Full Papers
doi.org/10.1002/cplu.202000739
ChemPlusChem
for the potential diastereoisomers and conformers arising from
the self-assembly of (�)-S8a (vide supra). To add insult to injury,
the insertion of one of the 1,5-DHN moieties into the cavity of
the supramolecule produces a loss of symmetry on the
corresponding inclusion complex, making the nuclei on the
upper and lower regions of the metallacyclic subunit chemically
non-equivalent but with, arguably, quite similar chemical shifts.
All these factors would certainly account for the complex
situation observed on the spectrum. Even though, a COSY
experiment recorded for the mixture rendered quite interesting
information. Specifically, groups of proton signals, rather than
individual nuclei, correlate in agreement with the proposed
structure (�)-S8a ·6NO3. Furthermore, following the same meth-
odology described for (�)-S5a,b, the Pt(II) analogue (�)-S8b ·6PF6
was self-assembled from (�)-L4 and M2, allowing for the
identification by means of HR-ESI-MS of a diagnostic peak at m/
z=1474.4163 for [S8b-2PF6]2+, with mass and isotopic distribu-
tion in good agreement with that calculated for the proposed
structure (Figure 3).
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
1
°
Figure 6. H-NMR (400 MHz, D2O, 60 C) spectra for: a) 2.5 mM L1, b) 2.5 mM
(�)-L3, c) (�)-S4a (2.5 mM (�)-L3, 2.5 mM L1 and 5 mM M1), d) DOSY NMR for
solution c).
With the information gathered on the self-assembly of the
homodimeric species (�)-S5a,b and (�)-S8a,b in our hands, as well
as that previously reported by our group on the structures of
[16]
non-functionalized metallacycles S2a,b and S7a,b
,
we then
proceeded to study the M1/M2-directed self-assembly involving
each of the designed matching pairs of complementary ligands.
As for the homodimeric species, the prospective analysis by
NMR spectroscopy of the targeted hetero pseudo[1]rotaxanes
would be as well conditioned by the structural features of the
ligands. However, in the case of the heterodimeric species, the
absence of syn-/anti- diasteromers, reduces the number of
potential isomers resulting from the self-assembly (Figure 5).
As previously observed for (�)-S5a, the reaction of a
stoichiometrically-adjusted mixture of (�)-L3 (2.5 mM), L1
relative chemical shifts, assigned with the aid of a COSY
experiment at 60 C, those showed the expected deshielding
[14]
°
of protons Ha, Hg and Hf in each of building blocks because of
coordination to M1 and CÀ H···π interactions with H4/H8, with
these protons of the guest part of (�)-L3 appearing massively
shielded by circa 4 ppm regarding the free ligand. The
remaining protons on the system appear consistently shielded,
because of the concurrence of the expected host-guest
°
interactions.
A DOSY experiment recorded at 60 C also
supported the formation of the heteromer (�)-S4a, with all the
signals for the building blocks diffusing as a whole (Figure 6d).
Furthermore, the M2-directed self-assembly of the matching
pair allowed us to obtain crucial HR-ESI MS data, in good
agreement with the formation of the self-sorted Pt(II) analogue
(�)-S4b (Figure 3).
1
(2.5 mM) and M1 (5 mM) in D2O, instantaneously produced a H
NMR with broad signals at r.t. Dilution experiments monitored
1
by H-NMR and UV-Vis for the mixture within the concentration
window 2.5–0.08 mM,[14] supported again the non-existence of
oligomeric species on the observed dynamic equilibrium.
Once more, VT 1H-NMR experiments allowed us to over-
come the initial observed coalescence, showing a single set of
signals for each of the ligands in very good agreement with the
Pd(II)-directed self-sorting of (�)-S4a. As shown in Figure 6, the
Regarding the study of the M1-directed self-sorting of the
matching pair L5/(�)-L4, the analysis of the results by means of
NMR spectroscopy was hampered by very similar reasons to
those obscuring the study of the homodimeric self-assembly of
1
°
experiment recorded at 60 C shows a situation of rapid
(�)-L4 itself (vide supra). In consequence, the H-NMR in D2O
interconversion between the DHN units inside/outside the
receptor, as well as all the potential up/down and endo/exo
isomers associated to the self-inclusion process. Regarding the
recorded for the mixture of the components needed for the
assembly of 5 mM (�)-S6a was certainly complex, to say the
least.[14]
In this situation, we found only subtle clues pointing out to
the heterodimeric self-assembly of the intended pseudo[1]
rotaxane. First, the NMR spectra for the mixture remained
unchanged within the 5–0.08 mM concentration window.
Furthermore, a linear correlation was found between the
absorbance attributed to the charge-transfer band (λmax
=
470 nm), and the concentration of components on the same
range, accounting for the non-occurrence of oligomeric species.
Additionally, the COSY NMR of the mixture allowed for the
observation of 1H–1H correlation peaks between groups of
signals that can be attributed (for instance, by analogy with
Figure 5. Schematic depiction of the potential isomers resulting from the Pd
(II)/PtII)-directed heterodimeric self-sorting of (�)-L3/L1 and (�)-L4/L5 (exem-
plified for (�)-L3/L1).
ChemPlusChem 2020, 85, 2672–2678
2675
© 2020 Wiley-VCH GmbH