Environmental sensitivity of Ru(II) complexes: The role of the accessory ligands

Article abstract of DOI:10.1021/ic201114u

A suite of Ru(II) complexes in which one ligand is pH responsive and the other two are varied in an effort to achieve improved photophysics has been synthesized and their potential as pH reporters assessed. The more general purpose of the study was to examine the role of the accessory ligands in heteroleptic reporter complexes and the degree to which such ligands can affect the performance of luminescent reporters. For this suite of complexes, judicious choice of the accessory ligand can alter both the pK_a* and the dynamic range of response. It was found that the emission color and brightness were influenced by pH, but the lifetimes were only weakly affected. Surprisingly, some accessory ligands which should have improved luminescent properties essentially turned off the pH response. Several possible reasons for this observation are explored. It is suggested, and density functional theory (DFT) calculations support, that the relative π* levels of the pH sensitive and the accessory ligands are critical.

Full text of DOI:10.1021/ic201114u

Article
pubs.acs.org/IC
Environmental Sensitivity of Ru(II) Complexes: The Role of the
Accessory Ligands
†
†
†
†
,†
Eileen N. Dixon, Michael Z. Snow, Jennifer L. Bon, Alison M. Whitehurst, Benjamin A. DeGraff,*
‡
‡
Carl Trindle, and James N. Demas
†Department of Chemistry, James Madison University, Harrisonburg, Virginia 22807, United States
‡Department of Chemistry, University of Virginia, Charlottesville, Virginia 22901, United States
*
S Supporting Information
ABSTRACT: A suite of Ru(II) complexes in which one ligand is pH responsive and the
other two are varied in an effort to achieve improved photophysics has been synthesized and
their potential as pH reporters assessed. The more general purpose of the study was to
examine the role of the accessory ligands in heteroleptic reporter complexes and the degree
to which such ligands can affect the performance of luminescent reporters. For this suite of
complexes, judicious choice of the accessory ligand can alter both the pK * and the dynamic
a
range of response. It was found that the emission color and brightness were influenced by
pH, but the lifetimes were only weakly affected. Surprisingly, some accessory ligands which
should have improved luminescent properties essentially turned off the pH response. Several
possible reasons for this observation are explored. It is suggested, and density functional
theory (DFT) calculations support, that the relative π* levels of the pH sensitive and the
accessory ligands are critical.
INTRODUCTION
Our simple starting point was that the impact of the
accessory ligands would relate to the degree to which the
various ligand π* orbitals mixed in the emitting electronic state
of the complexes. By selecting accessory ligands with known
useful effects on the emission, we hoped to enhance the
response of the core molecular reporter. To some degree this
hope was realized, but we also discovered unexpected pitfalls.
We chose Ru(II) as the central metal atom and 5-
■
The environmental sensitivity of many luminescent transition
metal complexes has stimulated considerable effort to under-
stand the origins of these effects. Attempts to obtain both a
fundamental understanding and to harness the effects to create
dyes which are useful in sensor applications are widely
1
reported. One of the more salutary aspects of these complexes
is the ability to subtly modify their structure to elicit useful
changes in photophysical behavior. Gross changes can be
2
diethylamino-1,10-phenanthroline (5-Et Nphen) as the envi-
2
ronmentally sensitive ligand. We then systematically changed
the accessory ligands guided by the prior work on the impact of
various ring substitutions on the photophysical properties of
the respective tris complexes of the accessory ligands. As is
often the case, our successes actually generated both answers
and more questions.
obtained by changing the central metal ion, but finer changes
can be obtained by modest alterations in the ligands. Thus, one
can modify the lifetime, emission maxima, and emission
intensity by suitable structural changes to the ligands of the
complex. Unfortunately, such tuning is still as much art as
science.
3
We have been involved in attempting to understand how
relatively small structural changes alter the environmental
sensitivity and photophysical properties of complexes whose
EXPERIMENTAL SECTION
General Information and Materials. Except as noted, all
chemicals were obtained from Sigma-Aldrich and used without further
■
lowest excited state derives from a metal-to-ligand charge4-
purification. The Ru(III)Cl
*·xH O was obtained from Johnson
3
2
transfer (MLCT) transition. The environmental driver is pH,
Matthey Co. cis-Ru(2,2′ bipyridine) Cl *·2H O was purchased from
2
2
2
and central metal ions have included Re(I), Ru(II), and
Strem Chemical Co. The ligands 1,10-phenanthroline (phen),
-chloro-1,10 phenanthroline (5-Clphen), 4, 7-dimethyl-1,10-phen-
5
Os(II). These ions offer the opportunity to create mixed
5
ligand, or heterolepic complexes in which the environmentally
sensitive ligand is partnered with other accessory ligands which
may alter the photophysical response in useful ways. In this
report we examine the role of the accessory ligand in pH
sensitive Ru(II) complexes. Because the analyte detection step
is quite simple and well understood we can focus cleanly on the
role of the accessory ligand. The degree to which these
accessory ligands can impact the response of the complex to its
anthroline(Me phen), 3,4,7,8-tetramethyl-1,10-phenanthroline-
2
(Me phen), 4,7-diphenyl-1,10-phenanthroline(φ -phen), 2,2′-
4
2
bipyridine(bpy), 4,4′-dimethyl-2,2′-bipyridine(Me bpy), and 4,4′-di-
2
phenyl-2,2′-bipyridine(φ -bpy) were purchased from GFS Chemical.
2
Instrumentation. NMR spectra were obtained using a Bruker
ERX 400 MHz instrument. UV−visible spectra were obtained with a
Received: May 25, 2011
Published: March 6, 2012
6
environment is, by comparison, little studied.
©
2012 American Chemical Society
3355
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365

Inorganic Chemistry
Article
Shimadzu UV-1601 spectrometer. Corrected room and low temper-
ature emission spectra were obtained using a Spex FluoroMax instru-
ment. Low temperature, 77 K, emission spectra were obtained using
CH OH/C H OH (1:4) as a glassing solvent. IR spectra were
was removed, and the material, 5-diethylamino, 6-hydroxy-5,6-
dihydro-1,10-phenanthroline used for the next step. Gas chromatog-
raphy showed a single product, but this intermediate was not
characterized.
Then 0.5 g of 5-diethylamino, 6-hydroxy-5,6-dihydro-1,10-phenan-
throline (1.86 mmol) was dissolved in dry, refluxing 1,4-dioxane. To
this was added in 3 portions, 1 g of NaH (60% in mineral oil)
3
2
5
recorded on a Midac FT-IR instrument. Lifetime measurements were
made using a locally constructed apparatus utilizing an LSI -VSL
nitrogen laser (λ = 337 nm, 3 ns pulse width) as the excitation source.
The decay curves were averages of 100 individual decays and analyzed
using a Marquardt based nonlinear least-squares fitting routine.
Quantum yields were based on the accepted value of 0.042 for
(25 mmol, 12 X excess) over 20 min. This mixture was refluxed for
4
h, and the reaction terminated by careful addition of 95% ethanol.
The solvent was removed, and the dark waxy product washed with
7
hexane. The product was dissolved in CH Cl , washed with water, and
2
2
Ru(bpy) Cl in degassed water and the appropriate refractive index
3
2
dried over anhydrous MgSO . The light tan material was shown by gas
4
corrections were applied.
1
1
chromatography to be >98% pure. H NMR spectra and elemental
Structure and purity were verified by H NMR, high resolution mass
spectra, elemental analysis, and UV−vis spectra. Elemental analyses
were done by Atlantic Microlabs, Norcross, GA, while mass spectral
analyses were performed by the Mass Spectrometer Facility, Chemistry
Department, North Carolina State University, Raleigh, NC. Assign-
analysis confirmed the desired product. Typical overall yields were
4
0% based on starting 1,10 phenanthroline. The major loss occurs in
the initial epoxidation step.
1
H NMR(400 MHz, CDCl ): δ 0.99 (t, J = 7.0 Hz, 6H), 3.15 (q, J =
3
7
.0 Hz, 4H), 7.16 (s, 1H), 7.42 (dd, J = 4.5 Hz, 1H), 7.50 (dd, J = 4.3
Hz, 1H), 7.99 (dd, J = 1.8 Hz, 1H), 8.56 (dd, J = 1.8 Hz, 1H), 8.94
dd, J = 1.5 Hz, 1H), 9.05 (dd, J = 1.8 Hz, 1H). Anal. Calcd for
C H N : C, 76.45; H, 6.84; N, 16.71. Found: C, 76.55; H, 6.93; N,
1
ments for H NMR spectra were facilitated by model complex spectra
using 5-methyl-1,10 phenanthroline as the proxy for the 5-
diethylamino-1,10 phenanthroline. The model complexes were
synthesized using unambiguous routes. For complexes with multiple
(
16
17
3
1
6.52.
,7-Dichloro-1,10 phenanthroline (4,7-Cl phen). This preparation
1
isomers and broken symmetry, an unambiguous H NMR assign-
4
2
1
1
ment could not be made. All complexes were purified by column
chromatography and gave materials for which thin layer chromatog-
raphy (TLC) showed only one spot with two different solvent
systems. Emission purity was established by lifetime measurements,
followed the procedures of Snyder and Freier. The final product was
characterized by H NMR and elemental analysis.
