Synthesis and Properties of Acridine and Acridinium Dye Functionalized Bis(terpyridine) Ruthenium(II) Complexes

DOI: 10.1002/ejoc.201800257

Full Paper

Chromophoric Ruthenium Complexes

Synthesis and Properties of Acridine and Acridinium Dye

Functionalized Bis(terpyridine) Ruthenium(II) Complexes

[

a]

[c]

[b]

[c]

Jens Eberhard,* Katrin Peuntinger, Roland Fröhlich, Dirk M. Guldi, and

Jochen Mattay*^[

a]

+

Abstract: We present first principle studies on the rational de- (Acr/MeAcr ) as well as the interconnecting linker unit (none,

+

sign of an acridine/N-methylacridinium dye (Acr/MeAcr ) substi- 4-phenyl, or 5-thien-2-yl) and investigated these aspects by

tuted terpyridine ligand to investigate if these chromophores steady-state/time-resolved spectroscopy, cyclic voltammetry,

can act as triplet-energy storage units in bis(terpyridine) ruth- X-ray structure analysis, and DFT calculations.

enium(II) complexes. We studied the influence of the dye form

1

. Introduction

successively separates the energy levels and increases the acti-

vation barrier of crossing to the MC state, respectively.

3

[1c,6]

Bis(terpyridine)ruthenium(II) complexes – [Ru(tpy) ]²– consti-

tute an important class of metal to ligand charge transfer

+

2

A less intuitive alternative is the use of organic dyes that

feature a matching triplet level leading to the so-called multi-

(

MLCT) dyes due to the very efficient population of their triplet

[4,7]

chromophore approach depicted in Scheme 1.

3

[1]

2+

excited state ( MLCT) after excitation. For a given [Ru(tpy)₂]

complex the luminescence lifetime is often the primary bench-

mark regarding its suitability for photosensitizing applica-

[

1c,2]

2+

tions.

However, most of the simple [Ru(tpy)₂] complexes

are non-luminescent at room temperature indicating a fast re-

3

laxation dynamics. To this end, rapid deactivation of the MLCT

3

state by metal-centered triplet ( MC) states, close in energy to

3

[1e,3]

the MLCT state, is inferred.

In this regard, it is important to

internalize that the effective energy gap (ΔE) between the

states is governed by the activation barrier and not the differ-

[

1c]

ence in zero-zero energy (E )

of the states minima on the

00

Scheme 1. Simplified Jabłoński diagram for systems with at least one addi-

tional ligand-centred triplet ( LC) state in energetic proximity to the MLCT

state of interest.

potential energy hypersurface.

Tremendous efforts have been made to improve this crucial

bottleneck in the photophysical properties of [Ru(tpy) ] as

3

2

+

2

[

1c,4]

summarized in several reviews.

include extending π-conjugation through introducing electron

Briefly, potential alternatives

The natural choice of anthracene (Ant) – the prototype poly-

cyclic aromatic hydrocarbon – as storage unit was subject of

[

5]

withdrawing groups (EWGs), on one hand, and the more gen-

eral approach of changing the field strength, i.e. by structural

changes of the ligands, on the other hand. The latter is meant

to realize a more octahedral coordination environment, which

detailed studies but the approach was unsuccessful for the

2+

[

Ru(tpy)₂] complex due to the lower lying triplet excited state

3

[8]

of anthracene ( Ant). A breakthrough in terms of the rational

linker design was the work by Hanan et al.

[4,7d,7e]

They maxi-

mized the π-conjugation through the terpyridine building block

3

and subsequent lowered the energy of the MLCT state to

[

a] Organische Chemie I, Fakultät für Chemie, Universität Bielefeld,

Universitätsstr. 25, 33501 Bielefeld, Germany

3

match the Ant state. Despite the great success with pyrimidyl-

(pm) or triazinyl-type of linkers in these designs, the synthetic

E-mail: mattay@uni-bielefeld.de

http://www.uni-bielefeld.de/chemie/emeriti/oc1-mattay/

b] Röntgenstrukturanalyse, Organisch-Chemisches Institut,

Westfälische Wilhelms-Universität Münster,

Corrensstr. 40, 48149 Münster, Germany

c] Physikalische Chemie I, Department of Chemistry and Pharmacy and

Interdisciplinary Center for Molecular Materials,

[7d,7e]

efforts are cumbersome.

Synthetically more accessible are

[

imidazolyl-based systems, as published by Baitalik et al., ranging

from lophine-based terpyridine assemblies to multichromo-

phoric designs by use of rigidified imidazolyl-connected anthra-

[9]

quinone or pyrene units.

Friedrich-Alexander-Universität Erlangen-Nürnberg,

Egerlandstr. 3, 91058 Erlangen, Germany

Supporting information and ORCID(s) from the author(s) for this article are

available on the WWW under https://doi.org/10.1002/ejoc.201800257.

Considering our own interest in photochemistry of func-

tional dyes and the availability of terpyridine building blocks,

we thought it would be reasonable to evaluate the multi-

[

10]

Eur. J. Org. Chem. 2018, 2682–2700

2682

Full Paper

chromophore approach with the highly interesting class of

+

[11]

1

0-methylacridinium dyes (MeAcr ).

Acridinium dyes attract considerable interest for their subtle

substituent-dependent charge transfer (CT) state. This renders

acridinium compounds attractive for basic studies on charge

[

12]

separation (CS).

light by the contrary opinion between Fukuzumi and Benniston

9-Arylacridinium dyes were put in the spot-

[

13]

+ [14]

et al.

Thus, several acridinium dyes, such as MeAcr ,

9-

and deriv-

were extensively studied in photocatalytic

redox-)applications and are currently experiencing a renais-

sance due to availability of cheap but powerful LED light sour-

+

[15]

+[16]

Phenyl-MeAcr , and especially 9-Mesity-MeAcr

[

17]

atives thereof

(

[

18]

ces.

The basic photophysical properties of 9-arylacridinium dyes

are well-known and their lowest triplet excited state level (T )

1

[

12a,19]

is generally found between 1.9–2.0 eV.

A good photo-

physical understanding exists also for acridine derivatives with,

for example, triplet excited states at energies comparable to

+

[19,20]

those of MeAcr derivatives (e.g. 1.88–1.95 eV).

Furthermore, the acridinium core is known for its reactive 9-

position, which has been explored in the context of pseudo-

[

21]

base formation

tems.

and switchability, i.e. in supramolecular sys-

[

22]

+

In regard to the cationic nature of the MeAcr dyes, studies

of quaternized pyridine acceptor units, such as methylviologen

MV ) in particular, are relevant to this work.

the methylene bridged heteroleptic [Ru(ttpy)(ttpy-MV)] sys-

tem had a charge-separated (CS) state at ca. 1.63 eV which

extends the overall excited state lifetime to 600 ns at

2+

[1e,23]

(

For instance

4

+

Scheme 2. Structures of compounds 1–5 and labeling scheme for ¹H NMR

spectroscopic assignments.

[

1e,23a]

200 K.

In addition to the aforementioned findings, our concept of

testing Acr/MeAcr dyes in terms of multichromophic behav- (see 28, Table 2).^[²⁶^]The geometric properties, viz. the conju-

iour is based on several incentives: On one hand, some of the gated character of the thienyl-group, was consequently verified

design principles towards long-lived photosensitizers are by reviewing the available X-ray structure data on thienyl-ter-

+

[

27]

fulfilled, namely i) π-conjugating through the linker, ii) triplet pyridines,

literature description about its unique proper-

3

[26,28]

excited state energy matching relative to the MLCT level of ties,

and own DFT calculations (see computational section).

Last but not least the smaller homologic systems are of high

and iii) placing an EWG by methylation of the importance to this work, namely 4′-pyridyl and 4′-pyridinium

acridine moiety to yield the electrophilic acridinium ion but terpyridines (py/py -tpy) with their corresponding metal ion

2+

π-extended [Ru(tpy) ] complexes (cf. [Ru(ttpy)₂] E₀₀(77 K)

2

[

2,24]

1.98 eV),

+

[

29]

[29b,29d,30]

[29b,30b]

with a potential drawback of photoinduced electron transfer complexes, such as, Fe, Ru

and Os.

4′-py/py -tpy complexes can be seen as a well investigated sub-

To characterize the influence of the cationic dye form, we category in the field of terpyridine complexes. The potentially

Therefore,

–

+

(

p-e T).

+

also synthesized complexes 1 and 4 with an acridine unit water-soluble 4-py -tpy complexes were focus for further

Scheme 2). These complexes are not only meant references research and also investigated by DFT calculations.^[³⁰^d^,³¹^]In re-

due to the similar triplet level of acridines (see above), but also spect to these reports, our study should be regarded as the

to shed light on possible p-e T phenomena.

Notably, in a next logical step towards the larger N-heterocyclic homologues.

conjugated assembly of the building blocks the electronic prop- We investigated some of these complexes for the suitability to

(

–

[25]

[

32]

erties and/or triplet levels of the assembly might be altered in act as sensitizer for singlet oxygen previously,

while the

comparison to the isolated building blocks.

synthesis and a more in-depth analysis of the ruthenium(II)

complexes is reported here.

On the other hand, as the linker-design might be crucial for

success, we focused on a thiophene unit as easy-to-implement

interconnecting group. The thienyl-group is known to contrib-

2

+

2. Results and Discussion

ute favorably to the photophysical properties of [Ru(tpy)₂] de-

rivatives: Albeit being electron-rich it increases the room tem-

2.1. Synthesis

[

26]

perature (RT) luminescence lifetime. A notable disadvantage

III

II

of this spacer is the lower redox potential of the Ru /Ru couple The synthesis for the different ligands was either straight for-

due to the electron donating character of the thienyl moiety ward, i.e. for AT ligand 8 (Scheme 3) or relied on aldehyde pro-

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2683

Full Paper

Scheme 3. Synthetic route to ligand 8.

Scheme 4. Synthetic route to ligands 19 and 20.

tection/deprotections step, starting from commercial available use of a MEM-protected 9-acridone (22) as electrophile

-bromobenzaldehyde (9) and thiophene-2-carbaldehyde (10), (Scheme 5). Up on acidic work up, necessary to achieve elimina-

4

to enable the well documented reactions of metalated linker tion of water of the formed 9-acridinol intermediate, the MEM-

blocks, such as, 11 and 12, with electrophilic acridin-9-one de- and the acetal protecting groups were cleaved in an one-pot

rivatives or 10-alkyl-9-aridinium salts as a key feature of all syn- procedure leading directly to the desired acridine-carbaldehyde

[

12a,15b,21e,22,33]

thetic routes presented here (Scheme 4).

The 23.

The synthetic strategies for the synthesis of 2,2′:6′,2′′-ter-

13, 84 %; 14, 68 %–99 %) while deprotection was possible with pyridines have been reviewed before^[¹^a^,¹^b^,³⁴^]Due to the facile

acetal-protected intermediates were obtained in good yields

(

excellent yields due to simple work-up (yield 90 %–93 %).

synthesis of aldehydes we followed a Kröhnke-type ap-

[

35]

For the synthesis of the non-methylated but thienyl-linked proach

although the isolated yield of this multi-condensa-

ligand ATT 24, a slightly different approach was necessary by tion reaction is often only moderate. Terpyridines 8, 17, 18, and

Scheme 5. Synthetic route to the ATT ligand (24) via aldehyde 23.

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2684

Full Paper

2

4 were isolated in 21 % – 36 % yield, which renders this step complexation. For example, the 3′/5′-protons on the 4′-substi-

as the bottle-neck of the synthetic route. Attempts to optimize tuted tpy ligand (labeled B3′) experience a downfield shift,

the reaction conditions for the terpyridine synthesis were un- while the 6/6′′-protons (labeled A6) shift upfield. Acridinium

successful.

and acridine heterocycles are easily distinguished by the gen-

+

The oxidation of acridanes 17 and 18 to the corresponding eral downfield shift of the signals for the MeAcr moiety (see

acridinium salts 19 and 20 was the final step of the synthetic Table S1) and the unique resonance of the methyl group usually

sequence and may be considered as deprotection. Hereby, the found between 4.93–5.04 ppm.

acridane moiety served as kind of a protecting group for the

highly basic conditions in Kröhnke-type condensation reac-

2

.2. X-ray Structure Elucidation

tions. Otherwise nucleophiles, such as, hydroxide ions or alco-

hols (viz. solvent) would add to the 9-position of the acridinium

moiety.

Crystals of the [Ru(MeAT)₂]⁴⁺complex (2) suitable for X-ray

structure elucidation were grown from a MeCN solution of

chromatographic purified material by slow diethyl ether vapor

diffusion. Except for 2, only crystals of poor diffraction quality

were obtained.

The molecular structure of 2 is depicted in Figure 1 and se-

lected structural parameters are compiled in Table 1. Structural

parameters of 2 are compared to the published X-ray structure

The bis(terpyridine)ruthenium(II) complexes were prepared

by standard methods with [Ru(dmso) Cl ] as the reagent-of-

4

2

[

1c,23a]

choice.

The desired complexes could usually be isolated

in moderate to good yields (41 % – 76 %; except 3, 17 %).

For the synthesis of the [Ru(MeAT)₂]⁴complex 2 we selected

+

[

7d,28,36]

a “chemistry-on-the complex” approach

by N-methyl-

ation of 1. Although it is known that N-methylation of uncom-

2+

[30c]

4+

[31a]

of [Ru(py-tpy)₂] (29)

and [Ru(Et-py-tpy)₂] (30).

