Yellow-light sensitization of a ligand photosubstitution reaction in a ruthenium polypyridyl complex covalently bound to a rhodamine dye

Article abstract of DOI:10.1039/c3dt52643g

The ruthenium complex [Ru(terpy)(bpy)(Hmte)]²⁺ ([1] ²⁺), where terpy is 2,2′;6′,2′′-terpyridine, bpy is 2,2′-bipyridine, and Hmte is 2-methylthioethan-1-ol, poorly absorbs yellow light, and although its quantum yield for the photosubstitution of Hmte by water is comparable at 570 nm and at 452 nm (0.011(4) vs. 0.016(4) at 298 K at neutral pH), the photoreaction using yellow photons is very slow. Complex [1]²⁺ was thus functionalized with rhodamine B, an organic dye known for its high extinction coefficient for yellow light. Complex [Ru(Rterpy)(bpy)(Hmte)]³⁺ ([2]³⁺) was synthesized, where Rterpy is a terpyridine ligand covalently bound to rhodamine B via a short saturated linker. [2]Cl₃ shows a very high extinction coefficient at 570 nm (44000 M^-1 cm^-1), but its luminescence upon irradiation at 570 nm is completely quenched in aqueous solution. The quantum yield for the photosubstitution of Hmte by water in [2]³⁺ was comparable to that in [1]²⁺ at 570 nm (0.0085(6) vs. 0.011(4), respectively), which, in combination with the much better photon collection, resulted in a higher photosubstitution rate constant for [2]³⁺ than for [1]²⁺. The energy of yellow photons is thus transferred efficiently from the rhodamine antenna to the ruthenium center, leading to efficient photosubstitution of Hmte. These results bring new opportunities for extending the photoactivation of polypyridyl ruthenium complexes towards longer wavelengths.

Full text of DOI:10.1039/c3dt52643g

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Yellow-light sensitization of a ligand
photosubstitution reaction in a ruthenium
polypyridyl complex covalently bound to a
rhodamine dye†
Cite this: Dalton Trans., 2014, 43,
4494
Azadeh Bahreman, Jordi-Amat Cuello-Garibo and Sylvestre Bonnet*
The ruthenium complex [Ru(terpy)(bpy)(Hmte)]2+ ([1]2+), where terpy is 2,2’;6’,2’’-terpyridine, bpy is 2,2’-
bipyridine, and Hmte is 2-methylthioethan-1-ol, poorly absorbs yellow light, and although its quantum
yield for the photosubstitution of Hmte by water is comparable at 570 nm and at 452 nm (0.011(4) vs.
0.016(4) at 298 K at neutral pH), the photoreaction using yellow photons is very slow. Complex [1]2+ was
thus functionalized with rhodamine B, an organic dye known for its high extinction coefficient for yellow
light. Complex [Ru(Rterpy)(bpy)(Hmte)]3+ ([2]3+) was synthesized, where Rterpy is a terpyridine ligand co-
valently bound to rhodamine B via a short saturated linker. [2]Cl3 shows a very high extinction coefficient
at 570 nm (44 000 M−1 cm−1), but its luminescence upon irradiation at 570 nm is completely quenched in
aqueous solution. The quantum yield for the photosubstitution of Hmte by water in [2]3+ was comparable
to that in [1]2+ at 570 nm (0.0085(6) vs. 0.011(4), respectively), which, in combination with the much
better photon collection, resulted in a higher photosubstitution rate constant for [2]3+ than for [1]2+. The
energy of yellow photons is thus transferred efficiently from the rhodamine antenna to the ruthenium
center, leading to efficient photosubstitution of Hmte. These results bring new opportunities for extend-
ing the photoactivation of polypyridyl ruthenium complexes towards longer wavelengths.
Received 23rd September 2013,
Accepted 5th December 2013
DOI: 10.1039/c3dt52643g
www.rsc.org/dalton
in the literature dealing with photo dynamic therapy
(PDT),22–24 light activation allows for controlling the amount
Introduction
Ruthenium polypyridyl complexes are known for their rich of reactive oxygen species produced locally, which may contri-
photochemistry, which often requires blue light irradiation.1–7 bute to limiting toxicity and side effects during chemotherapy.
In such complexes, photon absorption into a metal-to-ligand On the other hand, blue light irradiation in vivo has a rather
charge-transfer band (1MLCT) typically situated between 400 limited applicability for PDT since its tissue penetration is
3
and 500 nm leads to the corresponding MLCT state via inter- low.25,26 The fact that the MLCT band of most polypyridyl
system crossing. If the distortion of the octahedral coordi- ruthenium complexes is located in the blue region has been
nation geometry is sufficient to decrease the ligand field restricting, up to now, real phototherapeutic applications of
splitting energy, further thermal population of the metal-cen- these complexes. Thus, it is of great interest to make the
tered excited state (3MC) may result in ligand photosubstitu- photoactivation of ruthenium polypyridyl complexes possible
tion reactions.8–11 When performed in aqueous solution, such also with photons of longer wavelengths, without sacrificing
photoreactions lead to the formation of aqua metal complexes, the complex stability in the dark, which is an important
where one ligand of the coordination sphere is replaced by one requirement in photochemotherapy.
or two water molecules. Recently, these types of photoactive
One strategy, recently reviewed by Brewer et al.,27 is to
metal complexes have been proposed as light-activated drugs design complexes having their MLCT band at higher wave-
in phototherapy, as the aqua photoproducts may typically lengths. Such a strategy sometimes lowers the stability of the
interact with biomolecules and lead to significant cytotoxicity, complexes in the dark, but a few complexes have been pub-
whereas the initial complex may not.12–21 As has been shown lished that are reasonably stable in the dark and photoactive
using red light. A second strategy is the coordination of a fluo-
rescent ligand to the ruthenium center in order to sensitize
Leiden Institute of Chemistry, Gorlaeus Laboratories, Leiden University, P.O. Box
the metal complex with photons of higher wavelength.
9502, Leiden, 2300 RA, The Netherlands. E-mail: bonnet@chem.leidenuniv.nl
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
Mascharak and co-workers28–30 have used this strategy to bring
the sensitization of ruthenium nitrosyl compounds from the
c3dt52643g
4494 | Dalton Trans., 2014, 43, 4494–4505
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
UV to the visible region. Typically, direct coordination of the thioether Hmte ligand by an aqua ligand in the complex
fluorophore to ruthenium promotes merging of the absorption [Ru(terpy)(bpy)(Hmte)]2+ (compound [1]2+, where terpy is
band of both fragments, thus shifting light activation of the 2,2′;6′,2″-terpyridine, bpy is 2,2′-bipyridine, and Hmte is
metal center towards higher wavelengths.31 A third, somewhat 2-(methylthio)ethanol).39 The absorption spectrum of [1]2+
similar strategy is to link the fluorophore to the ruthenium extends up to 610 nm and slightly overlaps with the emission
complex via a non-conjugated linker and to use the “reverse” band of rhodamine B (λem = 570 nm) (Fig. 1b). The rhodamine
FRET effect.
B-functionalized analogue complex [2]3+ (Fig. 1c) may thus
Efficient Förster energy transfer (FRET) from a fluorophore allow energy transfer from the fluorophore to the ruthenium
to a ruthenium center is typically obtained when the 1MLCT center to occur, thus leading to efficient ligand photosubsti-
absorption band of the ruthenium complex overlaps with the tution. The high extinction coefficient of the organic dye may
emission band of the fluorophore. The efficiency of FRET is allow for more efficient photon collection and thus faster
also related to the distance between the fluorophore and the photosubstitution of Hmte when excited near 600 nm, com-
ruthenium center.32–34 When the maximum of the emission pared to complex [1]2+. In this work, the rate and quantum
spectrum of the dye is at a lower wavelength than the absorption yield for the photosubstitution of Hmte in the analogous
maximum of the ruthenium complex, forward FRET is ruthenium complexes [1]2+ and [2]3+ are compared between
obtained.35–37 However, for phototherapeutic applications, both yellow (570 nm) and blue (450 nm) light irradiation, in
photoactivation of the ruthenium complex via forward FRET, order to investigate the efficiency of photosensitization in the
i.e., with photons of low wavelength, is not suitable, and Ru-based ligand exchange process.
