Oxygen optical sensing by doping a rhenium complex into a silica matrix: Design, characterization and performance

Article abstract of DOI:10.1016/j.ica.2015.09.024

The present paper synthesized a diamine ligand having an oxadiazole group and its Re(I) complex to investigate their potential for oxygen sensing. Conjugation chain in this ligand was increased so that collision probability between excited state electrons

Full text of DOI:10.1016/j.ica.2015.09.024

Inorganica Chimica Acta 438 (2015) 212–219
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Oxygen optical sensing by doping a rhenium complex into a silica
matrix: Design, characterization and performance
Liu Daofu
School of Chemical and Materials Engineering, Huainan Normal University, Huainan 232001, Anhui, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 12 August 2015
Received in revised form 16 September 2015
Accepted 22 September 2015
Available online 9 October 2015
The present paper synthesized a diamine ligand having an oxadiazole group and its Re(I) complex to
investigate their potential for oxygen sensing. Conjugation chain in this ligand was increased so that col-
lision probability between excited state electrons and O2 molecules could be increased, favoring oxygen
sensing. This Re(I) complex was confirmed and analyzed by single crystal structure, density functional
theory calculation and photophysical measurement. It showed an emission peaking at 552 nm with
two long-lived emissive components. Using a silica matrix of MCM-41, oxygen sensing performance of
the resulting composite samples was analyzed and compared in detail. An optimal sample gave the
highest sensitivity of 5.65 with short response time of 8 s.
Keywords:
Re(I) complex
Oxygen-sensing
Photophysical features
MCM-41
Ó 2015 Elsevier B.V. All rights reserved.
Theoretical analysis
1. Introduction
especially Re(I)-based ones, have been proved attractive [13–16].
Theoretical analysis on them suggests that their occupied frontier
Composite materials combine and hold features of their each
component, and thus are gaining more research interests [1,2].
For a representative sample with organic–inorganic hybrid struc-
ture, its organic component usually focuses on sensing function
since organic probes can be conveniently modified, while its inor-
ganic component is usually designed as supporting matrix in virtue
of its high mechanical strength and stability [2,3]. Some hybrid
structures have been suggested so that each component, as well
as its feature, can be well preserved [3,4].
Molecular O2 is a life-supporting gas and thus considered as an
important target in chemical industry, food processing and envi-
ronmental protection [5,6]. Optical sensing systems for O2 quantifi-
cation have thus been intensively focused owing to their virtues of
no analyte consumption, simple operation procedure and needing
no sophisticated apparatuses [7,8]. Aiming at desired sensing per-
formance, some criteria should be considered. For example, sup-
porting matrix is responsible for dispersing probe molecules and
allowing smooth analyte diffusion. Thus, an ideal one should has
high diffusion coefficient, good photostability and compatibility
with probe [9–12]. A silica based supporting matrix MCM-41 has
been usually suggested among the numerous candidates [11,12].
For an ideal oxygen sensing probe, long lifetime and large distri-
bution of excited electrons are desired since they increase sensing
collision probability with analyte. Emissive metal complexes,
molecular orbitals (FMOs) have metal character, while their unoc-
cupied FMOs are essentially ligand p⁄ [13–16]. Large distribution of
excited electrons can be achieved by increasing ligand conjugation
plane. What’s more, electron-pulling group in conjugation plane
may decrease non-radiative decay probability, increasing emission
lifetime.
Enlightened by above consideration, we intent to design a dia-
mine ligand with a large conjugation plane and an electron-pulling
group, as shown by Scheme 1. Its corresponding Re(I) complex is
anticipated to be a promising probe for oxygen sensing. By doping
this Re(I) complex into a silica matrix MCM-41, oxygen sensing
composites are constructed. Corresponding sensing performance,
along with sensing mechanism, is investigated.
