9 CB[8]9 and to the cavity located between the
inner wall of CB[8] and an auxiliary guest.10 A few years ago, we
also reported that biaryl 1a binds to CB[7] despite its net negative
charge, albeit 103 times weaker than positive analog 1b (see Figure
1).11 We are now seeking to answer the following question: how do
a positive guest and a negative guest self-assemble in the presence
of CB[8]? Should one expect complete social self-sorting, i.e. the
exclusive formation of the hetero-ternary complex between all three
units, to maximize Coulombic interactions between the positive
and negative guests? Or to the contrary, are the overall charges of
the guests irrelevant to the self-sorting patterns? Biaryl 1a is one of
the rare negative guests with a reported binding affinity to a mem-
ber of the CB[n] family, therefore this seemingly straightforward
question has never been addressed before.
Figure 1. Biphenyl derivatives with reported binding affinities to
CB[7]. CB[n]-binding metal-ligand complexes relevant to this study.
Recently, our group showed that Fe(II) and Ir(III) bis-
terpyridine complexes 2 and 3, which bear 2-naphthyl groups and
ACS Paragon Plus Environment

Journal of the American Chemical Society
2-(2,3,5,6-tetrafluorophenyl) groups at their 4’-positions, respec-
Page 2 of 8
1
2
3
4
5
6
7
8
tively, undergo social self-sorting in the presence of CB[8], and
assemble into dynamic oligomers (2·CB[8]·3·CB[8])n with alter-
nating Fe and Ir metallic cores (see Figure 1).12 The macrocycle
forms hetero-ternary complexes with naphthyl units attached to the
Fe(II) complex and with the tetrafluorophenyl units of the Ir(III)
complex. The remarkable self-sorting pattern was attributed to
favorable quadrupole-quadrupole interactions between both aro-
matic units inside the cavity of CB[8].12 To answer the question
asked above, we will assess the self-sorting patterns of CB[8], posi-
tive Ir(III) complex [Ir•42]3+, positive Fe(II) complexes [Fe•52]2+,
[Fe•62]2+ (as controls) and doubly negative Fe(II) complex
[Fe•72]2- (see Figure 1; the preparation of all metal-ligand com-
plexes is presented in the supporting information section). The 2,6-
difluoro-4-tolyl substituent at the 4’-position of terpyridine ligands
5 – 7 was chosen to allow the monitoring of CB[n] binding by
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
19
both H and F nuclear magnetic resonance spectroscopy (NMR),
the latter nucleus being particularly valuable when the terpyridine
ligands are decorated with side chains resulting in more complex 1H
NMR spectra. For a clearer depiction of the intermolecular forces
at play, we start by assessing the conformation of the side chains in
complexes [Fe•62]2+ and [Fe•72]2- when the difluoroaryl unit is
encapsulated by CB[7]. This host is used for this portion of the
Figure 2. Complex [Fe•62]2+•CB[7]2 highlighting the possibility for
NH···O=C hydrogen bonding. 1H NMR spectra of (a) complex
[Fe•62]2+ and (b) assembly [Fe•62]2+•CB[7]2; 19F NMR spectra of (c)
complex [Fe•62]2+ and (d) assembly [Fe•62]2+•CB[7]2.
study,
as
discrete,
well-defined
ternary
complexes
[Fe•62]2+•CB[7]2 and [Fe•72]2-•CB[7]2 are formed, as opposed to
dynamic oligomers in the presence of CB[8].
RESULTS AND DISCUSSION
1H NMR spectra with water suppression were then recorded in a
water/deuterium oxide mixture (9:1) to examine the possibility of
hydrogen bonding between the amide side chains of complexes
[Fe•62]2+ and [Fe•72]2- and the carbonyl portal of CB[7] (see Fig-
ure 2). NH signals underwent upfield shifts (0.18 and 0.19 ppm at
Upon addition of CB[7] to complexes Fe•L2 (L = 5 – 7),
difluoroaryl hydrogens, as well at its methyl group, underwent the
expected upfield shift that characterizes encapsulation (0.95 and
0.40 ppm on average throughout the ligand 5 – 7 series; see com-
plex [Fe•62]2+•CB[7]2 and signals labeled 7 and 8 in Figure 2). To
the contrary, hydrogens at positions 3 and 3’’ of the terpyridine unit
underwent a very strong downfield shift (1.13 ppm on average), an
effect that we have attributed to CH···O=C hydrogen bonding with
the cabonyl portal of the macrocycle.12 19F NMR experiments
showed marked downfield shifts upon CB[7] encapsulation (2.81
ppm on average, see Figure 2). In all cases, host/guest exchanges
were slow on the 1H and 19F NMR time scales.