[

37]

plexed py-tpy ligands can be selective, the late-stage “chem-

istry-on-the complex” strategy was more appealing to us. Re-

cently, acridinone/acridine chemistry was even successfully im-

plemented in a multi-step “chemistry-on-the complex” reaction

2

+

[38]

sequence on [Ru(bpy) X] derivatives. The N-methylation of

2

2+

the different regioisomers of [Ru(py-tpy)₂] has been reported

previously but mixtures of non-, mono-, and the desired bis-

alkylated product were obtained.

[

29b,29c]

These remarks let us

consider the use of an autoclave with subsequent reaction tem-

peratures of 110 °C (b.p. MeI 42 °C). To this end, a clean conver-

sion was achieved and formation of any unidentified yellow by-

[

29c]

products

was prevented.

Figure 1. Structure of the [Ru(MeAT)

[I ]0.5·6 MeCN with thermal ellipsoids plotted at the 33 % probability level.

3

Important atoms are labeled while hydrogen atoms, solvent molecules, and

counterions are omitted for clarity.

2

]⁴⁺cation 2 in [Ru(MeAT)][PF

] -

6 3.5

More detailed background information to the synthetic route

and considerations thereof, respectively, are given in the Sup-

porting Information.

The complexes were fully characterized by means of NMR

and MS analysis (see Supporting Information). The success of

complex formation was corroborated by H NMR spectroscopy N8A–Ru–N8B axis (denoted c-axis) by shorter bond lengths and

by virtue of characteristic shifts of the proton resonances upon almost linear coordination (177.0°). Both parameters are in

The primary coordination axis of 2 is identified along the

1

Table 1. Comparison of selected structural parameters (distances [Å], angles ∠ [°]) of complexes 2 with structural related compounds as determined by X-ray

crystallography.

]⁴⁺

[Ru(py-tpy)

(29)

]²⁺

[Ru(Et-py-tpy)

(30)

]⁴⁺

[Ru(Ant-tpy)(Cl-pm-tpy)]

(36)

2+

[

(

Ru(MeAT)

2)

2

[

30c]

[31a]

[7e]

c-axis^[^a^]

b-axis

a-axis

1.973(5)

.981(5)

2.059(5)

.076(5)

2.053(5)

.068(6)

1.483(8)

.498(8)

1.494(9)

.493(8)

1.982(3)

2.082(3)

1.491(7)

–

1.976(3)

1.981(3)

2.075(3)

2.082(3)

2.076(4)

2.088(4)

1.476(6)

1.971(7)

1.977(7)

2.060(8)

2.072(8)

2.074(8)

2.090(7)

1.48(1)^[

1.49(1)^[

–

1

2

b]

Ctpy–Caryl

c]

1

+

N –Calkyl

1.49(2)

1

∠

ꢀ

c-axis

b-axis

a-axis

177.0(2)

157.7(2)

158.3(2)

61

180.0(1)

157.0(1)

35

177.8(1)

157.6(1)

157.8(1)

48

178.1(3)

158.2(3)

158.8(3)

[b]

0

76^[

c]

73

57

[

a] c-axis = principle coordination axis. [b] At pyrimidyl [pm] unit. [c] At 9-anthryl (Ant) unit.

Eur. J. Org. Chem. 2018, 2682–2700 www.eurjoc.org

2685

Full Paper

+

agreement with the bond length and angle of 4′-py/py -tpy 29

and 30.

In 2, the four dative bond lengths are marginally shorter in

comparison to 30 (0.006–0.020 Å), but the coordination angles

The nonlinearity along the coordination axis is passed to the are equal within 0.5° (see Table 1). For an in-depth discussion

II

substituents (measured as ∠ N26A–Ru–N26B), which is rather of the bond lengths and angles at the Ru center we refer to

pronounced in the case of 2 (163.9°), giving the overall curved the comparison of DFT calculated and experimental values at a

topography (see Figure 2). A similar but weaker bending was later stage.

also found for 29 (169.1°) and 30 (171.5°). Such a phenomenon

The dihedral angle ꢀ and the C_t_p_y–C_a_r_y_ldistance between

is common for compounds without crystallographic symmetry the MeAcr⁺moiety and the terpyridine unit agree well with

[

31a]

constrains.

Severe distortions were also found for 29 in a 1- examples incorporating the related 4′-(9-anthryl) residue, such

[

39]

2+

[7e]

D coordination polymer with AgNO (163.2°). A plausible ori- as [Ru(Ant-tpy)(Cl-pm-tpy)] (36). Thus, the similar orthogo-

3

gin was the asymmetry of anions and solvent within the lattice nal orientation in 2 (exp. 61°/73°) due to the steric demand of

+

which seems also applicable to our case due to the mixed anion the 1,8-H-atoms compared to the smaller homologous py/py -

occupation in the crystal lattice.

tpy complexes 29 and 30 does not surprise.

No noteworthy intermolecular packing features were found

for 2 in regard to the X-ray packing features reported for other

[

31a,40]

terpyridine metal complexes.

Overall, the packing in the

solid state shows only weak interactions between the outer pyr-

idyl rings (4.051 Å), but no interpenetration of the individual

4

+

[

Ru(MeAT)₂] cations (see Figure S118).

2.3. Electrochemical Analysis

All of the ruthenium(II) terpyridine complexes were electro-

chemically investigated in MeCN between –2.0 and +1.5 V vs.

+

III

II

Fc /Fc. In each case, a fully reversible one-electron Ru /Ru

redox process is observed in the anodic scans. The analysis of

the anodic scan in electrochemical investigations is particularly

important as the electron donor/acceptor-character of the 4′-

substituent is known to correlate with the observed potential

II

Figure 2. Packing diagram for 2 in the unit cell. The Ru center is plotted with

3 % van der Waals radius while hydrogen atoms, solvent molecules, and

3

III

II

[5,36a,41]

of the Ru /Ru redox couple

an estimate for the HOMO energy.

and may therefore serve as

counterions are omitted for clarity: a.) view along the a-axis; b.) turned 20°

to emphasize perspective view.

Table 2. Electrochemical data of compounds 1–5 and some reference systems for comparison.

1/2 [V] and tentative assignment

E

III

II

Compound

Ru /Ru

N-heterocyl. subst.

Terpyridine ligand

Ref.

centered reductions

PhMeAcr⁺

–

–0.93 (rev.)

–

–0.77 (rev.)

–0.89 (rev.)

–

this work

[

Ru(AT)

Ru(MeAT)

Ru(MeAPhT)

2

]²⁺

1

2

3

4

0.95 (rev.)

1.01 (rev.)

0.88 (rev.)

0.92 ± 0.02 (rev.

–1.55 (rev.)

–1.14 (rev.)

n.d.

–1.53 (rev.)

–1.15 P

–1.77 (rev.)

–1.34 (rev.)

n.d.

–1.71 (rev.)

–1.24 P

–1.83 (n.d.)

–1.90 (rev.)

n.d.

]⁴⁺

2

]⁴⁺

2

+

Ru(ATT)

2

]

–

Ru(MeATT)

2

]⁴⁺

5

–0.79 (rev.)

c

(irrev.)

c

(irrev.)

this work

2

+

[a]

[1e,29b,36a,41,46]

Ru(tpy)

Ru(ttpy)

2

]

25

27

28

29

30

31

32

33

34

35

37

)

–

–1.64 ± 0.03 (rev.

)

–1.85 ± 0.08 (rev.

–1.89 ± 0.08 (rev.

)

_]₂+

–2.34 (rev.^[^a^]

–2.24

[1e,2,41,46b,46c]

2

0.87 ± 0.03 (rev.

–1.64 ± 0.02 (rev.

)

_]2

+

₀_.₈₃b

_–₁_.₆₁b

[b]

[26]

Ru(th-tpy)

Ru(py-tpy)

Ru(Et-py-tpy)

Ru(Me-py-tpy)

2

–1.82

2+

[a]

–1.80 (rev.[a]₎

[29b]

[31a]

[29b]

[41]

2

]

0.95 (rev.

1.01 (rev.

1.03 (rev.

)

–1.54 (rev.

–1.56 (rev.

)

_]₄+

4

[a]

–1.09 (rev.^[^a^])

[a]

–1.79 (quasi-rev.^[^a^])

2

+

–1.06, –1.16 (rev.[a]₎

[a]

2

_]4

]

–1.79 (rev. )

+

– [a]

[a]

Ru(H

3

TP-tpy)

TP-ph-tpy)

2

> 1.22 (n.d.)

–1.16 (2 e

)

n.d.

_]4

+

[a]

–1.29/-1.36 (rev.[a]₎

–1.69 (rev.[a]₎

–1.92 (rev.[a]₎

_–₁_.₇₀₍_i_r_r_e_v_.[a]

)

[41]

3

2

0.91 (rev.

0.89 (rev.

)

4+

[a]a

[a]

–1.59 (quasi-rev.[a]₎

[23a]

[7d]

Ru(ttpy)(ttpy-MV)]

Ru(Ant-pm-tpy)

Ru(tpy-HImzPy)

)

–0.74 (rev. )

2

+

0.97 (rev.[a]₎

[a]

2

]

–

–1.46 (rev.

)

2

+

0.96 (quasi-rev.^[^a^])

–1.58 (quasi-rev.^[^a^])

–1.81 (quasi-rev.^[^a^])

–2.07 (irrev.^[^a^]

)

[9c]

2

Reduction and oxidation half-potentials E1/2 [= (Epeak,cathodic + Epeak,anodic)/2] and peak separation values (ΔE = Epeak,cathodic – Epeak,anodic) were obtained from

+

square wave and cyclic voltammetry (forward scan) respectively. All data is referenced on the Fc /Fc couple in MeCN with 0.1

6

M TBAPF as supporting

+

electrolyte at 21 ± 1 °C. Sweep rates were tested between 50–500 mV/s. Reversibility was estimated by the peak separation value ΔE compared to the Fc /Fc

internal reference. Conversion from/to other standard reference electrodes was achieved by the conversion tables.^[Values with a ±-sign are average values

of the cited references and the resulting variance. Abbrv.: n.d. = not determined. [a]As stated in cited reference(s). [b]Peak separations are stated to be between

47]

6

0–100 mV.

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2686

Full Paper

Here, the potential shifts from 0.92 V for pristine [Ru(tpy)₂]²⁺thienyl-linker does not establish itself. The MeAcr⁺reduction

(

25) to 0.95 V for 1. Upon quaternization by methylation, a dis- potential in 3, however, is shifted less pronounced (+40 mV) to

+

tinctive shift to about 1.0 V is observed for 2. These findings less negative potentials compared to PhMeAcr indicating

are in perfect agreement with the reported data for the py/py - again the electronically more decoupled properties of the linker

tpy complexes 30 and 31, respectively (Table 2).

+

[

31a]

group (Table 2, Figure 3).

The complexes containing a linker unit (3 – 5) possess lower

Changes are also apparent for the terpyridine centered re-

oxidation potentials and compare well with the redox proper- ductions below –1 V. For 2 the reductions are shifted by 410 mV

ties of their parent tolyl- and thienyl-terpyridine model com- more positive, to –1.14 and –1.34 V, compared to pristine

plexes, that is, [Ru(ttpy)₂]²(27) and [Ru(th-tpy)₂] (28), respec- [Ru(tpy)₂] (25). Unfortunately no clear statements can be

tively.

+

2+

+

made for MeAcr -containing complexes 3 and 5 because only

This result is rather unexpected as it infers that the electro- poorly resolved features due to a high cathodic peak current

chemical perturbation – viz. the EWG character of the acridin- (Ip,c) were observed at –1.15 V and –1.24 V following the first

+

ium moiety – towards the metal center is weak. It seems that MeAcr reduction at –0.79 ± 0.01 V. It is known that acridinium

+

effect stemming from the Acr/MeAcr unit are counterbalanced ions may give irreversible peak shapes when the anionic species

by those from the spacer. The oxidation in 4 and 5 is after all is formed from the radical intermediate at higher cathodic po-

slightly raised to 0.88 V from 0.83 V in [Ru(th-tpy)₂]²(28). A tential.

+

[42a]

The anionic species subsequently abstracts rapidly a

more pronounced influence of the cationic group on the linker- proton from the solvent to yield the acridane species, which

terpyridine assembly was expected due to the reported data may then adsorb on the electrode due to limited solubility in

4

+

[41]

[42a]

for [Ru(H TP-ph-tpy) ] derivative 33, namely an increase of polar solvents.

Therefore, the potential window was limited

3

2

III

II

+

about +30 mV for the Ru /Ru redox potential with respect to the reversible MeAcr reduction.

to 27 (Table 2). Moreover, terpyridine systems containing an

In contrast, acridine complexes 1 and 4 show the expected

innocent polycyclic aromatic fragment but an electron-poor pattern of multiple ligand-centered reductions at –1.54 ± 0.01,

[

7d,7e]

[9c]

pyrimidyl

(35, 36) or imidazolyl linker (37)

perform –1.76 ± 0.04, and – if observable in the electrochemical window

III

II

[1e,7e,23a]

clearly better in this regard as indicated by Ru /Ru redox proc- – at below –1.99 V.

esses at about 0.97 V.

