Solution aggregation of platinum(ii) triimine methyl complexes

Article abstract of DOI:10.1039/d0dt02190c

The NMR chemical shifts of [Pt(tpy)(CH3)](PF6) (1) and [Pt(mbzimpy)(CH3)](PF6) (2), where tpy = 2,2′;6′2′′-terpyridine and mbzimpy = 2,6-bis(N-methylbenzimidazol-2-yl)pyridine, in room-temperature DMSO-d6 displayed concentration dependence as a result of formation of dimers. Quantification of these dimers, expressed by equilibrium constant (K), shows a greater tendency of 2 to aggregate in solution. Structural conformations of these dimers were confirmed by 2D 1H-1H NOESY; the results explicitly suggest a head-to-tail stacking arrangement of molecules in dimers. This journal is

Full text of DOI:10.1039/d0dt02190c

Dalton
Transactions
View Article Online
View Journal
COMMUNICATION
Solution aggregation of platinum(II) triimine
methyl complexes†
Cite this: DOI: 10.1039/d0dt02190c
Vikas M. Shingade * and William B. Connick‡
Received 20th June 2020,
Accepted 20th July 2020
DOI: 10.1039/d0dt02190c
rsc.li/dalton
The NMR chemical shifts of [Pt(tpy)(CH3)](PF6) (1) and [Pt the self-assembly for applications such as chemosensing,
(mbzimpy)(CH3)](PF6) (2), where tpy = 2,2’;6’2’’-terpyridine and supramolecular chemistry, host–guest chemistry, and bio-
mbzimpy = 2,6-bis(N-methylbenzimidazol-2-yl)pyridine, in room- molecular interactions.19,21 Our interest in the class of plati-
temperature DMSO-d6 displayed concentration dependence as a num(II) triimine complexes stems from our discovery of lumi-
result of formation of dimers. Quantification of these dimers, nescent Pt(mbzimpy)Cl+ salts and related compounds that
expressed by equilibrium constant (K), shows a greater tendency of display interesting chemosensing properties as a result of
2 to aggregate in solution. Structural conformations of these interactions noted above.22–26 In principle, solid-state struc-
dimers were confirmed by 2D 1H–1H NOESY; the results explicitly tures may give some insight into the nature of stacking inter-
suggest
a
head-to-tail stacking arrangement of molecules in actions. However, given the relatively weak forces involved, it is
not obvious that aggregates formed in solution will necessarily
dimers.
retain their preferred conformation when precipitated. A
A well-known property of late 2nd and 3rd row d8-electron tran-
sition metal complexes is their tendency to self-assemble in
solutions, which can have dramatic consequences for the
colors and other spectroscopic properties of these systems.1–8
In the case of platinum(II) complexes with triimine chelates,
such as Pt(tpy)Cl+, the consensus view is that aggregation is
driven mainly by Pt⋯Pt and ligand π⋯π interactions.1–5,9
Several literature reports on terpyridine-based platinum(II)
complexes demonstrate the use of electronic absorption and
1H NMR spectroscopies to gain insight into the nature of
aggregates.1,10–16 The work of Romeo et al. using the aforemen-
tioned spectroscopies demonstrated that Pt(tpy)(CH3)+ forms
dimers in dilute aqueous solution and larger aggregates when
the concentration and/or ionic strength increases.10,12,14,15
Despite the fact that Pt(tpy)(CH3)+ and many other platinum(II)
compounds have been extensively investigated for their intri-
guing spectroscopic and photophysical properties,1,3,4,9,17–20
solution structures of these self-assemblies, which strongly
influence these properties, have remained elusive. Knowledge
of these structures is important to efforts aimed at exploiting
warning of the potential pitfalls of such assumptions was
highlighted in a recent study of the vapochromism of yellow
[Pt(tpy)Cl](ClO4), which turns red upon exposure to water
vapor.27 Loss of water vapor restores the original yellow color,
but this yellow material has a different packing arrangement
than crystals of [Pt(tpy)Cl](ClO4) precipitated from solution. To
better understand the preferred stacking conformation in such
complexes, herein, we have investigated the aggregation of two
model complexes in Scheme 1 in DMSO-d6 by means of NMR
and UV-vis absorption spectroscopy.