“reverse FRET” from
a fluorophore with an emission
maximum at a higher wavelength than that of the absorption
maximum of the ruthenium moiety is preferable.34 Etchenique
and co-workers recently introduced this strategy by coordinat-
Results
Synthesis
ing
a green-emitting, rhodamine B-functionalized nitrile
ligand to a chlorido-bis(bipyridine)ruthenium(II) compound. In order to couple a rhodamine B molecule to the 4′ position
The use of a saturated linker avoided orbital overlap between of the 2,2′;6′,2″-terpyridine (terpy) ligand, an ethanolamine
the organic dye and the ruthenium complex, and green light linker may seem at first sight appropriate. However, in basic
irradiation was shown to result in photosubstitution of the conditions the secondary amide bond resulting from coupling
nitrile ligand, thus releasing the fluorophore from the ruthe- between the primary amine of ethanolamine and the car-
nium complex.38
boxylic acid of rhodamine B cyclizes to a spirolactame, which
We report here a new photoactivatable system, in which leads to quenching of the fluorescence of the dye.40,41 Thus, a
coupling of the rhodamine B dye to the ruthenium center is secondary amine, 2-methylaminoethanol, was used instead,
realized at the 4′ position of a spectator terpyridine ligand because the resulting tertiary amide cannot be deprotonated
(Fig. 1). We recently reported the photosubstitution of the and does not cyclize into the spiro compound. The synthetic
Fig. 1 (a) Chemical structure of [Ru(terpy)(bpy)(Hmte)]2+ ([1]2+). (b) Absorption spectrum of compound [1]2+ (left axis) and emission spectrum of
rhodamine B (right axis). (c) Chemical structure of the rhodamine B-functionalized ruthenium complex [2]3+ and photochemical scheme.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 4494–4505 | 4495

View Article Online
Paper
Dalton Transactions
Scheme 1 Synthetic procedure towards compound 3 and [4]Cl. (a) KOH, DMSO (dry), heating at 60 °C for 3 h, overnight at r.t. Yield of 3: 87%. (b)
POCl3, C2H4Cl2 (dry), reflux, 5 h. (c) I: Et3N, CH3CN (dry), reflux, 14 h, II: KPF6 in water, III: chloride exchange DOWEX resin, acetone–H2O (1 : 1), 4 h,
r.t. Yield: 31% (from compound 3).
Fig. 2 Absorption (a) and emission (b) spectra of rhodamine B, rhodamine B-terpyridine conjugate [4]Cl, and rhodamine B-functionalized ruthe-
nium complex [2]Cl3 in MilliQ water at pH = 7. Excitation: 570 nm, slit width: 3 nm. The concentrations of the solutions used for emission measure-
ments were taken such that their absorbance values at 570 nm were identical in the three solutions (A570 = 0.23).
route towards ligand [4]Cl is shown in Scheme 1. In the first allowed removing the unreacted rhodamine B to afford com-
step, a literature procedure was modified42 to substitute the pound [4]Cl as a purple solid with an overall yield of 31%. The
chloride substituent of 4′-chloro-2,2′;6′,2″-terpyridine by UV-vis spectrum of [4]Cl in water (Fig. 2a and Table 1) showed
2-methylaminoethan-1-ol using KOH as a base, to form com- a red shift of about 14 nm (λabs = 569 nm) compared to rhoda-
pound 3. Two structural isomers, compounds 3 (O-bound) and mine B.
3′ (N-bound) (Scheme 1 and Fig. S1†), can be formed depending
Adapting known synthetic procedures39,44,45 the ruthenium
on the amount of base, the temperature, and the reaction time. complex [2]Cl3 was synthesized as shown in Scheme 2.
By using a low amount of KOH (2.8 eq.) and short reaction
1
times no side product 3′ was detected by H NMR of the crude
product, and compound 3 could be further functionalized.
Table 1 Spectroscopic data in MilliQ water for compounds [2]Cl3, [4]Cl,
and rhodamine B. Emission data were obtained upon excitation at
λ = 570 nm
In the second step, rhodamine B was coupled to 3 following
a modified literature procedure43 involving the acid chloride of
Compound
ε
(λmax
) (M−1 cm−1
)
λmax (abs) (nm)
λmax (em) (nm)
rhodamine B and 3 using Et3N as a base in acetonitrile. After
precipitation from water using PF6− as a counter ion, full water
solubility was recovered by anion exchange to Cl− using an
anion exchange resin. Column chromatography on silica gel
Rhodamine B
[4]Cl
[2]Cl3
120 000
74 000
44 000
555
569
570
576
586
585
4496 | Dalton Trans., 2014, 43, 4494–4505
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Scheme 2 Synthetic route towards ruthenium complexes [5]Cl, [6](PF6)2, and [2]Cl3. (a) MeOH, reflux, 7 h, yield: 54%. (b) I: bpy, LiCl, NEt3, EtOH–
H2O(3 : 1), reflux, 6 h. II: KPF6 in water. Yield: 40%. (c) I: Hmte, AgPF6 (2.6 eq.), acetone–H2O (5 : 3), reflux, 9 h. II: chloride exchange DOWEX resin,
acetone–H2O (1 : 1), 4 h, r.t. Yield: 43%.
Refluxing a mixture of ligand [4]Cl with RuCl3·3H2O in metha- 9.80 ppm (6A) appears at a different chemical shift compared
nol resulted in the paramagnetic complex [5]Cl. Product for- to that in [6](PF6)2 (10.28 ppm). The high resolution mass spec-
mation was followed by TLC and the final product was trum showed two peaks for the product at m/z = 360.45780
characterized by paramagnetic 1H NMR and ESI-MS spec- ([2]3+) and at m/z = 540.18289 ([2–H]2+). Overall, [2]Cl3 and its
trometry. The unpaired electron of the Ru(III) complex gener- analogous complex [2](BF4)2, which was synthesized as
1
ates short relaxation times, which shields the H–1H coupling reported previously,39 are soluble enough in water for studying
and thus results in broad NMR signals. This effect is signifi- their photophysical properties and the photosensitivity of their
cant for hydrogen atoms of the terpyridine moiety in [5]Cl that Ru–S bond.
are close to the paramagnetic ruthenium(III) center. Highly
upfield-shifted signals were observed in methanol-d4 at
Emission measurements and energy transfer
−1.43 ppm, −10.26 ppm, −10.71 ppm and −35.94 ppm for As reported by Etchenique et al. for a similar rhodamine–
T33″, T44″, T55″, or T66″. T3′ and T5′ are more remote from ruthenium system,38 the use of a short linker in [2]3+ was
the paramagnetic center and their signals appear at expected to allow at least some of the energy absorbed by rho-
10.90 ppm.46 The peaks in the 6.90–8.10 ppm region most damine B to be donated to the ruthenium center in the
likely correspond to the rhodamine B moiety and traces of the covalent dyad. The emission and absorption spectra of [2]Cl3
free ligand [4]Cl (see ESI, Fig. S3†). In the ESI-MS spectrum a were measured in water and compared to that of [4]Cl and rho-
peak at m/z = 938.2 for [5]+, and at 902.2 for [5–Cl–H]+ were damine B (Fig. 2b). All compounds absorb strongly in the
found that confirmed the formation of compound [5]Cl.
yellow region, with the extinction coefficient diminished in
In the second step, the complex [Ru(4)(bpy)(Cl)](PF6)2· [4]Cl and [2]Cl3, however, compared to rhodamine (Table 1). In
([6](PF6)2) was obtained as a purple solid via treatment of [5]Cl addition, the emission spectrum of the dyad [2]3+ shows
with 2,2′-bipyridine in the presence of EtN3 and LiCl in an almost complete quenching of the fluorescence of the rhoda-
ethanol–water mixture, followed by column chromatography mine moiety upon excitation at 570 nm. This effect is not
and precipitation with aqueous KPF6. Finally, the water observed with ligand [4]+, which keeps a significant part of the
soluble, potentially photosensitive ruthenium complex [Ru(4)- rhodamine fluorescence, characterized by an emission
(bpy)(Hmte)]Cl3·([2]Cl3) was synthesized by removal of the quantum yield φ4 of 0.12(2) (see Fig. 2b and the Experimental
chloride ligand in [6](PF6)2 using AgPF6 in the presence of an part). In the dyad [2]3+ the overlap integral between the emis-
−
excess of Hmte at elevated temperatures. The PF6 counter sion spectrum of the rhodamine fragment (modeled as [4]+)
ions were then exchanged using a chloride-loaded exchange and the absorbance spectrum of the ruthenium moiety
resin, to form the purple, water-soluble complex [2]Cl3. 1H (modeled as [1]2+) was calculated, from which a Förster dis-
NMR in methanol-d4 showed that the protons of the co- tance R0 = 20.5 Å was found (see the Experimental part and
ordinated Hmte ligand (3.46, 1.81, and 1.36 ppm) are shielded Fig. S10†). In classical photophysics R0 represents the distance
in [2]Cl3 compared to free Hmte (3.75, 2.80, and 2.30 ppm). at which Förster Resonance Energy Transfer (FRET) occurs
Moreover, the characteristic aromatic proton for [2]Cl3 at with an efficiency of 50% (see the Experimental part and ESI†).