2. Experimental
2.1. General information for reagents and apparatus
General reagents for ligand and complex synthesis are summa-
rized as follows. Starting chemical 2-(2H-tetrazol-5-yl)pyridine
(TPD) was synthesized following a reported method [17]. Other
compounds, such as Re(CO)5Br, 4-methylbenzoyl chloride, zinc
bromide, sodium azide and MCM-41, were commercially provided
by Yiquan Chemical Co. (Hangzhou) and used without further
purifications. Organic solvents were purified through standard
procedures before usage.
E-mail address: liudf363@163.com
http://dx.doi.org/10.1016/j.ica.2015.09.024
0020-1693/Ó 2015 Elsevier B.V. All rights reserved.

L. Daofu / Inorganica Chimica Acta 438 (2015) 212–219
213
Scheme 1. Synthesis and construction method for ligand POD, Re(I) complex Re(CO)3(POD)Br and Re(CO)3(POD)Br doped MCM-41.
Elemental analysis was taken on a Vario Element Analyzer.
NMR, MS, UV–Vis absorption and emission spectra were measured
using a Varian INOVA 300 spectrometer, a Agilent 1100 MS series/
AXIMA CFR MALDI/TOF MS spectrometer, a Shimadzu UV-3101PC
spectrophotometer and a Hitachi F-4500 fluorescence spectropho-
tometer, respectively. Single crystal data were collected on a
Siemens P4 single-crystal X-ray diffractometer with a Smart
(1H, d, J = 4.0). Anal. Calc. for C17H11BrN3O4Re: C, 34.76, H, 1.89,
N, 7.15. Found: C, 34.67, H, 1.69, N, 7.31%. MS m/z: [m]+ calc. for
17H11BrN3O4Re, 586.9; found, 587.0. Its single crystal (CCDC
C
1096376) XRD analysis will be discussed in detail later.
2.4. Construction of composite Re(CO)3(POD)Br doped MCM-41
CCD-1000 detector, using graphite-monochromated Mo K
a radia-
Sensing composites were constructed by doping Re(CO)3(POD)
Br into MCM-41 matrix with various doping concentrations,
following below procedure [11,12]. Different amounts of Re(CO)3
(POD)Br were weighed and separately dissolved into CH2Cl2
(5 mL) solutions under stirring. MCM-41 matrix (1 g per portion)
was added and stirred at room temperature for 30 min. Solid sam-
ple was then filtered off and washed with CH2Cl2 (10 mL ꢀ 3) to
give sensing composite as light yellow powder.
tion (50 kV, 30 A, 298 K). All hydrogen atoms were calculated.
Emission lifetimes were recorded by a two-channel TEKTRONIX
TDS-3052 oscilloscope, with pulsed Nd:YAG laser as excitation
source (k = 355 nm). Small-angle X-ray diffraction (SAXRD) was
finished by a Rigaku-Dmax 2500 diffractometer (k = 0.154 nm,
scanning step = 0.02°). Theoretical calculation was carried out
using GAMESS software at RB3LYP1/SBKJC level with Re(CO)3
(POD)Br single crystal as initial geometry. Graphical presentation
for FMOs was plotted by wxMacMolPlt with contour value of
0.025.
3. Results and discussion
2.2. Synthesis of diamine ligand POD
3.1. Single crystal structure of Re(CO)3(POD)Br
Diamine ligand POD was synthesized following a reported pro-
cedure [17]. The mixture of TPD (10 mmol), 4-methylbenzoyl chlo-
ride (11 mmol) and anhydrous pyridine was stirred under ice bath
for 30 min. Then this solution was heated to reflux and kept for
3 days. After cooling, plenty of water was added. The resulting
crude product was collected and purified on a silica gel column
(n-hexane: ethyl acetate = 50:1). 1H NMR (CDCl3): d 2.42 (3H, s),
7.28 (2H, d, J = 6.0), 7.44 (1H, t), 7.90 (1H, t), 8.12 (2H, d, J = 6.0),
8.33 (1H, d, J = 6.0), 8.80 (1H, d, J = 3.6). Anal. Calc. For
Re(CO)3(POD)Br is firstly investigated by its single crystal struc-
ture. As shown in Fig. 1A, Re(CO)3(POD)Br takes an octahedral
coordination mode, which is consistent with literatures [13–15].