25 °C in the case of complexes [Fe•62]2+ and [Fe•72]2-, respective-
ly) upon CB[7] binding. As hydrogen bonding typically results in
pronounced downfield shifts for the hydrogen nucleus of the do-
nor,13-15 strong NH···O=C interactions are unlikely to be present in
our system; we do remain cautious however, as NH···O=C hydro-
gen bonding resulting in upfield shifts have been observed in pro-
teins.16 Another diagnostic test for NH···O=C hydrogen bonding
that does not involve isotope labeling consists in monitoring NH
chemical shifts as a function of temperature. In proteins, weak tem-
perature gradients (–3 2 ppb/K) are symptomatic of hydrogen
bonding, while larger gradients (–7 3 ppb/K) indicate the contra-
ry. From a set of 793 amide protons in 14 different proteins, a 93%
probability of hydrogen bonding was detected for gradients weaker
than –4 ppb/K.17 NH chemical shifts were measured in complexes
[Fe•62]2+, [Fe•62]2+•CB[7]2, [Fe•72]2- and [Fe•72]2-•CB[7]2 at 4, 8
and 12 °C. Gradients were 4.6, 5.3, 5.3 and 3.9 ppb/K, respectively.
The pair of gradients for complexes [Fe•62]2+ and
[Fe•62]2+•CB[7]2 shows unambiguously that NH···O=C hydrogen
bonding is not significant, as the gradient increases in the presence
of CB[7]. To the contrary, the gradient measured for assembly
[Fe•72]2-•CB[7]2 is 1.5 ppb/K weaker than its free counterpart,
and falls just below the 4 ppb/K threshold. This suggests that
NH···O=C hydrogen bonding between the amide side chains and
the CB[7] rims, at least in some conformations of the guest, is a real
possibility. We can also rule out intramolecular hydrogen bonding
between the amide donors and the carboxylate acceptors of com-
plex [Fe•72]2-, as the gradient of the latter is stronger than the one
ACS Paragon Plus Environment

Page 3 of 8
Journal of the American Chemical Society
measured for complex [Fe•62]2+, that does not possess side chains
1
2
3
4
5
6
7
8
terminated with carboxylate groups.
The plausible interaction between the CB[7] portals and the am-
ide group of complex [Fe•72]2- would position the butyrate side
chain in a conformation prone to interactions with the outer wall of
CB[7] (see Figure 3). To test this hypothesis, we monitored the
chemical shifts of the side-chain methylene groups α – γ. While the
signals of hydrogens at position α remain steady upon addition of
CB[7] to the metal-ligand complex, hydrogens at position β and γ
undergo mild upfield shifts (0.09 and 0.01 ppm, respectively, see
Figure 3). While any shift should indicate proximity to the macro-
cycle, the former are typically upfield when hydrogens penetrate
the cavity of the macrocycle, or downfield when they reside close to
the carbonyl portal. Therefore in the present case, we propose that
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. (a) Portion of the 1H−1H NOESY spectrum of complex
[Fe•72]2-•CB[7]2 with a 0.25 s mixing time. Correlations between
hydrogens α and β of the butyrate side chain and the co-axial hydro-
gens of the CB[7] methylene rim (red and blue circles, respectively),
and between hydrogens α and β and the CB[7] co-equatorial meth-
ylene hydrogens (dashed orange and green circles, respectively). (b)
NOE crosspeak signal growth as a function of mixing time (same color
code). Proper volume integration could not be carried out for the
crosspeak highlighted with the dashed orange circle due to signal over-
lap. The reference NOE buildup (grey circles) corresponds to the
correlation between hydrogens 3 and 3’ of the terpyridine unit (see
Figure 2 for numbering).
one or more guest conformations position the β-CH2 groups, and
to a lesser extent γ-CH2 groups, near the outer methylene rim of
CB[7].
Yet, the most convincing evidence for the interaction between
the outer wall of CB[7] and the terpyridine side chain of complex
[Fe•62]2- is the impact of encapsulation on the pKa of its carboxylic
groups. pKa shifts of ammonium groups interacting with, and being
stabilized by the partially negative carbonyl portals of CB[7] have
been thoroughly documented21-29 (up to 5.2 pKa units).30 If the
carboxylate groups of complex [Fe•72]2- were to interact with and
be stabilized by the partially positive outer wall of CB[7], a negative
pKa shift for the conjugate carboxylic acid units would ensue. A
positive pKa shift would be more likely if the outer wall interactions
were insignificant, as the interaction between the carbonyl portals
of CB[7] and the terpyridine ligands would likely stabilize and
weaken the positive charge of the complex. The pKa of both free
and CB[7]-bound [Fe•72]2- complexes were determined precisely
by 1H NMR spectroscopy from the ratios of protonated and depro-
tonated species in an acetic acid/sodium acetate buffer (see sup-
porting information for details). pKa’s of 4.62 ( 0.01) and 4.52 (
0.01) were respectively obtained; the negative 0.10 ( 0.01) pKa
shift thus confirms the elusive interactions between the butyrate
side chains and the outer wall of CB[7]. To the best of our
knowledge, this negative shift, albeit modest but clearly outside the
margin of error, is unprecedented in CB[n]/guest recognition.