The hexafluorophosphate salt of the terpyridyl complex 1
was prepared from the reaction of PtClMe(SMe2)2 with terpyri-
dine followed by anion methathesis, as previously
described.28,29 2 was prepared in an analogous fashion. The
complexes were fully characterized, including assignment of
1
the H NMR chemical shifts using a combination of 2D COSY
Department of Chemistry, University of Cincinnati, P.O. Box 210172, Cincinnati, OH
45221-0172, USA. E-mail: vikas.shingade@gmail.com
†Electronic supplementary information (ESI) available: Synthesis, characteriz-
ation, NMR and absorption spectra, table of absorption maxima, equations and
calculations, and relevant plots and data fitting summaries. See DOI: 10.1039/
d0dt02190c
Scheme 1 Molecular structures and atom numbering for complexes 1
and 2. (PF6)− is the counterion.
‡Professor William B. Connick passed away in 2018.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Communication
Dalton Transactions
metry of an aggregate. Also, at equivalent concentration range,
for example at 0.09–35 mM, the Δδ35 value for a given a proton
of 1, excepting for H4′, is relatively lower than that for a similar
proton of 2. Besides, the Δδc values for Pt-methyl protons are
relatively higher than the Δδc values for rest of the protons due
to the stronger shielding resulted of the ring current. This
suggests its location near one of the rings of triimine ligand.
Further, it stands to reason that a head-to-head conformation
may not be the preferred way of stacking interaction in these
molecules.
Interestingly, at the lowest concentrations (≤10 mM), the 1H
NMR spectrum of 2 exhibited additional resonances that
display growing dominance with the lowering of concentration
(Fig. S2, ESI†). The identity of these resonances is consistent
with that of the free mbzimpy ligand, which indicates a com-
plete dissociation of mbzimpy from 2. The ligand dissociation
constant was estimated to be in the range of 3 × 10−6 to 3 ×
10−5 M, and it suggest that a stacking structure provides stabi-
lity to 2 from undergoing ligand dissociation in DMSO solu-
tion. Down to 0.09 mM, the spectra of 1 showed no evidence of
free tpy ligand, suggesting that the dissocation constant for
tpy is at least 3 × 103 times lesser than that for mbzimpy; a
contributing factor to this enhanced stability may be the more
optimal bite angle favored by the 6-membered pyridyl rings of
tpy, as compared to the 5-membered rings of mbzimpy.
Because the methyl protons were the most sensitive to the
concentration, methyl proton chemical shifts of 1 and 2
(Fig. 2) were fitted to a model describing the dynamic equili-
brium between a monomer (M) and an aggregate (Mn):
Fig. 1 Apparent molar absorptivities of 0.3 mM solutions of 1 ( ) and 2
(—) in DMSO. Inset: visible absorption spectra of 2 over the 0.3–32 mM
concentration range.
and NOESY experiments. Both complexes dissolve in DMSO to
give pale yellow solutions at low concentrations. However, the
solutions darken and become dark red as the solubility limits
are approached (1, 0.25 M; 2, 0.04 M). To better characterize
this behavior, the electronic absorption spectra were recorded
(Fig. 1 and Table S1 ESI†). The main features of the spectra of
such complexes have previously been discussed.1,10,22,30,31
Notably, characteristic vibronic structure of the triimine-cen-
tered 1(π → π*) transitions (ε ≈ 104 M−1 cm−1) occurs in the
300–370 nm range for 1 and 280–390 nm range for 2. A 1MLCT
[d(Pt) → π*(triimine)] band (ε ≈ 103 M−1 cm−1) occurs at
longer wavelengths, near 408 nm in the spectrum of 1 and at
468 nm in the spectrum of 2. With increasing concentration,
new features appear in the long wavelength region
(450–600 nm) of the spectra of 1 and 2 (Fig. 1 (inset), Fig. S6,
ESI†), which display broadening and increasing molar extinc-
tion coefficients. The non-Beer’s law behavior (or the hyper-
nM Ð Mn
ð1Þ
Fits to models involving a higher order aggregate (Mn, n > 2)
chromicity) of these features is indicative of formation of were inferior. The data were sufficiently well described (R2
>
aggregates. Furthermore, the origin of these features is consist- 0.999) that fits models involving more than one aggregate were
ent with the 1MMLCT transitions.1,13
not justified.