In conclusion, despite the findings for the anodic scans of

From the cathodic range one may derive the influence of the complexes 1–5, where no substantial impact of the annulated

+

metal-coordinating pyridyl-ring of the terpyridine moiety and Acr/MeAcr systems was notable in terms of oxidative electro-

the linker group by using the cationic acridinium moiety as a chemical properties, we concur with the reported conclusion

+

[42]

+

[29b]

probe. The electrochemistry of MeAcr

and 9-PhMeAcr is about the quaternization effect for the cathodic scans.

well characterized: Its reduction is reported to occur at Methylation affects the ligand-centered reductive processes at

[

15a,21a,33a,33e,42a,42b,43,44]

–

0.92 ± 0.01 V,

which conforms precisely the terpyridine moiety to a significantly larger extent than the

+

with our reference measurements (Table 2). Correspondingly, metal-centered oxidation processes. In general MeAcr -ter-

acridinium complexes 2, 3, and 5 feature an additional re- pyridines are even more easy to reduce than the H TP - and

versible reduction at about –0.9 to –0.7 V. These are tentatively py -tpy series and are, in turn, more comparable with the elec-

attributed to MeAcr -centered reductions.

Ru(MeAT)₂] (2) and [Ru(MeATT) ] (5) are 140–160 mV easier 34

to reduce than PhMeAcr . The low reduction potential of the

+

3

+

[15a,33e,42a,42b,45]

2+

trochemical behaviour of the methylviologen (MV ) system

4

+

4+

[23a]

+

[

(E_o_x= 0.89, E_r_e_d= –0.74 vs. Fc /Fc; see Table 2).

2

+

latter was surprising as the electron-donating character of the

2.4. Analysis of Electronic Absorption Spectra

The absorption spectra of 1–5 are shown in Figure 4 while the

spectra of ligands 8, 19, 20 and 24 are depicted in Figure 5.

Absorption maxima and molar absorptivities for the complexes

1–5 as well as the corresponding ligands and some references

compounds are reported in the Supporting Information. Close

inspections reveals that the complexes can be arranged by an

increased molar absorptivity and, furthermore, by the maxi-

mum of the MLCT transition found between 490 and 510 nm

2

+/4+

4+

in the order [Ru(MeAT) ]

(1, 2)

< [Ru(MeAPhT)₂]

2

+/4+

(3) < [Ru(MeATT)₂]

(4, 5) which correlates with the inter-

connecting linker group. In addition to the shift by the linker,

complexes of the MeAcr⁺series (2, 3, 5) show prominently a

small shift (6–8 nm) of MLCT maxima (λ_M_L_C_T) to lower energies

concomitant with broadening at the low energy edge of the

absorption compared to their Acr counterparts 1 and 4.

The UV/Vis spectra of 1–5 are superimpositions of the

Figure 3. Steady exemplary CV spectra of [Ru(AT)

2

]²⁺(1), [Ru(MeAT)

2

]⁴⁺(2)

2

+

[1a,1b]

4+

+

typical [Ru(tpy)₂]

spectra

and the corresponding

and [Ru(MeAPhT)

2

]

(3; only to –1.35 V) vs. PhMeAcr for comparison (top).

+

All spectra referenced to the Fc /Fc couple.

chromophores, that is, either the acridine or the acridinium

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2687

Full Paper

[

19]

(

S →S ).

All these features can be found in the appropriate

0

3

spectra of the complexes.

Owing to the lack of a MLCT transition, the uncomplexed

Acr/MeAcr⁺ligands reveal all of the aforementioned features

and compare well to UV/Vis spectra of acridine and the

+

PhMeAcr ions, respectively (see Figure 5).

The interconversion between acridine and acridinium species

can be monitored by protonation/deprotonation studies on

non-quaternized complexes such as 4. Titration of ATT complex

4

with aliquots of acid in MeCN (Figure 6) give rise to the 260,

3

60, and 450 nm absorption bands. The shape of the absorp-

tion spectra and the molar absorption coefficients are in very

good agreement with those seen for the quaternized variant 5.

The similarity of the methylated and the protonated complexes,

as shown in Figure 6, is in good agreement with the litera-

[

29a–c]

ture.

MLCT transition maximum becomes also apparent. Similar ob-

Here, the aforementioned 6 – 8 nm red-shift of the

+

[29b,30b,31a]

servations were made for the py/py -series.

Figure 4. Steady state UV/Vis spectra (left axis) and uncorrected 77 K lumines-

cence emission spectra (right axis, λex. at λMLCT, 10–12 μ ) of 1–5 in MeCN

solution and BuCN/MeCN (9:1) glassy solvent, respectively.

M

Figure 6. Changes in UV/Vis spectra of 5.8 μ

titration with 0.1 TFA (0 to 7.5 m , corrected for dilution). The spectra of

M [Ru(ATT)

2

]²⁺(4) in MeCN upon

M

quaternized complex 5 is included for comparison (dotted black line).

Figure 5. Steady state UV/Vis spectra of the prepared ligands 8, 19, 20 and

4 in MeCN solution with a magnification of the S and S transitions shown

in the inset. The spectra of acridine and PhMeAcr are included for compari-

son (solid grey and black lines, respectively).

2

1

+

2

_O_w_i_n_g_t_o_t_h_e_r_e_a_c_t_i_v_i_t_y_o_f_M_e_A_c_r+ to nucleophiles at the

-position (see Introduction), complexes 3 and 5 were studied

by means of titrations with sodium hydroxide (Figure 7 and

Figure 8) or sodium borohydride (not shown). In all cases, about

9

unit.^[¹²^a^,¹²^f^,¹²^e^,¹³^c^,²¹^c^,²¹^e^,⁴⁸^]Acridinium ions have two main fea- 2 equivalents nucleophile are sufficient to completely convert

tures in the near-UV and Vis-part of the electromagnetical spec- the complexes to the bisacridane species 3a or 5a, respectively,

trum: On one hand, a very narrow absorption at about 360 nm inferred from the changes in the absorption spectra. The spec-

–

1

–1

with ε = 13000 – 20500 L mol cm for the S →S transi- tral changes for this conversion are well documented in the

0

2

tion^[

12a,12d,12e,12f,13c,21e]

and, on the other hand, a broad, solvent literature.

[21d,22,50]

Please note, however, that the different

depended and slightly vibronical-structured transition at acridanes (e.g. –OH, –OR, –H) cannot be distinguished by ab-

around 425 nm for simple arene substituents with ε = 3000– sorption spectroscopy. Whereas dihydroacridanes are relatively

–

1

–1

9

000 L mol cm , for the S →S transition. Upon substitution stable (except spurious air oxidation), alkoxyacridanes and

0

1

+

with a donor-type aryl-group this band shifts to lower energies acridanols can be reverted to the Alkyl-Acr species by addition

[

12a,12d,12e,12f,13c,21e,48a,49]

by up to 50 nm and gains intensity.

In contrast, acridines have their lowest-energy local absorp- complexes 3a and 5a (not shown).

tion maximum at about 390 nm (S →S ) and two less pro- In summary, the following trends were established: i) The

nounced transitions at about 360 nm (S →S ) and 345 nm molar absorption coefficients nearly doubled for the complexes

of acid, which was also ensured for the in situ formed acridanol

0

1

0

2

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2688

Full Paper

]⁴⁺(3) in MeCN

Figure 8. UV/Vis titration of 9.1 μM [Ru(MeATT)

2

]⁴⁺(5) under same conditions

as for complex 3 (Figure 7) to form acridanol complex 5a.

Figure 7. Changes in UV/Vis spectra of 9.5 μ

upon addition of increasing amounts of 1 m

M

[Ru(MeAPhT)

aqueous NaOH solution to

2

M

form bis(acridanol) complex 3a (corrected for dilution; blue lines corresponds

–

to ≥ 2 equiv. OH ).

found. This is not caused by the rise of the underlying 425 nm

+

absorption band of the MeAcr chromophore (former S →S )

0

1

2

+

but by a d-Ru →π*

MLCT transition and we will refer to

Acr+

relative to the corresponding ligands. ii) All acridinium com-

pounds show an intense absorption at 261 nm while the acrid-

ine compounds show this absorption at 251 nm. iii.) the ter-

this phenomenon in the TD-DFT section.

pyridine ligand-centered (LC ) absorption features at about

2

tpy

2.6. Emission Properties

70 – 290 nm and 300 – 350 nm (Figure 4)^[¹^a^,¹^b^,⁹^c^]are hardly

affected by methylation as previously noted for 4′-(4-py)-tpy While RT luminescence could not be detected in MeCN, lumi-

[

29b]

examples.

iv) The high intensity absorption band at 360 nm nescence evolved at 77 K in a butyronitrile glassy matrix for all

serves as an additional marker to distinguish between Acr and studied complexes (see Figure 4 and Table 3). Complexes with-

+

MeAcr compounds. v) A small shift of λ_M_L_C_Tto lower energies out a linker, namely 1 and 2, emit at 664 nm (1.87 eV) while

due to the quaternization of the N-heterocycle could also be the thienyl-systems 4 and 5 emit at 672 nm (1.84 eV). The

Table 3. Compilation of luminescence data for compounds 1–5 and reference compounds at 77 K. The calculated Gibbs energy of the photoinduced electron

transfer (ΔG°CS) and the energy of the charge separated state (CS) is also given.

Compound

λ

em,77K. [nm] (eV)^[^a^]

τ [μs] ^[^b^,^c^]

CS [V]

ΔG°CS [eV]^[^d^]

Ref.

[

Ru(AT)

Ru(Me-AT)

Ru(Me-AOHT)

2

]²⁺

1

2

664 (1.87)

930, 94

2020, 115

20

9500, 1200

12

1320, 111

1430, 99

13

2.50

1.78

n.a.

1.77

n.a.

2.41

1.67

n.a.

2.55

2.48

2.44

2.49

0.60

–0.10

n.a.

–0.14

n.a.

0.56

–0.18

n.a.

this work

]⁴⁺

2

]⁴

+

2a 610 (2.03)

648 (1.91)

3a 628 (1.98)

2

Ru(Me-APhT)

Ru(MeA-OHPhT)

2

]⁴

+

3

]⁴⁺

2

2+

Ru(ATT)

Ru(Me-ATT)

Ru(MeA-OHTT)

2

]

4

5

672 (1.84)

]⁴⁺

2

+

2

]²

]

5a 659 (1.88)

5b 659 (1.88)

25 598 (2.07)

27 628 (1.97)

28 656 (1.89)

29 n.a.

+

Ru(MeA-HTT)

2

11

n.a.

0.48

0.51

0.55

0.60

0.37

this work

2+

[g]

[1e,29b,36a,41,46,53]

Ru(tpy)

Ru(ttpy)

Ru(th-tpy)

Ru(py-tpy)

2

]

11

11.0

_]₂+

[e]

[g]

[23a,51d,53]

[26]

2

_]2

+

[g]

2

16

2

+

[g]

[f, g]

[30b]

[31a]

[41]

2

]

n.a.

10.6

12.1

3500

n.a.

4

+

[g]

[f, g]

Ru(Et-py-tpy)

2

]

30 n.a.

2.10

> 2.44

2.20

n.a.

4

+

[g]

Ru(H

3

TP-tpy)

TP-ph-tpy)

2

]

32 622 (1.99)

33 631 (1.96)

35 694 (1.79)

37 646 (1.92)

> 0.45

_]4

+

[g]

[41]

3

2

0.23

2

+

[7d]

Ru(Ant-pm-tpy)

Ru(tpy-HImzPy)

2

]

n.a.

0.62

2

+

[g]

[9c]

2

]

2.54

[

a] ±3 nm for own data, independent of λex. [b] In BuCN/MeCN (9:1) solid matrix. Referenced to [Ru(bpy)

3 2

][Cl] λem. = 580 nm (≈ 620 nm shoulder; ratio:

≈

1:0.23), τ = 6 ± 1 μs (mono-exponential). [c] Composition of the amplitude; long-lived component is given first. Estimated error 15 %. [d] Calc. by use of

III

Equation (1) with CS = ERu /(II) – Eligand/ligand, reduced (see Table 2); E00 was estimated from λem,max. at 77 K. [e] At 90 K. [f] At 298 K. [g] Calc. from data given in

the reference.

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2689

Full Paper

apparently electronically more decoupled 3 emits at higher en-

ergies, that is, at 648 nm (1.91 eV). The typical lower-energy

3

vibronic shoulder of such MLCT emitters can only be implied

by a small feature or broadening in the low energy range of

the uncorrected luminescence spectra.

Initial experiments implied a dual emission behaviour for 2

–

610 and 664 nm – with different decay times. As a matter of

fact, it was traceable to acridane formation (see above) in the

employed solvents. Consequently, small amounts of acid en-

sured a uniform species present in all experiments.

Luminescence lifetimes for the complexes were obtained by

fitting the decays with bi-exponential fitting functions. The life-

times were found to be between 94–1200 μs for the shorter

3

lived MLCT component and between ca. 1 ms up to 9.5 ms for

3

the longer lived LC component.

To determine the effect of the chromophores we chemically

switched to the acridane compounds as established previously

(

3

see above). Indeed, in situ acridane formation, for instance for

to 3a, gave rise to a different luminescence profile with a

mono-exponential lifetime and emission maxima comparable

2+

to the pristine Ru(tpy) complexes such [Ru(ttpy) ] 27 (see

2

Figure 9 and Table 3). Direct examination of the ligand triplet

excited state was hampered by the lack of phosphorescence

from 19 and 20 in rigid matrix at 77 K even with subsequent

[

12a]

addition of halogenated solvents (10–20 % v/v).