1
1
H NMR(400 MHz, CDCl ): δ 7.66(d, J = 4.4 Hz, 2H), 8.12
3
(s, 2H), 8.81(d, J = 4.4 Hz, 2H). Anal. Calcd for C H N Cl : C,
1
2
6
2
2
that is, a single lifetime in CH OH, and an acceptable excitation
57.86; H, 2.43; N, 11.26. Found: C, 57.75; H, 2.36; N, 11.30.
3
8
spectra ratio which was flat to within 2%.
Complexes. Preparation of RuL Cl2 Intermediates. The
2
Electrochemical measurements of the first reduction potential for
heteroleptic complexes RuN_1−RuN_3 and RuN_5−RuN_7
were synthesized by first preparing the appropriate cis-
2+
2+
[
Ru(5-Et Nphen) ] and [Ru(4,7-Cl phen) ] were made using a
2
3
2
3
BAS Model CV-1B potentiostat with a Ag/AgCl reference electrode
and a glassy carbon working electrode. The measurements were done
in dried CH CN with 0.1 M Bu NClO (recrystallized) as added
electrolyte. Ferrocene was used as an internal standard. At scan rates of
5
Ru(II)L
2
Cl
2
complex where L is the accessory ligand. These
intermediates were prepared using a modification of the
9a
procedure reported by Meyer.
Typically 100 mg (0.416 mmol) of RuCl *H O, 2 equiv (0.832
mmol) of the accessory ligand, and 0.5 g of LiCl were refluxed in 5 mL
of DMF for 3 to 6 h. The reactions were followed by TLC (alumina
3
4
4
3
2
0 mV/s the waves were reversible.
Ligands and Complexes. Figure 1 shows the basic core complex
plates with CH Cl /CH CN (80:20) eluant). After the requisite
where L = the accessory ligand. All the complexes were prepared by
adapting well-known literature preparations as described below.
2
2
3
9
reaction time, the solvent volume was reduced, and a small amount of
water was added to the residue. This slush was filtered and washed
with water and diethylether. The dark amorphous solid was dried,
dissolved in CH Cl and placed on a neutral alumina column. Elution
2
2
with CH Cl containing increasing amounts of CH CN served to
2
2
3
isolate the desired dark maroon intermediate. Authentication was by
UV−vis with a strong MLCT band in the 500 nm range. Yields ranged
from 50 to 70%.
The synthesis of Ru(5-Et Nphen) Cl for RuN_7 was more
2
2
2
challenging as this ligand was readily oxidized by the Ru(III) at
DMF reflux temperature, a common problem for functionalized
1
2
ligands. For this complex, 100 mg of RuCl *H O (0.416 mmol) and
3
2
Figure 1. Basic Core Complex (L = Accessory Ligand).
0
.5 g LiCl were placed in a microwave tube with ∼3 mL of 95%
ethanol. The brown mixture was irradiated with 100 W for 5 min at
10 °C (CEM Discover system). To the resulting forest green solution
was added 209 mg (0.832 mmol) of 5-Et Nphen, and the mixture was
Ligands. 5-Diethylamino-1,10-phenanthroline. The pH sen-
sitive ligand was prepared in three steps using an extension of
the work of Kishnan and Shen.
1
10a
10b
2
refluxed for 2 h. The reaction was followed by TLC (alumina plates
with CH Cl /CH CN (80:20) eluant). The reaction was stopped after
To a rapidly stirred mixture of 450 mL of commercial Chlorox,
00 mL of water, 1.25 g of tetrabutylammonium hydrogen sulfate, and
50 mL of CH Cl at 20 °C was added 2.5 g of 1,10-phenanthroline
2
2
3
3
2
complete disappearance of the ligand. The reaction mixture contained
the tris complex, identified by its luminescence, the desired bis
complex (maroon color), and an additional complex which did not
move on the alumina plate. Solvent removal, followed by treat-
ment with aliquots of CH Cl gave a dark maroon solution. The un-
2
2
monohydrate (12.6 mmol). During the 3 h reaction time the pH was
maintained at 8.5−8.6, and the temperature maintained at 20 ± 1 °C.
At the end of the reaction, the CH Cl layer was separated, and its
2
2
2
2
volume reduced to ∼30 mL. Acetone was added until the formation of
the white precipitate was complete. The 5,6-epoxy-1,10 phenanthro-
line was collected via suction filtration and dried in air. Gas chroma-
dissolved solid, mostly LiCl, was removed by filtration, and the
solution placed on a short column of neutral alumina. Elution with
CH Cl and increasing amounts of CH CN separated the various
2
2
3
tography showed this product >97% pure and thus adequate for the
factions and allowed isolation of the Ru(5-Et Nphen) Cl . The yields
2
2
2
1
next step. Yields were ∼50%. H NMR(400 MHz, CDCl ): δ/ppm
were only fair (∼ 25%).
3
4
2
.62(s, 2H), 7.44(dd, J = 4.6, 7.9 Hz, 2H), 8.00(dd, J = 2.0, 7.8 Hz,
H), 8.91(dd, J = 2.0, 4.6 Hz, 2H).
The [Ru(5-Et Nphen) ]Cl was also collected from the column and
2
3
2
converted to the perchlorate salt as outlined below.
Next, 1.0 g (5.1 mmol) of the epoxide, 5 mL (3.54 g, 48 mmol) of
[Ru(5-Et Nphen)L ](ClO ) ; Method 1. This method was used for
2
2
4 2
diethylamine, and 10 mL of DMF were refluxed for 6 h. The solvent
RuN_1−RuN3, RuN_5 and RuN_6. A 100 mg portion of the
3
356
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365

Inorganic Chemistry
Article
appropriate RuL Cl2 intermediate and a 50% molar excess of 5-
Et Nphen were refluxed in 4:1 ethanol:water solvent until TLC-
2
8.39(dd, J = 1.3 Hz, 1H), 8.57−8.61(m, 4H), 8.70(dd, J = 1.3 Hz, 1H).
2
+
ESI-MS (m/z): (M /2), C H N Ru, Expected = 356.5920, Found =
40 33
7
(
alumina plates with CH Cl /CH CN (80:20) eluant) showed
356.5912. Anal. Cal.(%) for C H N O Cl Ru*2H O: C = 50.69, H =
2
2
3
40 33 7 8 2 2
complete conversion of the intermediate; usually 3 to 5 h. Dropwise
addition of this reaction mixture to a hot, stirred solution of saturated
LiClO was followed by removal of the ethanol via an N stream. The
3.93, N = 10.34. Found(%): C = 50.43, H = 3.90, N = 10.45.
1
[Ru(5-Et Nphen) ](ClO ) . (RuN_4). H NMR.(Mixture of geo-
2
3
4 2
+
4
2
metric isomers). ESI-MS (m/z): (M /2), C H N Ru, Expected =
48 51
9
mixture was cooled in an ice/water bath, and the resulting precipitated
perchlorate complex was isolated by suction filtration and washed with
cold water. The crude complex was dried at 65 °C and then dissolved
in either acetone or CH Cl . The complex solution was placed on a
427.6657, Found = 427.6647. Anal. Cal(%) for C H N -
48 51 9
O Cl Ru*2H O: C = 52.89, H = 5.08, N = 11.56. Found(%): C =
8
2
2
53.01, H = 4.98, N = 11. 48.
[Ru(5-Et Nphen)(Me phen) ](ClO ) . (RuN_5). 1H NMR (400
2
2
2
4
2
4 2
neutral alumina column and eluted with either acetone or CH Cl2
MHz, CD CN): δ 1.15(t, J = 7.0 Hz, 6H), 2.22(m, 12H), 2.75(m,
2
3
containing 4% CH OH and 1% triethylamine. The desired fraction
12H), 3.38(q, J = 7.0 Hz, 4H), 7.46(q, J = 5.0 Hz, 1H), 7.55(q, J = 5.3
Hz, 1H), 7.67(s, 1H), 7.70−7.76(m, 5H), 7.92(dd, J = 1.3 Hz, 1H),
3
was collected, concentrated and then added dropwise to a rapidly
stirred beaker of chilled diethylether. The resulting orange solid was
then collected via suction filtration. Yields ranged from 70−90%.
Warning! Perchlorates are potentially explosive and should be handled
with great care! Do not heat to dryness and use small amounts.
8
.33−8.35(dd, J = 1.3 Hz, 5H), 8.66(dd, J = 1.2 Hz, 1H). ESI-MS
+
(
m/z): (M /2), C H N Ru, Expected = 427.6657, Found =
48 49 7
4
27.6647. Anal. Cal(%) for C H N O Cl Ru*2H O: C = 54.39,
48 49 7 8 2 2
H = 5.04, N = 9.24. Found(%): C = 54.58, H = 5.00, N = 9.12.
[
Ru(5-Et Nphen)L ](ClO ) ; Method 2. For the complexes RuN_8
1
2
2
4 2
[Ru(5-Et Nphen)(Me phen) ](ClO ) . (RuN_6). H NMR(400
2
2
2
4 2
and RuN_11−13, the intermediate Ru(5-Et Nphen)(DMSO) Cl was
2
2
2
MHz, CD CN): δ 1.18(t, J = 7.0 Hz, 6H), 2.91(d, J = 2.5 Hz,
3
9
d
first prepared following a procedure similar to that of Wilkinson.