For 2, the resulting dimerization constant (K) obtained
based on the N-CH3 chemical shift was in excellent agreement
with that obtained from the variation in the Pt-CH3 chemical
Concentration dependence of the
shift (Fig. 2, Fig. S5; Tables S3 and S4, ESI†). Interestingly, K
chemical shifts
Each of the 1H NMR chemical resonances of 1 and 2 in DMSO-
d6 are concentration dependent (Fig. S1–S4 ESI†), shifting
monotonically upfield and broadening with increasing concen-
tration. These observations are consistent with a dynamic
equilibrium between the monomer and an aggregate with a
stacked geometry. It should be noted that the H NMR chemi-
for 2 is about ten times higher than that for 1 (Tables 1; and
S2–S4, ESI†), which suggests a greater tendency of 2 to aggre-
gate in solution. This conclusion also is consistent with the
relative solubilities of complexes, as well as the Δδc values over
similar concentration ranges (Scheme S1, ESI†).
1
cal shifts of the free ligands, tpy and mbzimpy, are invariant
over these concentration ranges, indicating that the free
ligands do not aggregate under these conditions. The total
shift (Δδc) for a given a proton over the investigated concen-
tration (c) range (1, 0.09–240 mM; 2, 0.09–35 mM) gives an
indication of the resonance’s sensitivity to concentration
(Scheme S1 ESI†). Interestingly, protons of 1 and 2 show a
Fig. 2 Pt-Methyl proton chemical shifts for 1 and 2 vs. concentration.
wide range of Δδc values, which is consistent with the notion
Solid black line shows best fit to a monomer–dimer dynamic equilibrium
that the deshielding is dependent on the specific stacking geo- model (R2 = 0.9998).
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Communication
Table 1 Dimerization constants determined from 1H NMR chemical
shifta and 560 nm electronic absorptionb data in DMSO-d6
size of the chloro group and the stronger σ-donor properties of
a methyl group, which are expected to enhance metal⋯metal
interaction.34
Complex
1H Resonance
K
a (M−1
)
K
b (M−1
)
ε )
b (M−1 cm−1
1
2
2
δ(Pt-CH3)
δ(Pt-CH3)
δ(N-CH3)
6.2 0.8
59 16
61 16
17
72
—
2
4
354
476
—
2
1
Molecular structure of a dimer
a 95% linear confidence intervals. Upper and lower bounds on lack-of-
fit confidence region are provided in ESI.† b Fitting of electronic
absorption data is provided in ESI.†
The 2D 1H–1H NOESY was employed to elucidate structural
conformations of a dimer of 1 and 2 in DMSO-d6. In order to
estimate distance restraints for weak NOEs,§ the NOESY
experiments also were performed on free ligands. The free
ligand NOE spectra (Fig. S7 and S9 ESI†) exhibit no signs of
NOEs between protons separated by ≥4.6 Å. The spectra also
suggest that free ligands have non-planar geometries in
solutions.
For the structural analysis of 1 and 2 dimers, the focus was
given in particular to the long-range (weak) NOEs. The dis-
crimination between intra- and intermolecular NOEs was
based upon the distance restraint of 5.0 Å above which any
sign of NOE was assessed for intermolecular interaction.