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 4494–4505 | 4497

View Article Online
Paper
Dalton Transactions
To evaluate the typical intramolecular distance between the monitored by 1H NMR spectroscopy in D2O. NMR samples
rhodamine fragment and the ruthenium moiety in [2]3+, a containing [2]Cl3 in degassed D2O were prepared, and the
theoretical model of the dyad was prepared in an extended samples were irradiated with blue (λe = 452 nm) or yellow light
1
conformation, and minimized by DFT (see the Experimental (λe = 570 nm) at room temperature. While the H NMR spec-
part). In the minimized structure the distance between the trum of a reference sample in the dark did not change, the
centroid of the central aromatic ring of rhodamine and the spectra of the irradiated samples showed the gradual dis-
nitrogen atom of the central pyridine ring of the terpyridine appearance of the starting compound [2]3+ (δ = 9.76 ppm for
ligand was found to be r = 12.1 Å. This distance is ∼1.7 shorter proton 6A, and δ = 3.48 ppm, 1.83 ppm, and 1.37 ppm for co-
than R0, which predicts efficient to very efficient (φFRET ∼ 96%) ordinated Hmte) and the formation of a single new ruthenium
FRET to occur from the rhodamine dye to the ruthenium frag- complex (δ = 9.61 ppm for proton 6A) and of the free Hmte
ment of [2]3+ upon irradiation at 570 nm. In the absence of ligand (at δ = 3.74, 2.66, and 2.01 ppm). Fig. 3 shows the evolu-
additional non-radiative decay, such energy transfer might tion of the 1H NMR spectra for proton 6A upon irradiation (the
lead to photosubstitution of the Hmte ligand.
complete spectra before and after irradiation are shown in
Fig. S6†). Mass spectra after irradiation were obtained for both
samples, and the peak found at 339.6 is characteristic for the
formation of [Ru(4)(bpy)(D2O)]3+. Integration of the protons 6A
for [2]3+ and [7]3+ indicated typically 40% photoconversion of
[2]3+ to [7]3+ after about 500 min irradiation. The present data
show that a substantial amount of Hmte is indeed photosub-
stituted, not only upon blue light irradiation but also upon
yellow light irradiation, which is absorbed by the rhodamine
dye more than by the ruthenium fragment (see below).
However, these NMR experiments could not provide
Photochemistry
In order to check whether emission quenching of the rhoda-
mine fragment in [2]3+ was indeed due to energy transfer to
ruthenium, the photoreactivity of the dyad [2]Cl3 was investi-
gated upon yellow and blue light irradiation. Irradiation with
1
blue light, i.e., in the MLCT band of the ruthenium complex,
is expected to lead to the photosubstitution of the Hmte
ligand by an aqua ligand, to form [Ru(4)(bpy)(OH2)]3+
(complex [7]3+, see Scheme 3). The formation of [7]3+ was first
Scheme 3 Photosubstitution of Hmte in [2]3+ by an aqua ligand to form [7]3+ upon blue light (λe = 452 nm) or yellow light (λe = 570 nm) irradiation
in aqueous solution.
Fig. 3 Evolution of the 1H NMR spectra of degassed D2O solution of [2]Cl3 upon irradiation with (a) blue light (λe = 452 nm, Δλ1/2 = 8.9 nm) or (b)
yellow light (λe = 570 nm, Δλ1/2 = 8.9 nm). Irradiation times are indicated for each spectrum. Conditions: total ruthenium concentration [Ru]tot
=
5.3 × 10−3 M, room temperature under argon.
4498 | Dalton Trans., 2014, 43, 4494–4505
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 4 Time evolution of the UV-vis spectrum of an aqueous solution of (a) [2]3+ and (b) [1]2+ irradiated with yellow light (λe = 570 nm). Conditions:
photon flux Φ = 5.3 × 10−9 Einstein s−1, irradiation pathlength = 3 cm, T = 298 K. Total ruthenium concentrations: (a) [Ru]tot = 3.4 × 10−5 M and
(b) [Ru]tot = 1.2 × 10−4 M.
quantitative information on the quantum efficiency of the photosubstitution rate constant depends on the photon flux Φ,
light-induced substitution reaction, as light intensities in the the extinction coefficient ελ of RuHmte at the irradiation wave-
e
irradiation setup were difficult to determine.
length, the total absorbance at the irradiation wavelength Ae,
e
In order to get quantitative information about the yellow the probability of absorbance of the photon (1–10−A ), the
and blue light-triggered release of Hmte from complex [2]3+, photosubstitution quantum yield φ, the irradiation pathlength
UV-vis experiments were performed in well-controlled L, and the irradiated volume V.
irradiation conditions. An aqueous solution of [2]Cl3 was
dnRuHmte dn
RuOH2
exposed to yellow light (570 nm) or blue light (452 nm)
ꢀ
¼
¼ kφ ꢁ nRuHmte
ð1Þ
dt
dt
shining from the top of a UV-vis cuvette placed inside the spec-
trophotometer (see Fig. S8). The UV-vis spectra were measured
perpendicular to the irradiating light beam, during light
irradiation. As shown in Fig. 4a, the absorption spectrum of
complex [2]3+ gradually evolved until a steady state was
obtained after 150 and 320 mintues of irradiation with yellow
and blue light, respectively. Isosbestic points at 380 nm,
460 nm, and 557 nm indicate the occurrence of only one
photochemical reaction. From the 1H NMR and mass spec-
trometry studies it is clear that extensive irradiation of [2]3+
leads to the full photoconversion into the aqua complex [7]3+
(RuOH2) (see ESI, Fig. S7†). Thus, in each experiment the con-
centration of [2]3+ and [7]3+ could be calculated from the
extinction coefficients of both species (see ESI†). Using eqn
(1), the photochemical substitution first-order rate constants
ln 2
kφ
t1=2
¼
ð2Þ
ꢀ
ꢁ
ελ L
AeV
e
kφ ¼ Φð1 ꢀ 10ꢀA
Þ
φ
ð3Þ
e
The number of moles of RuHmte remaining in solution,
nRuHmte, was plotted vs. the number of moles of photons Q
absorbed at time t since t = 0, by RuHmte (Fig. 5b and ESI†).
The photosubstitution quantum yields were obtained directly
from the slope of these plots; they were found to be 8.5(6) ×
10−3 and 9.2(7) × 10−3 for yellow and blue light irradiation,
respectively (Table 2). These values are similar, which demon-
strates that once absorbed, a yellow photon has almost the
same probability to lead to ligand photosubstitution as a blue
photon. This can be considered as a generalization to photo-
substitution reactions of Kasha’s rule, which states that the
emission quantum yield of a fluorophore is independent from
the excitation wavelength.
kφ570 and kφ452 could be obtained from the slope of a plot of ln
([RuHmte]/[Ru]tot) vs. irradiation time (Fig. 5a, I–II), where
[RuHmte] and [Ru]tot represent the concentration in [2]3+ and
the total ruthenium concentration in the solution, respectively.
Half-reaction times were calculated using eqn (2). The data are
reported in Table 2; they show that the photoconversion rate
upon yellow light irradiation, kφ570, was twice higher compared
to that obtained upon blue light irradiation (kφ452). Since the
photon flux values at 570 nm and 452 nm (Φ570 and Φ452) were
not equal, the rate constants kφ570 and kφ452 cannot be directly
compared, but the photosubstitution quantum yields have to
be calculated instead. As expressed in eqn (3), the
However, the quantity of RuOH2 formed in
a given
irradiation time depends on the amount of light absorbed by
the complex at the irradiation wavelength as well. In this
regard, the extinction coefficients of compound [2]3+ at
570 nm and 452 nm are very different (4.4(2) × 104 and 4.8(2) ×
103, respectively). Thus, in order to compare the photo-
substitution rates the extinction coefficients must be con-
sidered as well. Multiplying the extinction coefficient by the
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 4494–4505 | 4499

View Article Online
Paper
Dalton Transactions
Fig. 5 (a) Plots of ln([RuHmte]/[Ru]tot) vs. irradiation time; [RuHmte] represents the concentration in [2]3+ or [1]2+, and [Ru]tot the total ruthenium
concentration in the solution. The slope of each plot is −kφ (s−1). (b) Plots of the number of moles of RuHmte vs. the number of moles of photons
absorbed by RuHmte at time t, since t = 0; the slope is the photosubstitution quantum yield φ. (I) RuHmte = [2]3+, [Ru]tot = 3.4 × 10−5 M, yellow light
(λe = 570 nm). (II) RuHmte = [2]3+, [Ru]tot = 3.4 × 10−5 M, blue light (λe = 452 nm). (III) RuHmte = [1]2+, [Ru]tot = 1.2 × 10−4 M, yellow light
(λe = 570 nm). (IV) RuHmte = [1]2+, [Ru]tot = 1.2 × 10−4 M, blue light (λe = 452 nm). Photon fluxes: Φ570 = 5.3(8) × 10−9 Einstein s−1 and Φ452 = 3.0(6) ×
10−9 Einstein s−1
.