A
1
2H), 3.41(q, J = 6.8 Hz, 4H), 7.46−7.52(m, 5H), 7.58(q, J = 5.3 Hz,
2
00 mg portion of RuCl *H O (0.832 mmol) was refluxed for 10 min
3
2
1H), 7.67(s, 1H), 7.82−7.92(m, 5 H), 7.99(dd, J = 1.3 Hz, 1H), 8.37−
in 5 mL of dimethylsulfoxide. The solvent volume was reduced and
acetone added resulting in a copious precipitate of yellow/green solid.
The solid Ru(DMSO) Cl was collected and washed with acetone and
+
8
.40(m, 5H), 8.70(dd, J = 1.2 Hz, 1H). ESI-MS (m/z): (M /2), C H
44
N Ru, Expected = 384.6244, Found = 384.6249. Anal. Cal(%) for
7
41
4
2
C H N O Cl Ru*H O: C = 53.60, H = 4.40, N = 9.94. Found(%):
44
41
7
8
2
2
diethylether. Mp 192−194 °C (decomp). Yield ∼70%.
C = 53.48, H = 4.36, N = 10.05.
Ru(5-Et Nphen) ](CN) . (RuN_7). H NMR (Mixture of geometric
A 100 mg portion of Ru(DMSO) Cl (0.206 mmol) and 52 mg
1
4
2
[
2
2
2
(
0.206 mmol) of 5-Et Nphen were refluxed in 5 mL of DMF for
+
2
isomers). ESI-MS (m/z): (M+H) , C H N Ru, Expected 656.2032,
34
35
8
1
5 min. The solvent was removed, and the reaction product dissolved
in 15 mL of 95% ethanol. To this was added a 50% molar excess
0.618 mmol) of the appropriate accessory ligand, and the mixture
Found = 656.2031. Anal. Cal(%) for C H N Ru*2H O: C = 59.03,
34
34
8
2
H = 5.53, N = 16.19. Found(%): C = 58.85, H = 5.56, N = 16.13.
(
1
[
Ru(5-Et Nphen)(4,7-Cl phen) ](ClO ) . (RuN_8). H NMR(400
2 2 2 4 2
refluxed for 5 h. The reaction mixture was then added dropwise to a
MHz, CD CN):δ 1.14(t, J = 8.5 Hz, 6H), 3.48(q, J = 8.4 Hz, 4H),
3
stirred saturated solution of LiClO and treated as in Method 1.
4
7
1
8
.47(dd, J = 4.7, 5.2 Hz, 1H), 7.54(dd, J = 4.7, 5.2 Hz, 1H), 7.62(s,
H), 7.71−7.74(m, 4H), 7.80(d, J = 5.2 Hz, 1H), 7.91−7.99(m, 5H),
Ru(5-Et Nphen) (CN) . To 100 mg (0.148 mmol) of Ru(5-
2
2
2
Et Nphen) Cl dissolved in 30 mL of dry tetrahydrofuran (THF)
2
2
2
.37(d, J = 7.8 Hz, 1H), 8.55(s, 4H), 8.68(d, J = 8.4 Hz, 1H). ESI-MS
was added 31 mg (0.148 mmol) of anhydrous AgClO . This mixture
4
+
(
m/z): (M+H) , C H N Cl Ru, Expected 850.0360, Found =
40
30
7
4
was stirred at 50 °C for 3 h, cooled, and the AgCl precipitate was
removed via suction filtration. To the solution was added 200 mg of
8
2
50.0352. Anal. Cal(%) for C H N O Cl Ru*H O: C = 45.01, H =
40 29 7 8 6 2
.93, N = 9.18. Found(%): C = 45.10, H = 2.84, N = 9.05.
Ru(5-Et Nphen)(φ -phen) ](ClO ) . (RuN_11). 1H NMR(400
(
4.1 mmol, 27 X excess) NaCN dissolved in a minimum amount of
[
2
2
2
4 2
water. This mixture was refluxed for 4 h, cooled, and the solvent
MHz, CD CN): δ 1.18(t, J = 7.0 Hz, 6H), 3.41(q, J = 7.0 Hz, 4H),
3
removed. To the solid was added both water and CH Cl and stirred
2
2
7
3
1
.57−7.70(m, 26H), 7.67(s, 1H), 7.99(d, 5.3 Hz, 1H), 8.11−8.15(m,
until complete dissolution. The CH Cl layer was separated, washed
2
2
H), 8.16(d, 1.8 Hz, 4H), 8.22(dd, 5.6, 5.5 Hz, 2H), 8.44(d, 8.6 Hz,
twice with water, dried over anhydrous MgSO , and the solvent
4
+
H), 8.76(d, 8.5 Hz, 1H). ESI-MS (m/z): (M+H) /2 C H N Ru,
6
4
50
7
removed. The resulting orange/red solid was purified by column
chromatography using neutral alumina with CH Cl /CH CN mixtures
Expected = 509.1585, Found = 509.1582. Anal. Cal(%) for C H N -
6
4
49
7
2
2
3
O Cl Ru*H O: C = 62.28, H = 4.16, N = 7.94. Found(%): C = 62.52,
8
2
2
as the eluant. Yields were ∼50%.
H = 4.11, N = 8.13.
[Ru(5-Et Nphen)(5-Clphen) ](ClO ) . (RuN_12). H NMR (Mix-
Individual Complexes. [Ru(5-Et Nphen)(bpy) ](ClO ) .
2
2
4 2
RuN_1). H NMR (400 MHz, CD CN): δ 1.15(t, J = 7.0
1
1
(
Hz, 6H), 3.38(q, J = 7.0 Hz, 4H), 7.24(t, J = 7.0 Hz, 2H),
2
2
4 2
3
+
ture of geometric isomers). ESI-MS (m/z): (M+H) C H N Cl Ru,
40 32
7
2
Expected = 782.1140, Found = 782.1147 . Anal. Cal(%) for C H N -
7
7
8
8
1
.43(m, 2H), 7.56(d, J = 5.3 Hz, 2H), 7.60(q, J = 3.0 Hz, 1H),
.61(s, 1H), 7.68(q, J = 3.3 Hz, 1H), 7.82−7.87(m, 3H), 7.96−
.03(m, 3H), 8.06−8.10(m, 2H), 8.40(dd, J = 1.3 Hz, 1H),
.49(d, J = 8.0 Hz, 2H), 8.52(dd, J = 1.3 Hz, 2H), 8.72(dd, J =
40 31
7
O Cl Ru*H O: C = 48.11, H = 3.52, N = 9.81. Found(%): C = 47.89,
8
4
2
H = 3.21, N = 9.71.
Ru(5-Et Nphen)(φ -bpy) ](ClO ) . (RuN_13). H NMR(400 MHz,
1
[
2
2
2
4 2
+
CD CN): δ 1.16(t, 7.0 Hz, 6H), 3.38(q, 7.2 Hz, 4H), 7.38−7.51(m,
.3 Hz, 1H). ESI-MS (m/z): (M ), C H N Ru, Expected =
3
36
33
7
1
8
5
1H), 7.60(dd, 5.8 Hz, 2H), 7.72−7.89(m, 12H), 8.06(d, 5.8 Hz, 2H),
6
65.1841, Found = 665.1850. Anal. Cal.(%)for C H N O Cl Ru*-
36 33 7 8 2
H O: C = 49.09, H = 3.99, N = 11.11. Found(%): C = 48.78,
.12(d, 5.8 Hz, 1H), 8.15(d, 5.8 Hz, 2H), 8.26(d, 7.5 Hz, 1H), 8.35(d,
2
H = 4.10, N = 10.95.
.0 Hz, 1H), 8.57(s, 2H), 8.61(d, 5.0 Hz, 4H), 8.67(d, 8.4 Hz, 1H).
+
Ru(5-Et Nphen)(Me bpy) ](ClO ) . (RuN_2). 1H NMR (400
60 50
7
ESI-MS (m/z): (M+H) /2 C H N Ru, Expected = 485.1586, Found =
[
2
2
2
4 2
4
85.1583. Anal. Cal(%) for C H N O Cl Ru: C = 61.70, H = 4.23,
MHz, CD CN): δ 1.18(t, J = 7.0 Hz, 6H), 2.51(d, J = 3.0 Hz, 6H),
60 49 7 8 2
3
2.60(s, 6H), 3.44(q, J = 7.0 Hz, 4H), 7.21(dd, J = 7.5 Hz, 2H), 7.44(d,
N = 8.38. Found(%): C = 61.71, H = 4.21, N = 7.92.
J = 5.8 Hz, 2H), 7.65(dd, J = 1.8 Hz, 2H), 7.79(q, J = 5.3 Hz, 1H),
pH Titrations. Stock aqueous solutions of the various complexes
were made by stirring ∼100 mL of deionized water with an excess of
the appropriate complex for about 4 h at room temperature. The
solution was filtered through a Gelman 0.45 μ Acrodisc. The filtered
stock served as the basis for both intensity and lifetime measurements.
