D → f energy transfer in a series of IrIII/EuIII dyads: Energy-transfer mechanisms and white-light emission

ARTICLE

pubs.acs.org/IC

d f f Energy Transfer in a Series of Ir^III/Eu^IIIDyads: Energy-Transfer

Mechanisms and White-Light Emission^†

Daniel Sykes,^‡Ian S. Tidmarsh,^‡Andrea Barbieri,^§Igor V. Sazanovich,^‡Julia A. Weinstein,^‡and

Michael D. Ward*^,‡

^‡Department of Chemistry, University of Sheﬃeld, Sheﬃeld S3 7HF, U.K.

^§Istituto ISOFꢀCNR, Via P. Gobetti 101, 40129 Bologna, Italy

S Supporting Information

b

ABSTRACT: An extensive series of blue-luminescent iridium(III) complexes has

been prepared containing two phenylpyridine-type ligands and one ligand containing

two pyrazolylpyridine units, of which one is bound to Ir^IIIand the second is pendant.

Attachment of {Ln(hfac)₃} (Ln = Eu, Gd; hfac = anion of 1,1,1,5,5,5,-hexaﬂuoro-

pentanedione) to the second coordination site aﬀords Ir^III/Ln^IIIdyads. Crystal-

lographic analysis of several mononuclear iridium(III) complexes and one Ir^III/Eu^III

dyad reveals that in most cases the complexes can adopt a folded conformation

involving aromatic π stacking between a phenylpyridine ligand and the bis(pyra-

zolylpyridine) ligand, but in one series, based on CF₃-substituted phenylpyridine

ligands coordinated to Ir^III, the steric bulk of the CF₃group prevents this and a quite

diﬀerent and more open conformation arises. Quantum mechanical calculations well

reproduce these two types of “folded” and “open” conformations. In the Ir^III/Eu^IIIdyads, Ir f Eu energy transfer occurs with varying

degrees of eﬃciency, resulting in partial quenching of the Ir^III-based blue emission and the appearance of a sensitized red emission

from Eu^III. Calculations based on consideration of spectroscopic overlap integrals rule out any signiﬁcant contribution from F€orster

(dipoleꢀdipole) energy transfer over the distances involved but indicate that Dexter-type (exchange) energy transfer is possible if

there is a small electronic coupling that would arise, in part, through π stacking between components. In some cases, an initial

photoinduced electron-transfer step could also contribute to Ir f Eu energy transfer, as shown by studies on isostructural iridium/

gadolinium model complexes. A balance between the blue (Ir-based) and red (Eu-based) emission components can generate

white light.

’ INTRODUCTION

Nd-based emission, could occur by a Dexter-type mechanism

over a distance of 20 Å, facilitated by a long-conjugated oligo-

(phenylene) spacer in the bridging ligand.⁴

In recent years, the use of transition-metal (d-block) chromo-

phores as antenna groups to generate sensitized luminescence

from lanthanide(III) ions in d/f dyads has attracted substantial

attention from several research groups.^1ꢀ4Transition-metal

chromophores, especially polypyridyl complexes of d⁶(e.g., Ru^II,

Os^II, Re^I) and d⁸(especially Pt^II) metal ions, have many desirable

properties for use as energy donors to lanthanides instead of the

more usual organic chromophores. These advantages include (i)

intense absorption arising from fully allowed low-energy charge-

transfer bands, with a maximum that can be selected to be almost

anywhere in the UV/vis/near-IR region, (ii) long-lived excited

states, often of triplet metal-to-ligand charge-transfer (³MLCT)

character, which maximize the likelihood of energy transfer to the

Ln^IIIion, (iii) photochemical stability and kinetic inertness, and

(iv) luminescence of their own, which provides a basis to monitor

energy transfer to the lanthanide(III) acceptor.^1,2The majority

of the work in this area has involved Ln^IIIions with low-energy

excited states that emit in the near-IR region (Nd^III, Yb^III, Er^III, and

Pr^III), which is of considerable interest given the relevance of

near-IR luminescence to areas as diverse as biological imaging

and telecommunications. In some Ru^II/Nd^IIIdyads, we showed

recently that Ru f Nd energy transfer, resulting in sensitized

Generating sensitized luminescence from visible-emitting

lanthanides such as Eu^IIIis, however, more challenging in d/f

dyads because the higher energy of the emissive excited states

(⁵D₀, ca. 17 300 cm^ꢀ1) requires the use of transition-metal

chromophores as sensitizers, which themselves have a suﬃciently

high-energy excited state to act as the energy donors. For the

excited state of a d-block complex to act as an energy donor to

Eu^III, it must lie at least 1700 cm^ꢀ1above the ⁵D₀level to prevent

thermally activated back energy transfer at room temperature.⁵

This requires an excited-state energy of >19 000 cm^ꢀ1for the

donor. If a higher-lying excited state of Eu^IIIis populated as a

result of selection rule restrictions on energy transfer (see later),

then the donor must have an even higher energy content. Thus,

the energies required are considerably higher than is accessible

with the usual d⁶or d⁸metal complexes mentioned earlier; e.g.,

[Ru(bipy)₃]²⁺(bipy = 2,2⁰-bipyridine) has a ³MLCT energy of

ca. 17 000 cm^ꢀ1. However, much recent work on luminescent

Received: April 15, 2011

Published: October 05, 2011

r

2011 American Chemical Society

11323

dx.doi.org/10.1021/ic2007759 Inorg. Chem. 2011, 50, 11323–11339

|

Inorganic Chemistry

ARTICLE

Chart 1

iridium(III) complexes based on cyclometallating phenylpyri-

dine-type ligands has shown that, with judicious choice of ligands

and substituents, blue-emitting complexes with excited-state

energies well in excess of 19 000 cm^ꢀ1can be prepared;^6,7this

has been driven principally by their possible use in organic light-

emitting devices. These are ideal candidates for use in d/f dyads

to generate sensitized emission from Eu^III.

that iridium(III) complexes of this nature can be excited by a

two-photon process using near-IR excitation at 800 nm opens up

the possibility of using low-energy excitation—at wavelengths

where biological samples are relatively transparent—to excite the

iridium(III) chromophore and hence generate the important

Eu^III-based emission.¹⁰

The use of iridium(III) complexes as energy donors in d/f

dyads has received relatively little attention,³despite their ease of

synthesis and the straightforward tunability of the absorption/

emission properties over a wide range; in many of these cases, the

Ln^IIIunit is a near-IR emitter rather than a visible emitter such as

Eu^III.^3aꢀcIn this paper, we report the syntheses, structures, and

photophysical properties of a series of Ir^III/Eu^IIIdyads that

display Ir f Eu photoinduced energy transfer (hereafter abbre-

viated as “PEnT” to avoid confusion with “PET”, which is

commonly used for photoinduced electron transfer), and we

analyze the structural properties that are responsible for diﬀering

degrees of PEnT in the complexes.

Our interest in these is driven by possible applications in two

areas. First, if the extent of d f f energy transfer can be controlled

in these dyads, we can control the relative contributions of the

d-block emission (Ir, blue to green) and the Eu^IIIemission (red),

such that white-light emission can be generated from a single

molecule using a balance of blue/green and red luminescence

components, as demonstrated by De Cola et al. with an Ir^III/Eu^III

dyad^3eand by other groups using diﬀerent combinations of red

and blue emitters.⁸Second, the widespread use of Eu^IIIemission

in biological assays and for imaging purposes⁹requires that its

emission be sensitized at wavelengths that are nondestructive to

biological samples, which means that UV excitation should be

avoided. The use of these iridium(III) chromophores as energy

donors allows relatively low-energy excitation in the visible region

of the spectrum to be used. Additionally, the recent demonstration

’ RESULTS AND DISCUSSION

Design and Synthesis of Iridium(III) Complexes. To pre-

pare the Ir^III/Ln^IIIdyads, we required a series of bridging ligands

in which (i) the terminus binding to Ir^IIIhelps to generate the

desired high-energy excited state and (ii) the other terminus

binds to Ln^IIIions with high affinity. These criteria are well

met by the series of bridging ligands shown in Chart 1, which

comprise two pyrazolylpyridine chelating units attached to a

central aromatic spacer; conveniently, these ligands were avail-

able from different work in which they have been used as the

basis of the self-assembly of polyhedral coordination cages with

transition-metal ions.¹¹Complexes based on {Ir(phpy)₂}⁺frag-

ments (Hphpy = 2-phenylpyridine) but also containing five-

memberednitrogen-richheterocycles in theircoordination sphere,

such as triazole and pyrazole donors, have high-energy lumines-

cence due to the high-energy lowest unoccupied molecular orbital

(LUMO) of the azole ligand.^6,7The use of fluorinated phenylpyr-

idine derivatives conversely stabilizes the highest occupied molec-

ular orbital (HOMO; which is a combination of both Ir and phpy

components). Thus, the use of fluorinated phpy derivatives in

the {Ir(phpy)₂}⁺core, and a pyrazolylpyridine unit as the other

chelating fragment, results in a high HOMO/LUMO gap and

luminescence at the blue end of the visible spectrum.^6,7The

second N,N-chelating terminus of the bridging ligand binds to

{Ln(diketonate)₃} fragments bydisplacingtwo water ligands from

[Ln(diketonate)₃(H₂O)₂] units to generate an eight-coordinate

11324

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

Table 1. Crystal Parameters, Data Collection, and Reﬁnement Details for the Structures in This Paper

complex

[

²⁴Ir L^pPh] CH₂Cl₂

[

²⁴⁵Ir L^pPh] CHCl₃0.5H₂O

[

³⁵Ir L^pPh] CH₂Cl₂

3

b

formula

mol wt

C₄₇H₃₄Cl₂F₄IrN₉O₃

1111.93

C₄₇H₃₂Cl₃F₆IrN₉O_3.5

1191.37

C₄₇H₃₄Cl₂F₄IrN₉O₃

1111.93

T, K

150(2)

cryst syst

space group

a, Å

b, Å

c, Å

α, deg

β, deg

γ, deg

V, Å³

triclinic

P1

triclinic

P1

monoclinic

P2₁/c

21.2697(7)

11.2972(3)

18.2126(5)

90

94.062(2)

90

4365.3

9.9873(3)

14.3280(5)

16.9827(6)

113.757(2)

90.314(2)

93.876(2)

2217.75(13)

2

14.2790(19)

19.232(4)

19.724(3)

105.895(2)

110.582(2)

106.298(2)

4428.8(13)

4

Z

4

F, g cm^ꢀ3

cryst size, mm³

μ, mm^ꢀ1

data, restraints, param

ﬁnal R1, wR2^a

1.665

0.75 ꢁ 0.13 ꢁ 0.12

3.201

10 123/15/604

0.0509, 0.1440

1.787

0.44 ꢁ 0.10 ꢁ 0.03

3.278

19 972/181/1192

0.0574, 0.1590

1.692

0.30 ꢁ 0.27 ꢁ 0.02

3.253

7439/59/604

0.0428, 0.1179

²⁴Ir L^mBiph] 2.5CHCl₃

[

²⁴Ir L^pBiph] CHCl₃H₂O^b

[

²⁴Ir L^mPh] CH₂Cl₂

b

c

complex

[

3

formula

mol wt

C_54.5H_38.5Cl_7.5F₄IrN₉O₃

1401.52

C₅₃H₃₉Cl₃F₄IrN₉O₄

1240.48

C₄₇H₃₄Cl₂F₁₀IrN₈P

1194.89

T, K

150(2)

100(2)

150(2)

cryst syst

space group

a, Å

b, Å

c, Å

α, deg

β, deg

γ, deg

V, Å³

monoclinic

P2₁/n

15.7829(7)

20.9113(8)

18.3348(7)

90

104.315(2)

90

5863.3(4)

4

monoclinic

P2₁/c

19.1604(7)

12.2450(4)

23.0661(9)

90

113.843(2)

90

4949.9(3)

4

triclinic

P1

11.1126(3)

13.4765(3)

15.7590(4)

93.0620(10)

103.4640(10)

92.7780(10)

2287.40(10)

2

Z

F, g cm^ꢀ3

cryst size, mm³

μ, mm^ꢀ1

data, restraints, param

ﬁnal R1, wR2^a

1.588

0.28 ꢁ 0.27 ꢁ 0.23

2.682

13 426/88/716

0.0777, 0.2527

1.665

0.35 ꢁ 0.22 ꢁ 0.09

2.392

11 017/4/667

0.0458, 0.1480

1.735

0.20 ꢁ 0.15 ꢁ 0.10

3.156

10 217/1082/745

0.0380, 0.1080

^CF3Ir L^mPh] CH₂Cl₂0.5H₂O^c

[

²⁴⁵Ir L^pPhEu] CH₂Cl₂

d

complex

[

3

formula

mol wt

C₄₉H₃₇Cl₂F₁₂IrN₈O_0.5

1267.94

P

C₈₂H₃₉Cl₂Eu₂F₄₈IrN₈O₁₄

2839.23

T, K

150(2)

cryst syst

space group

a, Å

b, Å

c, Å

α, deg

β, deg

γ, deg

V, Å³

monoclinic

P2₁/n

19.0934(18)

13.1480(12)

21.542(2)

90

114.288(4)

90

4929.2(8)

4

triclinic

P1

13.7794(12)

19.1574(17)

19.4502(17)

93.6220(10)

92.0430(10)

106.0270(10)

4916.7(7)

Z

2

F, g cm^ꢀ3

cryst size, mm³

μ, mm^ꢀ1

data, restraints, param

ﬁnal R1, wR2^a

1.709

0.20 ꢁ 0.15 ꢁ 0.12

2.941

8416/578/665

0.0716, 0.2218

1.918

0.50 ꢁ 0.26 ꢁ 0.18

2.819

22 185/141/1396

0.0485, 0.1373

^aThe value of R1 is based on “observed” data with I > 2σ(I); the value of wR2 is based on all data. ^bThe anion is nitrate. ^cThe anion is

hexaﬂuorophosphate. ^dThe anion is [Eu(hfac)₄]^ꢀ.

11325

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

[Ln(diketonate)₃(NN)] complex with a high binding constant

in a noncompetitive solvent such as CH₂Cl₂(see Chart 1 for an

example).^1a,3f,3g,4

Table 2. Selected Bond Distances (Å) for the Crystal

Structures

²⁴Ir L^pPh

The mononuclear iridium(III) “complex ligands” were simply

prepared in good yield by reaction of the appropriate [Ir(phpy)₂-

(μ-Cl)]₂dimer with 2 equiv of the appropriate bridging ligand L

in a 1:2 ratio such that the Ir:L ratio was 1:1. Chromatographic

puriﬁcation separated the desired mononuclear complexes from

traces of the unwanted dinuclear complexes; the products were

isolated as their nitrate salts and were readily soluble in CH₂Cl₂

and other polar organic solvents. All complexes were satisfacto-

rily characterized by ¹H NMR spectroscopy, mass spectrometry,

and elemental analyses, and many have been structurally char-

acterized. The series of complexes prepared (Chart 1) covers a

range of diﬀerently substituted phenylpyridine ligands to aﬀord

diﬀerent excited-state energies for the Ir^IIIcenters and diﬀerent

lengths and conformations of the bridging ligands to aﬀord

diﬀering degrees of Ir f Eu PEnT.