Furthermore, the absence of some key NOE cross-peaks
offered undoubted assignments of some long-range NOEs to
inter- rather than intramolecular protons. For example, the
NOE spectrum of 1 (Fig. S8 ESI†) exhibits no sign of NOE
between Pt-CH3⋯H(4, 4′′) protons separated by distance (d) of
∼6.0 Å in monomer, however, medium NOE cross-peak
Fitting of the electronic absorption data using the same
model (eqn (1)),1,11 yielded comparable dimerization constants
(K), 17 2 M−1 for 1 and 72 4 M−1 for 2 (Table S5, Fig. S6
ESI†). This analysis also yields estimates for the molar absorp-
tivities of the dimer (1, 354 2 M−1 cm−1; 2, 476 1 M−1
cm−1), which are decidedly less than values observed for the
1MMLCT band of authentic Pt(tpy) dimers (2000–4000 M−1
cm−1),2,32,33 which consist of two covalently bridged Pt(tpy)
units resulting in a dimer with approximate C2v symmetry.
Faced with a similar discrepancy in the case of Pt(tpy)Cl+
dimerization in 0.1 M aqueous NaCl, Bailey et al. suggested
the possibility that a second dimer also forms in solution,
namely one supported by tpy π−π interactions but lacking
the Pt⋯Pt interactions necessary to give intensity to
a
1MMLCT band.1 Making a similar assumption of ε(MMLCT)
= 2000 M−1 cm−1 for the metal–metal (MM) dimer and
applying this treatment to our data (Table S6 ESI†), we
obtained estimates of the dimerization constants for the
ligand–ligand (ππ), and metal–metal (MM) supported dimers
(Kππ and KMM: 1, 13.7 M−1 and 3.0 M−1; 2 54.7 M−1 and 17.1
M−1). However, it should be emphasized that both the NMR
and UV-visible data are adequately modeled according to a
single dimer model (eqn (1)), and the justification for the
two dimer model rests entirely on expectations concerning
the molar absorptivity. Therefore, another possible expla-
nation for the comparatively low molar absorptivities derived
from the single-dimer model is that the oscillator strength of
the MMLCT band of C2h-symmetric head-to-tail dimers is
intrinsically weaker.
was observed between Pt-CH3⋯H(3′,
(d, ≥7.0 Å), which
4′, 5′)
suggests
protons. Similarly, a weak NOE cross-peak was observed
between Pt-CH3⋯H(3, (d, ∼6.2 Å). These results
intermolecular
interactions
between
these
3′′)
strongly indicate a head-to-tail fashioned interaction in a
dimer of 1.
In the case of 2, fewer NOE contacts helped to illustrate the
structure more clearly (Fig. 3). Similar to the 1, no signs of
NOEs were observed between Pt-CH3⋯H(6′, 6′′, 7′, 7′′) (d, >5.8 Å),
H(7′, 7′′)⋯H4 (d, ∼7.7 Å) and H(6′/6′′)⋯H(3/5,
Whereas, weak NOEs were observed between protons Pt-
(d, >7.5 Å).
4)
Comparison of the dimerization constants for 1 and 2 with
those previously determined15 K values for 1 and related com-
pounds were expected to provide insights into the factors influ-
encing dimerization. For example, K for 1 of 26(1) × 103 M−1 in
D2O by 1H NMR and 10(8) × 103 M−1 in a buffer (1 mM
Phosphate + 0.1 M NaCl(aq), pH = 7.0) by UV-vis absorption
techniques reveals a large difference between K values. This
difference is consistent with a notion that a lower dielectric
constant of DMSO disfavors formation of dicationic dimer
whereas intermolecular hydrophobic interactions favor stack-
ing in water. On the other hand, a survey of K values1,11,13 of Pt
(tpy)Cl+, 4(2) × 103 M−1 in water and 3(2) × 103 M−1 in 0.1 M
NaClaq, suggests that a substitution of chloro- ligand by
methyl- group increases a tendency of Pt-tpy complex to aggre-
Fig. 3 2D 1H–1H NOESY spectrum of 35 mM DMSO-d6 solution of 2.
gate in solution. This observation is in accord with the smaller Inset: partial 2D NOESY of 2.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Communication
CH3⋯H(3/5,
Dalton Transactions
(d, ≥7.0 Å), Pt-CH3⋯N-CH3 (d, ≥7.0 Å), and NMR. This work was supported by the National Science
4)
N-CH3⋯H(4′, 4′′, 5′, 5′′) (d, ∼6.0 Å). These NOEs strongly indicate Foundation (NSF Grants CHE-1152853 and CHE-1566438).
head-to-tail fashioned interactions of molecules in a dimer of
2 as well.