7
5
.85(s, 1H), 7.88−7.95(m, 3H), 8.18(d, J = 5.0 Hz, 1H), 8.37(d, J =
.3 Hz, 1H), 8.60(d, J = 8.3 Hz, 1H), 8.69(s, 2H), 8.74(s, 2H), 8.83(d,
+
J = 8.3 Hz, 1H). ESI-MS (m/z): (M ), C H N Ru, Expected =
40
41
7
7
21.2467, Found = 721.2462 . Anal. Cal(%) for C H N O Cl Ru*-
2
40 41 7 8 2
H O: C = 51.20, H = 4.62, N = 10.45. Found(%): C = 51.46, H =
The buffer solutions used employed H
PO , KH PO , K HPO , and
3 4 2 4 2 4
4
.60, N = 10.44.
K
3
PO in varying proportions. The buffer concentration was 0.1 M.
4
1
[
Ru(5-Et Nphen)(phen) ](ClO ) . (RuN_3). H NMR (400 MHz,
The pH of the buffers was determined using a Corning Model 440 pH
meter. Typically, 1 mL of the stock solution of complex and 3 mL of
buffer were combined for the intensity or lifetime measurement. There
was no detectable pH difference between the pure buffer solution and
2
2
4 2
CD CN): δ 1.26(t, J = 7.3 Hz, 6H), 3.04(q, J = 7.3 Hz, 4H), 7.49(q,
3
J = 5.3 Hz, 1H), 7.56−7.67(m, 6H), 7.81(dd, J = 1.3 Hz, 1H), 7.97(dd,
J = 1.0 Hz, 1H), 7.99−8.03(m, 2H), 8.05−8.08(m, 2H), 8.24(s, 4H),
3
357
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365

Inorganic Chemistry
Article
the test solution. The luminescence intensity and lifetime measure-
ments were made with air-saturated solutions.
Buffered solution lifetimes for all complexes, except RuN_7 at pH <
could be fit with a single exponential suggests this simple model is
adequate. The experimental data were fit with either eq 4 or 5 using
PSI Plot software. The fits were all satisfactory, gave internally
3
.5, were satisfactorily fit by a single exponential. The lifetime for
consistent values of pK * (i.e., intensity vs lifetime), and provided
chemically reasonable values.
a
RuN_7 was evaluated as a mean lifetime using eq 1.
On the basis of absorption changes with pH (see Figure 2A), eq 4
was modified for use with absorbance and used to calculate the ground
2
τM = ∑(A τ /∑ A τ )
i i
j j
(1)
state pK for several of the complexes.
a
Here A and τ i are the pre-exponential and lifetime for the i-th
i
component. ∑A τ is the sum of the pre-exponential-lifetime product
j
j
terms. Only 2 exponentials were required for satisfactory fit. The
necessity of multiple terms is attributed to the onset of protonation of
13
the CN ligands.
To test specific ion effects, a limited number of both intensity and
lifetime titrations were also done using a buffer system based on
H SO , KHSO , and K SO . This buffer yielded results indistinguish-
2
4
4
2
4
able from the phosphate buffer system. Tests were also run to ensure
the complexes were stable at low pH for the duration of our
measurements (Supporting Information, Figure S1).
Data Fitting/Analysis. Except as noted above, all the
luminescence decays observed in this study were satisfactorily fitted
with a single exponential decay with no evidence of a second
component. At the high buffer concentrations used this is not
unexpected and is consistent with the rapid exchange limit. In the fast
exchange limit, it is assumed that the excited state equilibrium between
−
HA* and A * is maintained. Under these conditions only a single
decay is seen, characterized by a single observed lifetime, τobs, which
implies a single observed decay rate constant kobs. Using Scheme 1, this
Scheme 1. Protonation Equilibrium for Ground and Excited
States
allows for very simple expressions for the observed intensity and
lifetime variations as a function of the fraction of each species
14
present.
max
max
Intensity:
Lifetime:
I
= f
= f
I
+ f I
A A
obs
HA HA
(2)
(3)
Figure 2. (a) Electronic absorbance spectra for [Ru(5-Et Nphen)-
2
2+
2+
(
bpy)2] at different pH in buffer solutions and Ru(bpy)3 in water.
k
k
+ f k
A A
obs
HA HA
2
+
(
b) Corrected emission spectra for [Ru(5-Et N phen)(bpy) ] at high
2 2
−
Here f HA and f are the fractions of HA* and A * present,
and low pH in buffer solutions.
A
max
respectively, I is the emission intensity when only that species is
present, and k and k are the decay rate constants for the HA* and
Density Functional Theory (DFT) Calculations. We modeled
the complex structures of RuN_1′ (simplified form of RuN_1 by
replacement of Et with Me) and the 5-Clphen species, RuN-12, using
HA
A
−
A * species. The fraction of each species present will depend on the
pH and pK * of the excited state of the complex. When this
a
1
5
dependence is incorporated into eqs 2 and 3, one obtains
Gaussian 09 software for computations with the familiar B3LYP
16
variant of the density functional, using the Stuttgart−Dresden, SDD,
1
7
basis which includes a pseudopotential for the metal and reduces
the number of electrons treated explicitly. For these large systems
this modest basis still included up to 581 members for RuN_12.
Counterions were not included in the modeling. Optimization of the
structures of the complexes and calculations of spectra incorporated a
pH−pKa*) −1
× IHAmax
I
= (1 + 10(
)
obs
[1 − (1 + 10(
pH−pKa*) −1
] × IAmax
+
)
(4)
(5)
k
= 1/τobs = (1 + 10(
[1 − (1 + 10(
pH−pKa*) −1
pH−pKa*) −1
)
× kHA
Polarizable Continuum Medium model of the solvent with the
18
obs
dielectric constant of water. Absorption spectra were estimated by TD-
+
)
] × kA
19
DFT with the B3LYP density functional and the same SDD basis.
The coordinates of the optimized geometries are provided in the
Supporting Information.
In the rapid exchange limit, these expressions are rigorously correct
and make no assumptions regarding the excitation wavelength, only
the limiting intensities or decay rates are required. Values for limiting
RESULTS AND DISCUSSION
max
■
I
and k were obtained from the appropriate data at the limits of the
Table 1 shows basic photophysical properties of the complexes
pH range used where both the intensity and the lifetime had reached a
constant value. Fitting the experimental data using a Marquardt
approach required only a single floating parameter, pK *, the pK of
in CH OH. All complexes exhibit the expected MLCT
3
absorption in the 450 nm region with ε values in line with
a
a
13
the excited state. The fact that all decays except for RuN_7 at low pH
similar complexes. In addition to the MLCT band, the
3
358
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365

Inorganic Chemistry
Article
Table 1. Photophysical Properties for RuL′L Complexes in CH OH; L′ = 5-Et N-1,10-phenanthroline, L = Accessory Ligand
2
3
2
Ru(L′)(L)2]2+
Abs λmax (nm)
ε (M−1 cm−1
1.83 × 104
2.33 × 10
1.52 × 104
)
a
Emis λmax (nm)
b
τ (μs)
c
k (O ) M s−1
−1
[
q
2
d
9
9
9
9
9
9
9
9
9
9
9
[
[
[
[
[
[
[
[
[
[
[
Ru(L′)(bpy) ]; RuN_1
453 , 424,370
609
609
597
594
612
603
631
632
613
604
624
0.824
1.06
1.45
0.884
1.58
1.26
1.40
2.02
4.86
1.12
1.52
2.0 × 10
3.2 × 10
3.8 × 10
3.6 × 10
4.4 × 10
3.7 × 10
3.0 × 10
3.7 × 10
2.2 × 10
1.6 × 10
1.7 × 10
2
d
4
Ru(L′)(Me bpy) ] ; RuN_2
456 , 426,350
2
2
d
Ru(L′)(phen) ]; RuN_3
448 , 421
2
d
4
Ru(5-Et Nphen) ]; RuN_4
456, 423, 369
2.13 × 10
2
3
Ru(L′)(Me phen) ]; RuN_5
423
2.08 × 104
4
2
d
4
Ru(L′)(Me phen) ]; RuN_6
448, 425
1.73 × 10
2
2
d
1.03 × 104
Ru(L′) (CN) ]; RuN_7
457, 360
2
2
4
Ru(L′)(4,7-Cl phen) ]; RuN_8
448
2.44 × 10
2
2
d
2.44 × 104
1.46 × 104
2.98 × 104
Ru(L′)(φ -phen) ]; RuN_11
459 , 433
2
2
d
Ru(L′)(5-Clphen) ]; RuN_12
450 , 423
2
d
Ru(L′)(φ -bpy) ]; RuN_13
471 , 443
2
2
a
b
c
d
ε for strongest peak. Corrected. Ar purged/CH OH. Strongest peak.
3
complexes also exhibited near UV bands attributable to π−π*
transitions of the ligands. The emissions were centered around
600 nm with a range of only about 40 nm among the various
λmax values. The lifetimes in purged CH OH, with the exception
3
of RuN_11, were in the 1 to 2 μs range and showed only
modest variation with changing spectator ligands. The emission
is readily quenched by oxygen with quenching rates near the
diffusion limit. This characteristic has an adverse impact on
their use as sensors, though the lower solubility of O in water,
2
20
∼
0.265 mM, ameliorates the problem to some degree.
Additionally, we have found that when the complexes are
embedded in supports such as polymers, the quenching is
2
1
further reduced.