Selected geometric parameters listed in Table 1 suggest that this
coordination field is a distorted one owing to the heterogeneous
ligands in Re(CO)3(POD)Br. It is noticed that the three Re–C bonds
are slightly different from each other even though they have the
same CO ligand. The two Re–N are slightly different from each
other. Most bite angles in Table 1 are close to 90°, which are similar
to literature values [13–15]. N–Re–N bite angle, however, is smal-
ler than those in tetrahedral field [13–16]. These ligands try to
achieve an optimal coordination with Re(I) center by adjusting
their geometry. There is roomy space in Re(CO)3(POD)Br coordina-
tion sphere, favoring oxygen attack.
C
14H11N3O: C, 70.87; H, 4.67; N, 17.71. Found: C, 70.93; H, 4.77;
N, 17.54%. MS m/z: [m]+ calc. for C14H11N3O, 237.1; found, 237.0.
2.3. Synthesis of complex Re(CO)3(POD)Br
Complex Re(CO)3(POD)Br was synthesized following a reported
procedure [15]. The mixture of POD (5 mmol), Re(CO)5Br (5 mmol)
and toluene (25 mL) was heated to reflux under N2 protection and
kept for 12 h. Solvent toluene was extracted by rotary evaporation.
The resulting crude product was purified on a silica gel column. 1H
NMR (300 MHz, CDCl3): d 2.56 (3H, s), 7.62 (1H, m), 7.73 (2H, m),
7.90 (1H, m), 8.31 (1H, t), 8.39 (1H, t), 8.42 (1H, d, J = 6.0), 8.95
A large conjugation plane is formed by a pyridine ring, an oxa-
diazole ring and a phenyl ring in POD ligand.
P-p attraction
between POD planes makes Re(CO)3(POD)Br molecules align in a
highly ordered mode, as shown in Fig. 1B. POD planes of these
molecules are nearly parallel to each other with minimal distance
of 3.338 Å, which confirms the face-to-face
them. Zhang and coworkers have confirmed that such
p
–
p
attraction between
-stacking is
p

214
L. Daofu / Inorganica Chimica Acta 438 (2015) 212–219
Fig. 1A. Molecular structure of Re(CO)3(POD)Br obtained from its single crystal.
the electronic nature of Re(CO)3(POD)Br [11,12]. Density functional
theory calculation is used here to explore its electronic transition
[13–16]. Tables 2 and 3 give percentage composition of the first
ten singlet excitations and their FMOs. Clearly, these onset elec-
tronic excitations are composed of transitions from occupied FMOs
to unoccupied ones. These occupied FMOs are Re d orbitals in nat-
ure, admixed with contributions from other ligands. On the other
hand, unoccupied FMOs are mainly composed of POD p⁄, with slim
contributions from other ligands and Re center. Electronic transi-
tions between occupied and unoccupied FMOs are thus assigned
as a mixed nature of metal-to-ligand-charge-transfer (MLCT) and
ligand-to-ligand-charge-transfer (LLCT), as shown in Fig. 2 [13].
In its first excited state, electrons are localized at POD p⁄ with life-
time as long as a few microseconds owing to heavy metal turbu-
lence effect. These long-lived excited electrons are vulnerable to
O2 attack, giving sensing signal.
3.3. Photophysical evaluation on probe Re(CO)3(POD)Br
3.3.1. Absorption and excitation spectra
Fig. 1B. Crystal packing mode of Re(CO)3(POD)Br.