Figure 10. a.) Differential absorption spectra upon fs flash photolysis (387 nm,

2

00 nJ/pulse) of 4 in deaerated MeCN with several time delays between 0.9

and 7500 ps. b.) Differential absorption spectra upon ns flash photolysis

355 nm, 6 mJ/pulse) of 4 in dearated MeCN with time delays of 0.19, 0.54,

(

2

.2 and 4.4 μs.

In case of 4 short lived as well as long lived transients were

observed, which did not decay on the time scale of 7.5 ns (Fig-

ure 10a). Therefore, transient absorption measurements were

also performed on the microsecond time regime (Figure 10b).

The depopulation of the initially prepared Franck–Condon ex-

cited state is double stage accompanied by the simultaneous

formation of new short and long lived transients. The two

lifetimes, which were derived from multiple wavelengths, are

Figure 9. Luminescence decay of 1–5 in BuCN/MeCN (9:1) glassy solvent at

7 K. The left inset shows a magnification of the decay curve for 2a, while

the right inset shows an expansion for the long-lived decay of 3.

7

4

.5–6 and 300 ± 30 ps, respectively. The short lived component

is likely to be associated with the formation and relaxation dy-

namics of the initially populated Frank–Condon excited state,

such as, linker group planarization or vibrational cooling and

3

finally MLCT equilibration on the tpy ligand. Therefore, the

2

.7. Time-Resolved Measurements

weak transient at about 640 nm is tentatively ascribed to the

To gather further insights into the excited state behaviour, 1, 2, formation of the reduced tpy-fragment based on evidence

2

+

2+

4

, and 5 were studied with fs and – if suitable – also with found for related [Ru(tpy) ] (25) and [Ru(ttpy)₂] (27) com-

2

[

1e,23a,46c,51]

ns absorption pump-probe spectroscopy. Photoexcitation of all plexes.

The longer lived component implies the sub-

complexes at 387 nm leads to a bleaching of the ground state sequent triplet-energy transfer to the acridine unit, forming an

3

absorption in the region from 450 to 570 nm (see Figure 10 extremely long lived ligand triplet state ( LC). Here, two very

and SI). This implies the depopulation of the ¹MLCT ground long lived transient features, at 420 and 700 nm, were observed

state. Simultaneously with the latter, we note the formation of (Figure 10b). The lifetime of the excited state was determined

a new transient that spans from 550 to 800 nm (see Figure 10a). from a mono-exponential fit across the spectrum to be

We rationalize the origin of the newly formed transient to the 1.09 ± 0.14 μs. The excited state is of triplet nature as the

3

MLCT excited state.

formed species is readily quenched by molecular oxygen (see

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

Full Paper

SI) and, as shown by us in a separate report,^[³²^]singlet oxygen where an unquenched 77 K emission was observed. The emis-

is formed quite effectively by 4 while [Ru(AT)₂]²complex 1, sion was attributed to the destabilization of the CS state. Simi-

+

–

without a thienyl-linker, is less efficient. The transient lifetime larly, no p-e T quenching has been observed in frozen solvents

2

+

[23a]

of this complex was determined to be 80 ns only.

for the tpy-prototype MV -system 34.

We are convinced that the competing p-e T processes are

+

–

The corresponding MeAcr complex 5 showed only a broad

and weak transient in the visible region centred at about frozen at 77 K for complex 2, 3, and 5 and the non-emissive

3

6

60 nm and an additional broad band at about 1100 nm (see thermal decay of the LC state as well as the intrinsic thermally

3

SI). The lifetime of the broad transient was determined to be activated decay through the MC state is slowed down to large

70 ± 20 ps. The spectra of the MeAcr complex 2 were very extend. The existence of a non-emissive triplet state was proven

+

1

similar to 5, but the transients were of lower intensity (see SI). for AT and ATT systems 1 and 4 by ns flash photolysis and

[

32]

Looking at the absorption time profiles, the linker-less complex singlet oxygen generation, both in line with the DFT findings

4

+

[

Ru(MeAT)₂] (2) decays mono-exponentially to the ground (see below). Taken together with the DFT data, these results

3

state with 613 ± 40 ps. In both cases, the very fast recovery of strongly suggest that this state is of LC nature as similarly de-

the ground state implies a fast quenching as typically observed duced for the 9-anthryl case before.

for energetic favourable p-e T processes and subsequent back

While the results from time-resolved spectroscopy indicated

electron transfer. The Gibbs energy of p-e T process can be that room temp. luminescence is unlikely to be achieved with

[

8]

–

[

12a,51c,52]

calculated by using Equation (1).

this class of ligands, we were interested in the rationalization of

the overall deactivation pathways as well as electronic features

and parameters, which are inaccessible in the experiments. Sim-

ΔG°_C_S= F [E°_R_u^I^I^I– E°_M_e_A_c_r_/_M_e_A_c_r_·] – E₀₀+ C

(1)

/(II)

+

[31a]

ilar as previously reported for the 4′-py/py -tpy systems,

we

where F is the Faraday constant; E°_R_u^I^I^I_/_R_uand E°_M_e_A_c_r_/_M_e_A_c_r_·are

the standard electrode potentials for a given reference elec-

trode; E₀₀is the MLCT triplet energy as estimated from the

II

focused on a computational chemistry approach as described

in the following section.

3

luminescence maximum of the low-temperature measurements

and C is the Coulomb correction term, which is in polar solvents

like MeCN not only relative small but for charge shift processes

2

.8. Computational Modeling by DFT

[

12a]

always zero.

2

.8.1. Ground State and Optical Spectra

–

In summary, p-e T processes are energetically favourable by

+

–

0.10 to –0.14 eV for the MeAcr -containing complexes 2, 3 and Terpyridine metal complexes are popular targets for density

, whereas these processes are endergonic for Acr-substituted functional theory (DFT) calculations in order to gain insight in

5

complexes (> 0.56 eV; see Table 3). The ΔG°_C_Svalues are compa- the geometrical features and the relative location of the energy

rable to other terpyridine complexes. Although the Coulomb levels. Several important aspects of our calculations may be

term may be be ignored in good conscience due to the pres- briefly highlighted here while further details about the method-

[

12a]

ence of charge,

the E₀₀value are still neglected throughout and therefore the

solvent and temperature effects concerning ology are described in the Supporting Information.

The ground state structural parameters for 1–5 are well re-

[

23a]

values shall be be considered as estimates.

of our low temperature results a study of a MV ruthenium(II) shown for the 4′-py/py -tpy systems previously.

rotaxane complex by Credi et al.

For discussion produced by B3LYP or PBE0 DFT calculations, which was also

2

+

[30d,31a]

Selected

[

23b]

is nevertheless helpful, geometrical key parameters for the AT-derived complexes 1 and

Table 4. Selected structural parameters (distances [Å], angles [°]) of complexes 1 and 2 in the ground state (S

0

) and for the different triplet states as obtained

by calculation on the B3LYP level of theory (and PBE0 evaluation for S

0

, respectively). For complex 2 the deviation to the experimental value as determined

by X-ray structure elucidation is given as Δexptl value.

[

Ru(AT)

2

]²⁺(1)

[Ru(MeAT)

2

]⁴⁺(2)

[

a]

[b]

³LC^[^a^]

2.018,

³MLCT^[^a^]

2.018,

2.016

³MC^[^a^]

2.201,

2.017

2.163

[a,c]

[b]

³LC^[^a^]

2.020,

2.015

2.126

³MLCT^[^a^]

2.039,

2.024

³MC^[^a^]

2.202,

2.015

2.405,

2.402

2.161,

2.162

1.496

Parameter

dist. c-axis

S

0

S

0

S

0

Δ

exptl

S

0

Δ

exptl

2.018

2.127

1.496

–

1.997

2.096

1.486

–

2.018

2.127

1.497

1.481

0.045

0.074

0.059

0.014

–0.013

1.997

2.096

1.488

1.468

0.016

0.037

0.020

–0.010

–0.025

2

.016

dist. b-axis

dist. a-axis

d.(CTpy–CAryl

2.128

2.127

1.493,

2.128

2.131,

2

.120

2.127,

2.404

1.493

–

2.127

2.113,

2.112

1.497,

1.462

1.481,

1.468

177.3

157.7

152.7

50.3,

2

.126

)

1.494,

1.467

–

1.496,

1.477

1.481,

1

–

.467

d.(NAcr–CMe

)

1.481

1

.475

∠c-axis

∠b-axis

∠a-axis

180.0

156.6

89.8

180.0

157.4

69.9

180.0

156.6

51.9,

180.0

156.4

156.6

56.7,

72.6

177.9

156.2

137.9

73.4,

74.3

179.9

156.6

88.6

2.9

179.9

157.6

157.5

83.2,

74.4

2.9

179.9

156.6

156.7

57.4,

84.1

178.5

138.9

156.3

78.9,

89.0

–1.1

–1.6

15.8,

27.5

–0.1

–0.7

10.3,

13.3

ꢀ

(aryl–tpy)

7

3.0

83.0

[

a] B3LYP//6-31G(d)/LANL2DZ calculation results. [b] PBE0//6-31G(d)/LANL2DZ calculation results. [c] Within 0.06 Å RMSD of a B3LYP//LANL2DZ-only calculation.

2691

Full Paper

2

are composed in Table 4 (S states) while data for 3–5 are was performed (see colour coding in Figure 11). The expected

0

compiled in the Supporting Information for reasons of space pattern of ligand based LUMOs and metal ion contribution for

see Table S6). the HOMO levels is clearly visible. Furthermore, similar as re-

Calculated N–Ru bond lengths were always found to be in ported,^[³¹^]we found the HOMO and the next lower MOs

(

between 2.018–2.128 Å for the ground state geometry with all (HOMO–1/–2) separated by only a small energy difference

three axes considered. Statistical evaluation of 119 [Ru(tpy)₂]- of ≈ 0.1 eV for complex 31 (Figure 11, left). The MOs of the

derivatives for their principal coordination axis (viz. c-axis) gave HOMO–1/–2 set itself are very close in energy (<< kT) and can

an average N–Ru bond length of 1.977 ± 0.013 Å (1σ; see SI). In be considered as degenerated. Also in accordance with previ-

[

30d]

4+

this regard the typical overestimation of the B3LYP functional ous reports,

to X-ray data

the non-alkylated complex [Ru(py-tpy)

is apparent but acceptable. In our case a the HOMO to HOMO–1/–2 difference was even smaller

direct comparison with X-ray data was only possible for com- (0.033 eV) while the overall frontier orbital positions were

]

(29)

2

[

6a,54]

plex 2 (see Table 4; Δ

values).

higher in energy.

exptl

In accordance with several reports^[⁶^a^,⁵³^,⁵⁴^]about the capabili-

With the compatibility verified, we were interested in the

II

ties of the PBE0 functional in comparison to B3LYP, we found outcome for our Ru systems 1–5: Next to the metal-based MOs,

+

the PBE0 functional also better suitable for describing the “ex- and in contrast to the py/py -tpy systems 29 and 31, respec-

act” geometry of the ground state. However, the N_A_c_r–C_M_ebond tively, a differentiation in almost pure terpyridine- and Acr/

+

lengths are always underestimated (interestingly worse for MeAcr -based MOs is prominent for complex 1–5 (Figure 11,

PBE0) and the experimentally found bending along the c-axis see Figure 12 for a graphical representation). Two sets of degen-

+

is not covered by both functionals.

erated MOs for the Acr/MeAcr chromophores are always in

With the reproduction of structural features reasonable close proximity to the frontier orbitals. For instance, a degener-

verified, we were interested in the electronic differences of ated set of MOs with π*_A_c_rcharacter compose the LUMO for

complexes 1–5 in comparison to published data for the 4′-py/ the acridinium compounds 2, 3 and 5. Thus methylation of the

+

[31,30d]

py -tpy complexes.

For this reason we reproduced B3LYP acridine moiety has a pronounced effect on such systems as

4

+

calculations for the non-alkylated complex [Ru(py-tpy) ] (29) the Acr -π* set is decreased low enough in energy to become

2

as well as [Ru(Me-py-tpy)₂]⁴(31) first to ensure a comparable the LUMO set instead of the usual π*

+

MOs. This deviates

notable from the py/py -systems where these LUMOs are in

tpy

+

setting.

To aid an in-depth analysis of the MOs an additional decon- principle pyridyl-based but mixed to considerable extend with

[

31a]

volution of each MO into tpy-ligand, linker, metal atom, and terpyridine MO contribution

4

(own calcd.: 48 % π*_P_y₊/44 %

′-substituent – either py/py⁺or Acr/MeAcr – contributions π*_t_p_y; see Figure 11).

+

Figure 11. Computed orbital energy level and MO composition for 1–5 in comparison to 29 and 31, respectively (left side) as calculated by DFT [degeneracy

threshold ΔEdeg. set manually to 0.0256 eV (= kT) to visualise otherwise overlaid MOs]. Colour scheme of the deconvolution: ruthenium ion (red), terpyridine

(

blue), pyridine/-ium (bright cyan) acridine/-ium (light green), phenyl- (purple), and thienyl-linker (yellow). Configurations of the first transition with meaningful

oscillator strength of TD-DFT calculations are indicated by grey lines. The circular dots mark the electrochemical determined HOMO (filled circle) and LUMO

+

[47]

(

open) obtained by the empirical interconversion relation of –4.5 eV on vacuum level ≈ 0 V on NHE scale (converted from the Fc /Fc standard by +0.630 V ).

2692

Full Paper

energies (Figure 11). Based on the orbital analysis the MeAcr⁺

unit in 2 is clearly the best overall electron acceptor in the

series as both LUMO and HOMO are lowered in energy, even

4

+

surpassing the [Ru(Me-py-tpy)₂] complex (31) theoretically.