Figure 2A shows the impact of protonation on the UV
spectra of [Ru(5-Et Nphen)(bpy) ](ClO ) . The spectrum of
2
2
4 2
[
Ru(bpy) ]Cl is also shown for comparison. The absorption
Figure 3. Emission intensity, corrected for both instrument response
3
2
2+
between 250−270 nm is assigned to the 5-Et Nphen ligand
on the basis of its singular absence in the model complex,
and absorbance, for [Ru(5-Et Nphen)(bpy) ] as a function of pH in
2
2 2
buffer solutions. Emission monitored at 595 nm.
[
Ru(bpy) ]Cl , and the pH dependence of this band. It is
3 2
clear that protonation of the amine group causes a shift in the
Table 2. pK and Dynamic Ranges for Ru(II) Complexes
a
energy levels of the 5-Et Nphen ligand. This effect was used to
2
excited
state pKa
dynamic
range
ground
state pKa
determine the pK of the ground state for a representative
alias
complex
a
sample of the complexes. The MLCT bands were also sensitive
to pH, though not as dramatically as the near UV (Supporting
Information, Figure S2). Qualitatively the relative change in
absorbance with protonation tracks the dynamic range
observed for the emission as a function of pH. Several of the
complexes, RuN_8 and RuN_11−RuN_13, showed very
muted luminescence response to pH, and their absorbance
changes with pH in the 230−350 nm region were also virtually
non existent.
RuN_1
RuN_2
RuN_3
RuN_4
RuN_5
RuN_6
[Ru(5-Et Nphen)(bpy) ]
3.1
3.7
3.3
4.1
3.9
3.8
4.5
2.4
4.4
2.9
3.0
6.2
1.9
1.8
2
2
(ClO4)2
[Ru(5-Et
Me bpy) ](ClO )
4 2
Nphen)
2
2
(
2
[Ru(5 Et Nphen)
2.3
2.2
2
(
phen) ](ClO )
2 4 2
[Ru(5-Et Nphen) ]
2
3
(ClO4)2
[Ru(5-Et Nphen)
2
(
Me phen) ](ClO )
4 2 4 2
2
[Ru(5-Et Nphen)(Me2
Figure 2B shows the impact of protonation on the emission
of the same complex, RuN_1. Protonation results in a red shift
in λmax and a decrease in the intensity of emission, consistent
with reports for other nitrogen based pH sensitive Ru(II)
phen) ](ClO )
2
4 2
RuN_7
RuN_8
[Ru(5-Et Nphen) ](CN)
2
3.5
2.9
2
2
a
[Ru(5-Et Nphen)(4,7-Cl2
N.R.
<1.3
2
phen) ](ClO )
2
4 2
2
2
complexes. The actual change in the integrated emission is
only ∼35%, but by judicious choice of monitoring wavelength,
the dynamic range, the ratio of single λ intensity at high and
low pH, can be significantly improved. Most of the complexes
gave useful pH-intensity response. This is illustrated in Figure 3
for RuN_1. Similar plots showed significant changes in the
emission with pH for all complexes except RuN_8 and
RuN_11−RuN_13. Table 2 summarizes the results.
a
RuN_11 [Ru(5 Et Nphen)
N.R.
<1.3
<1.3
<1.3
2
(φ
RuN_12 [Ru(5 Et
5-Clphen) ](ClO )
4 2
-phen)
2
](ClO )
2 4 2
a
Nphen)
2
N.R.
2
(
a
RuN_13 [Ru(5 Et Nphen)
N.R.
2
(
φ -bpy) ](ClO )
2 2 4 2
a
N.R. = not pH responsive.
Figure 4 shows a typical plot for the changes in emission
lifetime with pH. The lifetime response to pH changes was very
muted compared to that of the emission intensity. For none of
the complexes was the dynamic range over 2, and many were
nearly unresponsive. The small dynamic ranges encountered
using lifetime measurements made the accurate determination
of the pK * very difficult. For this set of complexes, using the
a
lifetime as an indicator of pH is not useful.
3
359
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365

Inorganic Chemistry
Article
by using a charged accessory ligand, as with RuN_7, the pK * is
a
incremented by >1 unit. However, a second surprise is that 5-Cl
phen, 4,7-Cl phen, φ -phen, φ -bpy accessory ligands effec-
2
2
2
tively mute the pH response of the complexes (Supporting
Information, Figures S3A−S3D). The latter two accessory
ligands were chosen because of the relatively high quantum
13
yields of their tris complexes. This finding forced us to
reevaluate our simple localized excitation model of the excited
state.
Ground State Mixing in Heteroleptic Ru(II) Com-
plexes. In the simplest or decoupled picture of the MLCT
electronic absorption process for a Ru(II) complex, an electron
from a d orbital of the metal ion is transferred to one of the π*
ligand orbitals. This is a vertical transition with no spin change,
resulting initially in a singlet excited state with a considerable
change in the dipole moment of the complex. Elegant fast laser
studies have revealed much about the temporal evolution of
prototype complexes during the post-absorption time win-
Figure 4. Emission lifetime as a function of pH for [Ru(5-
2+
Et Nphen)(Me phen) ] in buffer solutions.
2
4
2
2
5
dow. These studies suggest that the final thermally
equilibrated triplet emitting state is populated within a few
picoseconds and that under some circumstances significant
localization of the excitation can occur on a particular ligand.
This is shown qualitatively as an evolving time diagram in
Scheme 2. For tris complexes, the π* energy levels of all the
ligands are identical. The recent time-resolved absorption
Using the Fo
̈
rster cycle to describe the relationship between
a
the pK and pK *, we expect
a
pK * = pK + (0.625/T)(v − v
)
B
HB
(6)
a
a
where (v − v ) is the difference between the emission of the
B
HB
−
1
23
neutral and protonated forms in wavenumbers (cm ).
3
Because the emitting state is MLCT the emission data must
be used. This relationship is well followed for these complexes.
For example with RuN_1 the expected (v − v ) is 620 cm
versus an observed value of 612 cm . This relatively good
agreement suggests that protonation equilibrium is established
in the excited state. When the fast exchange limit applies, one
also expects
25c
studies seem to point to significant excitation delocalization
−1
in contrast to earlier work which, on balance, seemed to favor
B
HB
26
−1
localization.
For heteroleptic complexes the picture is more complex in
that the π* orbitals of the various ligands are only rarely and
accidentally the same energy. In this case, the simple decoupled
model would suggest that the observed absorption spectrum
would be the weighted sum of the allowed MLCT transitions
pK * = pH + log τHB/τ
B
(7)
a
i
27
associated with various component ligands. If, however, there
is some ground state mixing of the various ligand π* orbitals,
this simple approximation should give observably poor fits. This
latter statement is somewhat arbitrary as the MLCT absorption
for many tris Ru(II) complexes are not all that different.
where pH is the inflection point in the intensity versus pH
i
titration curve. However, as the lifetimes of the two forms for
these complexes differ by less than a factor of 2 under the most
propitious conditions, the pK * is effectively the pH .
The trends in pK * values determined using the emission
changes with pH are consistent with our earlier observations
using complexes with 5-carboxy-1,10-phenanthroline as the pH
sensitive element. One major surprise, however, is the
significant shift in the ground state pK when the free pH
sensing ligand, 5-diethylamino-1,10-phenanthroline is com-
plexed with Ru(II). We were not able to measure the pK of the
a
i
a
Most of our complexes were of the type RuL′L , where L′ was
2
the pH sensitive ligand 5-Et Nphen and the L′s were common
2
5
accessory ligands based on bpy or phen and their various ring
substituted analogues. Absorption spectra were obtained in
a
2+
2+
CH OH for RuL′ , our RuL′L2 complexes, and the related
3
3
2+
RuL3 complexes. All the spectra were normalized to the same
a
absorbance at the MLCT λmax . The expected spectrum for the
free ligand directly, but the pK of N,N-diethylaniline should be
a
2+
24
heteroleptic complex was then calculated as per RuL′L
(
=
2
a reasonable proxy. This aniline has a pK ∼ 6.6. While a shift
a
2 × RuL32+ + RuL′3 )/3. Because the normalization effec-
2+
to lower pK is expected because of the electron withdrawing
a
tively removes the differences in molar extinction of the com-
ponents, the final calculated spectra were scaled so that the
calculated and actual heteroleptic spectra matched at λmax . This
involved multiplication by a constant that was never greater
than ∼1.07.
power of the Lewis acid metal ion, our ground state pKa
measurements for these complexes fall in the range of ∼2 to 3,
which represents a shift of 4 orders of magnitude. As might be
expected from the shift in electron density accompanying an
MLCT excitation in which the transferred metal electron
The two spectra, actual and calculated, for several of our
heteroleptic complexes are shown in Figures 5A−5D. These
include both our best and worst fits as well as two non pH
responsive cases. Though our method is crude, the agreement
is actually quite good and suggests that this simple, isolated
chromophore model is a reasonable approximation for the
absorption process. However, the failure to obtain exact fits
may indicate some π* or other state mixing even in the ground
state for heteroleptic complexes in which the various π* orbitals
density is localized in the pH sensitive ligand, the pK * of the
a
excited states are about 1 to 2 orders of magnitude higher than
14a
the ground state values. Unfortunately, the excited state pKa*
values are well below the optimal values necessary for both
environmental and physiological monitoring.