Figure 3. DFT-optimized truncated structure of a possible confor-
mation for complex [Fe•72]2-•CB[7]2 (TPSS-D3(BJ) functional,
1
COSMO solvation term and def2-SVP basis sets). H NMR spectra of
complexes (a) [Fe•72]2- and (b) [Fe•72]2-•CB[7]2.
We then carried out quantitative two-dimensional nuclear Over-
hauser effect spectroscopy (NOESY) experiments to determine
through-space distances between the butyrate methylene groups
and the closest methylene rim of CB[7]. The distances can be ob-
tained from the growth of the cross-peaks measured upon increases
in mixing times, relative to the growth of a reference cross-peak
assigned to two hydrogens sitting at a known distance (see Figure
4).18-19 The distances between the “co-axial”20 inward-pointing
methylene hydrogen at the CB[7] rim closest to the α and β hy-
drogens of the butyrate side chain are at least 3.8 and 4.1 Å, respec-
tively, while the distance between the “co-equatorial”20 outer meth-
ylene hydrogen closest to the butyrate β-hydrogens is at least 4.3 Å.
Those distances are significantly longer than those measured on a
truncated DFT-optimized structure with a tight interaction be-
tween the carboxylate units and the outer wall of CB[7] (3.6, 2.3
and 2.4 Å, respectively; see Figure 3). While this structure is plausi-
ble, it is probably one of many conformations generated by the
flexible butyrate side chain.
The impact of the metal-ligand complex total charges on CB[7]
binding affinities were measured by isothermal titration calorimetry
(ITC). Remarkably, they were found to be very similar (6.0 0.2,
11.1 0.5 and 6.1 0.2 × 106 M-1 in the case of complexes
[Fe•52]2+, [Fe•62]2+ and [Fe•72]2-, respectively; see supporting
information section for binding isotherms). The mere 1.9-fold
preference for CB[7] binding to complex [Fe•62]2+, which bears
four amide substituents, compared to its unsubstituted analog
[Fe•52]2+ supports the lack of substantive hydrogen bonding be-
tween the CB[7] carbonylated rim and the amides of complex
[Fe•62]2+. The negligible difference in binding affinities between
complex [Fe•72]2- and the unsubstituted analog [Fe•52]2+ indicates
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 8
that the overall penalty for CB[7] binding to the overall negative
CB[7], as described above. Several cases of CB[8]-assembled oli-
gomers have been reported in the literature,39-43 a significant num-
ber of them exploiting the now iconic interaction between methyl-
viologen and 2-naphthoxy derivatives inside CB[8].44 However, to
the best of our knowledge, ([Fe•72]2-•CB[8])n represents the first
case of an all-negative CB[8]-assembled oligomer.
guest [Fe•72]2- is counterbalanced by the stabilization of the termi-
nal carboxylate groups by the partially positive outer methylene rim
of the macrocycle. One could also argue that in aqueous medium,
the major driving force for binding is the ejection of high-energy
water from the cavity of CB[7], and charges at remote substituents
only play a minor role, as solvation limits interactions with their
partially positive surroundings. These interactions are not insignifi-
cant though, as the pKa of both free and CB[7]-bound Fe•72 car-
boxylic groups are both lower than the pKa of acetic acid (4.62 and
4.52 vs. 4.76).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
While a few valuable examples of exclusion complexes have been
presented in the crystalline form,31-37 we showed here for the first
time an interaction between a guest and the convex outer wall of
CB[n] in aqueous solution. The interaction is weak though, as
explained above, and should not significantly impact self-sorting
patterns in the presence of CB[8].