3

Ir(1)ꢀC(221)

Ir(1)ꢀC(321)

Ir(1)ꢀN(211)

2.009(6)

2.022(6)

2.030(6)

Ir(1)ꢀN(311)

2.036(5)

2.150(5)

2.167(5)

Ir(1)ꢀN(122)

Ir(1)ꢀN(111)

²⁴⁵Ir L^pPh

3

Ir(1)ꢀC(321)

Ir(1)ꢀN(311)

Ir(1)ꢀC(221)

2.005(8)

2.013(7)

2.034(7)

Ir(1)ꢀN(211)

Ir(1)ꢀN(111)

Ir(1)ꢀN(122)

2.035(7)

2.139(6)

2.146(6)

³⁵Ir L^pPh

3

Ir(1)ꢀC(321)

Ir(1)ꢀC(221)

Ir(1)ꢀN(211)

2.013(10)

2.033(11)

2.046(7)

Ir(1)ꢀN(311)

Ir(1)ꢀN(122)

Ir(1)ꢀN(111)

2.058(8)

2.126(7)

2.140(7)

²⁴Ir L^mBiph

3

A note on the nomenclature scheme is appropriate here. The

Ir^IIItermini are denoted as ^HIr (based on unsubstituted phenyl-

pyridine ligands, i.e., bearing only H-atom substituents), ²⁴Ir

[based on 2-(2,4-diﬂuorophenyl)pyridine], ³⁵Ir [based on 2-

(3,5-diﬂuorophenyl)pyridine], ²⁴⁵Ir [based on 2-(2,4,5-triﬂuo-

rophenyl)pyridine], or ^CF3Ir {based on 2-[4-(triﬂuoromethyl)-

phenyl]pyridine}. Terminal {Eu(hfac)₃} or {Gd(hfac)₃} units

are denoted as Eu or Gd. The bridging ligands have the abbre-

viations shown in Chart 1 based on the nature of the aromatic

Ir(1)ꢀC(221)

Ir(1)ꢀC(321)

Ir(1)ꢀN(311)

2.013(11)

2.020(10)

2.054(9)

Ir(1)ꢀN(211)

Ir(1)ꢀN(122)

Ir(1)ꢀN(111)

2.056(8)

2.155(8)

2.165(9)

²⁴Ir L^pBiph

3

Ir(1)ꢀC(41)

Ir(1)ꢀC(21)

Ir(1)ꢀN(11)

1.992(7)

2.008(7)

2.047(6)

Ir(1)ꢀN(31)

Ir(1)ꢀN(61)

Ir(1)ꢀN(51)

2.054(6)

2.158(5)

2.169(5)

²⁴Ir L^mPh

spacer group. Thus, ²⁴Ir L^mPhis the mononuclear complex

3

[Ir{(2,4-F₂ꢀC₆H₃)pyridine}₂(L^mPh)](NO₃), and ²⁴Ir L^mPhEu

Ir(1)ꢀC(321)

Ir(1)ꢀC(221)

Ir(1)ꢀN(211)

2.000(5)

2.015(5)

2.049(4)

Ir(1)ꢀN(311)

Ir(1)ꢀN(122)

Ir(1)ꢀN(111)

2.051(4)

2.158(4)

2.160(5)

3

is the dinuclear adduct with a {Eu(hfac)₃} unit attached to the

secondary pyrazolylpyridine site. We use Ir (with no superscript

denoting substituents) as a generic label for all of the Ir^IIIunits.

Structures of Mononuclear Iridium(III) Complexes. Several

of the mononuclear complexes have been structurally character-

ized, and the structures will be grouped for discussion based on

the nature of the bis(pyrazolylpyridine) ligand L (see Tables 1

and 2). Figure 1 shows the structures of three complexes con-

taining L^pPhbut with different substituents on the terminal

^CF3Ir L^mPh

3

Ir(1)ꢀC(321)

Ir(1)ꢀC(221)

Ir(1)ꢀN(211)

2.003(14)

Ir(1)ꢀN(311)

Ir(1)ꢀN(111)

Ir(1)ꢀN(122)

2.051(11)

2.149(11)

2.178(11)

2.004(14)

2.035(11)

²⁴⁵Ir L^pPhEu

3

Ir(1)ꢀC(321)

Ir(1)ꢀC(221)

Ir(1)ꢀN(211)

Ir(1)ꢀN(311)

Ir(1)ꢀN(122)

Ir(1)ꢀN(111)

Eu(1)ꢀO(523)

2.015(6)

2.029(6)

2.044(5)

2.051(5)

2.146(5)

2.147(4)

2.336(4)

Eu(1)ꢀO(623)

Eu(1)ꢀO(422)

Eu(1)ꢀO(622)

Eu(1)ꢀO(522)

Eu(1)ꢀO(423)

Eu(1)ꢀN(131)

Eu(1)ꢀN(142)

2.347(4)

2.360(4)

2.373(4)

2.386(5)

2.403(5)

2.537(5)

2.556(5)

phenylpyridine ligands, viz., ²⁴Ir L^{pPh 245}Ir L^pPh, and ³⁵Ir L^pPh

.

,

3

Individual bond distances and angles around the Ir^IIIcenters are

unremarkable. The most interesting feature of these structures is

that in every case the pendant phenyl ring of L^pPhis approxi-

mately orthogonal to the coordinated pyrazolylpyridine group

from which it is pendant because of the torsion about the

intervening methylene spacer. This allows the pendant phenyl

ring, containing atoms C(151)ꢀC(156), to lie approximately

parallel to and stacked with the phenylpyridine ligand containing

N(211) and C(221): angles between the phenyl and phenylpyr-

idine mean planes are 7.9, 6.5, and 1.4°, respectively, and

distances are in the range 3.1ꢀ3.6 Å between the atoms of one

aromatic unit and the mean plane of the other. An alternative

and stacked with one of the ﬂuorinated phenylpyridine ligands at

the Ir^IIIcenter.

A comparison of the two structures obtained by incorporating

L^mPh(Figure 3) is particularly revealing. We can see how this

recurring π-stacking motif is made possible by the low steric

demands of those phenylpyridine ligands containing 2,4-diﬂuoro

and 3,5-diﬂuoro substituents on the phenyl ring; these ligands

are planar, which facilitates the stacking interaction. Use instead

of the bulkier CF₃substituent prevents the close approach of

another aromatic unit necessary for π stacking and so disrupts

this interaction. Thus, Figure 3a shows the structure of

view of the structure of ²⁴Ir L^pPhemphasizing this overlap is

3

in Figure 1d. This stacking will be facilitated by the electron-

deficient nature of the coordinated phpy ligand (due to the

pyridine ring, the fluorine substituents, and coordination to

Ir^III) and relies for steric reasons on the absence of out-of-plane

bulky substituents on the phenylpyridine ligand; we return to this

point later.

²⁴Ir L^mPh, which follows the pattern that we have seen in all of the

3

Figure 2 shows the structures of ²⁴Ir L^mBiphand ²⁴Ir L^pBiphin

previous examples (Figures 1 and 2), with the phenylene spacer

[containing atoms C(151)ꢀC(156)] lying parallel to and stacked

with the phenylpyridine ligand containing N(211) and C(221). In

3

which the bis(pyrazolylpyridine) ligands have a biphenyl spacer

separating the bidentate termini. Again, the ﬁrst phenyl ring pen-

dant from the coordinated pyrazolylpyridine group lies parallel to

contrast, ^CF3Ir L^mPhprovides the one instance that we have

3

11326

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

Figure 2. Molecular structures of the complex cations of (a) ²⁴Ir

3

L^mBiphand (b) ²⁴Ir L^pBiph. Color scheme as for Figure 1. Thermal ellip-

3

soids are at the 30% probability level.

conspicuously not stacked with the phpy ligand containing

N(311) and C(321), with phenyl ring C(151)ꢀC(156) and its

pendant pyrazolylpyridine group folded back on itself in a manner

quite diﬀerent from what is observed in every other case. This has

important consequences for energy transfer in Ir/Eu dyads based

on the ^CF3Ir chromophore (see later).

Photophysical Properties of Mononuclear Iridium(III)

Complexes. The photophysical properties of the iridium(III)

complexes are summarized in Tables 3 (UV/vis absorption) and

4 (luminescence) and contain little that is unexpected based

on the known properties of iridium(III) complexes that contain

(i) phenylpyridine ligands with varying patters of F-atom sub-

stituents and (ii) a chelating pyridylpyrazole ligand, both factors

of which combine to give a high-energy excited state and strong

emission in the blue region of the spectrum. The UV/vis absorp-

tion spectra showed, in addition to the expected strong absorp-

tions in the UV region associated with aromatic ligands, a less in-

tense but still strong band (ε between 3200 and 9400 M^ꢀ1cm^ꢀ1

)

in the 360ꢀ410 nm region associated with the formation of the

lowest-energy excited state that has a mixture of ligand-centered

(LC) πꢀπ*, and MLCT, character.

Figure 1. Molecular structures of the complex cations of (a) ²⁴Ir L^pPh

,

The nature of the aromatic group pendant from the coordi-

nated pyrazolylpyridine had a negligible eﬀect such that all com-

plexes with the same primary coordination sphere around Ir^III

have very similar luminescence properties. Thus, we see that the

series ^HIr L and ^CF3Ir L have slightly longer emission wave-

3

(b) ²⁴⁵Ir L^pPh, and (c) ³⁵Ir L^pPh. (d) Alternative view of the structure

3

of ²⁴Ir L^pPhemphasizing the overlap of the phenyl ring of L^pPhwith one

3

of the phenylpyridine ligands. Thermal ellipsoids are at the 30%

probability level.

3

lengths (477 and 481 nm, respectively) than the ²⁴Ir L and

3

observed where this π stacking between the aromatic core of the

pendant ligand and the {Ir(phpy)₂}⁺fragment does not occur

because of the steric bulk of the CF₃substituent (Figure 3b,c),

with the pendant phenyl group of L^mPhnow being oriented

to avoid an unfavorable steric interaction. This is emphasized in

Figure 3b, which shows how the phenyl ring C(151)ꢀC(156) is

²⁴⁵Ir L series (455 and 452 nm, respectively); the change in the

3

position of the F-atom substituents in the series ³⁵Ir L shifts the

3

emission maximum to 475 nm. All of the emission spectra

showed the usual series of well-resolved vibrational components,

with lifetimes (in aerated CH₂Cl₂) of ca. 200 ns for the ^HIr L

3

series and in the region 700ꢀ1600 ns for all of the other series

11327

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

Table 3. Summary of UV/Vis Spectral Data for the Iridium-

(III) Complexes in CH₂Cl₂

complex

λ_max/nm (10^ꢀ3ε/M^ꢀ1cm^ꢀ1

)

^HIr L^pPh

255 (58), 320 (14), 375 (4.9)

3

^HIr L^mPh

262 (69), 283 (53), 325 (18), 379 (9.4)

259 (75), 338 (10), 379 (6.0), 407 (4.5)

255 (71), 283 (43), 379 (5.0)

3

^HIr L^pBiph

3

^HIr L^mBiph

3

²⁴Ir L^pPh

253 (58), 279 (39), 315 (16), 361 (5.6)

251 (56), 280 (40), 360 (6.2)

3

²⁴Ir L^mPh

3

²⁴Ir L^pBiph

258 (55), 284 (38), 320 (13), 363 (3.8)

254 (57), 279 (35), 317 (11), 364 (4.1)

238 (53), 256 (55), 283 (35), 313 (21), 357 (4.4)

239 (58), 256 (65), 313 (20), 359 (3.4)

259 (52), 287 (35), 318 (15), 379 (4.7)

258 (50), 284 (36), 317 (15), 375 (5.1)

259 (45), 284 (31), 333 (5.9), 375 (3.1)

253 (58), 283 (39), 379 (4.9)

3

²⁴Ir L^mBiph

3

²⁴⁵Ir L^pPh

3

²⁴⁵Ir L^mBiph

3

³⁵Ir L^pPh

3

³⁵Ir L^mPh

3

³⁵Ir L^pBiph

3

^CF3Ir L^pPh

3

^CF3Ir L^mPh

253 (58), 282 (40), 379 (4.9), 407 (3.2)

258 (82), 286 (54), 340 (6.2), 382 (3.8)

253 (78), 279 (45), 379 (6.6), 408 (4.5)

3

^CF3Ir L^pBiph

3

^CF3Ir L^mBiph

3

component in the excited state, which is destabilized when the

solvent cannot repolarize to take account of the changed charge

distribution. From the emission maxima at 77 K, it is clear that

this series of complexes provides excited-state energies in the

range 21000ꢀ22000 cm^ꢀ1, which should be more than suﬃcient

to sensitize Eu^III. Although direct energy transfer to generate the

7

⁵D₀level of Eu^IIIfrom the F₀ground state is forbidden by the

relevant selection rules (see the later discussion), there is more

than enough excited-state energy available in these Ir^IIIﬂuoro-

phores to sensitize higher-lying levels of Eu^IIIsuch as ⁵D₁, which

can then generate the emissive ⁵D₀level by internal conversion.

Syntheses of Ir/Ln Dyads and a Structurally Characterized

Example. Preparation of the Ir L Ln (Ln = Gd, Eu) dyads was

3

achieved simply by combining the relevant Ir L mononuclear

3

complex with Ln(hfac)₃2H₂O in a CH₂Cl₂solution. Under

3

these conditions, the equilibrium shown in Chart 1 occurs in

which the chelating diimine ligand displaces the two water mole-

cules to give an eight-coordinate Ln(hfac)₃(diimine) unit. To

ensure that the reaction goes to completion (or very close to

completion), an excess of Ln(hfac)₃2H₂O can be used in a

3

spectroscopic titration; because our photophysical measure-

ments all involved selective excitation of the Ir^III-based chromo-

phore, the excess of unbound Ln(hfac)₃2H₂O present at the

3

end of the spectroscopic titrations is effectively invisible and does

not interfere with the results. This is a general approach that we

Figure 3. Molecular structures of the complex cations of (a) ²⁴Ir L^mPh

3

and (b) ^CF3Ir L^mPh. Color scheme as for Figure 1. (c) Alternative view

have used before.^1a,3f,3g,4In this case, we have used Eu(hfac)₃

3

of the structure of ^CF3Ir L^mPhemphasizing how the steric bulk of the

2H₂O to make the Ir/Eu dyads for the purpose of studying Ir f

3

triﬂuormethyl group prevents aromatic stacking between the phenyl ring

of L^mPhand the phenylpyridine ligands (cf. Figure 1d). Thermal

ellipsoids are at the 30% probability level.