Fig. 3 gives Re(CO)3(POD)Br UV–Vis absorption and excitation
spectra in CH2Cl2 (10 lM). Four absorption bands are observed,
peaking at 237 nm, 290 nm, 329 nm and ꢁ390 nm, respectively.
Table 1
The former three bands are similar to POD absorption shown in
Selected geometric parameters of Re(CO)3(POD)Br single crystal.
Fig. 3. They are thus attributed to ligand
p ?
p⁄ (ILCT) transition.
Bond length
Å
Bond angle
°
The last absorption band is a weak and broad one. No similar band
is observed in POD absorption. Bearing above theoretical analysis
result in mind, this new absorption band is assigned to MLCT/LLCT
absorption [13–15]. Conjugation chain in POD is much longer than
that in Phen, and Re(CO)3(POD)Br optical edge is supposed to be
moved to a long wavelength. Optical edge of Re(CO)3(POD)Br is
determined as ꢁ475 nm which is similar to that of Re(CO)3(Phen)
Re(01)–C(1)
Re(01)–C(2)
Re(01)–C(3)
Re(01)–Br(25)
Re(01)–N(11)
Re(01)–N(12)
1.890
1.911
1.903
2.615
2.239
2.169
N(11)–Re(01)–N(12)
N(11)–Re(01)–C(3)
N(12)–Re(01)–C(3)
N(11)–Re(01)–Br(25)
N(12)–Re(01)–Br(25)
C(1)–Re(01)–Br(25)
C(2)–Re(01)–Br(25)
C(1)–Re(01)–C(2)
73.20
91.65
95.78
86.25
84.69
92.81
89.82
88.54
Table 2
Percentage composition of the first ten singlet excitations calculated at RB3LYP1/
SBKJC level.
a
rigid structure which decreases structural distortion that
happens in excited state, showing emission blue shift, long lifetime
and high emission yield [16,18,19]. For sensing application,
Transitions
Energy (eV)
Composition
however, such
p-stacking should be eliminated since this highly
S0 ? S1
S0 ? S2
S0 ? S3
S0 ? S4
S0 ? S5
S0 ? S6
S0 ? S7
S0 ? S8
S0 ? S9
S0 ? S10
1.7748
2.0040
2.6563
2.7338
2.8541
3.0450
3.1814
3.2799
3.3152
3.6321
70–71 (98.6%)
69–71 (98.0%)
68–71 (98.6%)
70–72 (96.2%)
69–72 (97.6%)
67–71 (86.0%)/66–71 (11.6%)
70–73 (93.5%)
66–71 (56.4%)/69–73 (29.2%)/70–73 (3.9%)
69–73 (68.8%)/66–71 (24.6%)
68–72 (97.7%)
ordered array may block O2 diffusion and thus compromise sensing
procedure.
3.2. Electronic transitions of Re(CO)3(POD)Br
Considering that oxygen sensing procedure is finished through
an energy transfer collision between probe excited state and O2
molecules, it would be necessary to get a clear understanding on

L. Daofu / Inorganica Chimica Acta 438 (2015) 212–219
215
Table 3
Percentage composition of selected frontier molecular orbitals calculated at RB3LYP1/
SBKJC level.
MOs
Energy (eV)
Contribution
Re Br
0.6
CO
1.1
1.6
3.9
18.2
14.2
27.6
14.3
16.4
POD
73
72
ꢂ1.845
ꢂ2.253
ꢂ3.099
ꢂ5.627
ꢂ5.747
ꢂ6.566
ꢂ6.876
ꢂ6.993
0.3
0.4
98.0
96.7
88.8
4.4
1.3
5.3
71 (LUMO)
70 (HOMO)
69
68
67
66
2.0
35.5
29.0
61.4
33.3
36.8
41.8
48.0
0.7
32.0
34.0
8.9
10.2
20.4
12.0
Br, where Phen = 1,10-phenanthroline. We attribute the restricted
optical edge of Re(CO)3(POD)Br to its electron-pulling group of
oxadiazole [16].