+

Therefore, the conclusion for the py/py -tpy series could be

validated, namely, that quarternization of the pendent N-het-

erocycle lowers the energy of the lowest lying π* molecular

orbitals, thereby increasing the electron accepting ability of the

[

29c]

complexes.

With the electronic structure established, it is logical to de-

termine the possible optical transitions by TD-DFT. Starting

from our own geometries we were able to reproduce the TD-

+

[31]

DFT data for the py/py -tpy systems very well with the com-

bination of B3LYP functional and 6-31G(d)/LANDL2DZ mixed

basis sets. The PBE0 functional performed worse in TD-DFT cal-

culations with the same basis set combination (see Figure 13a,

upper trace), presumably due to the low HOMO energies (see

B3LYP vs. PBE0 calculated energy level diagram in the Support-

[

55]

ing Information).

Figure 12. Graphical representation (isodensity contour value 0.05 e Bohr^–³,

2 grid points Å^–¹) of the frontier molecule orbitals for [Ru(MeAT)

]

2+

(2).

1

2

In terms of linker influence one can state that the linker

+

group contributes only slightly to the Acr/MeAcr MOs which

justifies the overall view as an isolated terpyridine fragment

connected to a 9-arylacridin/ium unit in case of complexes 4

and 5 but especially 3.

Focussing on the metal-based orbitals the findings of previ-

[

30d,31a]

ous reports

can be confirmed and thereby be extended

+

to the Acr/MeAcr class of compounds. The MOs of typical

d_x_y/d_x_z/d_y_zshape are highly metal-based (61 – 75 %) and split

in an approximately degenerated subset and a singular MO (see

Figure 12 for 2, see SI for 1 and 3, 4, 5). The latter is separated

by only a small energy difference (max. 0.12 eV in 4) similarly

Figure 13. TD-DFT calculated transitions (coloured bars) for 2 starting with

different geometries compared to the λMLCT-region of the exp. spectrum

4

(

blue, only on top, scaling f/ε = 1 to 5 × 10 ). b.) λMLCT-region of the exp.

spectra and calculated transitions (B3LYP) for complexes 1, 3, 4, and 5 (scaling

+

4

as mentioned for the py/py -tpy systems.

f/ε = 1 to 2.5 × 10 ) for comparison. Colour coding as in Figure 4.

The e -type orbitals are calculated to be near the 0 eV poten-

g

tial energy frontier as described by other previously.^[³⁰^d^]This

Thus, we decided to perform all remaining TD-DFT calcula-

non-degenerated set of MOs can be found at similar positions tions with the B3LYP//6-31G(d)/LANDL2DZ combination on

+

for the Acr/MeAcr class of compounds and easily identified by B3LYP generated geometries. The numerical data of the TD-DFT

+

the 25 %–51 % metal character (Figure 11). Correspondingly, calculations for systems 1–5 as well as the py/py examples 29–

the ligand field splitting value (Δ ) averages to 5.33 ± 0.06 eV 31 are compiled in the Supporting Information.

0

for 1–5.

More importantly, we noted systematic differences in the re-

In summary, the DFT calculated MO energies agree qualita- sults by the choice of the starting geometry, namely optimized

+

tively well with comparable calculation for the py/py com- C vs. C symmetry vs. a PBE0 geometry. A graphical summary

1

2

plexes and with the electrochemical determined HOMO/LUMO for complex 2 is given in Figure 13a. This let us question the

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

Full Paper

cause of these discrepancies and we assumed that the dihedral

Noteworthy, all methylated systems show next to the

2

+

angle ꢀ, and only subsequently the symmetry, can be made d-Ru →π*tpy transition a second transition at lower energies

2+

responsible. Indeed, the corresponding d-Ru →π*

transi- due to a d-Ru →π*

transition (e.g, calcd. λ_M_L_C_T505 nm for

Acr+

tion becomes only active for non-orthogonal angle ranges be- 2, 497 nm for 3, 557 nm for 5), assuming to be the root cause

+

tween the Acr/MeAcr unit and the tpy fragment, which was of above established red-shift and broadening of MLCT band.

+

both given for the PBE0 C (ꢀ = 83.2/74.4°) and B3LYP C -geom- Moreover, the linker-extended MeAcr complexes 3 and 5 do

1

2

etry (ꢀ = 76.8°) but not for the optimized C geometries of not show the above mentioned conformation-related selection

1

complex 2 (see Figure 14) and 1, respectively (see Table 5 for a rule of the optical transition and have therefore usually high

2

+

compilation ꢀ₍_a_r_y_l_–_t_p_y₎values).

oscillator strengths for the d-Ru →π*

transition (see Fig-

Acr+

ure 13b).

Of course this raises the question which bands are to com-

pare with the experimentally visible MLCT transitions:

2

+

If the d-Ru →π*

transition is chosen the predicted

tpy

UV/Vis spectrum can be roughly aligned with the experimental

one. Qualitatively the same trends in position and oscillator

strength (as measure for the molar absorptivity) of the MLCT

transition can be reproduced. The calculated bands are, how-

ever, shifted to higher energies by 0.09–0.28 eV (Δλ = 17 to

5

4 nm at these wavelengths) in comparison with the experi-

mental results. This is a quite typical error for TD-DFT calcula-

[

46a]

tion of ruthenium(II) terpyridine systems.

2

+

If, however, the d-Ru →π*

transition is considered, the

Acr+

dependency of the spectral position on dihedral angle ꢀ needs

to be addressed for proper comparison as pointed out above

and also by others for conjugated terpyridine assemblies previ-

[

30d,56]

ously.

For the directly connected complexes 1 and 2 the d-

2

+

Ru →π*

transition can only considered as strong (f >> 0)

Acr+

Figure 14. a.) Calculated potential energy surface (PES) for different dihedral

angles ꢀ (filled symbols) and/or ω (open symbols) on one ligand fragment

of complexes 1–5. b.) Calculated λMLCT for the d-Ru²⁺→π*Acr transition (stars)

and corresponding oscillator strength (shaded bars) of the relaxed geome-

tries for MeAcr⁺complex 2.

at non-orthogonal angles for ω (= ꢀ in this case). To account

for the effects of the dihedral angle relaxed potential energy

surface (PES) scans of ω and ꢀ on one ligand of the complex

were performed and the position of λ_M_L_C_Tand oscillator

0

Table 5. Potential energy difference (diabatic Δ-SCF [eV]) of the triplet states for complex 1–5 in respect to the S state and change in geometric parameters

and spin density for the different calculated triplet states.

³MLCT

³LC

³MC

1

2

3

1

2

3

4

5 (#1)

5 (#2)

1

2

3

4

5

Δ-SCF

1.799

1.875

2.043

1.799

–0.001 –0.004

1.795

1.846 1.616

1.553

1.713

–0.005

1.824

0.806

1.836

0.813

1.805

0.802

1.778

0.816

1.790

0.813

[a]

Q

Ru

–0.001 –0.007 –0.003

–

0

–0.002 –0.002

.002

[

b]

RMSD ∠Ru–N

–0.063 –1.87

0.84 0.86

0.28C4′ n.s

0.26N1′

–1.94

0.86

n.s

0.001

not significant (n.s.)

not significant

–0.015 0.003 –0.102 0.011

–0.007

0.15

–6.63

1.72

not significant

–6.57

1.76

–6.02

1.76

–6.88

1.49

–6.53

1.76

Spin on Ru²

+[c]

Spin density

Terpyridine

fragment

[

c]

0.15C2′

0

–

.15C6′

Spin density

on linker

–

n.s.

–

0.31C2 0.17C2

0.26C4

0.34C2

0.25C4

–

not significant

[

c]

Spin density

n.s.

0.43C9

0.14C3

0.14C6

0.40C9

0.56C9 0.59C9

0.43 0.35

0.24C4 0.21C1

0.24C5 0.21C8

0.21C4

0.59C9 0.47

N

0.49C9

0.35

N

not significant

+

Acr/MeAcr

0.10

0.11c

0.11C8

N

0.39 0.34C9 0.37

N

0.32C9

0.14C2

0.14C7

[

c]

fragment

1

0.22C1 0.20C4

0.22C8 0.20C5

0

.12N0

0

.11C3

.11C6

0.21C5

[

a] Obtained as defined by Persson et al.^[⁶¹^b^]Q denotes the nuclear coordinate of the ruthenium(II) ion, briefly, the sum of the geometrical change Σ(Δdist.)

of the six Ru–N bond lengths (in Å) relative to the S conformation. [b] Denotes the root mean square deviation (RMSD) of the N–Ru–N angles (in °) of all

axes calculated and referenced to a standard, e.g. O symmetry. Due to the fact that bis(terpyridine) complexes generally do not show a perfect octahedral

coordination geometry we chose to reference relative to the corresponding S state geometry. [c] By Mulliken spin analysis (see SI for details). The numbering

scheme for the spin density position is based on the IUPAC nomenclature in view as isolated fragment.

0

[

6a]

h

0

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2694

Full Paper

strength was investigated (Figure 14). In our belief dihedral

changes on one ligand are a more realistic approach in view of

the often accounted C symmetry than a pseudo-C symmetric

1

2

change of both ligand fragments.

As a result, a medium to high rotation barrier was calculated

+

for the dihedral angle ω between Acr/MeAcr -fragment and the

9-aryl unit by relaxed PES scans for 1, 2 (ω = ꢀ) and 3 (see

Figure 14a).

For the sterically more crowded compounds 1 and 2 high

energies of ΔE_p_o_t= 70–79 kJ/mol were obtained while the sum

of thermally accessible conformations around the 90° minima

is quite large. This finding corroborates semiemperical calcula-

Figure 15. Structure of a forced planar conformation (ꢀ = 180°) for one ligand

4

+

of [Ru(MeAT)

2

]

(2) after relaxation (ΔEpot +70 kJ/mol).

[

12e,12f]

crowded

9-(2,6-dimethoxyphenyl)-1,8-dimethoxyacridinium

tions for 9-phenyl and 9-napth-1-yl-acridinium ions.

pre-

‡

[58]

salts by X-ray and NMR studies (ΔG 49–83 kJ/mol). The influ-

ence of the interplanar angle on the rotation barrier of such

annulated aromatic systems was pointed out in the study of

dicting a broad angle distribution (±30°) around the ca. 90°

minimum energy conformation.

Together with the wide angle range of ca. ±40° (equals

[

57]

9

-phenylanthracene previously.

The generation of a compu-

99.8 % Boltzmann population) a intrinsic broadening of the

tational artefact in our calculation seems therefore to be un-

likely.

transition of ca. 40 nm at room temp. (see Figure 14b) is im-

plied, whereas static TD-DFT calculation of the optimized geom-

etry would have been absolutely misleading in the relevance of

this transition. Nevertheless free rotation was still observed in

NMR spectroscopy at 298 K in agreement with measurements

More importantly the structural similarity of this rotamer to

3

the optimized geometry of LC state (see section 2.8.2) might

implicit a singlet-triplet conical intersection of the two hyper-

surfaces nearby, and consequently opening a route for radia-

tionless triplet relaxation, which will be discussed in further de-

tail in the following section.

‡

[57]

for structural related 9-phenylanthracenes (ΔG 88 kJ/mol).

+

For the MeAXT linker-systems 3 – 5 similar observations can

be made: Thienyl-systems (X = T) 4 and 5 have a small barrier

of rotation of ΔE_p_o_tabout 30–45 kJ/mol in regard to ω while 3

reflects the rotation barrier of a 9-phenylacridinium and

In summary, we agree with conclusions drawn by Constable

[31a]

et al.,

namely, the correct prediction of a red shift

[

29c,30d,31]

for the N-methylation is possible

but we deduced an

-

anthracene compounds (see Figure 14a). In regard to the sec-

additional root cause next to energetical lowering of the LUMO:

ond dihedral angle, namely ꢀ, a large angle distribution is easily

+

Thus, Acr/MeAcr systems show additional chromophore ab-

+

thermally accessible for the MeAXT linker-systems 3 – 5. The

sorptions resulting in intraligand (IL) π_A_c_r₊→π*_A_c_r₊-type of tran-

rotation between the tpy unit and the phenyl (X = Ph) or

thienyl-linker (X = T), respectively, is clearly not hindered

2

+

sitions as well as d-Ru →π*

MLCT transitions. In our case

Acr+

the latter displace the lowest energy “pure” MLCT d-

(

^Δ^Epot <18 kJ/mol). Corresponding TD-DFT calculations show

2+

+

Ru →π* -transition in all MeAcr complexes (2, 3, 5) and may

tpy

only small changes in the predicted wavelength and oscillator

therefore account for the observed red-shift as well as for the

markedly broadening of the MLCT band in the titration experi-

ments.

2+

strength of the d-Ru →π*

and d-Ru →π* transition in

Acr+

tpy

depedancy of ꢀ. For complex 3, however, by the coincidence

of having a restricted angle range for ω combined with a dou-

ble minima energy landscape for ꢀ (see Figure 14) the calcula- 2.8.2. Triplet States

tions capture the global minima of both dihedral angle popula-

Geometry optimizations without symmetry restraints become

particular important for the triplet states of these complexes

because the unpaired electrons may localize only on one frag-

tions well (S ꢀ = 70 ± 2°, ω = 36.5 ± 1° regardless of DFT func-

0

tional; see Table S6). This is reflected by a a very good agree-

^m^eⁿ^t^o^f^p^r^e^dⁱ^c^t^e^d^aⁿ^d^c^a^l^c^u^l^a^t^e^d^λMLCT (calcd. 497 nm, exp.

ment of the complex (viz. C symmetry). Mulliken spin-density

1

4⁹³ⁿ^m⁾^w^hⁱ^l^e^u^s^u^a^l^l^y^t^h^e^p^r^e^dⁱ^c^t^e^d^λMLCT transition for the

analysis of these calculations revealed that different triplet

calculated minima conformation is red-shifted by several

nanometers (Δλ = 16 – 47 nm at these wavelengths; see Fig-

ure 13b).

states were indeed obtained (Table 5, see Figure 16 and Fig-

[53,54a,59]

ure 17 for a graphical representation).