As CB[8] encapsulates two 2,6-difluoro-4-tolyl units pertaining
to two consecutive metal-ligand complexes in its cavity, consecu-
tive additions of the macrocycle to all three metal-ligand complexes
resulted in the partial formation of [Fe•L2]•CB[8]•[Fe•L2] ternary
complexes (L = 5 – 7) at 0.25 equivalent CB[8], concomitantly
with larger assemblies ([Fe•L2]•CB[8])2-3•[Fe•L2] at 0.50 – 0.75
equiv. CB[8], and finally long dynamic oligomers
([Fe•L2]•CB[8])n with 1.0 equivalent CB[8] (see Figure 5). We
note that guest encapsulation into CB[8] dramatically increases the
solubility of the host in aqueous medium. Both “π-π stacking” in-
teractions and CB[8] encapsulation result in large downfield shifts
for fluorine nuclei (4.6 ppm), compared to experiments carried out
with CB[7] (2.8 ppm on average). Signal labeled “a” and “b” at –
116.11 and –116.06 ppm (see Figure 5), respectively, account for
fluorine nuclei in free [Fe•72]2- complexes and unencapsulated
atoms in complexes ([Fe•72]•CB[8])n•[Fe•72] (n ≥ 1); signals “c”
and “d” at –111.31 and –111.52 ppm account for encapsulated
fluorine nuclei in ternary complexes [Fe•72]•CB[8]•[Fe•72] and
larger assemblies ([Fe•72]•CB[8])2-3•[Fe•72], respectively. The
differentiation between both assemblies is surprising, and must
indicate a significant structural change upon connection of a se-
cond CB[8] unit to ternary complex [Fe•72]•CB[8]•[Fe•72]. The
formation of longer oligomer results in the broadening of the signal,
with a maximum at –111.34 ppm (signal “e” in Figure 5).
Figure 5. 19F NMR spectra of complex [Fe•72]2- (1.0 mM) (a) in the
absence of CB[8], and in the presence of (b) 0.25 equiv., (c) 0.50
equiv. and (d) 1.0 equiv. CB[8] in deuterium oxide.
The self-sorting patterns of oligomers ([Fe•L2]•CB[8])n once
mixed with oligomer ([Ir•42]3+•CB[8])n could then be examined
1
19
by H and F NMR spectroscopy. The latter is a particularly useful
19
method to assess self-sorting as F NMR signals pertaining to 2,6-
difluoro-4-tolyl units in CB[8]-bound homodimers (referred to as
L•L; L = 5 – 7; see Figure 6) and in CB[8]-bound heterodimers
formed with the 4-tolyl substituent of ligand 4 (referred to as L•4)
do not overlap in most cases. L•4 signals are shielded by 0.56 and
0.37 ppm compared to L•L, for L = 5 and 6, respectively; 7•7 and
7•4 signals cannot be differentiated. Fortunately, in all cases, hy-
drogen nuclei at positions 3 of the 2,6-difluoro-4-tolyl unit in lig-
ands 5 – 7, and of the 4-tolyl unit in ligand 4 resonate at different
frequencies when homo- or hetero-dimers are formed inside
CB[8]. They can also be used to decipher self-sorting patterns, but
1
the complexity of H NMR spectra makes the task more arduous.
On average, hydrogens at position 3 of the 4-tolyl unit are deshield-
ed by 0.26 ppm when interacting with the difluorotolyl unit in het-
erodimers L•4 compared to homodimers 4•4 (see Figure 6).
Average molecular weights, and subsequently the average num-
ber of repeat metal-ligand units at a given monomer concentration
(1.0 mM), were determined by diffusion-ordered NMR spectros-
copy (DOSY), after calibration with all metal-ligand complexes and
their CB[7]-bound assemblies prepared in this study and in earlier
work12 (see SI section for details). Molecular weights M and diffu-
sion coefficients D are linked by the power law D ∝ M-m. For spher-
ical particles, the Stokes-Einstein equation affords a coefficient m
equal to 1/3, as long as molecular weights are proportional to the
particle volume. For non-spherical systems such as polymers, m
typically ranges from 0.3 to 0.6.38 Our calibration returns a coeffi-
cient m equal to 0.41 ( 0.03). Logarithms of diffusion coefficients
for oligomers ([Fe•L2]•CB[8])n (L = 5 – 7) were –10.25, –10.35
and –10.30, respectively, vs. –9.48 to –9.63 as free metal-ligand
complexes; this corresponds to 27, 43 and 28 repeat units at 1.0
mM monomer concentration, respectively. The longer assembly
formed upon interaction of complex [Fe•62]2+ with CB[8] com-
pared to the other metal-ligand complexes is likely due to a higher
binding affinity for the macrocycle. This trend is also observed with
Figure 6. 1H NMR spectra of (a) oligomer ([Fe•62]2+•CB[8])n, (b)
oligomer ([Fe•62]2+•CB[8])n in the presence of 1.0 equiv. oligomer
([Ir•42]3+•CB[8])n and (c) oligomer ([Ir•42]3+•CB[8])n. 19F NMR
spectra of (d) oligomer ([Fe•62]2+•CB[8])n and (e) oligomer
([Fe•62]2+•CB[8])n in the presence of 1.0 equiv. oligomer
([Ir•42]3+•CB[8])n. All spectra recorded in deuterium oxide; the con-
centration of complex [Fe•62]2+ is 1.0 mM.