Table 2 Photochemical data for the photosubstitution of Hmte by H2O in [2]3+ and [1]2+ in MilliQ water. Condition: T = 298 K, irradiation pathlength
= 3 cm, concentration in [2]3+: 3.4 × 10−5 M, concentration in [1]2+: 1.2 × 10−4
M
Ru complex
λe (nm)
ελ (M−1 cm−1
)
Ф (Einstein s−1
)
kφ (s−1
)
t(1/2) (min)
φ
ξ (φ × ελ )
e
e
[2]3+
[2]3+
[1]2+
[1]2+
570
452
570
452
44 000
4800
450
5.3(8) × 10−9
3.0(6) × 10−9
5.3(8) × 10−9
3.0(6) × 10−9
4.4(3) × 10−4
1.9(3) × 10−4
5.2(2) × 10−5
1.3(4) × 10−4
26(2)
59(2)
220(5)
89(3)
8.5(6) × 10−3
9.2(7) × 10−3
1.1(4) × 10−2
1.6(4) × 10−2
370(15)
44(8)
4.8(5)
6600
100(10)
photosubstitution quantum yield gives a value called the light irradiation (570 nm) the absorption band of [1]2+ at
photosubstitution reactivity (ξ),38 which best represents how 450 nm gradually disappeared to give rise to a new absorption
fast a photoreaction will occur under a given photon flux. Actu- maximum at a higher wavelength corresponding to [Ru(terpy)-
ally, eqn (3) simplifies into eqn (4) when the absorbance Ae is (bpy)(OH2)]2+ ([8]2+, see Fig. 4b). The first-order photosubstitu-
small compared to 1:
tion rate constant was obtained from the slope of the plots of
ln([RuHmte]/[Ru]tot) vs. irradiation time (Fig. 5a, III), and the
photosubstitution quantum yield was obtained as described
above (Fig. 5b, III). The photosubstitution quantum yield of
compound [1]2+ upon blue light irradiation was recently pub-
ꢀ
ꢁ
ꢀ
ꢁ
L
V
L
V
kφ ꢂ ln 10
Φελ φ ¼ ln 10
Φξ
ð4Þ
e
The calculated values of ξ are reported in Table 2. These lished by our group using different irradiation conditions.39
values show that for complex [2]3+ Hmte substitution is one For better comparison with [2]3+ we repeated the measurement
order of magnitude faster with yellow light than with blue under the same irradiation conditions as for [2]3+ (Fig. 5, IV).
light. In fact, ten times more moles of photoproduct ([7]3+) All photochemical data, including half-reaction times, are
were produced upon yellow light irradiation compared to blue reported in Table 2. Like for [2]3+ the photosubstitution
light irradiation at short reaction times. Quantitatively, the quantum yields for [1]2+ upon blue light and yellow light
higher molar absorptivity at 570 nm of complex [2]3+ due to irradiations were found very close to each other, i.e., 0.016(4)
the allowed character of the intraligand π–π* transition of the vs. 0.011(4), respectively. This result confirms our observations
rhodamine B moiety promotes intensive absorption of yellow on [2]3+, that once absorbed by [1]2+, yellow photons are able
photons compared to blue ones.
to lead to ligand photosubstitution as efficiently as blue
In order to evaluate the influence of the rhodamine B photons.
antenna on the photosubstitution of Hmte, similar irradiation
In order to compare the photoreactivity of different com-
experiments were performed on its analogue complex [1]2+, pounds one should compare their ξ values, which depends on
which does not have the fluorophore antenna. Upon yellow both the extinction coefficient (ελ) and the photosubstitution
4500 | Dalton Trans., 2014, 43, 4494–4505
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
quantum yield (φ). Although the photosubstitution quantum with the MLCT-based blue photon absorption in [2]3+, whereas
yields at 570 nm and 452 nm are comparable for both com- it contributes largely to yellow photon absorption.
plexes [1]2+ and [2]3+, the extinction coefficient at 570 nm (ε570
)
Considering on the one hand the almost full fluorescence
is two orders of magnitudes higher for [2]3+ than for [1]2+ due quenching of the rhodamine B moiety in [2]3+, and on the
to the presence of the yellow-absorbing dye, while the values of other hand the very similar photosubstitution quantum yields
ε452 are of the same order of magnitude for both complexes. upon blue (direct) and yellow (indirect) light irradiation,
As a result, under yellow light irradiation ξ is about two orders energy transfer from the rhodamine B moiety to the ruthe-
of magnitude higher for [2]3+ than for [1]2+, and it is still four nium center appears to be highly efficient in [2]3+. While from
times higher than that of [1]2+ under blue light irradiation. Etchenique’s work energy transfer was expected to occur in
Overall, at constant photon flux it is the different extinction [2]3+, it was not expected to be that efficient. Thus, non-
coefficients (ελ) that mostly influence the photosubstitution radiative decay probably occurs mostly from the 3MLCT state
rate constants for [1]2+ and [2]3+ at 450 or 570 nm, whereas of the ruthenium moiety rather than from the S1 excited state
their quantum yields poorly depend on irradiation wavelength. of the rhodamine B moiety. Although deeper photophysical
This result is similar to Kasha’s rule, which states that the studies would be needed to assess the exact nature of the
fluorescence quantum yield of a fluorophore is independent energy transfer mechanism, according to Etchenique’s work
from the irradiation wavelength.47 Indeed, like for fluoro- and to the low r/R0 ratio (see above) the energy transfer in [2]3+
phores where emission always occurs from the lowest singlet is expected to occur via reverse Förster Resonance Energy
excited state, for ruthenium complexes such as [1]2+ or [2]3+ Transfer (reverse FRET), i.e., the modest spectral overlap
photosubstitution is expected to occur from a ruthenium- between the emission of the FRET donor and the absorption
based 3MLCT state via thermal promotion to a nearby dissocia- of the ruthenium acceptor is compensated by the very short
3
3
tive MC state. Reaching the MLCT state can be done either distance r = 12.1 Å between both components in the dyad.
by direct excitation of the 1MLCT band of the ruthenium Other types of energy transfer mechanisms, such as Dexter’s,32
complex, or by excitation of the rhodamine dye followed by cannot be fully dismissed, although they would require direct
energy transfer to the ruthenium fragment. In the case of a orbital overlap between the donor and the acceptor, which,
direct excitation of the metal complex ([1]2+) yellow photons considering the saturated nature of the linker and the similar
need to be absorbed by vibrationally excited ground-state com- shapes of the absorption spectrum of [2]Cl3 and of [4]Cl,
plexes, to be able to lead to the 3MLCT excited state. Once there, seems unlikely.
non-radiative decay will occur with almost the same probability
From a pure photochemical point of view, the sensitization
as when the 3MLCT state is obtained by absorption of a blue of photosubstitution reactions might find application in
photons by a non-vibrationally excited ground state complex. In photoactivated chemotherapy (PACT), for which the practical
the case of indirect excitation of [2]3+ with yellow photons the efficiency of a given compound will depend on the amount of
3MLCT state is probably reached efficiently via absorption by photoproduct generated in a given irradiation time. Thus, at a
the rhodamine group, followed by energy transfer.