The impact on the excited state pK * of varying the accessory
a
5
ligands is similar to our earlier study. That is, if we take bpy or
phen as the baseline for our accessory ligands, pendent methyl
groups which are electron donating can raise the pK * by as
a
28
much as 0.5 units. If the formal charge on the Ru(II) is reduced
are not energetically widely separated.
3
360
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365

Inorganic Chemistry
Article
Scheme 2. Time Evolution of the Excited State after MLCT Absorption
2
+
Figure 5. Comparison of actual and calculated electronic absorption spectra in CH OH for (a) [Ru(5-Et Nphen)(Me bpy) ] ; (b) [Ru(5-
3
2
2
2
2+
2+
2+
Et Nphen)(Me phen) ] ; (c) [Ru(5-Et Nphen)(φ -phen) ] ; (d) [Ru(5-Et Nphen)(5-Clphen) ] .
2
2
2
2
2
2
2
2
Our approach to modeling the absorption spectra of these
complexes using time-dependent DFT (TD-DFT) methods
was limited to one-electron excitations that often under-
estimates excitation energies. The complexity of the systems
and limitations of our methods make the calculated spectra
below 300 nm of lower reliability. We examined both a pH
responsive (RuN_1′) and nonresponsive (RuN_12) complex
and there was reasonable qualitative agreement between the
calculated and observed spectra (Supporting Information).
Features well captured by the calculations include the
following: (1) General similarity of the form of the absorption
spectra of all species; first absorption band (MLCT) located
near 450 nm. (2) Noticeable intensity reduction was observed
with protonation of the responsive species, while a lesser
change was observed for the pH insensitive species. We also
note that the frequently used orbital pictures were not
particularly helpful in our effort to track the changes in
electron density accompanying excitation.
29
Excited State Mixing and Its Impact on Environmental
Sensitivity. The almost total lack (<10% intensity change,
3
361
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365

Inorganic Chemistry
7 nm λmax emission shift) of pH sensitivity for the complexes
Article
<
[
Ru(5-Et Nphen)(φ -phen) ](ClO ) , [Ru(5-Et Nphen)(5-Cl
2 2 2 4 2 2
phen) ](ClO ) , [Ru(5-Et Nphen)(4,7-Cl phen) ](ClO ) ,
2
4
2
2
2
2
4 2
and [Ru(5-Et Nphen)(φ -bpy) ](ClO was surprising. Re-
2
2
2
4
markably, the φ -phen and φ -bpy complexes have the highest
2
2
quantum yields and longest lifetimes of any of the suite.
On the basis of the localized excitation model (LEM), the
emission originates from the lowest energy excited state and
this state, for a heteroleptic complex, would involve the ligand
with the lowest energy π* orbital. If this ligand were the
accessory ligand, L, then the environmental sensitivity of the
complex would be muted unless protonation of the sensing
ligand reversed the relative π* energies. However, the fact that
several complexes containing accessory ligands with lower
energy π* orbitals than the pH sensing ligand gave acceptable
sensitivity suggested that this simple model may be inadequate.
Two modifications of the narrowly defined LEM can be
suggested to address our experimental observations. The first
retains the major features of the LEM but suggests that when
the difference in the π* energies of the two ligands, pH sensing
−1
and accessory, are not greatly different (i.e., < 500 cm ), then
at room temperature there could be a substantial population in
which the excitation is localized on the higher π* energy ligand.
For this model the response would be temperature dependent
(
TD-LEM).
A second model would invoke some measure of orbital
mixing in the excited state in which the emission originates
from an excited state of mixed orbital parentage (MOP). The
amount of mixing between the sensing ligand and accessory
ligand orbitals would depend, in part, on the relative energies of
their respective π* orbitals. The closer in energy, the greater the
expected degree of mixing.
Figure 6. Comparison of corrected, normalized low temperature
2
+
(
(
77 K) emission spectra in CH OH/C H OH (1:4) (a): [Ru(phen) ] ,
3 2 5 3
2
+
2+
1); [Ru(φ -phen) ] , (2); [Ru(phen) (φ -phen)] , (3). (b):
2
3
2
2
If the TD-LEM is correct, then at low temperature where
thermal access to the higher π* ligand is suppressed, the
emissions from the heteroleptic complexes should very closely
resemble those from the corresponding tris-complex with the
lower π* ligand.
2+
2+
[
Ru(phen) ] , (1); [Ru(φ phen) ] , (2); [Ru(5-Et Nphen)(φ -
3 2‑ 3 2 2
2+
phen) ] , (3).
2
Again, the general features of the emission are those of the tris
5
-Et Nphen complex.
2
We examined the emission for the prototype complex
As satisfying as these experiments are in suggesting that a
2
+
[
Ru(phen) (φ -phen)] at low temperature. The measured
2 2
mixed orbital excited state is present in our heteroleptic
complexes, the actual situation is more complex. We examined
the same cluster of complexes at room temperature, with
startlingly different result. At the higher temperature, we would
2
+
2+
λmax values for Ru(phen)3 and Ru(φ -phen) complexes at
2
3
−1
7
7 K in CH OH/C H OH glass are 17,700 cm and 16,300
3 2 5
−1
−1
cm respectively; ΔE ∼ 1,400 cm . Assuming these values
reflect the relative emitting state energies and are a proxy for
the relative positions of the respective π* orbitals, the proba-
bility of the emission involving a particular ligand assuming a
simple exponential population dependence leads to a 6.25 ×
still expect emission from the φ -phen ligand to dominate
2
according to the TD-LEM. However, the temperature depen-
dence of the relative mixing of the π* orbitals of the ligands is
unknown. In addition, there may be solvent relaxation and
conformational relaxations that affect this mixing. Further, the
change from a rigid to fluid environment as well as different
Franck−Condon factors due to higher temperature may
significantly change the shapes of the emission envelopes.
As Figures 7A and 7B show, room temperature, fluid solution
results in several changes. The entire spectra set are strongly
red-shifted and much of the sharpness is lost. Also, the low
temperature “phen” structure is lost, resulting in broad
featureless peaks of nearly Gaussian shape. Remarkably, the
spectra of the heteroleptic complexes are quite similar to the
1
0
1
0
advantage for the emission involving the φ -phen ligand;
2
neglecting any degeneracy terms. Thus, the low tempera-
2
+
ture emission spectrum for [Ru(phen) (φ -phen)] should
2
2
2
+
be quite similar to that of [Ru(φ -phen) ] based on the
2
3
TD-LEM.
As Figure 6A clearly shows, the TD-LEM fails to predict the
observed spectrum. We also tested this model using the pH
2
+
sensing complex [Ru(5-NH phen)(φ -phen) ] and compar-
2
2
2
ing it with the respective tris-homoleptic complexes. As Figure 6B
shows, the simple TD-LEM fails here as well. Inspection
suggests that the emitting excited state must involve orbitals
that are a mixture from both contributing ligands.
2+
[Ru(φ -phen) ] complex in both shape and λmax. Indeed, the
2 3
In Figure 6A, the effect of the φ -phen ligand is a modest red
spectral shift is only a few nm.
2
shift of the emission. The overall shape of the emission
spectrum more closely resembles that of the tris-phen complex.
The results of Figure 6B are quite similar, but the red shift is
How can we interpret this? On the basis of a Boltzman
distribution within the possible excited states, we expect the
heteroleptic spectra to most closely resemble the lowest π* tris
complex at lower, not higher temperatures. Yet, the LEM now
even greater because of the presence of 2 φ -phen ligands.
2
3
362
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365

Inorganic Chemistry
Article
the vertical T-S gaps are 1.723 eV for the nonprotonated and
1.685 eV for the protonated complex. This corresponds to a ver-
tical emission around 720 nm for the unprotonated and 737 nm
for the protonated (Scheme 3). While, as expected, these
Scheme 3. Potential Energy Curves for Ground and First
Excited States
absolute emission energies are too low, the relative ordering
results do support our observations that protonation should
result in a red shift of tens of nanometers for the pH sensitive
complexes and that the emission from the 5-Cl phen complex
will have a smaller pH response.
We calculate the zero point energy differences between the
S and T states would yield emission at 640 nm for the
0
1
unprotonated RuN_1′ and 660 nm for the protonated species.
We acknowledge that the energy differences predicted in these
calculations are small and our methods are approximate.
However, the calculations do support (1) a red emission shift
on protonation, and (2) variation in pH sensitivity depending
on the accessory ligand.
Figure 7. Comparison of corrected, normalized room temperature
2
+
emission spectra in CH OH (a): [Ru(phen) ] , (1); [Ru(φ -
3
3
2
2
+
2+
phen) ] , (2); [Ru(phen) (φ -phen)] , (3). (b): [Ru(5-
Et Nphen) ] , (1); [Ru(φ -phen) ] , (2); [Ru(5-Et Nphen)(φ -
3
2
2
2+
2+
2
3
2
3
2
2
2
+
phen) ] , (3).
2
There is a rich literature, earlier work drawing largely from
2
6,30
appears to work as the emission strongly resembles that of the
tris complex with the lower π* ligand. The MOP model has
many unknown parameters, but it is possible that the degree of
mixing, in addition to dependence on the π* energy separation,
may change with environment and/or temperature. A clear
choice among the various possible models is not possible from
this limited data. However, it does seem that the situation for
the emitting state(s) of these complexes is more complex than
the “pure” LEM approach.