Eu PEnT, and we used Gd(hfac)₃2H₂O to make analogous

3

Ir/Gd complexes as controls to see what effect (if any) the

presence of a nonquenching {Ln(hfac)₃} unit had on the photo-

physical properties of the complexes.

with partially ﬂuorinated ligands. Representative examples of

luminescence spectra are given in Figure 4. Quantum yields

varied from 0.08 (nonﬂuorinated ^HIr L series) to 0.31 (^CF3Ir L

Slow evaporation of a solution containing ²⁴⁵Ir L^pPhand

3

excess Eu(hfac)₃2H₂O in CH₂Cl₂aﬀorded crystals of the dyad

3

²⁴⁵Ir L^pPhEu (Figure 5). Signiﬁcantly, the general arrangement

3

series). In all cases, there was modest rigidochromism with

samples in a frozen MeOH/EtOH glass at 77 K having their

emission maximum blue-shifted by 5ꢀ10 nm compared to emis-

sion in the ﬂuid solution; this is associated with the charge-transfer

of the ²⁴⁵Ir L^pPhunit is not signiﬁcantly changed (cf. Figure 1b),

with the phenyl ring of the bridging ligand [C(151)ꢀC(156)]

lying parallel to and stacked with the phenylpyridine ligand con-

taining N(211) and C(221). However, the pendant pyrazolylpyridine

11328

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

Table 4. Summary of the Photophysical Data for the Mononuclear Iridium(III) Complexes and Their Ir/Ln Adducts^a

complex

λ_em/nm (RT) Φ (RT) τ/ns (RT) λ_em/nm (77 K) % Q (Gd)^b

τ/ns (Gd adduct) ^c

% Q (Eu) ^b

τ/ns (Eu adduct) ^c

^HIr L^pPh

477

0.08

198

194

196

205

467

0

11

0

195 (94), 76 (6)

47

86

11

80

189 (77), 46 (23)

3

^HIr L^mPh

198 (76), 106 (24)

213

194 (23), 23 (78)

172

3

^HIr L^pBiph

3

^HIr L^mBiph

11

220 (90), 120 (10)

209 (12), 33 (88)

3

³⁵Ir L^pPh

475

0.17

818

780

916

467

9

0

826

780

908

53

45

75

549

3

³⁵Ir L^mPh

820 (57), 266 (43)

306

3

³⁵Ir L^pBiph

22

3

^CF3Ir L^pPh

481

0.31

1086

961

472

5

1163

1165

1146

1148

33

30

25

22

727

3

^CF3Ir L^mPh

714 (60), 343 (40)

3

^CF3Ir L^pBiph

1081

1071

670

758

3

^CF3Ir L^mBiph

3

²⁴Ir L^pPh

455

0.13

703

698

744

759

450

84

47

25

59

369 (90), 48 (8), 11(2)^d

478 (90), 180 (9), 12 (1)^d

542 (70), 330 (27), 80 (3)^d

561 (57), 293 (41), 22 (2)^d

84

60

71

89

230 (57), 99 (25), 30 (19) ^d

286 (98), 52 (2)^d

307 (91), 51 (9)^d

3

²⁴Ir L^mPh

3

²⁴Ir•L^pBiph

²⁴Ir L^mBiph

337 (86), 34 (14)^d

3

²⁴⁵Ir L^pPh

452

0.17

1602

1653

447

85

56

795 (56), 334 (39), 29 (5)^d

998 (54), 430 (44), 69 (2)^d

73

86

508 (96), 67 (4)^d

408 (93), 39 (7)^d

3

²⁴⁵Ir L^mBiph

3

^aAll measurements were made in air-equilibrated CH₂Cl₂at room temperature unless otherwise speciﬁed. Quantum yield measurements have an

estimated uncertainty of (20%. ^bExtent of quenching of the Ir-based emission intensity in the Ir/Ln dyad at the end of the titration with the appropriate

[Ln(hfac)₃(H₂O)₂]. ^cWhere one lifetime value is given, the decay ﬁts satisfactorily to a single component (uncertainty (10%); where two or three

lifetime components are required to model the decay proﬁle, the percent contribution of each to the total emission is given in parentheses (see the main

text) and the uncertainties are larger. ^dFor these Ir/Gd and Ir/Eu adducts, the presence of PET in addition to PEnT makes the decay kinetics particularly

complicated and the presence of three components in some cases is necessary to achieve a reasonable ﬁt; these lifetime values are not treated

quantitatively in the main text because of their likely high uncertainty.

Figure 5. Molecular structure of the complex cation of ²⁴⁵Ir L^pPhEu.

3

Color scheme as for Figure 1. Thermal ellipsoids are at the 30%

probability level.

Figure 4. Luminescence spectra (CH₂Cl₂, room temperature) of (i)

^CF3Ir L^pPh, (ii) ³⁵Ir L^pPh, (iii) ^HIr L^pPh, (iv) ²⁴Ir L^pPhand (v) ²⁴⁵Ir

3

L^pPh. Solutions were isoabsorbing at the excitation wavelength such that

diﬀerences in the emission intensities reﬂect diﬀerences in the quantum

yields. Note the slightly lower energy emission for the ﬁrst three cases

compared to the last two.

which portions of Ln(hfac)₃2H₂O (Ln = Gd, Eu) were added

3

to the mononuclear Ir L complexes in CH₂Cl₂until no further

3

changes occurred to Ir-based emission. Formation of the dyads

Ir L Ln resulted in no significant changes in the lowest-energy

3

Ir-based MLCT absorptions, which extend out beyond 400 nm

(any spectral changes to the Ir units at shorter wavelengths

are masked by the strong absorption band associated with

the diketonate units, which results in strong absorbance at wave-

lengths shorter than 370 nm). For luminescence measurements, we

used excitation into this low-energy Ir-centered band at ca. 400 nm

such that there could be no direct excitation of the {Ln(hfac)₃} units.

unit is now part of the coordination sphere of an eight-coordinate

{Eu(hfac)₃(diimine)} unit, which has an approximately square-

antiprismatic geometry: one square face is formed by atoms

N(131), N(142), O(622), and O(623) and the other by atoms

O(422), O(423), O(522), and O(523). The Ir Eu separation

3 3 3

is 9.81 Å.

Photophysical Properties of Ir/Gd Dyads. As described

above, we performed a series of spectroscopic titrations in

Conversion of mononuclear Ir L complexes (typically ≈5 ꢁ

3

10^ꢀ5M^ꢀ1in a CH₂Cl₂solution) to Ir L Gd dyads by titration

3

11329

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

emission maxima 475ꢀ481 nm), as shown in Figure 4 and

Table 4. There is no reason why a higher-energy excited state

should be much more eﬀectively quenched by molecular

vibrations,¹²which suggests that fundamentally diﬀerent beha-

vior is occurring.

The most likely possibility is that PET is occurring from the

excited state of the Ir unit to the pyrazolylpyridine unit coordi-

nated to {Gd(hfac)₃} to give a short-lived Ir^IV/diimine^•ꢀcharge-

separated state, providing a new quenching pathway. This is

plausible given (i) the ease of oxidation of these complexes [we

measured the Ir^III/Ir^IVcouple to be in the range of 1.1ꢀ1.3 V vs

Fc/Fc⁺for the ﬂuorinated Ir complexes] and (ii) the well-known

ability of diimine-type ligands to undergo one-electron reduc-

tions when coordinated to electropositive metal ions; for exam-

ple, free 2,2⁰-bipyridine undergoes an irreversible reduction at

ꢀ1.86 V vs NHE, which shifts to ꢀ1.26 V vs NHE in

[Ru(bipy)₃]²⁺.^13,14We have observed this behavior before in

an anthracenyl/Gd(diimine) dyad in which coordination of a

{Gd(hfac)₃} unit to the diimine made the diimine unit a

suﬃciently good electron acceptor to accept an electron from

the excited state of the anthracene chromophore, generating a

transient anthracenyl^•+/diimine^•ꢀexcited state and quenching

anthracene ﬂuorescence.¹⁵Other Ir/Gd dyads recently pub-

lished by us likewise show quenching of Ir-based emission for

which we proposed the same explanation.^3gWe suggest that

similar behavior is occurring in the ²⁴Ir L Gd and ²⁴⁵Ir L Gd

Figure 6. Changes in luminescence spectra recorded during titration of

(a) ^HIr L^pPh(5.9 ꢁ 10^ꢀ5M) and (b) ²⁴⁵Ir L^pPh(7.8 ꢁ 10^ꢀ5M) with

3

[Gd(hfac)₃(H₂O)₂] in CH₂Cl₂to form the respective Ir L Gd dyads.

3

Note the much greater degree of quenching in the second case (see the

main text for discussion). The concentration of the iridium complex was

typically 10^ꢀ4ꢀ10^ꢀ5M; portions of [Gd(hfac)₃(H₂O)₂] were added

until there was no further change in Ir-based luminescence, which

typically required the addition of <10 equiv. The inset to part b shows

changes in the luminescence intensity at 480 nm as a function of added

[Gd(hfac)₃(H₂O)₂] and the ﬁt of the data to a 1:1 binding isotherm (see

the main text).

3

3 3

dyads, which have the highest-energy excited states of the series,

and that this accounts for the greater quenching of Ir-based

emission in those systems.

Time-resolved measurements of Ir-based emission support

this idea: in all cases where the {Gd(hfac)₃} unit causes

with Gd(hfac)₃2H₂O in a CH₂Cl₂solution resulted in some

signiﬁcant quenching of the Ir unit (i.e., the “anomalous” ²⁴Ir

3

loss of intensity of Ir-based emission to an extent depending on

the substituents on the phenylpyridine ligand attached to Ir^III

(representative examples are given in Figure 6). The changes in

the emission intensity during the titration ﬁtted in all cases to a

1:1 binding isotherm, as expected, with an association constant of

ca. 10⁵M^ꢀ1(the example shown in Figure 6b aﬀorded K_ass= 2 ꢁ

10⁵M^ꢀ1), which is typical for the formation of diimine adducts of

this type in chlorinated solvents as per the equilibrium shown in

Chart 1.^1a

L Gd and ²⁴⁵Ir L Gd series), the decay proﬁle of the residual

3

3 3

luminescence is complex and can only be ﬁtted to multiexpo-

nential decay with a minimum of three components, of which

one is short (tens of nanoseconds). This is consistent with the

presence of multiple conformers in solution that will have

diﬀerent rates of intercomponent PET. The shortest-lived com-

ponents are on the time scale of a few tens of nanoseconds, 1ꢀ2

orders of magnitude shorter than luminescence of the free

iridium complexes, indicative of electron-transfer rates on the

order of 10⁷ꢀ10⁸s^ꢀ1in these conformers. In all of these cases of

multiexponential decay, the errors associated with the lifetime

values for each component are much higher than usual and

the data for these complexes in Table 4 should therefore not be

overanalyzed, but the general behavior is clear. We note that, in

some other conformationally ﬂexible dyad molecules containing

saturated components in the bridging ligands, we and others have

likewise observed multiexponential decay kinetics associated

with diﬀerent molecular conformations.¹⁶

The complexes clearly split into two sets according to their

photophysical behavior. The sets that have the slightly lower-

energy Ir-based excited states, viz., based on ^HIr, ³⁵Ir, and ^CF3Ir,

all show little change in the Ir-based emission intensity when

the Ir L Gd dyads form, with a modest degree of quenching

3

(0ꢀ11% in most cases, with ³⁵Ir L^pBiphGd showing 22%

3

quenching); see Figure 6a for an example. This behavior is

unremarkable and can reasonably be ascribed to changes in the

vibrationalbehavior of the complexes when the bulky{Gd(hfac)₃}

unit binds such that nonradiative deactivation of the Ir-based

excited state becomes slightly more favorable. Gd^IIIis incapable

of acting as an energy acceptor from the excited states of these Ir

units because its lowest excited state lies in the UV region

(>30 000 cm^ꢀ1), hence its regular use as a control in experiments

of this nature.

In clear contrast, the other three series ^HIr L Gd, ³⁵Ir L Gd,

3

3 3

and ^CF3Ir L Gd, where this behavior is not apparent, show

3

much simpler time-resolved luminescence behavior, with a single

component, or one dominant component, for Ir-based emission

that is relatively little aﬀected by the presence of the {Gd(hfac)₃}

unit. On this basis, we can be sure that the quenching of Ir-based

luminescence in the ²⁴Ir L Gd and ²⁴⁵Ir L Gd series is not due

to a mechanical eﬀect associated with binding of the {Gd(hfac)₃}

fragment or it would occur equally in all ﬁve series of complexes.

This leaves a Ir* f (Gd-bound diimine) PET process in the

series ²⁴Ir L Gd and ²⁴⁵Ir L Gd as the most likely explanation

The other two series, ²⁴Ir L Gd and ²⁴⁵Ir L Gd, show quite

3

3 3

diﬀerent behavior, with much more substantial quenching of Ir^III-

based luminescence occurring (up to 85%; Figure 6b). In these

two series, the excited state of the iridium(III) chromophore is

signiﬁcantly higher in energy [room temperature (RT) emission

maxima at 452ꢀ455 nm] than in the ﬁrst set (room temperature

3

3 3

for the strong quenching.¹⁵

11330

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

Photophysical Properties of Ir/Eu Dyads. An exactly similar

series of experiments, adding portions of Eu(hfac)₃2H₂O to the

3

Ir L complexes (ca. 10^ꢀ5M in a CH₂Cl₂solution) to generate

3

dyads Ir L Eu, resulted in every case in partial quenching of Ir-

based emission to an extent comparable to, or greater than, what

3

was observed in the Ir L Gd dyads. Again, the complexes can be

3

split into the same two sets. In the series ^HIr L Eu, ³⁵Ir L Eu,

3

3 3

and ^CF3Ir L Eu, whose gadolinium analogues showed relatively

3

little quenching, the degree of quenching induced by the

presence of Eu^IIIwas much more substantial than that induced

by Gd^III. This is to be expected if a new quenching pathway (Ir f

Eu PEnT) becomes operative. For the other two series ²⁴Ir L

3

Eu and ²⁴⁵Ir L Eu, where quenching in the gadolinium(III)

3

analogues was more substantial, the degree of quenching of the

iridium luminophore induced by the presence of Eu^IIIis approxi-

mately comparable to what was induced by Gd^III, indicating that

the additional PEnT process is not faster than the PET process,

which will still be available.

In every case, the partial quenching of the Ir^IIIunit during

Figure 7. Changes in luminescence spectra recorded during titration of

(a) ²⁴Ir L^pPh(1.3 ꢁ 10^ꢀ4M) and (b) ^CF3Ir L^pPh(4.8 ꢁ 10^ꢀ5M) with

titration with Eu(hfac)₃2H₂O was accompanied by the appear-

3

ance of sensitized emission from Eu^IIIin the form of the usual

[Eu(hfac)₃(H₂O)₂] in CH₂Cl₂to form the respective Ir L Eu dyads.