Monitoring L•L and L•4 concentrations as a function of oligo-
mer ([Ir•42]3+•CB[8])n concentration allows the precise quantifi-
cation of the self-sorting pattern. While the terms “narcissistic” and
ACS Paragon Plus Environment

Page 5 of 8
Journal of the American Chemical Society
“social” were coined almost two decades ago45-46 to describe the
instantaneous on the experiment time scale,12 the precipitate is
much more likely to be a social aggregate of scrambled Fe- and Ir-
containing units interconnected with CB[8], than bundles of co-
precipitated narcissistic ([Ir•42]3+•CB[8])n and ([Fe•72]2-
•CB[8])n oligomers.
1
2
3
4
exclusive formation of homo- and heterodimers, respectively, we
propose here a scale that quantifies partial self-sorting patterns. We
consider equilibrium (1) between a pair of homodimers (L•L and
L’•L’) and their corresponding heterodimers L•L’ and L’•L.
5
6
7
!
!! ꢀ ∙ ꢀ + ! ꢀ! ∙ ꢀ! ⇌ !! ꢀ ∙ ꢀ! + !! ꢀ′ ∙ ꢀ
(1)
8
9
Its equilibrium constant K can then be readily converted into a
relative free energy term ΔG (see equation 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
!/!
ꢀ∙ꢀ! ꢀ!∙ꢀ
Δꢀ = −ꢀꢁ ln ꢀ = −ꢀꢁ ln
(2)
!/!
ꢀ∙ꢀ ꢀ!∙ꢀ!
Figure 7. (a) Concentration of metal-ligand complex [Fe•72]2- in solu-
tion as a function of the concentration of complex [Ir•42]3+, both rela-
Complexes L•L’ and L’•L each account for half the total con-
centration of heterodimer. The free energy term can thus be readily
obtained from equation 3.
tive to concentration of complex [Fe•72]2- in the absence of [Ir•42]3+
.
1H NMR spectra of solutions of (b) oligomer ([Fe•72]2-•CB[8])n, and
(c) oligomer ([Fe•72]2-•CB[8])n in the presence of 1.0 equiv. oligomer
([Ir•42]3+•CB[8])n.
ꢀ∙ꢀ! /!
Δꢀ = −ꢀꢁ ln
(3)
!/!
ꢀ∙ꢀ ꢀ!∙ꢀ!
Subsequent additions of positive oligomer ([Ir•72]3+•CB[8])n to
the aggregate triggered the reemergence of species in solution, as
We consider a self-sorting process to be fully social when ΔG < –
2.3 kcal/mol and fully narcissistic when ΔG > 2.3 kcal/mol. Those
thresholds correspond to 1:1:98 and 49:49:2 ratios of dimers L•L,
L’•L’ and L•L’, respectively, when equimolar amounts of species
L and L’ are combined with CB[8] at room temperature. A 1:1:2
1
showed by H NMR spectroscopy (see Figure 7, spectrum c). Re-
markably, the composition of the oligomer in solution remained
steady up to 1.2 equiv. ([Ir•42]3+•CB[8])n, with 31 ( 2)% of 4•4
homodimer and 69 ( 5)% of 7•4 heterodimer formation. The
average composition of the oligomer is thus ([Fe•72]2-
0.35•[Ir•42]3+0.65•CB[8])n (or a 1:2 ratio of Fe and Ir complexes this
time!), with a net charge of +1.3 per CB[8] unit. As the concentra-
tion of oligomer ([Ir•72]3+•CB[8])n is further increased, more
homodimer 4•4 is formed to the expense of heterodimer 7•4. The
solubility of the oligomers in water is thus minimal when their net
charge per CB[8] unit is close to neutral, within the –0.5 to +1.3
interval. Yet the most striking feature of the self-sorting process is
the absence of homodimer 7•7, even at low concentrations of oli-
gomer ([Ir•42]3+•CB[8])n (see Figure 7). Spectrum b shows the
signature signal for ([Fe•72]2-•CB[8])n (or 7•7) formation at 6.25
ppm, and the latter is unequivocally absent in spectrum c (see sym-
bol “×” in Figure 7). Therefore, self-sorting is fully social. As con-
trol experiments with terpyridine ligands bearing no side-chains or
neutral ones showed only partially social self-sorting with oligomer
([Ir•42]3+•CB[8])n, one can conclude that positive metal-ligand
complex [Ir•42]3+ do exhibit a strong preference for overall negative
complex [Fe•72]2- when undergoing CB[8] encapsulation. A 1:2
ratio of CB[8]-bound [Fe•72]2- and [Ir•42]3+ units, with the Fe
complex always followed by an Ir complex may either correspond
to (1) a distribution of sequences (Fe–Ir–)n, (Fe–Ir–Ir–)n and
(Fe–Ir–Ir–Ir–…)n, or (2) a unique (Fe–Ir–Ir–)n sequence. If the
first scenario were to be correct, the Fe/Ir ratio in solution would
steadily decrease as more oligomer ([Ir•42]3+•CB[8])n is added to
the solution, starting at a 1:1 ratio. As discussed above, this is not
the case: the Fe/Ir ratio in solution remains steady at 1:2 upon
addition of 0.6 to 1.2 equiv. ([Ir•42]3+•CB[8])n to oligomer
([Fe•72]2-•CB[8])n. We thus conclude that self-sorting is not only
exclusively social, but it also leads to the formation of a predomi-
nant (Fe–Ir–Ir–)n sequence.