The outcome was

depending on the initial geometry. This dependency was found

2

+

2+

Another aspect highlighted by the (TD)-DFT calculation was for other [Ru(tpy)₂]

the conformation at ≈ 0°/180° angle setting for ω (or ꢀ for 1 ously.

and 2). Upon rotation about ω the steric stress is avoided by complex 1 and 2 are given in Table 4 (columns LC, MC and

adopting an acridane-like (“butterfly”) conformation in the high

and [Ru(bpy)₃]

systems previ-

[

54a,59b,60]

Detailed values for the triplet geometries of

3

MLCT) while 3 – 5 are compiled in the Supporting Information

energy transition state and thereby giving up the energetically (Table S6). For a general impression the changes in bond

+

[6a]

favorable planarity of Acr/MeAcr ring systems (1, 159°; 2, 153°; lengths can be summarized in the parameter Q (= ΣΔdist.)

3

, 152°; 4, 165°; 5, 158°; see Figure 15).

and the root-mean-square deviation (RMSD) of the Ru–N coor-

The “butterfly” conformation in the transition state is slightly dination angles relative to the S

geometry which are compiled

0

+

favoured for the MeAcr complexes with the energetic order in Table 5.

3

2

> 3 > 5. A similar “butterfly” shape conformation was also

The MC state can be easily distinguished by its dissociative

found experimentally on a detailed study of sterically over- character due to drastic bond elongation (Table 4; max. 2.411 Å

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

Full Paper

2

+

for 5, see Table S6) on one N–Ru–N axis. For pristine [Ru(tpy) ]

2

[

6a,61]

(

25), a length of 2.382 Å was calculated.

In view of the Q

3

value the MC state manifests itself in differences of > 0.80 Å in

Q and a RMSD of > 6.0° in coordination angles. Severe out of

plane bending of the lateral 2′,6′-pyridyl units on one ligand

fragment is also noticeable (≈ 14°, see Figure 16 and Figure 17).

The spin density is mainly located on the metal center (Table 5).

and the values agree very well with observations described in

[

6a,61a]

previous publications.

Structural parameters of the 4′-sub-

stituents and of the chromophore units are, however, only

3

barely changed in the MC state. They resemble the S geome-

0

try in most parameters (i.e. C_T_p_y–C_A_r_y_ldistance; see Table 4)

which is in agreement with the triplet localization on the metal

center.

3

The geometry in the MLCT state is more affected in this

regard as both – metal center and chromophore – show moder-

ate distortions (see Table 4 and Table 5). Despite numerous at-

3

tempts no stable MLCT structure could be found for complex

4

and 5 due the intrinsic low dihedral angle ꢀ between the

3

thienyl and tpy fragment which subsequently leads only to LC

states in optimizations.

Figure 16. Calculated structures (B3LYP//6-31G(d)/LANL2DZ, CPCM solvent

3

LC states could be found for all complexes under study and

3

model for MeCN) for the

S

0

,

LC, MLCT, and MC triplet states of

3

the Q and RMSD parameters are compiled in Table 5. The LC

states of the Acr/MeAcr chromophore are of interest for energy

4+

[

2

Ru(MeAT) ]

(2) and corresponding spin density isocontour plot (purple,

.02 e Bohr ; 6 grid points Bohr ). Some structural parameters are also high-

+

–3

–1

0

lighted, viz. exceptional N–Ru bond lengths (red), 2′,6′-pyridyl units out-of-

plane bending (green), and parameters of the 4′-substituent (black), respec-

tively.

storage as outlined in the aim of this work but on the other

hand also may serve as deactivation pathway if too low in en-

3

ergy. Most obvious for the LC state are the geometrical

+

changes of the Acr/MeAcr moiety itself, i.e. significant shorter

C_M_e–N_A_c_rand C_A_c_r–C_A_r_y_lbond lengths. In several cases a 3LC

geometry could be found which resembles the “butterfly”-

shape of the ground state rotamers (Figure 17, cf. Figure 15).

The calculated Δ-SCF energies of the triplet states are sum-

marized in Table 5 while a graphical representation is depicted

in Figure 18 together with the experimental determined emis-

sion maxima (see Table 6).

3

Most trends are captured well by this direct approach: MC

states are predicted to be energetically close but below the

3

MLCT states except for complex 1 where this level is slightly

3

higher in energy. Furthermore the LC and MLCT state are pre-

dicted to be isoenergetic for complex 1, therefore ideally fulfill-

ing the requirement for an efficient multichromophoric assem-

bly. However, as already shown in the previous section, this

does not hold true experimentally.

The prediction for complex 2 is also favourable and would

still enable a multichromophoric behaviour (ΔE = 0.080 eV =

–1

–

6

45 cm ) if p-e T is neglected. Similarly the relative trend to

higher energy emission for 3 is qualitatively correct predicted

but overestimated in energy (calcd. 606 nm, exp. 648 nm). As

3

already mentioned before, a stable MLCT states could not be

found for complexes 4 and 5, respectively. However, even with-

3

out exact knowledge of the MLCT energy these systems have

3

apparently LC states too low in energy to be well-suited for

the multichromophore approach.

It is important to note that most calculated energies differen-

ces are in the order of magnitude of the estimated DFT error at

this level of theory. To obtain nevertheless an estimate about

Figure 17. Calculated structures (B3LYP//6-31G(d)/LANL2DZ, CPCM solvent

3

4+

model for MeCN) for the LC, MLCT, and MC triplet states of [Ru(MeATT)

2

]

(

5) and corresponding spin density isocontour plot. Colour scheme and iso-

3

contour values as in Figure 16.

the quality of the calculated Δ-SCF energies for the MLCT state

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2696

Full Paper

of complex 4 and 5 can be estimated to be in the range of

1

.80–1.84 eV. Interestingly this holds also true for 3 which was

overestimated by the Δ-SCF approach (see above).

3

. Conclusions

A family of terpyridine ligands incorporating either an acridine

or a methylacridinium chromophore and different linker groups

(none, phenyl, thienyl) have been synthesized by different

routes. Linker-extended acridinium–terpyridine ligands have

been synthesized by the use of acridane intermediates as pro-

tected acridinium moieties. The desired terpyridine ligands and

their ruthenium(II) complexes have been synthesized by stan-

dard methods, except for complex 2, which was directly methyl-

ated starting from 1 by a “chemistry on the complex” approach.

The physicochemical properties were investigated by a com-

bination of experimental and theoretical approaches. The elec-

trochemical and absorption features of the complexes are in

+

line with previous findings for the py/py -type series (29–

[

30b,30c,31a]

31

). To our surprise, no substantial positive impact of

+

the introduction of the Acr/MeAcr moiety was notable while

vice versa for the MeAcr complexes 2, 3, and 5 the reduction

potential of the MeAcr was less negative.

+

Figure 18. Energy of the experimentally determined emission maxima (black

Therefore, the well-known principle that properties of indi-

vidual units must not necessary be reflected in the de novo

molecular assembly holds true. This is not only eminent by con-

trary electrochemical expectations but also established by

HOMO/LUMO and triplet calculations – especially in ligand dy-

bars with FWHM indicated as fading grey area) and the calculated Δ-SCF

energies of the different triplet states ( MLCT, red; LC, blue; ³MC, green) for

3

complexes 1–5. Note that no MLCT states could be found for complexes 4

_a_n_d₅_._T_h_e_e_n_e_r_g_y_f_o_r_t_h_e_“_b_u_t_t_e_r_f_l_y_”_-_l_i_k_e_A_c_r_/_M_e_A_c_r+ high energy rotamer of

the ground state is also indicated (dark grey). Please note the different scaling

and the arising axis break for the energy scale in this area.

3

namics of the LC states (discussed below). In this regard the

chosen connectivity by use of the C-9 position to retain symme-

try was unexpectedly unfortunate.

Table 6. Singlet–triplet energy difference in λ [nm] (eV) by Δ-SCF (= ΔEdiabatic

calculations and by TD-DFT S –T (= ΔEvertical) transition calculations.

)

0

1

Complex 2, 3, and 5 showed in DFT calculations a lowered

LUMO level of almost pure π*_A_c_r₊MO composition. This leads

to pronounced and broader MLCT transitions in electronic ab-

sorption spectra. The latter was verified by conversion of the

acridine complexes 1 and 4 to the corresponding protonated

compounds at low pH but also through in situ destruction of

the acridinium chromophore by formation of acridanes for com-

plex 3 and 5. By using this conversion technique in low-temper-

ature luminescence measurements, the proposed electronic in-

_Δ_-_S_C_F₃_M_L_C_T_a_Δ_E_v_e_r_tb

_Δ_E_v_e_r_tc

Compound

Exp.

[

Ru(AT)

Ru(MeAT)

Ru(MeAPhT)

Ru(ATT)

2

]²⁺

1

2

3

4

5

689 (1.80)

661 (1.87)

606 (2.04)

not found

667 (1.86) 668 (1.86) 664 (1.87)

667 (1.86) 665 (1.86) 664 (1.87)

657 (1.89) 649 (1.91) 648 (1.91)

680 (1.82) 672 (1.84) 672 (1.84)

690 (1.80) 679 (1.83) 672 (1.84)

_]₄+

2

]⁴⁺

2

+

2

]

Ru(MeATT)

2

]⁴⁺

[

a] ΔE(SCF) calculated from optimized B3LYP geometries. [b] First 30 S

geometry. [c] Same as before but starting

geometry. Note that the lowest energy transition provided here

0 n

–T

transitions calculated on B3LYP S

from PBE0 S

was always well separated from other S

2); ΔE = +0.432 eV). However, no statement about oscillator strengths can

0

+

fluence of the Acr/MeAcr moiety on the ruthenium(II) complex

2

4+

0

n

–T transitions (e.g. for [Ru(MeAT) ]

3

fragment could be proven. The MLCT emission lifetimes of

(

be made as these transitions are spin-forbidden and spin-orbit coupling is

neglected throughout by the calculations.

1–5 compare favorably with current benchmark systems in the

[7d,7e]

milliseconds regime at 77 K.

The chemical conversion to

acridane derivatives reverts the emission properties solely to

we performed additional TD-DFT S –T₁calculations starting the corresponding Ru(tpy) -fragment of the complex. However,

from the optimized S B3LYP as well as PBE0 S geometries. The a photophysical distinction between acridine and acridinium

0

2

0

results are compiled in Table 6 and show a quite remarkable compounds was hardly possible for 1–5 at this temperature.

correlation – especially for the PBE0 calculations. The good This is understandable by the similar triplet energies, while

3

correlation presumably has two reasons: On the one hand the competitive decay mechanism, especially MC-mediated relaxa-

3

–

geometries of the emitting MLCT state show only moderate tion and p-e T, are slowed down or suppressed in the rigid

distortions in Q and RMSD values in comparison to the S state matrix. Time-resolved measurements on selected complexes

0

3

(

see Δ-SCF approach; Table 5) and on the other hand simple suggest a fast formation and decay of the MLCT states, pre-

error cancelation due to calculation of the difference can be sumably via p-e T mechanism in the case of 5 or formation of

assumed. Therefore, a direct S –T TD-DFT calculation, under long-lived but non-emissive LC state in the case of 1 and 4.

assumption of the T state to be of MLCT in nature, appears

–

3

0

1

3

In view of our overall design principle this means that i) for

1

3

to be a valid choice. By this methodology the MLCT energies complexes 1 and 2 the stabilizing effect of a conjugated

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2697

Full Paper

+

4

′-substituent on the Ru(tpy) -fragment is missing due to the tional inhibited system at the Acr/MeAcr fragment, such as a

2

+

steric bulk of the Acr/MeAcr units; ii) for linker-extended com- 2,6-methylphenyl spacer, seems especially appealing for the

plexes some positive effects of the Acr/MeAcr fragment are long-lived [Ru(MeAPhT)₂] complex 3 to prevent the proposed

counterbalanced by our choice of a conjugated linker – espe- deactivation channel of the LC state by rotation of the moiety

+

4+

3

cially for the thienyl complexes 4 and 5; iii) in case of the around the C-9 position. The rotation towards the terpyridine

4

+

[53]

[

Ru(MeAPhT)₂] complex (3) theoretical and experimental data fragment, however, shall not be hindered.

ii) Under the as-

suggest an electronically more decoupled linker group, namely pect that the C-9 position as connecting point was misfortunate

+

an isolated MeAcr chromophore and a 4′-p-tolylterpyridine li- but “simple” acridine or acridinium dyes would still be a viable

gand (ttpy). This manifests itself in an exceptional long lifetime option for the multichromophore approach, it may be of inter-

3

of 9.5 ms at 77 K and a high energy MLCT emission compared est to evaluate either different or rigidified points of attachment

to the other complexes. It underlines the crucial interplay of but also linker units, such as 1H-1,2,3-triazolyl, which may

+

electron donor character and electronic decoupling of the 9- partially decouple the Acr/MeAcr unit while stabilizing the ter-

[

56]

arene substituent for acridinium dyes but also for the 4′-posi- pyridine fragment.