The NOE data of 2 was further analyzed by applying a
Notes and references
dimer model in various stacking arrangement of molecules.
For this analysis, we used 5 Å as the upper bound of distance
restraints for any sign of NOE contacts. The modeling results
§NOE strength (distance restraint): strong (<2.5 Å), medium (<3.7 Å), weak
(<5 Å).
indicate that all NOE constraints of 2 are well satisfied in an
anti-parallel displaced conformation of a dimer with mole-
cules rocking along x- and y-directions in a symmetrical
fashion within a molecular plane. Moreover, the model shows
dominance of ligand⋯ligand interactions with no appreciable
Pt⋯Pt contact. On the other hand, as expected, a model with
face-to-face interactions violates NOE constraints. Thus, our
NOE based structural model implicitly supports the formation
of 2 dimer in a head-to-tail geometry.
Similar to 2, the NOE data of 1 are consistent with an anti-
parallel arrangement of molecules in a dimer. However, the
detection of fewer intermolecular NOE contacts limits our
efforts to show that ligand⋯ligand interactions dominate in a
dimer of 1 as well. Fitting results of 560 nm absorptions (vide
supra), however, are more conclusive in this regard. From the
spectroscopy and fitting results discussed above, it is apparent
that the aggregation is relatively stronger in 2, and the struc-
ture of a dimer is controlled by multiple types of inter-
actions,35 mainly ligand⋯ligand and Pt⋯Pt interactions. To
further support this assessment we note that the 13C chemical
1 J. A. Bailey, M. G. Hill, R. E. Marsh, V. M. Miskowski,
W. P. Schaefer and H. B. Gray, Inorg. Chem., 1995, 34, 4591–
4599.
2 J. A. Bailey, V. M. Miskowski and H. B. Gray, Inorg. Chem.,
1993, 32, 369–370.
3 V. M. Miskowski and V. H. Houlding, Inorg. Chem., 1989,
28, 1529–1533.
4 V. M. Miskowski, V. H. Houlding, C. M. Che and Y. Wang,
Inorg. Chem., 1993, 32, 2518–2524.
5 V. H. Houlding and V. M. Miskowski, Coord. Chem. Rev.,
1991, 111, 145–152.
6 I. S. Sigal and H. B. Gray, J. Am. Chem. Soc., 1981, 103,
2220–2225.
7 K. R. Mann, J. G. Gordon II and H. B. Gray, J. Am. Chem.
Soc., 1975, 97, 3553–3555.
8 V. W.-W. Yam, K. M.-C. Wong and N. Zhu, J. Am. Chem.
Soc., 2002, 124, 6506–6507.
9 K. M.-C. Wong and V. W.-W. Yam, Acc. Chem. Res., 2011, 44,
424–434.
shift for the Pt-CH3 group of 2 was observed relatively upfield 10 G. Arena, G. Calogero, S. Campagna, L. M. Scolaro,
due to the ring current (δ, −25.6 ppm for 2 at 29.4 mM and
−5.14 ppm for 1 at 240 mM) at nearly equivalent dimer :
V. Ricevuto and R. Romeo, Inorg. Chem., 1998, 37, 2763–
2769.
monomer ratios of the complexes in dmso-d6. On the other 11 K. W. Jennette, J. T. Gill, J. A. Sadownick and S. J. Lippard,
hand, metal⋯metal interaction is noticeably weak in 1 and 2 J. Am. Chem. Soc., 1976, 98, 6159–6168.
dimers. All these results suggest that the ligand⋯ligand inter- 12 R. Romeo, L. M. Scolaro, M. R. Plutino and
action majorly drives aggregation in our complexes, and it
A. Albinati, J. Organomet. Chem., 2000, 593–594, 403–
shapes a structural geometry of a dimer.
408.