3

7

⁵D₀ꢀ F_nline series. Given that selective excitation of the

For typical conditions, see the caption to Figure 6. The most intense Eu-

based emission peak at 615 nm is truncated to allow the rest of the

spectra to be seen clearly.

iridium(III) chromophore was used, this sensitized emission

can only arise from Ir f Eu EnT (either directly in the series

^HIr L Eu, ³⁵Ir L Eu, and ^CF3Ir L Eu or indirectly via an initial

3

3 3

electron-transfer step, as described above, in the series ²⁴Ir L

core of the bridging ligand was engaged in π stacking with one of

the phenylpyridine ligands (cf. Figures 1ꢀ3). This stacking is

preserved in the structure of the dyad ²⁴⁵Ir L^pPhEu (Figure 5),

3

Eu and ²⁴⁵Ir L Eu). Two representative examples are shown in

3

Figure 7; these are for the formation of ²⁴Ir L^pPhEu (top) and

3

^CF3Ir L^pPhEu (bottom). In the former case, the Ir-based

implying that bonding of a {Ln(hfac)₃} unit at the secondary

pyrazolylpyridine site need not disturb this disposition of the

bridging ligand, although more than one conformation in solu-

tion is, of course, likely, as mentioned above for the Ir/Gd dyads.

Given the dominance of a Dexter-type pathway for PEnT, on the

basis of our calculations (see below), we suggest that this aro-

matic stacking provides a pathway for weak intercomponent elec-

tronic coupling, which facilitates the PEnT process: the involve-

ment of the phenylpyridine ligands in the excited state supports

this. In contrast, the series ^CF3Ir L Eu provides the only cases in

3

emission intensity is 84% quenched in the dyad, whereas in the

second, it is only 37% quenched, indicative of slower PEnT,

which is less able to compete with radiative deactivation from the

Ir^IIIexcited state. In both sets of luminescence spectra, however,

the rise in the intensity of the characteristic Eu^III-based emission

lines as the titration proceeds, and the associated reduction in the

intensity of Ir^III-based emission, which becomes partially quen-

ched, are clear. This observation illustrates the general situation

that the extent of Ir f Eu PEnT is (i) incomplete in solution¹⁷

and (ii) highly variable. The PEnT rate will depend on electronic

factors such as donor/acceptor spectroscopic overlap, which

depends on the exact properties of the Ir^IIIdonor because the

acceptor is constant in every case, and electronic coupling, in the

case of Dexter-type PEnT. It will also depend on steric factors, in

particular the molecular conformation and intercomponent separa-

tion; and it will depend on the energy-transfer mechanism. This is a

complex issue here because, in addition to the usual possibilities of

F€orster- and Dexter-type mechanisms for Ir f Eu PEnT,^1bthere is

also the possibility of the initial electron-transfer step to generate a

charge-separated Ir^IV/diimine^•ꢀstate (cf. the behavior of some of

the Ir/Gd dyads) whose collapse by back electron transfer would

provide the energy needed to sensitize Eu^III.¹⁵These issues are

discussed in more detail in the next section.

3

which no such π stacking is present because of the steric bulk of

the CF₃groups (cf. the crystal structure in Figure 3b,c) and the

pendant aromatic groups of the bridging ligand L are not in close

electronic contact with the coordinated phenylpyridine ligands,

which participate in excited-state formation. Consequently, the

PEnT process is less eﬃcient, and quenching of the Ir donor by

the {Eu(hfac)₃} unit is diminished.

Time-resolved measurements show the same general picture,

with more complicated decay kinetics for the ²⁴Ir L Eu and

3

²⁴⁵Ir L Eu series, which required ﬁtting to at least two and often

3

three components. The other three series of dyads never required

more than two components to get a good ﬁt to the Ir-based

luminescence decay kinetics, and in many cases, one component

was suﬃcient. As mentioned earlier, the errors associated with

extracting individual lifetimes, especially for minor components,

of multiexponential decays are large, and the data should not be

overinterpreted. We concentrate therefore on the ^HIr L Eu,

The one set of data that does stand out is that in the

^CF3Ir L Eu series the extent of Ir f Eu PEnT is noticeably

3

poorer than that in all of the other series, averaging only ca. 30%

quenching of Ir-based emission across the series of four com-

plexes with diﬀerent bridging ligands. We suggest that this is due

to the molecular conformation shown by the crystal structures

(Figures 1ꢀ3). In all other series of complexes discussed here,

with Ir cores having a diﬀerent pattern of F-atom substituents,

the phenylpyridine ligands are planar, and in every case, the

crystal structures revealed a conformation in which the aromatic

3

³⁵Ir L Eu, and ^CF3Ir L Eu series for quantitative analysis of the

3

3 3

energy-transfer rates.

We start with the ^HIr L Ln series as a well-behaved example.

3

The parent complex ^HIr L^pPhhas an emission lifetime of 198 ns;

in the dyad ^HIr L^pPhGd, at the end of the titration with [Gd-

3

(hfac)₃(H₂O)₂], the emission has two components (195 and

76 ns), with the longer-lived component at the end of the

11331

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

titration likely to arise from traces of free ^HIr L^pPh, with no

well-behaved single-exponential behavior with τ_qvalues of 727,

670, and 758 ns, respectively, aﬀording from eq 1 k_EnTvalues of

5.2 ꢁ 10⁵, 6.2 ꢁ 10⁵, and 4.5 ꢁ 10⁵s^ꢀ1, respectively. For the

remaining complex ^CF3Ir L^mPhEu, Ir-based emission had two

3

{Gd(hfac)₃} bound.

In ^HIr L^pPhEu, there are also two components to Ir-based

3

luminescence decay with τ = 189 and 46 ns. We can estimate

the Ir f Eu EnT rate constant from eq 1, where τ_uis the

“unquenched” emission lifetime of the Ir donor and τ_qis the

partially quenched emission lifetime in the Ir/Eu dyad.

3

components (714 and 343 ns): assuming that these correspond

to diﬀerent conformers of ^CF3Ir L^mPhEu, the k_EnTvalues are 5.4 ꢁ

3

10⁵and 2.1 ꢁ 10⁶s^ꢀ1, respectively. It is clear that across the same

bridging ligands these Ir f Eu energy-transfer rates are at least an

order of magnitude slower from the ^CF3Ir unit than from the ^HIr

unit, which has a more or less identical excited-state energy, and

across the same bridging ligands, which we ascribe to the absence of

aromatic stacking in the complexes and the consequent lack of

electronic coupling between the components, which would inhibit

PEnT if it operates by the Dexter mechanism.

k_EnT¼ 1=τ_qꢀ 1=τ_u

ð1Þ

The most obvious change in emission lifetimes is in the shorter-

lived component in each case, which is reduced from 76 ns in

^HIr L^pPhGd to 46 ns in ^HIr L^pPhEu. Using these ﬁgures, the

3

value of k_EnTderived from eq 1 will reﬂect the additional

quenching due to the possibility of Ir f Eu EnT, which is absent

for the Ir/Gd dyads. Thus, we have τ_u= 76 ns and τ_q= 46 ns,

aﬀording an Ir f Eu energy-transfer rate constant, k_EnT, of 9 ꢁ

Alternative estimates of the Ir f Eu PEnT rate constant, k_EnT

,

can be obtained from the observed reduction in the Ir-based

emission intensity by using the following equation:

10⁶s^ꢀ1

.

The pair of complexes ^HIr L^mPhLn (Ln = Gd, Eu) is similarly

k_EnT¼ 1=τ_u½ðI_u=IÞ ꢀ 1ꢂ

ð2Þ

3

well-behaved, with lifetimes of 198 and 106 ns in the former case

and 194 and 23 ns in the latter, with the 23 ns component

dominating and only a small contribution from the 194 ns

component (which probably just corresponds to a trace of free

where τ_uis the luminescence lifetime of the unquenched donor

and I_u/I is the ratio between the luminescence intensity of the

unquenched donor, I_u, and the Ir-based luminescence intensity

of the lanthanide adduct, I. As discussed above, the three series of

dyads ^HIr L Eu, ³⁵Ir L Eu, and ^CF3Ir L Eu showed only little

^HIr L^mPh). This gives a higher PEnT rate constant of 3.4 ꢁ

3 _ꢀ1

10⁷s , consistent with a shorter Ir Eu separation because of

3 3 3

3

3 3

the presence of a m-phenylene spacer rather than a p-phenylene

spacer.

(or no) Ir-based luminescence quenching upon binding to a

{Gd(hfac)₃} fragment, whereas substantial quenching was in-

duced by binding to a {Eu(hfac)₃} fragment. Thus, in these cases,

we can consider that the observed donor luminescence quench-

ing is due to Ir f Eu PEnT processes that sensitize Eu-based

emission. By application of the equation above and by use of the

values reported in Table 4 [% Q(Eu)], we obtain k_EnTvalues of

4.5 ꢁ 10⁶, 1.4 ꢁ 10⁶, and 4.5 ꢁ 10⁵s^ꢀ1for ^HIr L^pPhEu,

Likewise for ^HIr L^mBiphEu, taking values for τ_qand τ_uof 33

3

and 120 ns (Table 4), we arrive at k_EnT= 2.5 ꢁ 10⁷s^ꢀ1. For the

ﬁnal pair in this series, Ir L^pBiphLn (Ln = Gd, Eu), Ir-based

H

3

luminescence from both could be ﬁtted acceptably to single-

exponential decay (213 and 172 ns, respectively). However,

given the likelihood, based on the ﬁrst three examples, of two

(or more) lifetime components being present, it is likely that

these are weighted average lifetime values for two similar lumi-

nescence components that cannot be deconvoluted; therefore,

these numbers may not correspond to any single species, and

using them for calculation of the EnT rates would be misleading.

This leaves us with a set of three complexes in which the Ir f Eu

energy-transfer rate constant can be calculated with conﬁdence

3

³⁵Ir L^pPhEu, and ^CF3Ir L^pPhEu, respectively, involving the

3

same bridging ligand (L^pPh) in each case. These values are similar

to those obtained by lifetime analysis (vide supra) and, most

importantly, they reﬂect the same trend in the series, with

PEnT in the dyad ^CF3Ir L^pPhEu being slowest because of the

3

absence of aromatic stacking between the components as shown

crystallographically.

to be on the order of 10⁷s^ꢀ1

.

The two series ²⁴Ir L Eu and ²⁴⁵Ir L Eu again show more

3

3 3

The series ³⁵Ir L Eu is also readily interpretable in the same

complex behavior (as did their gadolinium analogues) with at

least two and sometimes three lifetime components for Ir-based

emission. As discussed above, the properties of the ²⁴Ir L Gd

3

way and aﬀords k_EnTvalues of 6.2 ꢁ 10⁵s^ꢀ1, 2.5 ꢁ 10⁶s^ꢀ1, and

1.4 ꢁ 10⁶s^ꢀ1via L^pPh, L^mPh, and L^pBiph, respectively; again we

see that changing from L^pPhto L^mPh, which is expected to shorten

3

and ²⁴⁵Ir L Gd model complexes suggest that an initial elec-

3

the Ir Eu separation, increases the EnT rate constant, but

tron-transfer step from the excited Ir unit to the diimine fragment

coordinated to Gd^IIIis likely, and this will also be possible in the

²⁴Ir L Eu and ²⁴⁵Ir L Eu series: trying to separate this from

3 3 3

changing L^pPhto the longer ligand L^pBiphalso (and unexpec-

tedly) increases k_EnT. Compared to the ^HIr L Eu series, how-

3

3 3

ever, the k_EnTvalues are about an order of magnitude slower, with

all being in the vicinity of 10⁶s^ꢀ1

The series ^CF3Ir L Ln also provides readily interpretable

the additional contributions of F€orster and Dexter energy trans-

fer and the possible presence of diﬀerent conformers, and come

up with meaningful numbers for energy-transfer rates, is there-

fore not possible. We simply note that Ir f Eu PEnT also occurs

in these dyads with substantial quenching of Ir-based emission

(60ꢀ90%) by the Eu^IIIfragment and the associated appearance

of sensitized Eu^IIIemission. On the basis of a comparison of

luminescence intensities (eq 2), a similar conclusion is reached

for the two series ²⁴Ir L Eu and ²⁴⁵Ir L Eu because the lumi-

.

3

luminescence lifetime data, and the time-resolved studies con-

ﬁrm the picture that was apparent from the steady-state measure-

ments of intensity quenching, viz., that Ir f Eu PEnT is

noticeably poor in this series because of the lack of aromatic

stacking (cf. the crystal structures in Figure 3b,c). In all four cases,

the parent iridium complex has a luminescence lifetime of ca.

1 μs, which increases by about 10% in the ^CF3Ir L Gd adducts

3

3 3

nescence quenchings induced by the formation of the gadoli-

nium and europium adducts are comparable. Thus, the additional

contribution of F€orster or Dexter-type PEnT processes cannot

be separated from the initial electron-transfer step and might be

considered negligible.

3

(despite the ca. 5% loss of the emission intensity in each case). In

these cases, therefore, there is negligible PET to the {Gd-

(diimine)} fragment. In three of the four ^CF3Ir L Eu complexes

3

(with L^pPh, L^pBiph, and L^mBiph), residual Ir-based emission shows

11332

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

Table 5. Parameters Used To Evaluate Possible Energy-Transfer Pathways within the Ir L^pPhEu Dyads

3

)

3

a

b

c

dyad

k_EnT(s^ꢀ1

)

η_EnT

J_F(cm³M^ꢀ1

)

R_c(Å)

k_F(s^ꢀ1

J_D(cm)

H (cm^ꢀ1

)

3

^HIr L^pPhEu

44.8 ꢁ 10⁵

13.8 ꢁ 10⁵

4.5 ꢁ 10⁵

0.47

0.53

0.33

2.69 ꢁ 10^ꢀ20

2.95 ꢁ 10^ꢀ20

2.83 ꢁ 10^ꢀ20

2.76 ꢁ 10^ꢀ20

2.73 ꢁ 10^ꢀ20

2.6

3.0

3.3

2.8

3.0

2.39 ꢁ 10³

1.35 ꢁ 10³

1.78 ꢁ 10³

2.14 ꢁ 10³

1.28 ꢁ 10³

1.18 ꢁ 10^ꢀ4

1.25 ꢁ 10^ꢀ4

1.12 ꢁ 10^ꢀ4

1.46 ꢁ 10^ꢀ4

1.38 ꢁ 10^ꢀ4

0.18

0.10

0.06

3

³⁵Ir L^pPhEu

3

^CF3Ir L^pPhEu

3

²⁴Ir L^pPhEu

3

²⁴⁵Ir L^pPhEu

3

^aEstimated from donor luminescence intensity quenching (see the text and eq 2). ^bCalculated for the stacked conﬁguration at d = 9.3 Å (Figure 8). ^cThis

is the electronic coupling required for k_D= k_EnT, i.e., for the Dexter mechanism to dominate the EnT process.

Calculations of Possible Energy-Transfer Pathways. The

nature of possible energy-transfer pathways in d/f dyads of this

type has recently been reviewed.^1bWe consider initially the cases

of F€orster (dipole/dipole) EnT and Dexter-type EnT based on

double electron exchange via bridging ligand orbitals.

The use of available spectroscopic quantities, i.e., the emission

spectra of the Ir L donor models and the absorption spectrum of

3

the Eu³⁺ion, allows us to calculate the spectral overlaps J_Fand J_D.