Recent reports show interesting ap-

tion in terpyridine ligands in such complexes. By this, complex proaches towards utilizing typical acridine reactions directly on

[

38]

3

is heading towards a multichromophoric behaviour but the complexes.

+

lower reduction potential of the PhMeAcr unit in the complex

(

Last but not least we do not see the occurrence of a slightly

–0.89 V vs. –0.93 V for PhMeAcr ) combined with the lack of exergonic p-e T process for acridinium dyes as major obstacle

+

–

influence towards the HOMO led this complex still be suscepti- at first instance, as by such, switchability of long-lived lumines-

–

ble to quenching via p-e T (ΔG° = –0.14 eV).

cence by pH change might be feasible.

Another aspect highlighted by the (TD)-DFT calculations but

underestimated in our design principles was the dihedral twist

angle between the individual aryl fragments (ꢀ or ω). Both are

Experimental Section

of course key parameters in view on electronic conjugation. Procedures and references for the synthesis of known compounds

Stabilization via increased planarization was supposed to act on and detailed data on all new compounds and corresponding ruth-

enium(II) complexes 1–5 are given in the Supporting Information.

the 4′-substituent of the tpy-unit once the excited state is

The SI holds also detailed information regarding the instrumenta-

formed. However, even in ground state steric stress at the

_′_-_p_o_s_i_t_i_o_n_,_i_n_d_u_c_e_d_b_y_r_o_t_a_t_i_o_n_o_f_t_h_e_A_c_r_/_M_e_A_c_r+ moiety

around its C₂axis, is avoided by adopting an acridane-like

“butterfly”) conformation as transition state at low angles. The

tion for routine characterization, electrochemistry, and photophysi-

4

cal measurements (absorption, transient absorption and low temp.

luminescence) as well as employed computational methods and the

instrument and software used to derive X-ray structure information.

(

structural similarity of this rotamer to the optimized geometry

3

CCDC 1421582 (for 2) contains the supplementary crystallographic

data for this paper. These data can be obtained free of charge from

The Cambridge Crystallographic Data Centre.

of LC state suggest a route for radiationless triplet relaxation

at similar nuclear coordinates. In turn, it underlines the impor-

tance to account for the conformational dynamics in the study

of any sophisticated designed [Ru(tpy)₂]²complex. This was

also important to identify the major transitions as some of

these transitions were found to depend distinctly on the

dihedral angles between the fragments – especially for linker-

less complexes 1 and 2.

+

Acknowledgments

Financial support by the Deutsche Forschungsgemeinschaft

(SFB 613) is gratefully acknowledged. We furthermore thank the

regional computing centre of the University of Cologne (RRZK)

for providing the CPU time on the DFG-funded supercomputer

The “ordinary” non-emissive deactivation channel of Ru(tpy)

2

3

complexes into the MC states could not be probed experimen- “CHEOPS” as well as the support. The authors would also like

tally but the existence of ³MC states close in energy (1.78– to thank Dr. H. Görner (Max-Planck-Institut für Chemische Ener-

1.84 eV) is also implicit by the DFT calculations. As pointed giekonversion, Mülheim a. d. Ruhr) for a cold finger dewar. J. E.

out in the introduction the design goal of a multichromophoric would like to express his gratitude to Dr. J. A. Rudd for helpful

3

system does not directly influence the MC energy level and an suggestions and discussion. Special thanks go to S. Pelzer,

3

energetically close positioning to the MLCT or LC states must S. Solyntjes, and M. Werges for their efforts in synthesis.

not increase the chance for fast quenching.

From all the detailed studies within this work the message Keywords: Functional organic materials · Ruthenium ·

can be summarized plainly: While we still assume our rational Ligand design · Electronic structure · Energy transfer ·

design principle is a valid choice, the electronic interplay of the Dyes/Pigments

assembly together with the structural dynamics of such larger

annulated ring systems in terms of distortions and rotational

degrees of freedom was underestimated.

[

1] a) U. S. Schubert, H. Hofmeier, G. R. Newkome, Modern Terpyridine Chem-

istry, Wiley-VCH, 2006; b) U. S. Schubert, A. Winter, G. R. Newkome, Ter-

pyridine-based Materials for Catalytic, Optoelelectronics and Life Science

Applications, Wiley-VCH, Weinheim, 2011; c) L. Hammarström, O. Johans-

son, Coord. Chem. Rev. 2010, 254, 2546–2559; d) E. C. Constable, Chem.

Soc. Rev. 2007, 36, 246–253; e) J.-P. Sauvage, J.-P. Collin, J. C. Chambron,

S. Guillerez, C. Coudret, V. Balzani, F. Barigelletti, L. de Cola, L. Flamigni,

Chem. Rev. 1994, 94, 993–1019.

Further studies, such as temperature-depended measure-

_m_e_n_t_s[

3,51c,51d]

are required as the exact nature of some of the

photophysical processes could not be fully deduced due to the

complicated interplay between the fragments. Therefore, two

options for further work eventuates: i) The synthesis of a rota-

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2698

Full Paper

[

2] O. Johansson, M. Borgström, R. Lomoth, M. Palmblad, J. Bergquist, L.

Hammarström, L. Sun, B. Åkermark, Inorg. Chem. 2003, 42, 2908–2918.

3] A. Amini, A. Harriman, A. Mayeux, Phys. Chem. Chem. Phys. 2004, 6, 1157–

1633–1640; e) Y.-C. Lin, C.-T. Chen, Org. Lett. 2009, 11, 4858–4861; f) Y.-

K. Yang, J. Tae, Org. Lett. 2006, 8, 5721–5723; g) X. Yang, J. Walpita, D.

Zhou, H. L. Luk, S. Vyas, R. S. Khnayzer, S. C. Tiwari, K. Diri, C. M. Hadad,

F. N. Castellano, A. I. Krylov, K. D. Glusac, J. Phys. Chem. B 2013, 117,

15290–15296.

1164.

4] a) E. A. Medlycott, G. S. Hanan, Chem. Soc. Rev. 2005, 34, 133–142; b)

E. A. Medlycott, G. S. Hanan, Coord. Chem. Rev. 2006, 250, 1763–1782.

5] M. Maestri, N. Armaroli, V. Balzani, E. C. Constable, A. M. W. Car-

gill Thompson, Inorg. Chem. 1995, 34, 2759–2767.

6] a) T. Österman, M. Abrahamsson, H.-C. Becker, L. Hammarström, P. Pers-

son, J. Phys. Chem. A 2012, 116, 1041–1050; b) T. Österman, P. Persson,

Chem. Phys. 2012, 407, 76–82.

[22] a) L. Grubert, W. Abraham, Tetrahedron 2007, 63, 10778–10787; b) L.

Grubert, H. Hennig, W. Abraham, Tetrahedron 2009, 65, 5936–5944; c) A.

Vetter, W. Abraham, Org. Biomol. Chem. 2010, 8, 4666–4681.

[23] a) J.-P. Collin, S. Guillerez, J.-P. Sauvage, F. Barigelletti, L. de Cola, L. Flami-

gni, V. Balzani, Inorg. Chem. 1991, 30, 4230–4238; b) G. J. E. Davidson,

S. J. Loeb, P. Passaniti, S. Silvi, A. Credi, Chem. Eur. J. 2006, 12, 3233–3242.

2

+ 3

[

7] a) M. Hissler, A. Harriman, A. Khatyr, R. Ziessel, Chem. Eur. J. 1999, 5,

[24] Compare systems reported by Hanan et al.: [Ru(pm-tpy) ]

MLCT E₀₀

2

3

3366–3381; b) N. D. McClenaghan, Y. Leydet, B. Maubert, M. T. Indelli, S.

(77 K) 1.83–1.87 eV, Ant E₀₀(77 K) 1.85 eV.

Campagna, Coord. Chem. Rev. 2005, 249, 1336–1350; c) X.-y. Wang, A.

Del Guerzo, R. H. Schmehl, J. Photochem. Photobiol. C 2004, 5, 55–77; d)

J. Wang, Y.-Q. Fang, L. Bourget-Merle, M. I. J. Polson, G. S. Hanan, A. Juris,

F. Loiseau, S. Campagna, Chem. Eur. J. 2006, 12, 8539–8548; e) J. Wang,

E. A. Medlycott, G. S. Hanan, F. Loiseau, S. Campagna, Inorg. Chim. Acta

[25]

+

A brief survey on comparable pristine compounds, viz. PhAcr and

Ru(ttpy) ] , results in a Gibbs free energy value of –0.18 eV for the p-

2

e T process.

26] S. Encinas, L. Flamigni, F. Barigelletti, E. C. Constable, C. E. Housecroft,

2+

[

–

[

E. R. Schofield, E. Figgemeier, D. Fenske, M. Neuburger, J. G. Vos, M. Zehn-

der, Chem. Eur. J. 2002, 8, 137–150.

27] A CCDC database search performed for the parent 4′-(thiophen-2-yl)-2,

2007, 360, 876–884.

[

8] a) G. Albano, V. Balzani, E. C. Constable, M. Maestri, D. R. Smith, Inorg.

Chim. Acta 1998, 277, 225–231; b) E. C. Constable, D. R. Smith, Supramol.

Chem. 1994, 4, 5–7.

[

2′:6′2′′-terpyridine gave 21 hits with CDCC No. 163154–55, 693514,

694383–89, 724618, 725923, 778308–09, 790398–400 799943–44,

810489, and 822321 considered.

9] a) C. Bhaumik, S. Das, D. Maity, S. Baitalik, Dalton Trans. 2012, 41, 2427–

2438; b) D. Mondal, M. Bar, S. Mukherjee, S. Baitalik, Inorg. Chem. 2016,

[

28] E. C. Constable, R. W. Handel, C. E. Housecroft, A. Farran Morales, B. Ven-

tura, L. Flamigni, F. Barigelletti, Chem. Eur. J. 2005, 11, 4024–4034.

29] a) E. C. Constable, A. M. W. Cargill Thompson, J. Chem. Soc., Dalton Trans.

5

5, 9707–9724; c) D. Maity, C. Bhaumik, D. Mondal, S. Baitalik, Inorg.

Chem. 2013, 52, 13941–13955; d) S. Karmakar, D. Maity, S. Mardanya, S.

Baitalik, Inorg. Chem. 2014, 53, 12036–12049.

[

1992, 2947–2950; b) E. C. Constable, A. M. W. Cargill Thompson, J. Chem.

[

10] a) T. Schröder, S. N. Sahu, J. Mattay in Topics in Current Chemistry, Vol.

19 (Eds.: M. Albrecht, F. E. Hahn, D. Ajami), Springer, Heidelberg, 2012;

Soc., Dalton Trans. 1994, 1409–1418; c) J. E. Beves, E. L. Dunphy, E. C.

Constable, C. E. Housecroft, C. J. Kepert, M. Neuburger, D. J. Price, S.

Schaffner, Dalton Trans. 2008, 386–396; d) J. E. Beves, D. J. Bray, J. K.

Clegg, E. C. Constable, C. E. Housecroft, K. A. Jolliffe, C. J. Kepert, L. F.

Lindoy, M. Neuburger, D. J. Price, S. Schaffner, F. Schaper, Inorg. Chim.

Acta 2008, 361, 2582–2590; e) J. E. Beves, E. C. Constable, S. Decurtins,

E. L. Dunphy, C. E. Housecroft, T. D. Keene, M. Neuburger, S. Schaffner,

J. A. Zampese, CrystEngComm 2009, 11, 2406–2416.

3

b) F. Wehmeier, J. Mattay, Beilstein J. Org. Chem. 2010, 6, 53; c) F. Weh-

meier, J. Mattay, Beilstein J. Org. Chem. 2010, 6, 54.

[

11] Or A⁺as shorthand, such as, in formulas of complexes.

12] a) J. Hu, B. Xia, D. Bao, A. Ferreira, J. Wan, G. Jones II, V. I. Vullev, J. Phys.

Chem. A 2009, 113, 3096–3107; b) A. C. Benniston, A. Harriman, Chem.

Soc. Rev. 2006, 35, 169–179; c) J. Lappe, R. J. Cave, M. D. Newton, I. V.

Rostov, J. Phys. Chem. B 2005, 109, 6610–6619; d) S. A. Jonker, F. Ariese,

J. W. Verhoeven, Recl. Trav. Chim. Pays-Bas 1989, 108, 109–115; e) G.

Jones II, M. S. Farahat, S. R. Greenfield, D. J. Gosztola, M. R. Wasielewski,

Chem. Phys. Lett. 1994, 229, 40–46; f) H. van Willigen, G. Jones II, M. S.

Farahat, J. Phys. Chem. 1996, 100, 3312–3316.

[

30] a) E. C. Constable, E. L. Dunphy, C. E. Housecroft, W. Kylberg, M. Neubur-

ger, S. Schaffner, E. R. Schofield, C. B. Smith, Chem. Eur. J. 2006, 12, 4600–

4

610; b) E. C. Constable, C. E. Housecroft, A. M. W. Cargill Thompson, P.

Passaniti, S. Silvi, M. Maestri, A. Credi, Inorg. Chim. Acta 2007, 360, 1102–

110; c) E. C. Constable, C. E. Housecroft, M. Neuburger, S. Schaffner, F.

1

[

13] a) A. C. Benniston, A. Harriman, J. W. Verhoeven, Phys. Chem. Chem. Phys.