ratio returns ΔG = 0.
Oligomer ([Ir•42]3+•CB[8])n was found to undergo partially so-
cial self-sorting when allowed to scramble in the presence of oligo-
mers ([Fe•52]2+•CB[8])n and ([Fe•62]2+•CB[8])n. The ratio of
homo- and heterodimers could be obtained by fitting their 19F
NMR signals with a sum of two Gaussian functions, and by inte-
grating the latter (see signals 6•6 and 6•4 in Figure 6, spectra d
and e). ΔG terms are –0.25 0.03 and –0.29 0.03 kcal/mol, re-
spectively. The quadrupole-quadrupole interaction between 4-tolyl
and 2,6-difluoro-4-tolyl units is thus preferred to the geometric
mean of the tolyl/tolyl and difluorotolyl/difluorotolyl interactions.
The ΔG terms also show that short amide side-chains linked to the
terpyridine units do not significantly affect self-sorting patterns, at
least when total charges are not affected.
To answer the question proposed in the introduction (namely,
how do positive and negative building blocks self-organize in the
presence of CB[8]), we added aliquots of positive oligomer
([Ir•42]3+•CB[8])n to negative oligomer ([Fe•72]2-•CB[8])n, and
we monitored self-sorting patterns. The concentration of oligomer
([Fe•72]2-•CB[8])n in solution steadily decreased after each addi-
1
19
tion, as observed by H and F NMR spectroscopy, thereby indi-
cating aggregation. ([Ir•42]3+•CB[8])n was never detected in solu-
tion, thereby confirming that the precipitate is an aggregate of
scrambled oligomers containing both iron and iridium complexes.
Complete precipitation of oligomer ([Fe•72]2-•CB[8])n was ob-
tained in the presence of approximately 0.5 equiv.
([Ir•42]3+•CB[8])n. A plot of the amount of ([Fe•72]2-•CB[8])n in
solution as a function of added iridium content is linear (see Figure
7a); the slope of the linear regression (2.38 0.17) represents the
number of [Fe•72]2- units that precipitate together with each equiv-
alent of complex [Ir•42]3+. The composition of the precipitated
oligomer is thus ([Fe•72]2-0.70•[Ir•42]3+0.30•CB[8])n (or a near 2:1
ratio of Fe and Ir complexes) with a net charge of –0.5 per CB[8]
unit, counterbalanced by butyrate lithium countercations. As
scrambling between CB[8] and metal-terpyridine complexes is
CONCLUSIONS
An elusive interaction between the butyrate side-chains of com-
plex [Fe•72]2- and the convex outer wall of CB[7] was detected in
aqueous medium. The interaction led to a subtle, yet significant,
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 8
(4)
9342.
(5)
Florea, M.; Nau, W. M. Angew. Chem., Int. Ed. 2011, 50, 9338-
negative pKa shift (0.10) for the butyric acid unit when put in con-
tact with a neighboring CB[7] host. Despite its doubly negative
charge, this metal-ligand complex not only formed tight assemblies
with CB[7] upon encapsulation of its 4’-substituents, but also as-
sembled into dynamic oligomers in the presence of CB[8]. Fully
social self-sorting is observed between CB[8], positive metal-ligand
complex [Ir•42]3+ and negative complex [Fe•72]2-, which self-
assemble into hetero-ternary complexes exclusively. While the
addition of positive oligomer ([Ir•42]3+•CB[8])n to negative oli-
gomer ([Fe•72]2-•CB[8])n triggered the precipitation of a mixed
Fe/Ir aggregate in a 7:3 ratio at low Ir concentration, further en-
richment in Ir oligomer content returned Fe units into solution, as
part of an oligomer with a preferred (Fe–Ir–Ir–)n sequence.