According to the F€orster theory,¹⁸the critical radius, R_c, deﬁned

as the interchromophoric distance at which the F€orster EnT

rate constant, k_F, equals the intrinsic deactivation constant of

the donor, k_in= 1/τ_u, can be derived as well. The main

parameters used to evaluate the EnT features in the Ir L^pPhEu

3

dyads are summarized in Table 5: for calculations, we have

concentrated on the series of complexes with the same bridging

ligand L^pPh

.

The extremely small values obtained for J_Freﬂect the low

interaction between emission (donor) and absorption (acceptor)

dipoles, mainly due to the extremely small absorption coeﬃcient

of the Eu³⁺ion (ε_max≈ 0.08 M^ꢀ1cm^ꢀ1), which, in turn, result in

a very short critical radius, R_c. According to the obtained R_c

values, a distribution of EnT rate constants k_Fcan be calculated as

a function of the interchromophoric distance d. Geometry

optimization performed on the ²⁴⁵Ir L^pPhEu and ^CF3Ir L^pPh

3

Eu complexes using molecular modeling¹⁹show the existence of

minimum-energy structures (Figure 8) that nicely reproduce the

crystal structures. In the more compact arrangement that occurs

with those phenylpyridine ligands containing only nonhindering

in-plane substituents (such as ²⁴⁵Ir), we see a folded conforma-

tion with a π-stacking interaction between the phenylpyridine

ligand and the central phenyl moiety of the bridge (Figure 8a),

and the two metal centers are ca. 9.3 Å apart (cf. 9.8 Å in the

crystal structure of ²⁴⁵Ir L^pPhEu, vide infra). In contrast in the

3

unfolded structure that is predicted for ^CF3Ir L^pPhEu, in which

3

this π-stacking interaction is not sterically possible, this distance

increases to ca. 13.3 Å (Figure 8b).

Figure 8. Results of molecular modeling studies showing minimum-

energy conformations of (a) ²⁴⁵Ir L^pPhEu and (b) ^CF3Ir L^pPhEu.

Note that these are in good agreement with the crystal structures (vide

supra), with the sterically hindering CF₃group in the latter case resulting

in a more extended conformation with no interligand stacking involving

the phenylpyridine ligands.

The EnT rate constants predicted by the F€orster theory¹⁸(k_F)

for these complexes, calculated for the shortest interchromopho-

ric distance (taken as d_IrꢀEu= 9.3 Å), are around 2 ꢁ 10³s^ꢀ1, i.e.,

2 orders of magnitude lower than the experimentally observed

rate constants (Table 5). Any greater values of d_IrꢀEuwould

reduce the k_Fvalues still further. This deﬁnitely rules out any

signiﬁcant contribution of a PEnT transfer mechanism based on

through-space dipoleꢀdipole interactions; the calculated critical-

transfer radius for PEnT by this mechanism is just ≈3 Å in all

cases. In principle, higher-order multipolar contributions (e.g.,

dipoleꢀquadrupole) can contribute to PEnT in cases where the

dipole/dipole mechanism is inoperative because fꢀf transitions

have signiﬁcant transition quadrupole moments;²⁰however, this

has a higher (d^ꢀ8) distance dependence and, signiﬁcantly, we

3

were able to explicitly rule it out an earlier set of Ru^II/Nd^IIId/f

dyads.⁴

In the case of the double-electron-exchange (Dexter-type)

model,^21ꢀ23the calculated overlap integrals, J_D, are on the order

of 1 ꢁ 10^ꢀ4cm (Table 5) for all complexes. For this mechanism

to be the dominant component of the Ir f Eu EnT process (i.e.,

for k_D= k_EnT), an electronic coupling term H of between 0.06

11333

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

Figure 9. Time-resolved emission proﬁle measured at 615 nm for

²⁴Ir L^pPhEu in CH₂Cl₂using 355 nm excitation. Note the three

domains: (a) rapid decay of the residual Ir-based emission; (b) grow-

in of Eu-based emission on the microsecond time scale; (c) normal slow

decay of Eu-based emission (see the main text). The inset shows an

expansion of the 0ꢀ5 μs region to make components a and b clearer.

Figure 10. TA spectrum of ²⁴Ir L^pPhGd in degassed CH₂Cl₂at room

temperature using 355 nm excitation, reconstructured from a global ﬁt

3

and corresponding to the spectrum of the component with a decay

lifetime of 1.2 μs (see the main text). The spectra of ²⁴Ir L^pPhand

3

²⁴Ir L^pPhEu are identical in appearance and diﬀer only in their decay

3

rates, pointing to a common excited state being observed in each case

and 0.18 cm^ꢀ1is necessary (Table 5). Such a modest electronic

interaction might easily be provided by the ꢀCH₂ꢀC₆H₄ꢀ

CH₂ꢀ bridge connecting the two pyrazolylpyridine side ligands.

It is interesting to note that on this basis we would expect the

lowest H value to occur for ^CF3Ir L^pPhEu, where the presence

(see the main text).

(as calculated earlier), which is fully allowed by the Dexter mechan-

ism, followed by slow nonradiative decay to the emissive ⁵D₀level

and subsequent luminescent decay. Importantly, all of the Ir

energy donor units used in this work (Chart 1) have excited-state

energies of >21000 cm^ꢀ1(Table 4), more than enough to popu-

late the ⁵D₁level of Eu^III(ca. 19 000 cm^ꢀ1) with no signiﬁcant

back energy transfer at room temperature.

3

of the bulky ꢀCF₃substituents on the phenylpyridine ligands

prevents the π-stacked interaction observed in the other two

complexes, ^HIr L^pPhEu and ³⁵Ir L^pPhEu, where the Ir f Eu

3

EnT process is faster. As discussed above, this again implicates

the role of the π interaction in making available an additional

electronic communication pathway to facilitate the vectorial

PEnT process.

Transient Absorption (TA) Measurements. For the systems

where we have suggested that a PET step is operative, viz., the

complex series ²⁴Ir L Gd and ²⁴⁵Ir L Gd in which the pre-

3

3 3

We note that, although luminescence from Eu^IIIoriginates

from the ⁵D₀state, direct population of this level from the ground

⁷F₀state is forbidden. The relevant selection rules are |ΔJ| = 0, 1

for Dexter-type energy transfer (with the speciﬁc case of J = J⁰= 0

being excluded, which applies to the ⁵D₀r ⁷F₀transition) and

|ΔJ| = 2, 4, 6 for multipolar energy transfer.²⁴Although selection

rules can be relaxed by various means, this issue still needs to be

addressed when we consider the energy-transfer process. Accord-

ingly, we performed time-resolved measurements at 615 nm—

the peak of the Eu-based emission—for complex ²⁴Ir L^pPh

sence of the {Gd(hfac)₃} fragment results in substantial quench-

ing of Ir-based emission, we have performed TA measurements

on representative members of the series to see if any evidence for

a charge-separated Ir^IV(diimine^•ꢀ)Gd intermediate could be

3 3 3

found. The examples used were the free iridium complex ²⁴Ir

3

L

^pPhand its adducts ²⁴Ir L Gd and ²⁴Ir L Eu (Figure 10). The

3

3 3

TA spectrum of ²⁴Ir L Gd is illustrated in Figure 10, but the

3

spectra of free ²⁴Ir L^pPhand ²⁴Ir L Eu were essentially identical

3

3 3

and Figure 10 can be used as a basis for the discussion of all three

complexes.

3

Eu on the microsecond time scale. We could clearly identify

(Figure 9) three diﬀerent lifetime components at this wave-

length. These are (i) a rapid decay of ca. 0.15 μs, corresponding

to the weighted average of the rapid, multiexponential Ir-based

decay at this wavelength (cf. Table 4), (ii) a slow increase with

τ = 1.2 μs, corresponding to a grow-in of the Eu-based emission

intensity, and (iii) the expected slow Eu-based emission decay

with τ = 0.65 ms, typical for Eu^IIIemission in a nonprotic

solvent.

First, the TA spectrum of ²⁴Ir L^pPhon its own was measured

3

in deoxygenated CH₂Cl₂using excitation at 355 nm. The result

shows a new absorption at ca. 400 nm and a weaker region of

new absorbance in the red region between 600 and 700 nm,

which must be associated with the Ir-based excited state. The

strong negative feature between 450 and 600 nm arises from

intense stimulated emission from the Ir center. The decay of

the excited-state absorptions monitored at various wave-

lengths gave a lifetime of 3.40 ((0.14) μs. We measured the

The signiﬁcant component here is the rise time of 1.2 μs for

Eu-based emission. If the Eu^IIIion were being excited directly

into the emitting ⁵D₀level by Ir f Eu PEnT, this rise time would

match the rapid decay time of the Ir^IIIenergy donor, yet it is

much slower, implying the presence of an intermediate state.

What we observe is entirely consistent with population of the ⁵D₁

level of Eu^III, which is fully allowed by Dexter-type energy

emission lifetime of ²⁴Ir L^pPhin degassed CH₂Cl₂for compar-

3

ison purposes and obtained under these conditions dual-

exponential decay with lifetime components of 2.33 ((0.05)

and 3.36 ((0.05) μs, which we ascribe to diﬀerent conformers

of the ﬂexible complex (cf. a value of 703 ns in air-equilibrated

CH₂Cl₂; Table 4).²⁶These two components are suﬃciently

similar to one another that they are not resolved as separate com-

ponents in the TA data, which have lower signal/noise ratios,

but the agreement between these lifetimes and the value of 3.40

((0.14) μs obtained from TA data conﬁrms that the features in

the TA spectrum are associated with the same excited state that

generates luminescence.

7

transfer from the F₀ground state (the selection rule requires

that ΔJ = 1),²⁴in agreement with our calculations suggesting that

Dexter-type PEnT is feasible with only a very small electronic

coupling required. It is known that nonradiative decay of the ⁵D₁

5

level of Eu^IIIto the emissive D₀level is slow and occurs on a

microsecond time scale,²⁵so we are clearly observing Ir f Eu

In the presence of excess [Gd(hfac)₃(H₂O)₂], such that all of

5

PEnT to the D₁level of Eu^IIIon a time scale of ≈10^ꢀ7

s

²⁴Ir L^pPhwas converted to ²⁴Ir L^pPhGd, the appearance of the

3

11334

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

TA spectrum in deoxygenated CH₂Cl₂is essentially identical

(Figure 10) with the same regions of new absorption at

≈350ꢀ450 and 600ꢀ700 nm and the same negative emission

peak (on which the characteristic vibrational ﬁne structure of

Ir-based emission is clearly resolved). All of this is consistent with

formation of the normal Ir-based LC/MLCT excited state, and

there are no additional features that can be ascribed to a separate

charge-separated state. The kinetic behavior is diﬀerent, how-

ever, with two lifetime components of 1.2 ((0.1) μs and 60

((10) ns under these conditions, indicative of partial quenching

by electron transfer, as discussed earlier. Luminescence decay

obtained under the same conditions ﬁts to three emission

components with lifetimes of 1.43 ((0.05) μs, 0.56 ((0.02)

μs, and 48 ((1) ns. Allowing for the poorer signal/noise ratio of

the TA data, we can say that the 1.2 ((0.1) μs component

obtained from TA data corresponds to an unresolved combina-

tion of the ﬁrst two luminescence components, with the short

components in each case [60 ((10) ns by TA and 48 ((1) ns by

emission] also agreeing well with each other. Note that in the

titration measurements involving the formation of ²⁴Ir L^pPhGd

Figure 11. (a) Luminescence of samples of ²⁴Ir L^pPhtitrated with

3

[Eu(hfac)₃(H₂O)₂] at diﬀerent points during the titration: (i) pure

²⁴Ir L^pPhshowing the characteristic blue emission; (ii) after the

3

in air-equilibrated CH₂Cl₂we could resolve Ir-based emission

components with lifetimes of 369 and 48 ns (Table 4).²⁶Again,

the presence of multiple decay components in Ir-based decay for

²⁴Ir L^pPhGd is indicative of multiple conformations, with

addition of 0.1 equiv of [Eu(hfac)₃(H₂O)₂], i.e., a mixture containing

ca. 90% ²⁴Ir L^pPhand 10% ²⁴Ir L^pPhEu; (iii) after the addition of 2

3

equiv of [Eu(hfac)₃(H₂O)₂] when the red Eu-based emission domi-

nates. (b) Luminescence spectrum of the white-light-emitting mixture in

part a(ii), i.e., recorded at the point during the titration when lumines-

cence is closest to white.

3

diﬀerent Ir Gd separations, being present.

3 3 3

The adduct ²⁴Ir L^pPhEu in deoxygenated CH₂Cl₂gave a TA

3

spectrum nearly identical with that of ²⁴Ir L^pPhGd, showing the

3

signature of the Ir-based excited state. The two Ir-based lifetime

decay components were 0.45 ((0.04) and 0.99 ((0.10) μs.

Under the same conditions, time-resolved emission measure-

ments showed two decay components with lifetimes of 0.37

((0.01) and 0.76 ((0.03) μs, in reasonable agreement with the

TA lifetimes (in aerated CH₂Cl₂, these are reduced to 99 and

230 ns; Table 4).²⁶Any additional shorter-lived components in

the TA spectrum that might be expected based on the data in

Table 4 are too short to be detected with this setup. In this TA

spectrum, we could also detect, as a contribution to the global ﬁt,

a weak negative feature at 615 nm corresponding to stimulated

Eu-based emission (see the Supporting Information, Figures S1

and S2). However, the weakness of this and the relatively high

signal-to-noise ratio precluded a detailed kinetic analysis at this

wavelength, other than to observe that the rate of growth of this

feature is approximately consistent with the 1.2 μs value for the

rise time of Eu^III-based emission that was obtained from lumi-

nescence measurements (see earlier).

ﬂexible complex, but the charge-separated state is then decaying

too quickly to detect with our instrument.

Balance of Blue and Red Luminescence Components To

Give White Emission. As mentioned earlier, an appropriate

balance of blue/green and red luminescence components has

been used by several research groups to generate white-light

emission from a single molecule,^7e,8a,8band we were interested in

seeing if the same effect could be achieved using these Ir/Eu

dyads. We found in every case that luminescence of these dyads

quickly becomes dominated by the red component from the

Eu^IIIcenter during the titration, i.e., as free Ir L complex

3

becomes converted to Ir L Eu. This is understandable on the

3

basis of the very high quantum yield of red luminescence for

[Eu(hfac)₃(diimine)] complexes in a CH₂Cl₂solution (g50% in

some cases),²⁷and the partial quenching of the Ir-based emission

component in the Ir L Eu dyads, which will lead to loss of the

3

blue component. We found that we could obtain good white-

light emission from a sample of ²⁴Ir L^pPhto which ≈0.1 equiv of

3

Thus, the TA spectra are showing in every case the signature of

Eu(hfac)₃(H₂O)₂has been added, i.e., a mixture consisting of ca.

the Ir-based LC/MLCT excited state. If a charge-separated

90% free ²⁴Ir L^pPhand 10% of the dyad ²⁴Ir L^pPhEu. The

3

Ir^IV(diimine^•ꢀ)Ln state is providing a quenching pathway

progression of the emission color from blue through white to red

3 3 3

in ²⁴Ir L^pPhGd and ²⁴Ir L^pPhEu (in addition to the direct

as this titration of Eu(hfac)₃(H₂O)₂into ²⁴Ir L^pPhproceeds is

3

energy transfer that operates in the latter complex), which seems

likely,¹⁵then we cannot detect it by TA measurements. This

implies that any charge-separated state is short-lived (<20 ns)

because of rapid back electron transfer. If the forward electron-

shown in Figure 11; at the point at which it is most white, the CIE

coordinates are x = 0.34, y = 0.32.