Schaper, Inorg. Chem. Commun. 2006, 9, 616–619; d) M. G. Lobello, S.

Fantacci, A. Credi, F. de Angelis, Eur. J. Inorg. Chem. 2011, 2011, 1605–

2008, 10, 5156–5158 and references cited therein; b) S. Fukuzumi, H.

Kotani, K. Ohkubo, Phys. Chem. Chem. Phys. 2008, 10, 5159–5162 and

references cited therein; c) A. C. Benniston, A. Harriman, P. Li, J. P. Rostron,

J. W. Verhoeven, Chem. Commun. 2005, 2701–2703; d) S. Zilberg, Phys.

Chem. Chem. Phys. 2010, 12, 10292–10294.

1

613; e) C. Shen, P. Wang, J. E. Beves, Polyhedron 2016, 103, 241–247.

[

31] a) E. C. Constable, M. Devereux, E. L. Dunphy, C. E. Housecroft, J. A. Rudd,

J. A. Zampese, Dalton Trans. 2011, 40, 5505–5515; b) A. L. Kaledin, Z.

Huang, Q. Yin, E. L. Dunphy, E. C. Constable, C. E. Housecroft, Y. V. Geletii,

T. Lian, C. L. Hill, D. G. Musaev, J. Phys. Chem. A 2010, 114, 6284–6297.

32] J. Eberhard, K. Peuntinger, S. Rath, B. Neumann, H.-G. Stammler, D. M.

Guldi, J. Mattay, Photochem. Photobiol. Sci. 2014, 13, 380–396.

[

14] K. Suga, K. Ohkubo, S. Fukuzumi, J. Phys. Chem. A 2003, 107, 4339–4346.

15] a) K. Ohkubo, S. Fukuzumi, Org. Lett. 2000, 2, 3647–3650; b) K. Suga, K.

Ohkubo, S. Fukuzumi, J. Phys. Chem. A 2005, 109, 10168–10175; c) K.

Ohkubo, K. Suga, S. Fukuzumi, Chem. Commun. 2006, 2018–2020; d) X.

Fang, Y.-C. Liu, C. Li, J. Org. Chem. 2007, 72, 8608–8610.

16] a) A. G. Griesbeck, M. Cho, Org. Lett. 2007, 9, 611–613; b) K. Ohkubo, K.

Mizushima, R. Iwata, S. Fukuzumi, Chem. Sci. 2011, 2, 715–722; c) K. Oh-

kubo, A. Fujimoto, S. Fukuzumi, Chem. Commun. 2011, 47, 8515–8517;

d) D. J. Wilger, N. J. Gesmundo, D. A. Nicewicz, Chem. Sci. 2013, 4, 3160–

[

33] a) S. Fukuzumi, K. Ohkubo, Y. Tokuda, T. Suenobu, J. Am. Chem. Soc.

2

1

000, 122, 4286–4294; b) K. Lehmstedt, F. Dostal, Ber. Dtsch. Chem. Ges.

939, 72, 804–806; c) T. Suzuki, Y. Yoshimoto, T. Takeda, H. Kawai, K.

[

Fujiwara, Chem. Eur. J. 2009, 15, 2210–2216; d) A. R. Katritzky, W. H.

Ramer, J. Org. Chem. 1985, 50, 852–856; e) H. Kawai, T. Takeda, K. Fuji-

wara, M. Wakeshima, Y. Hinatsu, T. Suzuki, Chem. Eur. J. 2008, 14, 5780–

3165.

5

793.

[

17] A. Joshi-Pangu, F. Lévesque, H. G. Roth, S. F. Oliver, L.-C. Campeau, D.

Nicewicz, D. A. DiRocco, J. Org. Chem. 2016, 81, 7244–7249.

18] M. H. Shaw, J. Twilton, D. W. C. MacMillan, J. Org. Chem. 2016, 81, 6898–

926.

19] K. Kasama, K. Kikuchi, Y. Nishida, H. Kokubun, J. Phys. Chem. 1981, 85,

148–4153.

20] a) K. Kikuchi, Y. Hattori, C. Sato, H. Kokubun, J. Phys. Chem. 1990, 94,

039–4042; b) Ò. Rubio-Pons, L. Serrano-Andrés, M. Merchán, J. Phys.

[

34] a) A. M. W. Cargill Thompson, Coord. Chem. Rev. 1997, 160, 1–52; b) M.

Heller, U. S. Schubert, Eur. J. Org. Chem. 2003, 2003, 947–961.

35] a) F. Kröhnke, Synthesis 1976, 1–24; b) I. Sasaki, Synthesis 2016, 48, 1974–

6

1

992.

36] a) E. C. Constable, A. M. W. Cargill Thompson, D. A. Tochter, M. A. M.

Daniels, New J. Chem. 1992, 16, 855–867; b) R. Passalacqua, F. Loiseau,

S. Campagna, Y.-Q. Fang, G. S. Hanan, Angew. Chem. Int. Ed. 2003, 42,

1608–1611; Angew. Chem. 2003, 115, 1646.

4

Chem. A 2001, 105, 9664–9673.

[

21] a) D. Zhou, R. Khatmullin, J. Walpita, N. A. Miller, H. L. Luk, S. Vyas, C. M.

Hadad, K. D. Glusac, J. Am. Chem. Soc. 2012, 134, 11301–11303; b) A. J.

Ackmann, J. M. J. Fréchet, Chem. Commun. 1996, 605–606; c) J. W. Bun-

ting, V. S. F. Chew, S. B. Abhyankar, Y. Goda, Can. J. Chem. 1984, 62, 351–

[37] W. Goodall, J. A. G. Williams, J. Chem. Soc., Dalton Trans. 2000, 2893–

2895.

[38] J.-F. Lefebvre, D. Saadallah, P. Traber, S. Kupfer, S. Gräfe, B. Dietzek, I.

Baussanne, J. de Winter, P. Gerbaux, C. Moucheron, M. Chavarot-Kerlidou,

M. Demeunynck, Dalton Trans. 2016, 45, 16298–16308.

354; d) B. Zhou, K. Kano, S. Hashimoto, Bull. Chem. Soc. Jpn. 1988, 61,

Eur. J. Org. Chem. 2018, 2682–2700 www.eurjoc.org

2699

Full Paper

[

39] J. E. Beves, E. C. Constable, C. E. Housecroft, C. J. Kepert, D. J. Price,

CrystEngComm 2007, 9, 456–459.

40] a) E. C. Constable, C. E. Housecroft, E. A. Medlycott, M. Neuburger, F.

Reinders, S. Reymann, S. Schaffner, Inorg. Chem. Commun. 2008, 11, 805–

Benniston, V. Grosshenny, A. Harriman, R. Ziessel, Angew. Chem. Int. Ed.

1994, 33, 1884; Angew. Chem. 1994, 106, 1956–1958; c) A. C. Benniston,

G. M. Chapman, A. Harriman, S. A. Rostron, Inorg. Chem. 2005, 44, 4029–

4

036; d) L. Hammarström, F. Barigelletti, L. Flamigni, M. T. Indelli, N. Ar-

maroli, G. Calogero, M. Guardigli, A. Sour, J.-P. Collin, J.-P. Sauvage, J. Phys.

Chem. A 1997, 101, 9061–9069; e) E. A. Alemán, C. D. Shreiner, C. S.

Rajesh, T. Smith, S. A. Garrison, D. A. Modarelli, Dalton Trans. 2009, 6562–

808; b) J. McMurtrie, I. Dance, CrystEngComm 2009, 11, 1141–1149; c) J.

McMurtrie, I. Dance, CrystEngComm 2010, 12, 3207–3217.

41] P. P. Lainé, F. Bedioui, E. Amouyal, V. Albin, F. Berruyer-Penaud, Chem. Eur.

J. 2002, 8, 3162–3176.

42] a) N. W. Koper, S. A. Jonker, J. W. Verhoeven, C. van Dijk, Recl. Trav. Chim.

Pays-Bas 1985, 104, 296–302; b) P. Hapiot, J. Moiroux, J. M. Saveant, J.

Am. Chem. Soc. 1990, 112, 1337–1343; c) S. Fukuzumi, M. Nishimine, K.

Ohkubo, N. V. Tkachenko, H. Lemmetyinen, J. Phys. Chem. B 2003, 107,

[

6

577.

52] J. Mattay, Angew. Chem. Int. Ed. Engl. 1987, 26, 825–845; Angew. Chem.

987, 99, 849.

[

1

[

53] K. E. Spettel, N. H. Damrauer, J. Phys. Chem. A 2014, 118, 10649–10662.

54] a) H. A. Meylemans, N. H. Damrauer, Inorg. Chem. 2009, 48, 11161–11175;

b) J. Wang, F.-Q. Bai, B.-H. Xia, L. Sun, H.-X. Zhang, J. Phys. Chem. A 2011,

12511–12518.

1

15, 1985–1991.

[

43] S. Fukuzumi, K. Ohkubo, T. Suenobu, K. Kato, M. Fujitsuka, O. Ito, J. Am.

Chem. Soc. 2001, 123, 8459–8467.

[

55] Test calculations with the PBE0 functional showed a qualitatively equal

MO ordering but the occuppied MO energies shifted 0.27–0.30 eV lower

in energy, while unoccupied orbitals stayed almost isoenergetically to

B3LYP calculated eigenvalues [e.g ΔE (LUMO) 0.004–0.030 eV; see Sup-

porting Information].

[56] A. Wild, C. Friebe, A. Winter, M. D. Hager, U.-W. Grummt, U. S. Schubert,

Eur. J. Org. Chem. 2010, 2010, 1859–1868.

[57] K. Nikitin, H. Müller-Bunz, Y. Ortin, J. Muldoon, M. J. McGlinchey, Org.

+

[

44] In MeCN solution referenced to Fc /Fc; back-referenced from other stan-

dards by the values given in Ref. [47] if necessary.

[

45] Dimerisation of the formed radicals is only observed for the parent, viz.

non-substituted, 9-acridinium ion.

46] a) P. G. Bomben, K. C. D. Robson, P. A. Sedach, C. P. Berlinguette, Inorg.

Chem. 2009, 48, 9631–9643; b) M. Beley, J.-P. Collin, J.-P. Sauvage, H.

Sugihara, F. Heisel, A. Miehé, J. Chem. Soc., Dalton Trans. 1991, 3157–

Lett. 2011, 13, 256–259.

3

159; c) J. T. Hewitt, P. J. Vallett, N. H. Damrauer, J. Phys. Chem. A 2012,

[58] B. Laleu, C. Herse, B. W. Laursen, G. Bernardinelli, J. Lacour, J. Org. Chem.

2003, 68, 6304–6308.

1

16, 11536–11547.

[

47] V. V. Pavlishchuk, A. W. Addison, Inorg. Chim. Acta 2000, 298, 97–102.

48] a) G. Jones II, D.-X. Yan, D. J. Gosztola, S. R. Greenfield, M. R. Wasielewski,

J. Am. Chem. Soc. 1999, 121, 11016–11017; b) G. Jones II, D.-X. Yan, S. R.

Greenfield, D. J. Gosztola, M. R. Wasielewski, J. Phys. Chem. A 1997, 101,

[59] a) C. Kreitner, E. Erdmann, W. W. Seidel, K. Heinze, Inorg. Chem. 2015, 54,

11088–11104; b) P. P. Lainé, F. Loiseau, S. Campagna, I. Ciofini, C. Adamo,

Inorg. Chem. 2006, 45, 5538–5551.

[60] J. Romanova, Y. Sadik, M. R. Ranga Prabhath, J. D. Carey, P. D. Jarowski,

4

939–4942.

49] G. Jones II, D. Yan, J. Hu, J. Wan, B. Xia, V. I. Vullev, J. Phys. Chem. B 2007,

11, 6921–6929.

50] a) H. Kawai, T. Takeda, K. Fujiwara, T. Suzuki, Tetrahedron Lett. 2004, 45,

289–8293; b) H. Kawai, T. Takeda, K. Fujiwara, T. Suzuki, Tetrahedron Lett.

004, 45, 8289–8293.

J. Phys. Chem. C 2017, 121, 2333–2343.

[

[61] a) O. A. Borg, S. S. M. C. Godinho, M. J. Lundqvist, S. Lunell, P. Persson, J.

Phys. Chem. A 2008, 112, 4470–4476; b) M. Abrahamsson, M. J. Lundqvist,

H. Wolpher, O. Johansson, L. Eriksson, J. Bergquist, T. Rasmussen, H.-C.

Becker, L. Hammarström, P.-O. Norrby B. Åkermark, P. Persson, Inorg.

Chem. 2008, 47, 3540–3548.

1

8

2

[

51] a) A. C. Benniston, V. Grosshenny, A. Harriman, R. Ziessel, Angew. Chem.

Int. Ed. Engl. 1994, 33, 1884–1885; Angew. Chem. 1994, 106, 1956; b) A. C.

Received: February 13, 2018

Eur. J. Org. Chem. 2018, 2682–2700

www.eurjoc.org

2700

Article Doi

Synthesis and Properties of Acridine and Acridinium Dye Functionalized Bis(terpyridine) Ruthenium(II) Complexes

DOI: 10.1002/ejoc.201800257

Source and publish data:

Authors:

Article abstract of DOI:10.1002/ejoc.201800257

Full text of DOI:10.1002/ejoc.201800257

Products guided by the article

R&D Labs maybe for 105114-53-2

Relevant to this article

Hot Product