1
2
3
4
5
6
7
8
Biedermann, F.; Uzunova, V. D.; Scherman, O. A.; Nau, W. M.;
De Simone, A. J. Am. Chem. Soc. 2012, 134, 15318-15323.
(6)
577.
(7)
(8)
Sci. 2016, 7, 3569-3573.
(9)
(10) Rabbani, R.; Masson, E. Org. Lett. 2017, 19, 4303-4306.
(11) Joseph, R.; Masson, E. Org. Biomol. Chem. 2013, 11, 3116-
Nau, W. M.; Florea, M.; Assaf, K. I. Isr. J. Chem. 2011, 51, 559-
Koner, A. L.; Nau, W. M. Supramol. Chem. 2007, 19, 55-66.
Ling, X.; Saretz, S.; Xiao, L.; Francescon, J.; Masson, E. Chem.
Lu, X.; Isaacs, L. Angew. Chem., Int. Ed. 2016, 55, 8076-8080.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3127.
(12) Joseph, R.; Nkrumah, A.; Clark, R. J.; Masson, E. J. Am. Chem.
Soc. 2014, 136, 6602-6607.
(13) Berglund, B.; Vaughan, R. W. J. Chem. Phys. 1980, 73, 2037-
2043.
(14) Haushalter, K. A.; Lau, J.; Roberts, J. D. J. Am. Chem. Soc.
1996, 118, 8891-8896.
(15) Yao, L.; Grishaev, A.; Cornilescu, G.; Bax, A. J. Am. Chem. Soc.
2010, 132, 10866-10875.
(16) Wagner, G.; Pardi, A.; Wuethrich, K. J. Am. Chem. Soc. 1983,
105, 5948-5949.
(17) Cierpicki, T.; Otlewski, J. J. Biomol. NMR 2001, 21, 249-261.
(18) Baleja, J. D.; Moult, J.; Sykes, B. D. J. Magn. Reson. 1990, 87,
375-384.
(19) Mugridge, J. S.; Zahl, A.; van Eldik, R.; Bergman, R. G.;
Raymond, K. N. J. Am. Chem. Soc. 2013, 135, 4299-4306.
(20) Term coined by Prof. Werner Nau in private communications.
(21) Praetorius, A.; Bailey, D. M.; Schwarzlose, T.; Nau, W. M. Org.
Lett. 2008, 10, 4089-4092.
(22) Saleh, N. i.; Koner, A. L.; Nau, W. M. Angew. Chem., Int. Ed.
2008, 47, 5398-5401.
(23) Shaikh, M.; Dutta Choudhury, S.; Mohanty, J.; Bhasikuttan, A.
C.; Nau, W. M.; Pal, H. Chem. - Eur. J. 2009, 15, 12362-12370.
(24) Mohanty, J.; Bhasikuttan, A. C.; Nau, W. M.; Pal, H. J. Phys.
Chem. B 2006, 110, 5132-5138.
(25) Wang, R.; Macartney, D. H. Org. Biomol. Chem. 2008, 6,
1955-1960.
(26) Wyman, I. W.; Macartney, D. H. Org. Biomol. Chem. 2010, 8,
247-252.
(27) Saleh, N. i.; Al-Handawi, M. B.; Bufaroosha, M. S.; Assaf, K. I.;
Nau, W. M. J. Phys. Org. Chem. 2016, 29, 101-106.
(28) Gavvala, K.; Koninti, R. K.; Sengupta, A.; Hazra, P. Phys. Chem.
Chem. Phys. 2014, 16, 2823-2826.
(29) Li, S.; Yin, H.; Wyman, I. W.; Zhang, Q.; Macartney, D. H.;
Wang, R. J. Org. Chem. 2016, 81, 1300-1303.
(30) Barooah, N.; Mohanty, J.; Pal, H.; Bhasikuttan, A. C. J. Phys.
Chem. B 2012, 116, 3683-3689.
(31) Cui, X.; Zhao, W.; Chen, K.; Ni, X.-L.; Zhang, Y.-Q.; Tao, Z.
Chem. - Eur. J. 2017, 23, 2759-2763.
(32) Shen, F.-F.; Zhao, J.-L.; Chen, K.; Hua, Z.-Y.; Chen, M.-D.;
Zhang, Y.-Q.; Zhu, Q.-J.; Tao, Z. CrystEngComm 2017, 19, 2464-2474.
(33) Ni, X.-L.; Xiao, X.; Cong, H.; Zhu, Q.-J.; Xue, S.-F.; Tao, Z. Acc.
Chem. Res. 2014, 47, 1386-1395.
The next step in this study consists in evaluating the impact of
third party targets, such as proteins, on the self-sorting properties of
these dynamic oligomers, whose sequence is expected to be altered
to maximize interactions between the ligand side-chains and the
target. Progress will be reported in due course.