given light intensity the photosubstitution quantum yield does
not matter too much, but it is the photosubstitution reactivity
ξ, which also takes the extinction coefficient into account, that
should be considered. On the other hand, it cannot be forgot-
ten that functionalization of a light-activatable metallodrug
Discussion
The covalent binding of a rhodamine B dye to the terpy ligand with large, flat aromatic dye is expected to change many bio-
of the ruthenium complex in [2]3+ leads to rather efficient logical properties of the complex such as its lipophilicity,
photosensitization, as photosubstitution upon yellow light uptake mechanism, and/or mechanism of cytotoxicity. In the
irradiation became faster even compared to blue light end, only compounds that combine good uptake, a low toxicity
irradiation of the parent complex [1]2+. Sensitization seems to in the dark, a high toxicity after ligand substitution, and a
occur via energy transfer from the rhodamine B sensitizer to high photosubstitution reactivity, might be interesting for
the ruthenium complex, as reported by Etchenique.38 By using medicinal purposes.
a short saturated linker, the attachment of rhodamine B to the
ruthenium complex occurs without mixing the orbitals of the
dye and that of the ruthenium complex. Thus, we assume that
Conclusions
the spectrum of [1]2+ is a good model for the contribution of
the ruthenium moiety to the spectrum of [2]3+, i.e., that the Our data show that yellow photons that do not seem to have
excited states of the rhodamine B part and of the ruthenium enough energy to populate the 1MLCT state of [1]2+ or [2]3+
part in [2]3+ are not too much affected by each other. By com- lead, once absorbed, to photosubstitution of Hmte with
paring the extinction coefficient of [2]3+ with that of [1]2+ in almost the same quantum efficiency as that achieved with blue
Table 2, it appears that only 1% of the yellow photons are photons. Thus, for this family of ruthenium compounds
absorbed by the ruthenium-centered 1MLCT band in [2]3+, Kasha’s rule remains valid, i.e., the quantum efficiency of
while this fraction goes up to 73% for blue photons. In fact, photosubstitution reactions does not depend on the energy of
the presence of rhodamine B does not significantly interfere the incoming photons. However, for practical applications
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 4494–4505 | 4501

View Article Online
Paper
Dalton Transactions
irradiating photosensitive complexes such as [1]2+ far down Physik. The eluent for column chromatography purification of
their absorption band does render photon collection less compound [6](PF6)2 was prepared by mixing MeCN, MeOH,
efficient. Upon covalent attachment of an organic dye with and H2O in a 66 : 17 : 17 ratio, followed by addition of solid
high molar absorptivity (here rhodamine B for yellow photons) NaCl until saturation was reached.
the photon collection problem was solved, and for complex
Synthesis
[2]3+ efficient energy transfer from the organic dye to the ruthe-
nium center was observed. The resulting photosubstitution
Compound 3. 2-Methylaminoethanol (45 mg, 0.60 mmol)
reactivity under yellow light irradiation became even higher was added to a suspension of powdered KOH (94 mg,
than that of compound [1]2+ under blue light irradiation, due 1.7 mmol) in dry DMSO (2 mL). The mixture was stirred for
to the much improved collection of yellow photons.
30 min at 60 °C. 4′-Chloro-2,2′:6′,2″-terpyridine (160 mg,
To conclude, it may be noted that sensitizing the ruthe- 0.600 mmol) was added and the mixture was stirred at 60 °C
nium complex with dyes absorbing at still higher wavelengths, for 3 h and then overnight at r.t. Then, the mixture was poured
i.e., up in the red region, might become increasingly difficult. onto water (60 mL). The aqueous phase was extracted with
The efficiency of energy transfer is expected to decrease when DCM (3 × 30 mL) and the organic phases were combined and
the spectrum overlap between the emission of the dye and the dried over MgSO4. DCM was evaporated under reduced
MLCT band of the ruthenium complex becomes smaller, as a pressure and the product was left 24 h under high vacuum at
result of which sensitization might not remain possible with 40 °C to remove trace amounts of DMSO. Compound 3 was
dyes that absorb too far in the red region. In the extreme case obtained as a pale yellow oil (160 mg, 0.520 mmol, 87% yield).
of negligible spectral overlap, the photoreactivity of the metal 1H NMR (300 MHz, CD3OD, 298 K, see Fig. S1† for proton attri-
center and the emission of the fluorophore are expected to bution) δ (ppm) 8.61 (d, J = 4.8 Hz, 2H, T66″), 8.54 (d, J =
decouple. In such a case, the absorbed photons would lead 8.0 Hz, 2H, T33″), 7.96–7.87 (m, 4H, T44″ + T3′ + T5′), 7.41
either to ligand photosubstitution, or to fluorescence, depend- (ddd, J = 7.5, 4.8, 1.1 Hz, 2H, T44″), 4.29 (t, J = 5.2 Hz, 2H, α),
ing on the irradiation wavelength. Such systems might find 3.00 (t, J = 5.2 Hz, 2H, β), 2.46 (s, 3H, γ). 13C NMR (75 MHz,
potential application in molecular imaging, for example to CD3OD, 298 K) δ (ppm) 168.39 (T4′), 158.32 (T1), 157.03 (T1′),
probe the position of a ruthenium complex and follow its fate, 150.09 (T66″), 138.68 (T3′, T5′), 125.43 (T44″), 122.91 (T33″),
either in biological or in artificial systems.18,48
108.35 (T44″), 68.18 (α), 50.84 (β), 35.85 (γ). High resolution
ES-MS m/z (calc): 307.15589 (307.15516, [M + H]+).
Compound [4]Cl. Following a literature procedure43 phos-
phorus oxychloride (60.0 µL, 0.657 mmol) was added to a solu-
tion of rhodamine B (150 mg, 0.313 mmol) in dry 1,2-
dichloroethane (5 mL). The mixture was refluxed for 5 h. The
Experimental section
General
1H and 13C NMR spectra were recorded using a Bruker solvent was evaporated under reduced pressure and the crude
DPX-300 spectrometer; chemical shifts are indicated in ppm mixture was immediately re-dissolved in dry CH3CN (10 mL).
relative to TMS. Electrospray mass spectra were recorded on a Et3N (131 µL, 0.939 mmol) and compound 3 (96 mg,
Finnigan TSQ-quantum instrument by using an electrospray 0.31 mmol) were added and the mixture was refluxed for 14 h.
ionization technique (ESI-MS). High resolution mass spec- The solvent was evaporated under reduced pressure at 30 °C
trometry was performed using a Thermo Finnigan LTQ Orbi- and the crude product was dissolved in water and filtered to
trap mass spectrometer equipped with an electrospray ion remove any solid. The product was precipitated by addition of
source (ESI) in positive mode (source voltage 3.5 kV, sheath KPF6, filtered, washed with H2O, and dried in a desiccator at
gas flow 10, capillary temperature 275 °C) with resolution R = ambient pressure over silica gel blue for 4 h. Exchange of the
60.000 at m/z = 400 (mass range = 150–200) and dioctylphtha- PF6 counter anions with Cl− was achieved by stirring an
−
late (m/z = 391.28428) as “lock mass”. UV-vis spectra were acetone–water solution (1 : 1) of the product with the Cl−
obtained using a Varian Cary 50 UV-vis spectrometer. Emis- exchange resin DOWEX 22 (2.0 g) for 4 h. The resin was fil-
sion spectra were obtained using Shimadzu RF-5301 spectro- tered, acetone was evaporated under reduced pressure at
fluorimeter. The irradiation setup was a LOT 1000 W xenon arc 22 °C, and water was removed using a freeze drier. The
lamp, fitted with a 400FH90-50 Andover standard cutoff filter product was purified by column chromatography on silica gel
and a Andover 450FS10-50 (λe = 452 nm, Δλ1/2 = 8.9 nm) or a (CHCl3–MeOH, 10% to 20% of MeOH). Solvents were evapor-
570FS10-50 (λe = 570 nm, Δλ1/2 = 8.9 nm) interference filter. ated under reduced pressure and compound [4]Cl was
DMSO and dichloroethane were dried over CaSO4 and distilled obtained as a purple solid (75 mg, 0.097 mmol, 31%). 1H NMR
before use. CH3CN was dried using a solvent dispenser Pure- (300 MHz, CD3OD, 298 K, see Fig. S2† for proton attribution) δ
Solve 400. 4′-Chloro-2,2′:6′,2″-terpyridine49 and [Ru(terpy)(bpy)- (ppm) 8.77–8.70 (m, 4H, T33″, T66″), 8.06 (td, J = 7.7, 1.8 Hz,
39
(Hmte)](BF4)2 ([1](BF4)2 were synthesized following literature 2H, T44″), 7.84–7.72 (m, 3H, 5R, 4R, 3R), 7.70 (s, 2H, T3′, T5′),
procedures. AgPF6, LiCl, KPF6 and the anionic exchange resin 7.54 (ddd, J = 7.5, 4.8, 1.2 Hz, 2H, T44″), 7.47 (dd, J = 6.5,
DOWEX 22 were purchased from Sigma-Aldrich. Triethylamine 1.0 Hz, 1H, 5R), 7.34 (d, J = 9.6 Hz, 2H, 10R′, 1R′), 7.01 (dd, J =
was purchased from Acros; KOH and POCl3 were purchased 9.6, 2.5 Hz, 2H, 2R′, 9R′), 6.44 (d, J = 2.4 Hz, 2H, 4R′, 7R′), 3.84
from Merck; and rhodamine B was purchased from Lambda (t, J = 4.3 Hz, 2H, α), 3.74 (t, J = 4.3 Hz, 2H, β), 3.41 (dd, J =
4502 | Dalton Trans., 2014, 43, 4494–4505
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
13.4, 6.6 Hz, 8H, δ), 2.96 (s, 3H, γ), 1.14 (t, J = 7.1 Hz, 12H, ε). 512.15646 (512.15650 [M − 2PF6]2+). UV-vis: λmax (ε in L mol−1
13C NMR (75 MHz, CD3OD, 298 K) δ (ppm) 171.10 (CvO), cm−1) in 9 : 1 acetone–H2O: 570 nm (58 × 103).