In order to make a “pure” Ir L Eu dyad a better white-light

3

emitter, it would be necessary to decrease the eﬃciency of the

Ir f Eu EnT process such that, under selective excitation of the

Ir component, the ratio of blue-to-red emission components is

increased to give a better balance. In principle, this can be

accomplished using ^CF3Ir as the donor because of the reduced

Dexter EnT associated with the more open conformation, as

discussed above. However, the altered pattern of substituents on

the phenylpyridine ligands that slows down the PEnT process

also shifts Ir-based emission to lower energy (²⁴Ir, λ_em= 455 nm;

^CF3Ir, λ_em= 481 nm) such that it no longer gives white light when

transfer process to generate the Ir^IV(diimine^•ꢀ)Ln state in

3 3 3

²⁴Ir L^pPhGd, which would be responsible for quenching of Ir-

3

based emission, were also fast, we would not expect to see Ir-

based emission components with lifetimes as long as 1.43 μs. The

implication is that the charge-separated state is forming slowly

because the gradient for forward PET is marginal. The observa-

tion of partially quenched lifetime components of 1.43 ((0.05)

μs, 0.56 ((0.02) μs, and 48 ((1) ns in ²⁴Ir L^pPhGd is

3

ascribable to diﬀerent PET rates in diﬀerent conformers of the

11335

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

combined with the red from Eu-based emission. The closest that

can be obtained, from an approximately 1:1 mixture of free

^CF3Ir L^pBiphand ^CF3Ir L^pBiphEu that arises half-way through

L^{mBiph 30}

;

the dimers [Ir(phpy)₂(μ-Cl)]₂based on unsubstituted 2-phe-

nylpyridine and the various fluorinated analogues 2-(2,4-difluorophe-

nyl)pyridine, 2-(3,5-difluorophenyl)pyridine, 2-(2,4,5-trifluorophenyl)-

pyridine, and 2-[4-(trifluoromethyl)]phenylpyridine;³¹iridium(III)

3

titration of Eu(hfac)₃(H₂O)₂into ^CF3Ir L^pBiph, is yellow in color

3

complexes ²⁴Ir L^pBiphand ³⁵Ir L^{pBiph 3g}and Eu(hfac)₃(H₂O)₂.³²The

;

(CIE coordinates x = 0.37, y = 0.48). The combination of the ²⁴Ir

and {Eu(hfac)₃} ﬂuorophores to give white-light emission is

clearly feasible, but correct balancing of the two components will

require bridging ligands that give poorer Ir f Eu EnT; these

studies are in progress.

3

crystal structure of [²⁴Ir L^pBiph] CHCl₃H₂O (Figure 2b) was re-

3

ported in an earlier communication^3gbut is also included here for

completeness.

Electrospray mass spectrometry (ESMS) spectra were recorded using

a Micromass LCT instrument; H NMR spectra were recorded on a

1

Bruker Avance-2 400 MHz instrument. UV/vis absorption spectra were

measured on a Cary 50 spectrophotometer and luminescence spectra on

a Jobin-Yvon Fluoromax 4 ﬂuorimeter using air-equilibrated CH₂Cl₂

solutions at room temperature. Ir-based emission lifetimes derived from

the luminescence titrations in Table 4 were measured using the time-

correlated single-photon-counting technique with an Edinburgh Instru-

ments “Mini-τ” luminescence lifetime spectrometer, equipped with a

405 nm pulsed diode laser as an excitation source and a Hamamatsu-

H5773-03 photomultiplier tube (PMT) detector; the lifetimes were

calculated from the measured data using the supplied software. Lumi-

nescence quantum yields (Φ in Table 4) were calculated by comparing

areas of corrected luminescence spectra on an energy scale, from

isoabsorbing solutions, following the method described by Demas and

Crosby³³and using fac-[Ir(ppy)₃] (ppy = anion of 2-phenylpyridine) as

a standard.³⁴

Flash photolysis experiments were performed on a home-built

setup. The samples were excited at 355 nm with the third harmonic

(355 nm) of a Q-switched Nd:YAG laser LS-2137U (LOTIS TII).

The energy of excitation pulses delivered to the sample was ca. 2.5 mJ,

at 10 Hz repetition rate and 7 ns pulse width. A 150 W Xe arc lamp

(Hamamatsu) was used as a probe light source. The probe light

was detected through a SPEX MiniMate monochromator by a

custom-built detector unit, based on a FEU-118 PMT. The detector

current output was coupled into a Tektronix TDS 3032B digital

oscilloscope and subsequently transferred to a computer. The same

setup was used for the time-resolved emission measurements in the

microsecond time domain, with the only diﬀerence being a blocked

probe lamp.

’ CONCLUSIONS

A range of Ir/Eu dyads has been prepared in which an Ir^IIIunit

with high-energy luminescence acts as an energy donor to Eu^IIIto

generate sensitized luminescence following Ir f Eu energy trans-

fer. The extent of energy transfer varies substantially between the

diﬀerent dyads because of changes in the intercomponent distance

and the conformational ﬂexibility of the molecules, with an energy-

transfer eﬃciency of up to 90% occurring in some cases. There is

a clear correlation of energy-transfer eﬃciency with the mole-

cular conformation, as shown by crystallographic studies, with the

majority of the dyads able to adopt a folded conformation in which

the donor and acceptor components are indirectly linked by π

stacking between ligand fragments. In the one series where this is

sterically impossible, as conﬁrmed by crystallography and also by

calculation, the Ir f Eu energy-transfer eﬃciency is notably low.

Multiexponential decay kinetics for Ir-based emission in the Ir/Eu

dyads and the related Ir/Gd complexes can be ascribed to the

presence of diﬀerent conformers in solution.

In the related Ir/Gd dyads prepared as controls, two of the ﬁve

series show surprisingly large quenching of Ir-based emission

following binding of the {Gd(hfac)₃} fragment, even though

PEnT to Gd^IIIis not possible. We ascribe this to a PET process

from the excited state of the iridium chromophore to the pendant

diimine unit, which becomes easily reduced upon coordination

to Gd^IIIto give a Ir^IV(diimine^•ꢀ)Ln charge-separated state;

3 3 3

however, this is too short-lived to be detected by nanosecond

time scale TA spectra. This initial PET process could indirectly

contribute to PEnT in the Ir/Eu dyads if the energy released by

back electron transfer is suﬃcient to sensitize Eu^III.

All ﬂash photolysis and microsecond time-resolved emission experi-

ments were performed with the deoxygenated samples, degassed by the

freezeꢀpumpꢀthaw technique, and saturated with argon, unless noted

otherwise. Quartz cells of 1 cm path length were used.

Calculations of donor/acceptor spectroscopic overlap inte-

grals for the Ir/Eu dyads and consideration of the requirements

for F€orster and Dexter-type energy transfer unambiguously show

that F€orster energy transfer is not possible over the distances

involved, but Dexter-type energy transfer is possible provided a

small electronic coupling across the bridging ligand is present, to

which the aromatic π stacking observed in many of the crystal

structures could contribute. The intermediacy of the Eu^{III 5}D₁

level as an initial energy acceptor, which has ΔJ = 1 from the

ground state and whose population is therefore allowed by

Dexter PEnT, is entirely consistent with this. Thus, Dexter-type

PEnT is possible in all of the Ir/Eu dyads, and in two of the ﬁve

series, an initial electron-transfer mechanism is also possible.

An appropriate balance between the blue emission of the highest-

energy iridium chromophores and the sensitized red Eu-based

emission leads to white-light emission with CIE coordinates of

x = 0.34, y = 0.32.

Analysis of the time-resolved data to obtain decay lifetimes was

performed using Igor Pro software (WaveMetrics, Inc.). The decay

kinetics were ﬁtted to the exponential decay law using the least-squares

algorithm built into Igor Pro. Global ﬁtting was applied to analyze

simultaneously the decay kinetics obtained for numerous spectral points,

which enabled us to reconstruct the shape of TA spectra and consider-

ably increase the reliability of the lifetime values.

Syntheses of Iridium(III) Complexes. All complexes were

prepared in the same general way; the method described here for

^HIr L^pPhis typical. A solution of L^pPh(0.033 g, 85 μmol) was dissolved

3

in dry CH₂Cl₂/MeOH (3:1, v/v) under N₂. To this was added a solution

of [Ir(phpy)₂(μ-Cl)]₂(0.040 g, 33 μmol) in the minimum amount of

CH₂Cl₂. The mixture was stirred and heated to 50 °C overnight in the

dark. The mixture was cooled to room temperature and the solvent

removed under reduced pressure. A saturated aqueous KPF₆solution

(20 cm³) was added, and the resulting two-phase mixture was shaken

vigorously and then separated; the aqueous residue was further extracted

with several portions of CH₂Cl₂(3 ꢁ 30 cm³). The combined organic

fractions (containing the crude complex as its hexafluorophosphate salt)

were dried using sodium sulfate, and the solvent was removed. The

crude yellow powder was purified by column chromatography on silica

’ EXPERIMENTAL DETAILS

General Details. The following compounds were prepared accord-

ing to published methods: the bridging ligands L^{pPh 28}

,

L^{mPh 29}

,

and

gel using MeCN and 1% aqueous KNO₃; complex ^HIr L^pPhwas the

3

11336

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

second yellow band to come off the column. Fractions containing the

pure product were combined and reduced in volume; excess KNO₃was

precipitated by the addition of CH₂Cl₂and filtered off. Evaporation of

(2H, m), 6.02 (2H, d), 5.47 (1H, d), 5.28 (2H, s), 5.07 (1H, d). ESMS:

m/z 1001 (M + H⁺). Anal. Calcd for C₄₆H₃₀N₈F₆IrNO₃CH₂Cl₂: C,

3

47.6; H, 2.7; N, 10.6. Found: C, 47.7; H, 2.7; N, 10.3.

the resultant solution to dryness afforded pure ^HIr L^pPhas its nitrate

²⁴⁵Ir L^mBiph. H NMR (400 MHz, CDCl₃): δ 8.69 (1H, d), 8.64

1

3

salt. Characterization data are given below.

(1H, d), 8.22 (2H, m), 8.11 (1H, d), 7.94 (1H, d), 7.78ꢀ7.60 (6H, m),

7.56 (1H, d), 7.41 (1H, t), 7.36ꢀ7.20 (8H, m), 7.10 ꢀ7.04 (2H, m),

6.92 (1H, d), 6.76ꢀ6.68 (2H, m), 6.51ꢀ6.45 (1H, m), 6.21ꢀ6.17 (2H,

m), 5.47ꢀ5.43 (3H, m), 5.15 (1H, d). ESMS: m/z 1077 (M + H⁺). Anal.

^HIr L^pPh. ¹H NMR (400 MHz, CDCl₃): δ 8.68 (1H, d), 8.41 (1H, d),

3

8.06ꢀ7.99 (3H, m), 7.84 (1H, d), 7.79ꢀ7.67 (4H, m), 7.59ꢀ7.53 (2H,

m), 7.52 (1H, d), 7.40ꢀ7.35 (2H, d), 7.26ꢀ7.21 (2H, m), 7.16ꢀ7.06

(3H, m), 7.02 (1H, d), 7.0ꢀ6.96 (1H, t), 6.87ꢀ6.79 (6H, m), 6.21 (1H,

m), 6.01 (1H, d), 5.97 (2H, d), 5.29ꢀ5.25 (3H, m), 5.13 (1H, d). ESMS:

Calcd for C₅₂H₃₄N₈F₆IrNO₃CH₂Cl₂: C, 52.0; H, 3.0; N, 10.3. Found:

3

C, 52.4; H, 3.3; N, 10.0.

m/z 893 (M + H⁺). Anal. Calcd for C₄₆H₃₆N₈IrNO₃CH₂Cl₂: C, 54.3;

³⁵Ir L^pPh. ¹H NMR (400 MHz, CDCl₃): δ 8.68 (1H, d), 8.57 (1H,

3

H, 3.7; N, 12.1. Found: C, 54.8; H, 3.7; N, 12.3.

d), 8.15ꢀ8.05 (3H, m), 7.82ꢀ7.73 (4H, m), 7.67ꢀ7.62 (3H, m),

7.38ꢀ7.24 (5H, m), 7.15 (1H, dt), 7.04 (1H, d), 6.92ꢀ6.83 (4H, m),

6.62 (1H, d), 6.38ꢀ6.29 (2H, m), 5.98 (2H, d), 5.49 (1H, d), 5.27 (2H,

s), 5.15 (1H, d). ESMS: m/z 965 (M + H⁺). Anal. Calcd for C₄₆H₃₂-

^HIr L^mPh. ¹H NMR (400 MHz, CDCl₃): δ 8.67 (1H, d), 8.33 (1H,

3

d), 8.04ꢀ7.94 (3H, m), 7.84ꢀ7.68 (4H, m), 7.61ꢀ7.56 (3H, m),

7.48ꢀ7.43 (4H, m), 7.31ꢀ7.21 (3H, m), 7.09 (1H, t), 7.01ꢀ6.94

(4H, m), 6.9ꢀ6.82 (4H, m), 6.23 (1H, d), 6.11 (1H, s), 6.07 (1H, d),

5.90 (1H, d), 5.18ꢀ5.13 (3H, m), 5.01 (1H, d). ESMS: m/z 893 (M + H⁺).

N₈F₄IrNO₃CH₂Cl₂: C, 50.8; H, 3.1; N, 11.3. Found: C, 51.1; H, 3.4;

3

N, 11.5.

Anal. Calcd for C₄₆H₃₆N₈IrNO₃CH₂Cl₂: C, 54.3; H, 3.7; N, 12.1.

³⁵Ir L^mPh. ¹H NMR (400 MHz, CDCl₃): δ 8.65 (1H, d), 8.54 (1H,

3

Found: C, 54.1; H, 3.9; N, 11.8.

d), 8.11 (1H, t), 8.02 (1H, d), 7.95 (1H, d), 7.77ꢀ7.72 (4H, m),

7.65ꢀ7.47 (4H, m), 7.39ꢀ7.21 (5H, m), 7.13 (1H, q), 7.02ꢀ6.88 (5H,

m), 6.44ꢀ6.34 (2H, m), 6.15 (1H, s), 5.97 (1H, d), 5.30 (1H, d), 5.19

(2H, s), 5.04 (1H, d). ESMS: m/z 965 (M + H⁺). Anal. Calcd for

C₄₆H₃₂N₈F₄IrNO₃0.5CH₂Cl₂0.5H₂O: C, 51.8; H, 3.2; N, 11.7.