Seeking insight into the nature of the emitting state, we
carried out a series of calculations. First, we estimated the
energies of the π* levels for some of our accessory ligands as
well as a proxy for our pH sensitive ligand, again with the
B3LYP/SSD model (see Supporting Information). While the
calculations for the free ligands showed a significant difference
between the π* levels of the sensing and accessory ligands,
there was no obvious distinction between the pH sensitive and
pH muting accessory ligands. Thus, examining the π* levels of
the uncomplexed ligands was not particularly helpful.
static and Resonance Raman measurements
and more
25
recently fast kinetic studies, which provides insight into the
question of orbital mixing in Ru(II) heteroleptic complexes.
The salient results suggest the following: (1) The first reduc-
tion potential for the appropriate tris complex is a good proxy
for the energy of the ligand π* state. (2) When the energy
difference between the π* levels is large (i.e., >0.2 V), then the
localized excitation model works well and the emission
emanates from a virtually pure state involving the ligand with
the lowest π* energy. (3) As the π* levels of the various ligands
approach each other, mixing occurs and a MOP model is more
appropriate. Thus for our systems, if the lowest π* energy level
of the accessory ligand is similar to or below that of the sensing
ligand, the accessory ligand will not be “innocent” and may
have a significant impact on the observed environmental sensi-
tivity of the complex.
Table 3 shows the quantum yields, radiative and nonradia-
tive rate constants for our suite of complexes as well as compar-
ison values for the appropriate tris complexes. One chooses
accessory ligands to enhance some aspect of the reporter
complex, and using [Ru(5-Et Nphen)(phen) ](ClO ) as a
We next modeled the emission which we assumed originated
from the lowest triplet state; Eem ∼ E(T ) − E(S ′), where
1
0
2
2
4 2
E(S ′) is the energy on the ground state singlet surface at the
reference, we find that ring substitution by CH , φ, and Cl all
0
3
lowest-energy geometry of the triplet state. For the 5-chloro
complex these energies are virtually identical for both
unprotonated and protonated species, 1.733 and 1.730 eV,
respectively. However, for the pH responsive system, RuN_1′,
enhance the quantum yield, largely through an increase in kr.
These parameters give no clue that φ and Cl substitution will
essentially render the complex pH insensitive. However, the
first reduction potentials, a proxy for the π* energy levels of
3
363
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365

Inorganic Chemistry
Article
Table 3. Quantum Yields, Decay Rate Constants, and Reduction Potentials for RuL′L22+ and RuL32+ Complexesa
RuL′L 2+
φ
b
RuL′L2
b
kr s−1
b
knr s−1
φ RuL3
kr s−1
knr s−1
1st red. (V)
c
2
4
4
4
4
4
4
4
4
4
4
11.5 × 105
9.0 × 105
6.7 × 105
10.8 × 10
6.3 × 105
d
7.7 × 104
9.0 × 104
4.2 × 104
7.9 × 105
9.62 × 105
21.8 × 105
e
(
(
(
(
(
(
(
(
(
(
(
bpy)2
0.048
0.050
0.024
0.044
0.087
0.070
0.042
0.049
0.24
5.8 × 10
4.7 × 10
1.7 × 10
5.0 × 10
5.5 × 10
5.6 × 10
2.6 × 10
2.4 × 10
5.0 × 10
4.7 × 10
0.089
−1.35
−1.43
−1.35
−1.43
f
f
Me bpy)2
0.073
2
d
e
phen)2
0.019
5
5-Et Nphen)2
2
d
6.7 × 104
20.2 × 105
e
Me phen)2
0.032
<−1.52
4
5
d
4
5
e
Me phen)2
7.4 × 10
0.061
7.2 × 10
11.0 × 10
−1.52
2
5
5-Et Nphen)(CN)
6.0 × 10
2
2
4,7-Cl phen)2
4.7 × 105
0.052
3.0 × 104
5.48 × 105
0.99 × 105
−1.15
−1.22
−1.22
−1.25
2
5
d
4
g
φ2-phen)2
5-Clphen)2
φ2-bpy)2
1.6 × 10
0.37
5.7 × 10
d
5.2 × 104
1.3 × 105
5
e
0.053
0.16
8.4 × 105
5.5 × 105
0.035
14.4 × 10
5.0 × 105
4
f
f
10.3 × 10
0.20
a
b
c
d
L′ = 5-Et Nphen, L = accessory ligand. Values obtained in CH OH at 25 °C. Values obtained in CH CN at 25 °C. These values taken from
2
3
3
ref 3b in EtOH/CH OH (4:1) at 20 °C. They represent upper limits as φ for Ru(bpy) Cl is reported as 0.045 for CH OH and 0.069 for EtOH.
3
3
2
3
e
f
g
Ref 3b. Ref 25b. Ref 31.
the ligands, do help provide an explanation for our results.
Thus, when the first reduction potential of the accessory ligand
is more negative than that of the sensing ligand, the complex is
pH sensitive, excitation is more localized on the sensing ligand,
and the dynamic range is enhanced . Even when the potentials
are similar, (i.e., ΔE < 0.1 V), sufficient mixing or thermal
equilibration occurs such that the π* orbital of the sensing
ligand is well represented in the emitting state wave function,
conferring pH sensitivity. However, for the Cl and φ
substituted accessory ligands, the π* orbitals are significantly
below that of the sensing ligand, and the excitation is largely
localized on a pH insensitive accessory ligand.
ASSOCIATED CONTENT
Supporting Information
Further details are given in Figures S1−S4 and Tables S1−S3.
This material is available free of charge via the Internet at
http://pubs.acs.org.
■
*
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: degrafba@jmu.edu.
Notes
■
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
CONCLUSIONS
It is a pleasure to acknowledge the assistance of the North
Carolina State University Chemistry Department’s Mass
Spectroscopy Facility in obtaining the ES/MS data for our
complexes. E.D., M.S., J.B., A.M.W., and B.A.D. wish to
acknowledge the Dreyfus Foundation through its Senior
Scientist Mentoring Grants Program for valuable support of
this research. J.N.D. and B.A.D. wish to acknowledge the
National Science Foundation (Grant CHE-04-10061) for
partial support of this work. C.T. is grateful to the Body
Foundation for computer equipment.
■
Most environmentally sensitive luminescent complexes rely on
a modified ligand to perform the analyte recognition and
register this event by changes in the luminescence. Many of
these complexes also contain additional accessory ligands
which can be used to enhance various aspects of the lumine-
scence (i.e., improved quantum yield, color shifts, or lifetime
enhancement). The ability to tune the response of Ru(II),
Os(II), and Ir(III) complexes through judicious choice of
accessory ligands is one the main reason for selecting this type
of complex for sensor applications. This work again confirms
REFERENCES
■
that significant changes in both dynamic range and pK * can be
(1) Extensive reviews include: (a) Wolfbeis, O. S. Anal. Chem. 2008,
80 (12), 4269−4283. (b) Lowery, M.; Fakayode, S. O.; Geng, M. L.;
Baker, G. A.; Wang, L.; McCarrol, M. E.; Patonay, G.; Warner,
I. M. Anal. Chem. 2008, 80 (12), 4551−4574. (c) De Silva, A. P.;
Gunaratne, H. Q. N.; Gunnlaugsson, T.; Huxley, A. J. M.; McCoy, C. P.;
Rademacher, J. T.; Rice, T. E. Chem. Rev. 1997, 97, 1515−1566.
2) DeGraff, B. A.; Demas, J. N. Coord. Chem. Rev. 2001, 211 (1),
317−335.
3) (a) Kalyanasundaram, K. Photochemistry of Polypyridine and
a
obtained by judicious choice of accessory ligand. Use of
accessory ligands to enhance environmental response is a
widely applicable strategy regardless of the target analyte.
However, there is a caveat. If the first reduction potential of the
accessory ligand, a proxy for the energy of the π* level, is well
below that of the sensing ligand, the response of the complex
to the environmental change or analyte may be significantly
muted. The simple localized excitation model is a good starting
point and largely true for ligands of rather different π* energies.
However, as the π* levels approach each other, mixing clearly
occurs and a more complex model is required. Theory suggests,
and our observations support, that ultimately it is the energy
difference between the π* levels of the protonated and non-
protonated ligands in comparison with the π* level of the
accessory ligand, coupled with the energy gap law that deter-
mines whether the system is or is not environmentally (i.e., pH)
responsive.
(
(
Porphyrin Complexes; Academic Press: San Diego, CA, 1992. (b) Juris,
A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.; Von Zelewsky,
A. Coord. Chem. Rev. 1988, 84, 85−277.
(
4) (a) Lowe, M. P.; Mathieu, C. E.; Parker, D.; Senanayake, P. K.;
Kataky, R. Inorg. Chem. 2001, 40, 5860−5867. (b) Ellerbrock, J. C.;
McLoughlin, S. M.; Baba, A. I. Inorg. Chem. Commun. 2002, 5, 555−
5
59. (c) Gunnlaugsson, T.; Leonard, J. P.; Senechal, K.; Harte, A. J.
J. Am. Chem. Soc. 2003, 125, 12062−12063. (d) Vicente, M.; Bastida,
R.; Lodeiro, C.; Macias, A.; Parola, A. J.; Valencia, L.; Spey, S. E. Inorg.