167.99 (3R, 8R), 158.80, 158.25, 157.00, 156.87, 156.30, 150.30
Compound [2]Cl3. [6](PF6)2 (30 mg, 0.023 mmol) and AgPF6
(T66″), 138.82 (T55″), 137.59, 133.31 (2R′ + 9R′), 131.73 + (15 mg, 0.060 mmol) were dissolved in a 3 : 5 acetone–H2O
131.65 (3R + 2R + 4R), 131.17, 130.99 (5R), 128.75 (T5′), 125.78 mixture (8 mL). To this solution was added Hmte (156 μL,
(T44″), 122.89 (T33″), 115.19 (1R′ + 10R′), 114.43, 108.06 (T′3), 1.80 mmol). The mixture was refluxed under argon for 9 h in
97.19 (4R′ + 7R′), 68.13 (α + β), 46.80 (δ), 40.53 (γ), 12.78 (ε). the dark, after which it was filtered hot over celite. Acetone was
High resolution ES-MS m/z (calc): 731.37096 (731.37041 [M]+). removed under reduced pressure upon which the crude
UV-vis: λmax (ε in L mol−1 cm−1) in pure H2O: 569 nm (74 000). product with PF6− counter ions precipitated in water. It was fil-
−
Anal. Calcd for C46H47ClN6O3·CHCl3·H2O: C, 62.39; H, 5.57; tered, washed and dried. PF6 ions were exchanged by Cl− by
N, 9.29. Found: C, 61.77; H, 5.75; N, 9.68.
stirring a 1 : 1 acetone–water solution (20 mL) of the crude
Compound [5]Cl. Compound [4]Cl (120 mg, 0.156 mmol) product [2](PF6)3 with ion-exchange resin DOWEX 22 (30 mg)
and RuCl3·3H2O (41 mg, 0.16 mmol) were dissolved in MeOH for 4 h. After filtration of the resin, acetone was evaporated
(20 mL) and refluxed for 7 h under argon. The mixture was under reduced pressure, and water was removed using a freeze
first cooled down to room temperature, and then cooled in an drier machine to afford [2]Cl3 as a reddish purple powder
1
ice bath for 30 min and overnight in the fridge. The precipitate (12 mg, 0.011 mmol, 43%). H NMR (300 MHz, CD3OD, 298 K,
was filtered off and air dried to yield [5]Cl as a dark purple see Fig. S5† for proton attribution) δ (ppm) 9.80 (d, J = 6.1 Hz,
powder (83 mg, 0.075 mmol, 54%). 1H NMR (300 MHz, 1H, 6A), 8.81 (d, J = 8.1 Hz, 1H, 3A), 8.57 (t, J = 8.7 Hz, 2H, 1R′
CD3OD, 298 K, see Fig. S3† for proton attribution) δ (ppm) + 3B), 8.39 (m, 2H, 10R′ + 4A), 8.0–8.05 (m, 4H, 5R + 9R′ + 7R′
10.90 (s, T3′,T5′), 8.07–7.88 (m, 3H), 7.69 (d, J = 6.9 Hz, 2H), + 5A), 7.93 (t, 2H, 4B + 2R′), 7.86–7.73 (m, 5H, 3R + T33″ + T3′
7.55 (d, J = 9.4 Hz, 2H), 7.01 (d, J = 9.7 Hz, 3H), −1.43 (s, T33″/ + T5′), 7.56 (m, 1H, 2R), 7.48–7.32 (m, 4H, 4R′ + 4R + T4 + T4″),
T44″/T55″), −10.26 (s, T33″/T44″/T55″), −10.71 (s, T33″/T44″/ 7.27 (d, J = 7.2 Hz, 1H, 5B), 7.20–7.07 (m, 3H, 6B + T55″), 6.92
T55″), −35.94 (s, T66″). ES-MS m/z (calc): 938.2 (938.2 [M − (d, J = 4.1 Hz, 2H, T6 + T6″), 4.46 (d, J = 5.5 Hz, 2H, α), 3.80 (t,
Cl]+), 902.2 (902.5 [M − 2Cl − H]+).
2H, β), 3.69 (q, 8H, δ), 3.46 (d, J = 5.7 Hz, 2H, HO–CH2), 3.25 (s,
Compound [6](PF6)2. [5]Cl (78 mg, 0.080 mmol), 2,2′-bipyri- 3H, γ), 1.81 (t, J = 5.8 Hz, 2H, CH2-S), 1.43–1.36 (s, 3H, S-CH3),
dine (13 mg, 0.083 mmol), and LiCl (5.0 mg, 0.12 mmol) were 1.28 (t, J = 6.9 Hz, 12H, ε). 13C NMR (75 MHz, CD3OD, 298 K) δ
mixed in a 3 : 1 EtOH–H2O mixture (15 mL) and the solution (ppm) 173.90 (CvO), 168.19 (3R,8R), 159.51, 159.34, 159.30,
was degassed with argon for 5 min, after which Et3N (15 µL, 159.12, 158.99, 158.96, 157.22, 154.52 (6A), 153.41, 140.05,
0.10 mmol) was added. The reaction mixture was refluxed 139.95, 139.09, 137.18, 135.81, 135.50, 133.83, 133.34, 133.24,
under argon for 6 h, and then it was filtered hot over celite. 131.33, 131.16, 129.59, 128.90, 127.22, 126.20, 125.81, 124.98,
The filtrate was evaporated under reduced pressure. Column 115.35, 114.86, 112.95, 97.32, 60.46 (α), 47.05 (β), 46.10 (δ),
chromatography purification was performed over silica gel 46.08 (S-CH3), 39.53 (γ), 38.51 (OH-CH2), 38.08(CH2-S), 12.83
(eluent: MeCN–MeOH–H2O, 66 : 17 : 17: saturated in NaCl, Rf = (ε). High resolution ES MS m/z (calc): 360.45788 (360.45780
0.5). The solvent was evaporated, then the crude product was [M − 3Cl]3+), 540.18291 (540.18289 [M − 3Cl − H]2+). UV-vis:
dissolved in water (50 mL) and precipitated by adding KPF6 λmax (ε in L mol−1 cm−1) in pure H2O: 570 nm (44 × 103).