ASSOCIATED CONTENT
Supporting Information
Preparation and characterization of metal-ligand complexes [Fe•L2] (L
= 5 – 7), their precursors, their assemblies with CB[7] and CB[8], as
well as complex [Ir•42]3+ and oligomer ([Ir•42]3+•CB[8])n. Are also
included the binding isotherms of CB[7] to complexes [Fe•L2], the
determination of the pKa of complexes [Fe•72]2- and [Fe•72]2-•CB[7]2,
and the determination of oligomer length by DOSY experiments. The
Supporting Information is available free of charge on the ACS Publica-
tions website.
AUTHOR INFORMATION
Corresponding Author
* masson@ohio.edu
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
We are grateful to the National Science Foundation (grant CHE-
1507321), the American Chemical Society Petroleum Research
Fund (grant 56375-ND4) and Ohio University for their financial
support. MR and KK were also supported in part by a fellowship
from the Condensed Matter and Surface Science Program (CMSS)
at Ohio University. We also thank Dr. Roymon Joseph for his valu-
able and frequent input in the course of this study, as well as Dr.
Alexandar L. Hansen, from the Campus Chemical Instrument Cen-
ter at The Ohio State University, for recording the DOSY and
NOESY spectra discussed therein, and offering training to MR and
KK.
(34) Chen, K.; Kang, Y.-S.; Zhao, Y.; Yang, J.-M.; Lu, Y.; Sun, W.-Y.
J. Am. Chem. Soc. 2014, 136, 16744-16747.
(35) Zhang, F.; Yajima, T.; Li, Y.-Z.; Xu, G.-Z.; Chen, H.-L.; Liu, Q.-
T.; Yamauchi, O. Angew. Chem., Int. Ed. 2005, 44, 3402-3407.
(36) Gao, Z.-W.; Feng, X.; Mu, L.; Ni, X.-L.; Liang, L.-L.; Xue, S.-F.;
Tao, Z.; Zeng, X.; Chapman, B. E.; Kuchel, P. W.; Lindoy, L. F.; Wei, G.
Dalton Trans. 2013, 42, 2608-2615.
(37) Chen, K.; Liang, L.-L.; Liu, H.-J.; Tao, Z.; Xue, S.-F.; Zhang, Y.-
Q.; Zhu, Q.-J. CrystEngComm 2012, 14, 8049-8056.
(38) Shimada, K.; Kato, H.; Saito, T.; Matsuyama, S.; Kinugasa, S. J.
Chem. Phys. 2005, 122, 244914.
REFERENCES
(1)
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.
(2)
Inoue, Y.; Gilson, M. K. J. Am. Chem. Soc. 2011, 133, 3570-3581.
(3) Ko, Y. H.; Hwang, I.; Lee, D.-W.; Kim, K. Isr. J. Chem. 2011,
51, 506-514.
Moghaddam, S.; Yang, C.; Rekharsky, M.; Ko, Y. H.; Kim, K.;
(39) Appel, E. A.; Biedermann, F.; Rauwald, U.; Jones, S. T.; Zayed,
J. M.; Scherman, O. A. J. Am. Chem. Soc. 2010, 132, 14251-14260.
ACS Paragon Plus Environment

Page 7 of 8
Journal of the American Chemical Society
(40) del Barrio, J.; Horton, P. N.; Lairez, D.; Lloyd, G. O.;
Toprakcioglu, C.; Scherman, O. A. J. Am. Chem. Soc. 2013, 135, 11760-
11763.
(41) Liu, J.; Tan, C. S. Y.; Lan, Y.; Scherman, O. A. Macromol.
Chem. Phys. 2016, 217, 319-332.
(44) Kim, H.-J.; Heo, J.; Jeon, W. S.; Lee, E.; Kim, J.; Sakamoto, S.;
Yamaguchi, K.; Kim, K. Angew. Chem., Int. Ed. 2001, 40, 1526-1529.
(45) Shivanyuk, A.; Rebek, J., Jr. J. Am. Chem. Soc. 2002, 124,
12074-12075.
1
2
3
(46) Taylor, P. N.; Anderson, H. L. J. Am. Chem. Soc. 1999, 121,
4
5
6
(42) Ma, X.; Tian, H. Acc. Chem. Res. 2014, 47, 1971-1981.
(43) Yin, J.-F.; Hu, Y.; Wang, D.-G.; Yang, L.; Jin, Z.; Zhang, Y.;
Kuang, G.-C. ACS Macro Lett. 2017, 6, 139-143.
11538-11545.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
TOC Figure
44x23mm (600 x 600 DPI)
ACS Paragon Plus Environment