In summary, the 2D NOE spectra of 1 and 2 provide the 13 M. G. Hill, J. A. Bailey, V. M. Miskowski and H. B. Gray,
first definitive evidence of a head-to-tail dimer formation in Inorg. Chem., 1996, 35, 4585–4590.
fluid solution. Further, fittings of the NOE and the electronic 14 R. Romeo, G. Arena, L. Monsù Scolaro and
absorption data support the notion that the forces that are sta-
bilizing these dimers are dominated by ligand⋯ligand inter-
R. F. Pasternack, Nuovo Cimento Soc. Ital. Fis., D, 1994, 16D,
1523–1528.
actions in lower dielectric solvents. In the case of 2, the inter- 15 G. Arena, L. M. Scolaro, R. F. Pasternack and R. Romeo,
molecular interactions stabilizing the dimer are decidedly Inorg. Chem., 1995, 34, 2994–3002.
stronger. Also, a dimeric structure gives chemical stability to 2 16 A. R. Petersen, R. A. Taylor, I. Vicente-Hernandez,
from undergoing decomposition.
J. Heinzer, A. J. P. White and G. J. P. Britovsek,
Organometallics, 2014, 33, 1453–1461.
17 A. Y.-Y. Tam, W. H. Lam, K. M.-C. Wong, N. Zhu and V.
W.-W. Yam, Chem. – Eur. J., 2008, 14, 4562–4576.
18 D. R. McMillin and J. J. Moore, Coord. Chem. Rev., 2002,
229, 113–121.
Conflicts of interest
There are no conflicts to declare.
19 I. Eryazici, C. N. Moorefield and G. R. Newkome, Chem.
Rev., 2008, 108, 1834–1895.
20 K. M.-C. Wong, W.-S. Tang, B. W.-K. Chu, N. Zhu and V.
W.-W. Yam, Organometallics, 2004, 23, 3459–3465.
Acknowledgements
We thank Dr Keyang Ding (NMR facility manager at the 21 A. J. Goshe, I. M. Steele and B. Bosnich, J. Am. Chem. Soc.,
University of Cincinnati) for expert technical assistance with
2003, 125, 444–451.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Communication
22 L. J. Grove, J. M. Rennekamp, H. Jude and W. B. Connick, 28 A. W. Addison, S. Burman, C. G. Wahlgren, O. A. Rajan,
J. Am. Chem. Soc., 2004, 126, 1594–1595.
23 L. J. Grove, A. G. Oliver, J. A. Krause and W. B. Connick,
Inorg. Chem., 2008, 47, 1408–1410.
T. M. Rowe and E. Sinn, J. Chem. Soc., Dalton Trans., 1987,
2621–2630.
29 G. S. Hill, M. J. Irwin, C. J. Levy, L. M. Rendina and
R. J. Puddephatt, Inorg. Synth., 1998, 32, 149–153.
24 E. J. Rivera, C. Barbosa, R. Torres, L. Grove, S. Taylor,
W. B. Connick, A. Clearfield and J. L. Colon, J. Mater. 30 I. Mathew and W. Sun, Dalton Trans., 2010, 39, 5885–5898.
Chem., 2011, 21, 15899–15902.
25 J. R. Kumpfer, S. D. Taylor, W. B. Connick and
31 V. M. Shingade, L. J. Grove and W. B. Connick, Dalton
Trans., 2020, 49, 9651–9661.
S. J. Rowan, J. Mater. Chem., 2012, 22, 14196– 32 E. M. A. Ratilla, B. K. Scott, M. S. Moxness and
14204. N. M. Kostic, Inorg. Chem., 1990, 29, 918–926.
26 V. M. Shingade, Ph.D. Dissertation, University of 33 K. M.-C. Wong, N. Zhu and V. W.-W. Yam, Chem. Commun.,
Cincinnati, 2016. 2006, 3441–3443.
27 S. D. Taylor, A. E. Norton, R. T. Hart, M. K. Abdolmaleki, 34 W. B. Connick, R. E. Marsh, W. P. Schaefer and H. B. Gray,
J. A. Krause and W. B. Connick, Chem. Commun., 2013, 49,
9161–9163.
Inorg. Chem., 1997, 36, 913–922.
35 C. Janiak, J. Chem. Soc., Dalton Trans., 2000, 3885–3896.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.