^HIr L^pBiph. ¹H NMR (400 MHz, CDCl₃): δ 8.67 (1H, d), 8.49 (1H,

3

d), 8.06 (1H, t), 8.00ꢀ7.96 (2H, m), 7.85 (1H, d), 7.80ꢀ7.66 (4H, m),

7.61ꢀ7.58 (2H, m), 7.53ꢀ7.44 (5H, m), 7.38ꢀ7.32 (3H, m),

7.28ꢀ7.08 (6H, m), 7.03ꢀ6.81 (6H, m), 6.30ꢀ6.26 (1H, m),

6.11ꢀ6.04 (3H, m), 5.47 (2H, s), 5.26 (1H, d), 5.09 (1H, d). ESMS:

m/z 969 (M + H⁺). Anal. Calcd for C₅₂H₄₀N₈IrNO₃0.5H₂O 0.5

3

Found: C, 51.9; H, 3.2; N, 12.1.

³⁵Ir L^pBiph. ¹H NMR (400 MHz, CDCl₃): δ 8.66 (2H, m), 8.16 (1H,

3

t), 8.05 (1H, d), 7.98 (1H, d), 7.80ꢀ7.73 (5H, m), 7.61ꢀ7.52 (3H, m),

7.48 (2H, d), 7.43ꢀ7.20 (7H, m), 7.15 (3H, m), 7.00 (1H, t), 6.97 (1H,

d), 6.86 (1H, d), 6.42 (1H, t), 6.36 (1H, t), 6.15 (2H, d), 5.48 (2H, s),

5.40 (1H, d), 5.12 (1H, d). ESMS m/z 1041 (M + H⁺). Anal. Calcd for

CH₂Cl₂: C, 58.2; H, 3.9; N, 11.6. Found: C, 58.2; H, 3.7; N, 11.8.

^HIr L^mBiph. ¹H NMR (400 MHz, CDCl₃): δ 8.63 (1H, d), 8.48 (1H,

3

d), 8.06 (1H, t), 8.00 (1H, d), 7.94 (1H, d), 7.83 (1H, d), 7.76ꢀ7.66

(4H, m), 7.60ꢀ7.56 (3H, m), 7.44ꢀ7.18 (10H, m), 7.13ꢀ7.09 (2H, m),

7.01ꢀ6.92 (3H, m), 6.85ꢀ6.70 (4H, m), 6.26ꢀ6.20 (2H, m),

6.05ꢀ5.99 (2H, m), 5.48 (2H, s), 5.27 (1H, d), 5.16 (1H, d). ESMS:

C₅₂H₃₆N₈F₄IrNO₃0.5CH₂Cl₂: C, 55.0; H, 3.2; N, 11.0. Found: C,

3

55.1; H, 3.4; N, 11.1.

^CF3Ir L^pPh. ¹H NMR (400 MHz, CDCl₃): δ 8.70 (1H, d), 8.50 (1H,

m/z 969 (M + H⁺). Anal. Calcd for C₅₂H₄₀N₈IrNO₃0.5CH₂Cl₂: C,

3

d), 8.12ꢀ8.03 (3H, m), 7.93 (1H, d), 7.82 (2H, t), 7.75 (1H, d), 7.69

(1H, d), 7.63 (1H, d), 7.61ꢀ7.51 (3H, m), 7.40 (1H, d), 7.32ꢀ7.16

(6H, m), 7.08ꢀ6.99 (3H, m), 6.86 (2H, d), 6.30 (1H, s), 6.11 (1H, s),

5.90 (2H, d), 5.29ꢀ5.23 (3H, m), 5.07 (1H, d). ESMS: m/z 1029 (M +

58.7; H, 3.8; N, 11.7. Found: C, 59.1; H, 3.8; N, 11.7.

²⁴Ir L^pPh. ¹H NMR (400 MHz, CDCl₃): δ 8.68 (1H, d,), 8.50 (1H,

3

d), 8.24 (1H, d), 8.10 (2H, m), 8.01 (1H, d), 7.78 (2H, t), 7.68 (2H, t),

7.59ꢀ7.19 (9H, m), 7.00 (1H, d), 6.86 (3H, d), 6.51 (1H, dt), 6.41 (1H,

dt), 5.96 (2H, d), 5.61 (1H, dd), 5.45 (1H, d), 5.27 (2H, s), 5.08 (1H, d).

H⁺). Anal. Calcd for C₄₈H₃₄N₈F₆IrNO₃CH₂Cl₂: C, 50.1; H, 3.1; N,

3

ESMS: m/z 965 (M + H⁺). Anal. Calcd for C₄₆H₃₂N₈F₄IrNO₃

10.7. Found: C, 50.5; H, 3.0; N, 11.1.

3

^CF3Ir L^mPh. ¹H NMR (400 MHz, CDCl₃): δ 8.65 (1H, d), 8.47 (1H,

CH₂Cl₂: C, 50.8; H, 3.1; N, 11.3. Found: C, 50.4; H, 3.4; N, 11.5.

3

²⁴Ir L^mPh. H NMR (400 MHz, CDCl₃): δ 8.65 (1H, br d), 8.45

1

d), 8.10ꢀ8.05 (2H, m), 7.96ꢀ7.92 (2H, m), 7.85 (1H, t), 7.76ꢀ7.65

(4H, m), 7.61ꢀ7.50 (5H, m), 7.40 (1H, d), 7.29ꢀ7.21 (4H, m),

7.16ꢀ7.12 (2H, m), 6.99ꢀ6.96 (2H, m), 6.85 (1H, t), 6.33 (1H, s),

6.19 (1H, s), 6.05 (1H, s), 5.80 (1H, d), 5.18ꢀ5.14 (3H, m), 4.94 (1H,

3

(1H, d), 8.25 (1H, d), 8.09 (1H, t), 8.03 (1H, d), 7.91 (1H, d),

7.84ꢀ7.63 (6H, m), 7.56 (1H, d), 7.46ꢀ7.40 (2H, m), 7.32 (1H, t),

7.22 (2H, t), 7.08 (1H, t), 7.00ꢀ6.90 (3H, m), 6.58ꢀ6.46 (2H, m), 6.02

(1H, s), 5.97 (1H, d), 5.64 (1H, d), 5.46 (1H, d), 5.30 (1H, m), 5.18

(2H, s), 5.04 (1H, d). ESMS: m/z 965 (M + H⁺). Anal. Calcd for

d). ESMS: m/z 1029 (M + H⁺). Anal. Calcd for C₄₈H₃₄N₈F₆IrNO₃

3

0.5CH₂Cl₂: C, 51.3; H, 3.2; N, 11.0. Found: C, 51.5; H, 2.8; N, 10.8.

^CF3IrL^pBiph. ¹H NMR (400 MHz, CDCl₃): δ 8.67 (1H, d), 8.60 (1H,

d), 8.12 (1H, t), 8.08 (1H, d), 7.97 (2H, t), 7.87 (1H, t), 7.77ꢀ7.64 (5H,

m), 7.60 (1H, d), 7.55ꢀ7.53 (2H, m), 7.40ꢀ7.38 (3H, m), 7.38ꢀ7.36

(3H, m), 7.31ꢀ7.10 (8H, m), 6.97 (1H, d), 6.39 (1H, s), 6.16 (1H, s),

6.05 (2H, d), 5.48 (2H, s), 5.23 (1H, d), 5.04 (1H, d). ESMS m/z 1105

(M + H⁺). Anal. Calcd for C₅₄H₃₈N₈F₆IrNO₃0.5H₂O 0.5CH₂Cl₂: C,

C₄₆H₃₂N₈F₄IrNO₃CH₂Cl₂: C, 50.7; H, 3.1; N, 11.3. Found: C, 50.3;

3

H, 3.2; N, 11.3.

²⁴Ir L^pBiph. ¹H NMR (400 MHz, CDCl₃): δ 8.67 (1H, s), 8.59 (1H,

3

d), 8.25 (1H, d), 8.15ꢀ8.11 (2H, m), 7.99 (1H, d), 7.83 (1H, t),

7.78ꢀ7.64 (5H, m), 7.53ꢀ7.22 (10H, m), 7.09 (2H, d), 7.01ꢀ6.97 (2H,

m), 6.54 (1H, t), 6.46 (1H, t), 6.03 (2H, d), 5.68 (1H, d), 5.49ꢀ5.41

(4H, m), 5.12 (1H, d). ESMS: m/z 1041 (M + H⁺). Anal. Calcd for for

3

53.7; H, 3.3; N, 10.3. Found: C, 53.5; H, 3.1; N, 10.5.

^CF3Ir L^mBiph. ¹H NMR (400 MHz, CDCl₃): δ 8.62ꢀ8.58 (2H, m),

C₅₂H₃₆N₈F₄IrNO₃0.5CH₂Cl₂: C, 55.0; H, 3.2; N, 11.0. Found: C,

3

8.11 (1H, t), 8.06 (1H, d), 7.96ꢀ7.80 (4H, m), 7.71ꢀ7.66 (3H, m), 7.61

(1H, d), 7.58 (1H, d), 7.43ꢀ7.36 (3H, m), 7.32ꢀ7.24 (6H, m),

7.22ꢀ7.16 (3H, m), 7.11 (1H, d), 6.97ꢀ6.91 (3H, m), 6.81 (1H, t),

6.38 (1H, s), 6.07 (1H, s), 6.02 (1H, s), 5.97 (1H, d), 5.49 (2H, s), 5.34

(1H, d), 5.11 (1H, d). ESMS: m/z 1105 (M + H⁺). Anal. Calcd for

C₅₄H₃₈N₈F₆IrNO₃H₂O 0.5CH₂Cl₂: C, 53.3; H, 3.4; N, 10.3. Found:

55.0; H, 3.3; N, 11.0.

²⁴Ir L^mBiph. ¹H NMR (400 MHz, CDCl₃): δ 8.64 (1H, s), 8.56 (1H,

3

d), 8.25 (1H, d), 8.12 (1H, t), 8.07 (1H, d), 7.96 (1H, d), 7.83ꢀ7.74

(3H, m), 7.69ꢀ7.58 (4H, m), 7.41 (1H, t), 7.36ꢀ7.19 (9H, m), 7.06

(1H, t), 6.93 (1H, d), 6.72 (1H, t), 6.54 (1H, t), 6.33 (1H, t), 6.14ꢀ6.10

(2H, m), 5.65 (1H, m), 5.49ꢀ5.39 (4H, m), 5.18 (1H, d). ESMS: m/z

3

1041 (M + H⁺). Anal. Calcd for C₅₂H₃₆N₈F₄IrNO₃CH₂Cl₂: C, 53.6;

C, 53.5; H, 3.0; N, 10.0.

3

H, 3.2; N, 10.6. Found: C, 54.0; H, 3.4; N, 10.9.

X-ray Crystallography. Crystals were removed from the mother

liquor, coated with oil, and transferred to a stream of cold N₂on the

diffractometer as quickly as possible to prevent decomposition due to

solvent loss. All structural determinations were carried out on a Bruker

²⁴⁵Ir L^pPh. H NMR (400 MHz, CDCl₃): δ 8.69ꢀ8.65 (2H, m),

1

3

8.21ꢀ8.14 (3H, m), 8.00 (1H, d), 7.79ꢀ7.71 (4H, m), 7.56ꢀ7.21 (7H,

m), 7.09 (1H, t), 6.99 (1H, d), 6.90 (2H, d), 6.83 (1H, t), 6.77ꢀ6.53

11337

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

SMART-APEX2 diffractometer using graphite-monochromated Mo Kα

radiation (λ = 0.710 73 Å) from a sealed tube source. After integration of

the raw data and before merging, an empirical absorption correction was

applied (SADABS)³⁵based on a comparison of multiple symmetry-

equivalent measurements. The structures were solved by direct methods

and refined by full-matrix least squares on weighted F²values for all

reflections using the SHELX suite of programs.³⁶Pertinent crystal-

lographic data are collected in Table 1; selected bond distances (from

the metal coordination spheres) are given in Table 2. In general, the

solutions and refinement were straightforward, although in some cases,

scattering was weak because of one or all of (i) rapid desolvation of

the crystal and loss of crystallinity, (ii) disorder of the anions and

solvent molecules, and (iii) rotational disorder associated with the

CF₃groups; this necessitated the use of geometric restraints on

disordered components and in some cases the use of isotropic

thermal displacement parameters for particular atoms. Full details

are given in the individual CIFs.

set by the GNORM = 0.0 and DDMIN = 0.0 parameters. Cartesian

coordinates for the two calculated structures shown in Figure 8 are in the

Supporting Information (Tables S1 and S2).

’ ASSOCIATED CONTENT

S

Supporting Information. X-ray crystallographic data in

b

CIF format, ﬂash photolysis information on ²⁴Ir L^pPhEu in

degassed CH₂Cl₂, and coordinates for structures calculated by

molecular modeling. This material is available free of charge via

the Internet at http://pubs.acs.org.

3

’ AUTHOR INFORMATION

Corresponding Author

*E-mail: m.d.ward@sheﬃeld.ac.uk.

Calculation of Energy-Transfer Rates. For a general energy-

transfer process, the intramolecular rate constant can be obtained from

eq 1 (based on lifetime measurements) or eq 2 (based on emission inten-

sity measurements); see the main text. The efficiency of the intramo-

lecular energy-transfer process (η_EnT) can be evaluated by using eq 3:

’ ACKNOWLEDGMENT

We thank the EPSRC (UK) and the Italian CNR (Project PM.

P04.010, MACOL) for ﬁnancial support and Dr. Andrew Beeby

(University of Durham, Durham, U.K.) for valuable discussions.

η_EnT¼ k_EnT=ðk_EnTþ k_inÞ

ð3Þ

’ DEDICATION

^†Dedicated to our collaborator and colleague Francesco Barigel-

letti on the occasion of his retirement.

in which k_in= 1/τ_u, i.e., the intrinsic deactivation rate constant of the

unquenched donor luminophore, and k_EnTwas defined earlier.

For calculation of the energy-transfer rates based on the F€orster and

Dexter mechanisms, corrected donor-emission spectra and acceptor-

absorption spectra on a wavenumber scale were used. Computations

of the rate constants and the relevant spectroscopic overlap integrals

were performed by using home-developed routines for MATLAB 5.2

(The MathWorks, Inc.). The rate constant for energy transfer according

to the F€orster (k_F) and Dexter (k_D) mechanisms, respectively, were

evaluated by employing eqs 4 and 5, respectively:

’ REFERENCES

(1) Reviews: (a) Ward, M. D. Coord. Chem. Rev. 2007, 251, 1663. (b)

Ward, M. D. Coord. Chem. Rev. 2010, 254, 2634. (c) Chen, F. F.; Chen,

Z.-Q.; Bian, Z.-Q.; Huang, C.-H. Coord. Chem. Rev. 2010, 254, 991. (d)

Faulkner, S.;Natrajan, L. S.; Perry, W. S.;Sykes, D.Dalton Trans. 2009, 3890.