(∼1 g). After filtration, washing with water and drying in a
Emission measurements
desiccator at ambient pressure over silica gel blue for 5 h, com-
pound [6](PF6)2 was obtained in 40% yield as a dark purple Three stock solutions of rhodamine B (solution A, 2.4 mg in
powder (41 mg, 0.031 mmol). 1H NMR (300 MHz, CD3OD, 50 mL H2O, 1.0 × 10−4 M), compound [4]Cl (solution B, 3.8 mg
298 K, see Fig. S4† for proton notation) δ (ppm) 10.28 (d, J = 5.6 in 50 mL H2O, 1.0 × 10−4 M) and compound [2]Cl3 (solution C,
Hz, 1H, 6A), 8.79 (d, J = 8.2 Hz, 1H, 3A), 8.51 (d, J = 8.1 Hz, 3H, 1.2 mg in 10 mL H2O, 1.0 × 10−4 M) were prepared. 150 µL of
10R′ + 1R′ + 3B), 8.32 (t, J = 8.1 Hz, 1H, 4A), 8.07–7.91 (m, 5H, stock solution A, 100 µL of solution B, or 120 µL of solution C
2R′ + 9R′ + 7R′ + 5R + 5A), 7.89–7.68 (m, 6H, T3′ + T5′ + 3R + 4B was transferred into a quartz cuvette and diluted to 3 mL by
+ T33″), 7.47 (d, J = 7.6 Hz, 1H, 2R), 7.44–7.31 (m, 5H, 4R′ + 4R adding H2O using a micropipette (final concentrations: A′:
+ 5B + T44″), 7.13–7.01 (m, 3H, 6B + T55″), 6.71 (d, J = 2.4 Hz, 5.0 × 10−6 M, B′: 3.3 × 10−6 M, C′: 4.0 × 10−6 M). The absor-
2H, T66″), 4.02 (t, J = 4.5 Hz, 2H, α), 3.88 (d, J = 4.5 Hz, 2H, β), bance of each solution was measured (A570 = 0.23 for all solu-
3.45 (m, 8H, δ), 3.05 (s, 3H, γ), 1.31 (t, J = 12.9 Hz, 12H, ε). 13C tions). Emission spectra were recorded with the same
NMR (75 MHz, CD3OD, 298 K) δ (ppm) 13C NMR (75 MHz, excitation parameters (λe = 570 nm).
CD3OD, 298 K) δ 171.25 (CvO), 166.26 (3R,8R), 160.55, 160.48,
Calculation of the emission quantum yield for [4]Cl. The
159.89, 159.07, 159.03, 158.19, 157.10, 153.86 (6A), 153.78, relative method was applied to obtain the emission quantum
152.85 (4R′), 138.42 (4R′ + 5R), 137.72 (4A), 137.64, 136.70 yield of [4]Cl (φ4). Rhodamine B was used as a reference
(T33″), 133.46 (T44″), 132.43 (2R), 131.99 (T3′), 131.89 (T5′), sample (φref = 0.31 in water).38 The emission spectra of rhoda-
130.98 (4B), 129.85 (3R), 128.74 (4R), 128.57 (5B), 127.96 (5A), mine B and [4]Cl were recorded in optically diluted aqueous
127.43 (6B), 125.09 (10R′ + 1R), 124.85 (3B), 124.58 (3A), 115.40 solutions (A570 ≤ 0.1); the integrated intensities of the emis-
(T55″), 114.40, 110.89 (2R′ + 9R′), 97.76 (6T + 6″T), 69.91(α + β), sion spectra, Dref and D4, respectively, were calculated, to
48.15 (δ), 46.98 (γ), 13.04 (ε). High resolution ESI-MS m/z (calc): afford the emission quantum yield of [4]Cl, φ4, according to
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 4494–4505 | 4503

View Article Online
Paper
Dalton Transactions
1
the equation:
the tube in the dark under argon. The H NMR of the sample
was measured as a reference, and irradiation at 452 nm or
570 nm was started at T = 298 K using the beam of a LOT
1000 W xenon arc lamp filtered with an Andover filter at the
appropriate wavelength, and arriving on the side of the NMR
ꢀ
ꢁꢀ
ꢁ
D4
Dref
Aref
A4
φ4 ¼ φref
where A is the absorbance at the excitation wavelength (λe
=
570 nm). The subscripts ref and 4 refer to the reference tube (see ESI, Fig. S9†). After 220 minutes, 310 minutes, and
(Rhodamine B) and measured sample ([4]Cl), respectively. 480 minutes of irradiation at 452 nm, or 170 minutes,
The value found was φ4 = 0.12(2).
320 minutes, and 530 minutes at 570 nm, 1H NMR spectra
Calculation of the Förster distance R0 and Förster efficiency were measured. A reference sample was also prepared at the
φFRET for the dyad [2]Cl3. The Förster distance R0 (in Å) for the same concentration, and kept in the dark for comparison of
FRET couple made of a rhodamine donor (modeled as [4]+) their 1H NMR spectra. Neither of these reference samples
and a ruthenium acceptor (modeled as [1]2+) was calculated showed any observable conversion in the dark.
using the equation:
UV-vis experiments. 1 mL of a stock solution D of com-
pound [2]Cl3 (1.2 mg in 10 mL H2O, 1.0 × 10−4 M) or 0.8 mL of
a stock solution E of [1](BF4)2 (1.7 mg in 5 mL H2O, 4.5 ×
10−4 M) was transferred into a UV-vis cuvette. The volume of
the solution was completed to 3 mL with H2O (using a micro-
pipette) in the dark (final concentration: D′: 3.4 × 10−5 M, E′:
1.2 × 10−4 M). The UV-vis spectrum of each sample was
measured and afterwards the sample was irradiated at 452 nm
or 570 nm using the beam of a LOT 1000 W xenon arc lamp fil-
tered by an Andover bandpass filter, and directed into an
2.5 mm diameter optical fiber bundle bringing the light verti-
cally into the cuvette, i.e., perpendicular to the horizontal
optical axis of the spectrophotometer (see ESI, Fig. S8†). After
each irradiation period (varying from 1 min to 3 min depend-
ing on the samples) a UV-vis spectrum was measured until a
total irradiation time of 350 minutes and 82 minutes was
reached, for D′ and E′, respectively. The concentrations in
[RuHmte] ([2]3+ or [1]2+) and [RuOH2] ([7]3+ or [8]2+) were deter-
mined by deconvolution, knowing the extinction coefficients
of both species (see ESI†). The evolution of ln([RuHmte]/
[Ru]tot) was plotted as a function of irradiation time, and from
the slope S of these plots kφ at λe = 452 nm and λe = 570 nm
ð
1=6
R0 ¼ 9:78 ꢃ 103ðk2nꢀ4φ4 FDðλÞεAðλÞλ4dλÞ
where k2 represents the relative orientation in space of the
transition dipoles of the donor and acceptor (k2 = 0.667), n is
the reflective index of the solution (1.3), φ4 = 0.12 is the emis-
sion quantum yield of the donor ([4]Cl) in the absence of the
acceptor, FD(λ) is the donor emission intensity at the wave-
length λ, dimensionless, and normalized to an area of 1, and
εA is the extinction coefficient of the ruthenium acceptor at the
wavelength λ. In this equation the unit of λ and εA are cm and
cm−1 M−1, respectively. The overlap integral curve is shown in
Fig. S10,† the overlap integral JDA was 3.05 × 10−15 cm3 M−1
and R0 was calculated to be 20.5 Å.
,
From the DFT model of the dyad [2]3+ (see below) a distance
r = 12.1 Å was found between the centroid of the central aro-
matic ring of rhodamine, and the nitrogen atom of the central
pyridine ring of the terpyridine ligand. Using r as the distance
between the donor and the acceptor in the dyad, a FRET
efficiency φFRET = 0.96 was calculated according to the
equation:
were determined to be 1.9(3) × 10−4 s−1 and 4.4(3) × 10−4 s−1
,
1
for [2]3+, respectively, and 1.3(4) × 10−4 and 5.2(2) × 10−5 s−1
for [1]2+, respectively. Knowing the photon flux and probability
φFRET
¼
6
ð1 + ðr=R0Þ Þ
e
of photon absorption 1–10−A , where Ae is the absorbance of
DFT calculation
the solution at the excitation wavelength λe, the number of
From the X-ray structure of [1](PF6)2 a 3D model of the dyad
[2]3+ was built using the MOLDEN software.50 An extended
configuration of the linker was chosen to maximize the dis-
tance between the ruthenium and rhodamine fragments. The
initial geometry was minimized by DFT using the B3LYP func-
tional and LANL2DZ as the basis set for all elements as
described in the GAMESS-UK package.51 In the final geometry
the centroid Cg1 of the C76–C77–C82–O83–C84–C136 aromatic
ring was calculated using mercury from CCDC, and the Cg1–
N8 distance was measured to be r = 12.1 Å. The Z-matrix of the
minimized structure is given in the ESI,† and a picture of the
model prepared with MOLDEN is given as Fig. S11.†
moles of photons Q absorbed at time t by RuHmte since tirr
=
0 was calculated. Plotting nRuHmte (the number of moles of
RuHmte complex [1]2+ or [2]3+) vs. Q gave a straight line in
each case. The slope of this plot directly corresponds to the
quantum yield of the photosubstitution reaction. The values
for the photosubstitution quantum yields were 9.2(3) × 10−3
and 8.5(3) × 10−3, respectively, for [2]3+ and 1.6(4) × 10−2 and
1.1(4) × 10−2, respectively, for [1]2+, at λe = 452 nm or λe
570 nm, respectively (see ESI†).