Chem. 2003, 42, 6768−6779. (e) Wong, K. M-C.; Tang, W.-S.; Lu,
X.-X.; Zhu, N.; Yam, V. W-W. Inorg. Chem. 2005, 44, 1492−1498.
(f) Ressalan, S.; Iyer, C. S. P. J. Lumin. 2005, 111, 121−129.
3
364
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365

Inorganic Chemistry
Article
(
g) Tzeng, B.-C.; Chen, B.-S.; Chen, C.-K.; Chang, Y.-P.; Tzeng,
(c) Ferguson, J.; Krausz, E. R.; Maeder, M. J. Phys. Chem. 1985, 89,
1852−1854. (d) Kitamura, N.; Kim, H.-B.; Kawanishi, Y.; Obata, R.;
Tazuke, S. J. Phys. Chem. 1986, 90, 1488−1491. (e) DeArmond, M. K.;
Myrick, M. L. Acc. Chem. Res. 1989, 33, 364−370. (f) Yabe, T.; Orman,
L. K.; Anderson, D. R.; Yu, S.-C.; Xu, X.; Hopkins, J. B. J. Phys. Chem.
1990, 94, 7128−7132.
W.-C.; Lin, T.-Y.; Lee, G.-H.; Chou, P.-T.; Fus, Y.-J.; Chang, A. H-H.
Inorg. Chem. 2011, 50, 5379−5388.
(
5) Higgins, B.; DeGraff, B. A.; Demas, J. N. Inorg. Chem. 2005,
4
4 (19), 6662−6669.
(
6) In contrast, the impact of accessory ligands on the emission of
cyclometalated Ir(III) complexes has received considerable study. See
for example: Li, J.; Djurovich, P. I.; Alleyne, B. D.; Yousfuddin, M.;
Ho, N. N.; Thomas, J. C.; Peters, J. C.; Bau, R.; Thompson, M. E.
Inorg. Chem. 2005, 44, 1713−1727 , and references therein.
(27) (a) Smalley, S. J.; Waterland, M. R.; Telfer, S. G. Inorg. Chem.
2009, 48, 13−15. (b) Juris, A.; Campagna, S.; Balzani, V.; Germaud,
G.; von Zelewsky, A. Inorg. Chem. 1988, 27, 3652−3655. (c) Juris, A.;
Barigelletti, F.; Balzani, V.; Belser, P.; von Zelewsky, A. Inorg. Chem.
1985, 24, 202−206. (d) Barigelletti, R.; Juris, A.; Balzani, V.; Belser, P.;
von Zelewsky, A. Inorg. Chem. 1983, 22, 3335−3339.
(
7) Caspar, J. V.; Meyer, T. J. J. Am. Chem. Soc. 1983, 105, 5583−
590.
8) Sacksteder, L.; Demas, J. N.; DeGraff, B. A. Inorg. Chem. 1989,
8, 1787−92.
9) (a) Sullivan, B. P.; Salmon, D. J.; Meyer, T. J. Inorg. Chem. 1978,
7, 3334−3341. (b) Giordano, P. J.; Bock, C. R.; Wrighton, M. S.
5
(
2
(
(28) (a) Chen, Y.-J.; Xie, P.; Heeg, M. J.; Endicott, J. F. Inorg. Chem.
2006, 45, 6282−6297. (b) Kumar, C. V.; Barton, J. K.; Gould, I. R.;
Turro, N. J.; Van Houten, J. Inorg. Chem. 1988, 27, 648−651.
(29) Dreuw, A.; Head-Gordon, M. J. Am. Chem. Soc. 2004, 126,
4007−4016.
1
J. Am. Chem. Soc. 1978, 100, 6960−6965. (c) Baggot, J. E.; Gregory, G.
K.; Pilling, M. J.; Anderson, S.; Seddon, K. R.; Turp, J. E. J. Chem. Soc.,
Faraday Trans. 2 1983, 79, 195−210. (d) Evans, I. P.; Spencer, A.;
Wilkinson, G. J. Chem. Soc., Dalton Trans. 1973, 204−209.
(30) (a) Elliot, C. M.; Hershenhart, E. J. J. Am. Chem. Soc. 1982, 104,
7
519−7526. (b) Rillema, D. P.; Allen, G.; Meyer, T. J.; Conrad, D.
Inorg. Chem. 1983, 22, 1617−1622. (c) Mabrouk, P.; Wrighton, M. S.
Inorg. Chem. 1986, 25, 526−531. (d) Wacholtz, W. F.; Auerbach, R. A.;
Schmehl, R. H. Inorg. Chem. 1986, 25, 227−234.
(
e) Zakeeruddin, S. M.; Nazeeruddin, Md. K.; Humphry-Baker, R.;
Gratzel, M.; Shklover, V. Inorg. Chem. 1998, 37, 5251−5259.
10) (a) Kishnan, S.; Kuhn, D. J.; Hamilton, G. A. J. Am. Chem. Soc.
(31) Skarda, V.; Cook, M. J.; Lewis, A. P.; McAuliffe, G. S. G.;
(
Thompson, A. J.; Robbins, D. J. J. Chem. Soc., Perkin Trans. II 1984,
1
977, 99, 8121−8125. (b) Shen, Y.; Sullivan, B. P. Inorg. Chem. 1995,
1
309−1311.
3
4, 6235−6236. (c) Riklin, M.; Tran, D.; Bu, X.; Laverman, L. E.;
Ford, P. C. J. Chem. Soc., Dalton Trans. 2001, 1813−1819.
(
11) Snyder, H. R.; Freier, H. E. J. Am. Chem. Soc. 1946, 68, 1320−
322.
12) Freedman, D. A.; Evju, J. K.; Pomije, M. K.; Mann, K. R. Inorg.
Chem. 2001, 40, 5711−5715.
13) Kalyanasundram, K. Photochemistry of Polypyridine and Porphyrin
Complexes; Academic Press: San Diego, CA, 1992; Chapter 6.
14) (a) Vos, J. G. Polyhedron 1992, 11 (18), 2285−2299. (b) Wang,
R.; Vos, J. G.; Schmehl, R. H.; Hage, R. J. Am. Chem. Soc. 1992, 114,
964−1970.
15) Frish, M. J., et al. Gaussian 09, Revision A.02; Gaussian, Inc:
Wallingford, CT, 2009.
16) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785−89.
b) Becke, A. D. J. Chem. Phys. 1992, 96, 2155−2160. (c) Becke, A. D.
1
(
(
(
1
(
(
(
J. Chem. Phys. 1992, 92, 9173−9177. (d) Becke, A. D. J. Chem. Phys.
1
993, 98, 1372−1377. (e) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−
5652.
(
17) Andrae, D.; Haeusserman, U.; Dolg, M.; Preuss, H. Theor. Chem.
Acc. 1990, 77, 123−141.
18) Tomasi, J.; Menucci, B.; Cammi, R. Chem. Rev. 2005, 105,
(
2
(
(
999−3093.
19) Furche, F.; Ahlrichs, R. J. Chem. Phys. 2002, 117, 7433−7447.
20) Murov, S. L. Handbook of Photochemistry; Marcel Dekker:
New York, N.Y., 1973; p 89.
21) Clarke, Y.; Xu, W.; Demas, J. N.; DeGraff, B. A. Anal. Chem.
000, 72, 3469−3475.
22) (a) Sun, H.; Hoffman, M. Z. J. Phys. Chem. 1993, 97, 5014−
018. (b) Casalboni, F.; Mulazzani, Q. G.; Clark, C. D.; Hoffman,
M. Z.; Orizondo, P. L.; Perkovic, M. W.; Rillema, D. P. Inorg. Chem.
997, 36, 2252−2257.
23) Ireland, J. F.; Wyatt, P. A. H. Adv. Phys. Org. Chem. 1976, 12,
31−153.
24) Handbook of Chemistry and Physics, 71st ed; Lide, D. R., Ed.;
CRC Press: Boca Raton, FL, 1990; pp 8−33.
25) (a) Damrauer, N. H.; Cerullo, G.; Yeh, A.; Boussie, T. R.;
(
2
(
5
1
(
1
(
(
Shank, C. V.; McCusker, J. K. Science 1997, 275, 54−61. (b) Damrauer,
N. H.; Boussie, T. R.; Devenney, M.; McCusker, J. K. J. Am. Chem. Soc.
1
997, 119, 8253−8268. (c) Damrauer, N. H.; McCusker, J. K. J. Phys.
Chem. A 1999, 103, 8440−8446. (d) Curtright, A. E.; McCusker, J. K.
J. Phys. Chem. A 1999, 103, 7032−7041. (e) Monat, J. E.; McCusker,
J. K. J. Am. Chem. Soc. 2000, 122, 4092−4097.
(
26) (a) Bradley, P. G.; Kress, N.; Hornberger, B. A.; Dallinger, R. F.;
Woodruff, W. H. J. Am. Chem. Soc. 1981, 103, 7441−7446. (b) Kober,
E. M.; Sullivan, B. P.; Meyer, T. J. Inorg. Chem. 1984, 23, 2098−2104.
3
365
dx.doi.org/10.1021/ic201114u | Inorg. Chem. 2012, 51, 3355−3365