(2) Representative recent examples: (a) Nonat, A. M.; Quinn, S. J.;

Gunnlaugsson, T. Inorg. Chem. 2009, 48, 4646. (b) Ni, J.; Zhang, L.-Y.;

Chen, Z.-N. J. Organomet. Chem. 2009, 694, 339. (c) Peng, G.; Qiu,

Y.-C.; Liu, Z.-H.; Liu, B.; Deng, H. Cryst. Growth Des. 2010, 10, 114.

(d) Eckes, F.; Bulach, V.; Guenet, A.; Strassert, C. A.; De Cola, L.;

Hosseini, M. W. Chem. Commun. 2010, 46, 619. (e) Nonat, A. M.; Allain,

C.; Faulkner, S.; Gunnlaugsson, T. Inorg. Chem. 2010, 49, 8449.

(f) Kadjane, P.; Platas-Iglesias, C.; Ziessel, R.; Charbonniere, L. J. Dalton

Trans. 2009, 5688. (g) Li, X.-L.; Zhang, K.-J.; Li, J.-J.; Cheng, X.-X.;

Chen, Z.-N. Eur. J. Inorg. Chem. 2010, 3449.

(3) (a) Mehlstaubl, M.; Kottas, G. S.; Colella, S.; De Cola, L. Dalton

Trans. 2008, 2385. (b) Li, D.; Chen, F.-F.; Bian, Z.-Q.; Liu, Z.-W.; Zhao,

Y.-L.; Huang, C.-H. Polyhedron 2009, 28, 897. (c) Chen, F.-F.; Bian,

Z.-Q.; Lou, B.; Ma, E.; Liu, Z.-W.; Nie, D.-B.; Chen, Z.-Q.; Bian, J.; Chen,

Z.-N.; Huang, C.-H. Dalton Trans. 2008, 5577. (d) Chen, F.-F.; Bian,

Z.-Q.; Liu, Z.-W.; Nie, D.-B.; Chen, Z.-Q.; Huang, C.-H. Inorg. Chem.

2008, 47, 2507. (e) Coppo, P.; Duati, M.; Kozhevnikov, V. N.; Hofstraat,

J. W.; De Cola, L. Angew. Chem., Int. Ed. 2005, 44, 1806. (f) Tart, N. M.;

Sykes, D.; Sazanovich, I.; Tidmarsh, I. S.; Ward, M. D. Photochem.

Photobiol. Sci. 2010, 9, 886. (g) Sykes, D.; Ward, M. D. Chem. Commun.

2011, 47, 2279. (h) Lian, P.; Wei, H.; Zheng, C.; Nie, Y.; Bian, J.; Bian, Z.

Dalton Trans. 2011, 40, 5476.

k_F¼ 8:8ꢁ10^ꢀ25k²jJ_F=n⁴τd⁶

ð4Þ

k_D¼ 4π²H²J_D=h

ð5Þ

2

in which k²= /₃is the statistical orientation factor, ϕ is the photo-

luminescence quantum yield of the donor, n is the refractive index of the

solvent, τ is the excited-state lifetime of the donor, d is the donorꢀaccep-

tor distance (taken as the metalꢀmetal separation calculated from mole-

cular modeling), H is the electronic coupling term, and h is Planck’s

constant. The spectral overlap integrals J_Fand J_Dwere calculated from

the emission spectrum of the donor, D(ν), and the acceptor absorption

spectrum in molar absorptivity units, A(ν), by using eqs 6 and 7.

̅

Z

DðνÞ AðνÞ=ν ⁴dν

̅

Z

J_F

¼

ð6Þ

ð7Þ

DðνÞ dν

̅

Z

DðνÞ AðνÞ dν

̅

(4) Lazarides, T.; Sykes, D.; Faulkner, S.; Barbieri, A.; Ward, M. D.

Chem.—Eur. J. 2008, 14, 9389.

(5) Parker, D. Coord. Chem. Rev. 2000, 205, 109.

Z

J_D

DðνÞ dν AðνÞ dν

̅

(6) (a) Flamigni, L.; Barbieri, A.; Sabatini, C.; Ventura, B.; Barigelletti,

F. Top. Curr. Chem. 2007, 281, 143. (b) Chen, Z.-Q.; Bian, Z.-Q.; Huang,

C.-H. Adv. Mater. 2010, 22, 1534.

(7) (a) Mydlak, M.; Bizzarri, C.; Hartmann, D.; Sarfert, W.; Schmid,

G.; De Cola, L. Adv. Funct. Mater. 2010, 20, 1812. (b) Orselli, E.; Kottas,

G. S.; Konradsson, A. E.; Coppo, P.; Fr€ohlich, R.; De Cola, L; van Dijken,

A.; B€uchel, M.; B€orner, H. Inorg. Chem. 2007, 46, 11082. (c) Sajoto,

T.; Djurovich, P. I.; Tamayo, A. B.; Oxgaard, J.; Goddard, W. A.;

Molecular Modeling. Optimized and fixed ground-state geome-

tries in vacuo for the Ir L Eu complexes were obtained using the

3

Sparkle-PM6 semiempirical quantum mechanical method^19a,bwith

built-in parameters, as implemented in the MOPAC 2009 for Windows

package,^19cusing the Gabedit GUI.^19dAn eigenvector following routine was

used as the geometry optimization method, with a termination condition

11338

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Inorganic Chemistry

ARTICLE

Thompson, M. E. J. Am. Chem. Soc. 2009, 131, 9813. (d) He, L.; Duan,

L.; Qiao, J.; Wang, R.; Wei, P.; Wang, L.; Qiu, Y. Adv. Funct. Mater. 2008,

18, 2123. (e) Orselli, E.; Albuquerque, R. Q.; Fransen, P. M.; Fr€ohlich,

R.; Janssen, H. M.; De Cola, L. J. Mater. Chem. 2008, 38, 4579.

(8) (a) Law, G.-L.; Wong, K.-L.; Tam, H.-L.; Cheah, K.-W.; Wong,

W.-T. Inorg. Chem. 2009, 48, 10492. (b) He, G.-J.; Guo, D.; He, C.;

Zhang, X.-L.; Zhao, X.-W.; Duan, C.-Y. Angew. Chem., Int. Ed. 2009, 48,

6132. (c) Kalinowski, J.; Cocchi, M.; Virgili, D.; Tattori, V.; Williams,

J. A. G. Adv. Mater. 2007, 19, 4000.

(9) (a) Allain, C.; Faulkner, S. Future Med. Chem. 2010, 2, 339.

(b) B€unzli, J.-C. G. Chem. Rev. 2010, 110, 2729. (c) Pandya, S.; Yu, J.-H.;

Parker., D. Dalton Trans. 2006, 2757.

(10) Natrajan, L. S.; Toulmin, A.; Chew, A.; Magennis, S. W. Dalton

Trans. 2010, 39, 10837.

(c) Gonc-alves e Silva, F. R.; Malta, O. L.; Reinhard, C.; G€udel,

H.-U.; Piguet, C.; Moser, J. E.; B€unzli, J.-C. G. J. Phys. Chem. A 2002,

106, 1670.

(25) (a) Nardi, E.; Yatsiv, S. J. Chem. Phys. 1962, 37, 2333. (b) Bhaumil,

M. L.; Nugent, L. J. J. Chem. Phys. 1965, 43, 1680. (c) Buono-Core, G. E.; Li,

H.; Marciniak, B. Coord. Chem. Rev. 1990, 99, 55. (d) Zhao, X.; Fei, H.; Tian,

K.; Li, T. J. Lumin. 1988, 40ꢀ41, 286.

(26) The luminescence lifetimes obtained in deoxygenated CH₂Cl₂are

longer than those obtained for the same compounds during the titration

studies in air-equilibrated CH₂Cl₂(Table 4), which is typical behavior

associated with the absence of quenching of a triplet excited state by ³O₂.

(27) (a) De Silva, C. R.; Maeyer, J. R.; Wang, R.; Nichol, G. S.;

Zheng, Z. Inorg. Chim. Acta 2007, 360, 3453. (b) Osawa, M.; Hoshino,

M.; Wada, T.; Hayashi, F.; Osanai, S. J. Phys. Chem. A 2009, 113, 10895.

(c) Hasegawa, Y.; Yamamuro, M.; Wada, Y.; Kanehisa, N.; Kai, Y.;

Yanagida, S. J. Phys. Chem. A 2003, 107, 1697.

(11) Ward, M. D. Chem. Commun. 2009, 4487.

(12) (a) Engelman, R.; Jortner, J. Mol. Phys. 1970, 18, 145.

(b) Siebrand, W. J. Chem. Phys. 1967, 47, 2411. (c) Cummings, S. D.;

Eisenberg, R. J. Am. Chem. Soc. 1996, 118, 1949.

(13) (a) Belser, P.; von Zelewsky, A. Helv. Chim. Acta 1980, 63,

1675. (b) Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.;

von Zelewsky, A. Coord. Chem. Rev. 1988, 84, 85.

(28) Argent, S. P.; Adams, H.; Riis-Johannessen, T.; Jeﬀery, J. C.;

Harding, L. P.; Ward, M. D. J. Am. Chem. Soc. 2006, 128, 72.

(29) Argent, S. P.; Adams, H.; Harding, L. P.; Ward, M. D. Dalton

Trans. 2006, 542.

(30) Paul, R. L.; Couchman, S. M.; Jeﬀery, J. C.; McCleverty, J. A.;

Reeves, Z. R.; Ward, M. D. Dalton Trans. 2003, 845.

(31) (a) Sprouse, S.; King, A. K.; Spellane, P. J.; Watts, R. J. J. Am.

Chem. Soc. 1984, 106, 6647. (b) Coppo, P.; Plummer, E. A.; De Cola, L.

Chem. Commun. 2004, 1774. (c) Ragni, R.; Plummer, E. A.; Brunner, K.;

Hofstraat, J. W.; Babudri, F.; Farinola, G. M.; Naso, F.; De Cola, L.

J. Mater. Chem. 2006, 16, 1161.

(14) We cannot directly measure the reduction potential of the

pyrazolylpyridine unit coordinated to Gd^IIIin the Ir L Gd dyads

3

because the competitive conditions required for electrochemical mea-

surements (large excess of base electrolyte) would result in dissociation

of the {Gd(hfac)₃} unit from the pyrazolylpyridine ligand.

(15) Lazarides, T.; Alamiry, M. A. H.; Adams, H.; Pope, S. J. A.;

Faulkner, S.; Weinstein, J. A.; Ward, M. D. Dalton Trans. 2007, 1484.

(16) (a) Easun, T. L.; Alsindi, W. Z.; Deppermann, N.; Towrie, M.;

Ronayne, K. L.; Sun, X.-Z.; Ward, M. D.; George, M. W. Inorg. Chem.

2009, 48, 8759. (b) Easun, T. L.; Alsindi, W. Z.; Towrie, M.; Ronayne,

K. L.; Sun, X.-Z.; Ward, M. D.; George, M. W. Inorg. Chem. 2008, 47,

5071. (c) Frank, M.; Nieger, M.; V€ogtle, F.; Belser, P.; von Zelewsky, A.;

De Cola, L.; Balzani, V.; Barigeletti, F.; Flamigni, L. Inorg. Chim. Acta

1996, 242, 281.

(32) Richards, M. F.; Wagner, W. F.; Sands, D. E. J. Inorg. Nucl.

Chem. 1968, 30, 1275.

(33) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 991.

(34) (a) Namdas, E. B.; Ruseckas, A.; Samuel, I. D. W. J. Phys. Chem.

B 2004, 108, 1570. (b) King, K. A.; Spellane, P. J.; Watts, R. J. J. Am.

Chem. Soc. 1985, 107, 1431.

(35) Sheldrick, G. M. SADABS: A program for absorption correction

with the Siemens SMART system; University of Gottingen: Gottingen,

Germany, 1996.

(17) In contrast, in a microcrystalline sample of ²⁴Ir L^pPhEu, Ir-

3

based emission is completely quenched and only the characteristic

emission spectrum of europium(III) can be seen following excitation

of the iridium(III) chromophore. However, in a condensed phase, addi-

tional quenching pathways can operate, and this is not directly compar-

able to behavior of molecules in a dilute solution.

(36) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64, 112.

(18) F€orster, Th. Discuss. Faraday Soc. 1959, 27, 7.

(19) (a) Stewart, J. J. P. J. Mol. Modeling 2007, 13, 1173. (b) Freire,

R. O.; Simas, A. M. J. Chem. Theory Comput. 2010, 6, 2019. (c) James,

J. P.; Stewart MOPAC2009; Stewart Computational Chemistry: Color-

ado Springs, CO, 2008; http://OpenMOPAC.net. (d) Allouche, A.-R.

J. Comput. Chem. 2011, 32, 174.

(20) (a) Scholes, G. D. Annu. Rep. Phys. Chem. 2003, 54, 57. (b)

Malta, O. L. J. Lumin. 1997, 71, 229.

(21) Dexter, D. L. J. Chem. Phys. 1953, 21, 836.

(22) (a) One of the reviewers pointed out that pure Dexter PEnT is

of necessity short range as it is predicated on direct orbital overlap

between donor and acceptor components. However, the involvement of

bridging ligand orbitals in a superexchange process can extend the

eﬀective range over which the double-electron exchange can occur, and

PEnT involving a superexchange process across extended bridging

ligands is often referred to as Dexter-type [see refs 4 and 22b,c for

examples]. Superexchange between d- and f-block ions across a bridging

ligand has been the subject of theoretical analysis in a Cr^III/Yb^IIIsystem

(ref 23). (b) Schlike, B.; Belser, P.; De Cola, L.; Sabbioni, E.; Balzani, V.

J. Am. Chem. Soc. 1999, 121, 4207. (c) Benniston, A. C.; Harriman, A.; Li,

P.; Sams, C. A. J. Am. Chem. Soc. 2005, 127, 2553.

(23) Mironov, V. S.; Chibotaru, L. F.; Ceulemans, A. Phys. Rev. B

2003, 67, 014424.

(24) (a) de Sꢀa, G. F.; Malta, O. L.; de Mello Donegꢀa, C.; Simas, A. M.;

Longo, R. L.; Santa-Cruz, P. A.; da Silva, E. F., Jr. Coord. Chem. Rev. 2000,

196, 165. (b) Hebbink, G. A.; Grave, L.; Woldering, L. A.; Reinhoudt,

D. N.; van Veggel, F. C. J. M. J. Phys. Chem. A 2003, 107, 2483.

11339

dx.doi.org/10.1021/ic2007759 |Inorg. Chem. 2011, 50, 11323–11339

Article Doi

D → f energy transfer in a series of IrIII/EuIII dyads: Energy-transfer mechanisms and white-light emission

DOI: 10.1021/ic2007759

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ic2007759

Full text of DOI:10.1021/ic2007759

Products guided by the article

R&D Labs maybe for 1337536-24-9

Relevant to this article

Hot Product