=
Acknowledgements
Irradiation experiments
NWO-CW is kindly acknowledged for a Veni grant to S.B.
NMR measurements. [2]Cl3 (3.8 mg, 3.2 μmol) was weighed Prof. E. Bouwman is also wholeheartedly acknowledged for
into an NMR tube and degassed D2O (0.60 mL) was added to fruitful scientific discussions and support.
4504 | Dalton Trans., 2014, 43, 4494–4505
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
26 R. Weissleder and V. Ntziachristos, Nat. Med., 2003, 9, 123–
References
128.
1 C. R. Hecker, P. E. Fanwick and D. R. McMillin, Inorg. 27 S. L. H. Higgins and K. J. Brewer, Angew. Chem., Int. Ed.,
Chem., 1991, 30, 659–666. 2012, 51, 11420–11422.
2 V. Balzani, G. Bergamini, S. Campagna and F. Puntoriero, 28 M. J. Rose, M. M. Olmstead and P. K. Mascharak, J. Am.
in Photochemistry and Photophysics of Coordination Com-
pounds I, ed. V. C. S. Balzani, 2007, vol. 280, pp. 1–36.
3 B. A. McClure, E. R. Abrams and J. J. Rack, J. Am. Chem.
Soc., 2010, 132, 5428–5436.
4 P. S. Wagenknecht and P. C. Ford, Coord. Chem. Rev., 2011,
255, 591–616.
5 E. Baranoff, J. P. Collin, J. Furusho, Y. Furusho, A. C. Laemmel
and J. P. Sauvage, Inorg. Chem., 2002, 41, 1215–1222.
6 A. A. Rachford and J. J. Rack, J. Am. Chem. Soc., 2006, 128,
14318–14324.
Chem. Soc., 2007, 129, 5342–5343.
29 M. J. Rose, N. L. Fry, R. Marlow, L. Hinck and
P. K. Mascharak, J. Am. Chem. Soc., 2008, 130, 8834–8846.
30 N. L. Fry, J. Wei and P. K. Mascharak, Inorg. Chem., 2011,
50, 9045–9052.
31 N. L. Fry and P. K. Mascharak, Acc. Chem. Res., 2011, 44,
289–298.
32 J. P. Sauvage, J. P. Collin, J. C. Chambron, S. Guillerez,
C. Coudret, V. Balzani, F. Barigelletti, L. Decola and
L. Flamigni, Chem. Rev., 1994, 94, 993–1019.
7 S. Bonnet, J. P. Collin, J. P. Sauvage and E. Schofield, Inorg. 33 K. Sienicki and W. L. Mattice, J. Chem. Phys., 1989, 90,
Chem., 2004, 43, 8346–8354. 6187–6192.
8 P. Mobian, J. M. Kern and J. P. Sauvage, Angew. Chem., Int. 34 P. Woolley, K. G. Steinhauser and B. Epe, Biophys. Chem.,
Ed., 2004, 43, 2392–2395. 1987, 26, 367–374.
9 J. Van Houten and R. J. Watts, J. Am. Chem. Soc., 1976, 98, 35 M. H. Lim and S. J. Lippard, Inorg. Chem., 2004, 43, 6366–
4853–4858.
6370.
10 P. C. Ford, Coord. Chem. Rev., 1982, 44, 61–82.
11 S. Bonnet and J.-P. Collin, Chem. Soc. Rev., 2008, 37, 1207–
1217.
36 E. K. Kainmuller, E. P. Olle and W. Bannwarth, Chem.
Commun., 2005, 5459–5461.
37 M.-J. Li, K. M.-C. Wong, C. Yi and V.-W. Yam, Chem.–Eur. J.,
2012, 18, 8724–8730.
12 R. B. Sears, L. E. Joyce, M. Ojaimi, J. C. Gallucci,
R. P. Thummel and C. Turro, J. Inorg. Biochem., 2013, 121, 38 O. Filevich, B. Garcia-Acosta and R. Etchenique, Photochem.
77–87. Photobiol. Sci., 2012, 11, 843–847.
13 B. S. Howerton, D. K. Heidary and E. C. Glazer, J. Am. 39 A. Bahreman, B. Limburg, M. A. Siegler, E. Bouwman and
Chem. Soc., 2012, 134, 8324–8327. S. Bonnet, Inorg. Chem., 2013, 52, 9456–9469.
14 R. N. Garner, J. C. Gallucci, K. R. Dunbar and C. Turro, 40 H. N. Kim, M. H. Lee, H. J. Kim, J. S. Kim and J. Yoon,
Inorg. Chem., 2011, 50, 9213–9215. Chem. Soc. Rev., 2008, 37, 1465–1472.
15 S. Betanzos-Lara, L. Salassa, A. Habtemariam, O. Novakova, 41 M. Beija, C. A. M. Afonso and J. M. G. Martinho, Chem. Soc.
A. M. Pizarro, G. J. Clarkson, B. Liskova, V. Brabec and
P. J. Sadler, Organometallics, 2012, 31, 3466–3479.
16 A. J. Gomes, P. A. Barbougli, E. M. Espreafico and
E. Tfouni, J. Inorg. Biochem., 2008, 102, 757–766.
Rev., 2009, 38, 2410–2433.
42 S. Burazerovic, J. Gradinaru, J. Pierron and T. R. Ward,
Angew. Chem., Int. Ed., 2007, 46, 5510–5514.
43 J. del Marmol, O. Filevich and R. Etchenique, Anal. Chem.,
2010, 82, 6259–6264.
17 F. Barragan, P. Lopez-Senin, L. Salassa, S. Betanzos-Lara,
A. Habtemariam, V. Moreno, P. J. Sadler and V. Marchan, 44 R. A. Leising, S. A. Kubow, M. R. Churchill, L. A. Buttrey,
J. Am. Chem. Soc., 2011, 133, 14098–14108.
18 R. E. Goldbach, I. Rodriguez-Garcia, J. H. van Lenthe,
J. W. Ziller and K. J. Takeuchi, Inorg. Chem., 1990, 29,
1306–1312.
M. A. Siegler and S. Bonnet, Chem.–Eur. J., 2011, 17, 9924– 45 C. A. Bessel, J. A. Margarucci, J. H. Acquaye, R. S. Rubino,
9929.
J. Crandall, A. J. Jircitano and K. J. Takeuchi, Inorg. Chem.,
1993, 32, 5779–5784.
46 M.-P. Santoni, A. K. Pal, G. S. Hanan, A. Proust and
B. Hasenknopf, Inorg. Chem. Commun., 2011, 14, 399–402.
19 N. J. Farrer, L. Salassa and P. J. Sadler, Dalton Trans., 2009,
10690–10701.
20 H. M. Chen, J. A. Parkinson, S. Parsons, R. A. Coxall,
R. O. Gould and P. J. Sadler, J. Am. Chem. Soc., 2002, 124, 47 M. Kasha, Discuss. Faraday Soc., 1950, 14–19.
3064–3082.
48 A. Bahreman, B. Limburg, M. A. Siegler, R. Koning,
A. J. Koster and S. Bonnet, Chem.–Eur. J., 2012, 18, 10271–
10280.
21 D. Crespy, K. Landfester, U. S. Schubert and A. Schiller,
Chem. Commun., 2010, 46, 6651–6662.
22 A. Juarranz, P. Jaen, F. Sanz-Rodriguez, J. Cuevas and 49 E. C. Constable and M. D. Ward, Dalton Trans., 1990, 1405–
S. Gonzalez, Clin. Transl. Oncol., 2008, 10, 148–154. 1409.
23 A. P. Castano, P. Mroz and M. R. Hamblin, Nat. Rev. Cancer, 50 G. Schaftenaar and J. H. Noordik, J. Comput. Aided Mol.
2006, 6, 535–545.
Des., 2000, 14, 123–134.
24 A. M. Bugaj, Photochem. Photobiol. Sci., 2011, 10, 1097–1109.
25 P. Zhang, W. Steelant, M. Kumar and M. Scholfield, J. Am.
Chem. Soc., 2007, 129, 4526–4527.
51 M. F. Guest, I. J. Bush, H. J. J. Van Dam, P. Sherwood,
J. M. H. Thomas, J. H. Van Lenthe, R. W. A. Havenith and
J. Kendrick, Mol. Phys., 2005, 103, 719–747.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 4494–4505 | 4505