Despite of their high absorption coefficients, the former three
absorption bands are ineffective on exciting Re(CO)3(POD)Br emis-
sive center, as shown in Fig. 3. The last weak absorption band,
however, is an effective excitation. This fact can be explained as
follows. As above mentioned, the former three absorption bands
Fig. 3. UV–Vis absorption (abs.) and excitation (ex.) spectra of Re(CO)3(POD)Br and
POD in CH2Cl2 (10 lM).
are ligand
are much higher than that of MLCT transition [19]. These ligand
p⁄ transitions, however, have to experience a series of system
p
?
p⁄ transitions whose molar distinction coefficients
3.3.2. Emission spectra and lifetime
Re(CO)3(POD)Br emission spectrum in CH2Cl2 (10 lM) has a
p
?
maximum at 552 nm with FWHM of 76 nm, as shown in Fig. 4,
where FWHM means full width at half maximum. There are no
vibronic progressions. Stokes shift between absorption edge
(475 nm) and emission maximum (552 nm) is 77 nm. These factors
suggest that Re(CO)3(POD)Br emissive center is a charge-transfer-
derived one, which is consistent with our theoretical calculation
crossing procedures before transferring their energy to the emis-
sive center, leading to their poor excitation efficiency. On the other
hand, MLCT state can directly transfer its energy to the emissive
center since the emissive center is derived from this MLCT state,
showing a high coefficient.
Fig. 2. FMOs of Re(CO)3(POD)Br (up, HOMO; down, LUMO).

216
L. Daofu / Inorganica Chimica Acta 438 (2015) 212–219
spectrum, as shown in Fig. 4. This 3LC level is much higher than
3MLCT state, which denies the possibility of electronic configura-
tion transition between 3MLCT state to 3LC state. As a consequence,
geometric relaxation of MLCT excited state should be claimed as
the major reason leading to Re(CO)3(POD)Br’s unsatisfactory
emission yield. To improve emission yield, complex geometric
relaxation should be limited through ligands having large steric
hindrance. Our first guess is to explore multidentate ligands so that
geometric relaxation can be restrained without compromising O2
diffusion.
Kr
Kr þ Knr
U
¼
ð1Þ
ð2Þ
ð3Þ
1
s
¼ Kr þ Knr
A1s2 þ A2s2
A1s11 þ A2s22
s
¼
Fig. 4. Emission spectrum of Re(CO)3(POD)Br in CH2Cl2 (10 lM) and POD at 77 K
(kex = 355 nm). Inset: emission decay dynamics of Re(CO)3(POD)Br in CH2Cl2
(10 M).
l
3.4. Morphology and characterization of Re(CO)3(POD)Br doped MCM-
41
result [13]. Emission maximum and FWHM of a reference complex
Re(CO)3(Phen)Br have been reported as 554 nm and 90 nm by Si
and coworkers, respectively [15]. It is observed that the structural
relaxation that happens in Re(CO)3(POD)Br excited state is limited.
We attribute its causation to the roomy coordination sphere in Re
(CO)3(POD)Br and the electron-pulling group in POD ligand. To be
more specific, for a crowded environment, the strong steric
hindrance between ligands leads to distortion and tension. Many
vibration levels are thus generated, which facilitates non-radiative
decay. In a roomy coordination sphere, ligands are freely extended,
their vibration levels are thus decreased, favoring radiative decay
of emissive center.
Aiming at fluent and smooth analyte diffusion, probe is usually
immobilized in supporting matrix. Meanwhile, self-quenching
between probe molecules can be minimized [11,12]. MCM-41
has been proved as a promising supporting matrix for oxygen sens-
ing purpose owing to its highly ordered tunnels [11,12]. In this ini-
tial effort, we choose MCM-41 as supporting matrix to explore
oxygen sensing performance of Re(CO)3(POD)Br. Three doping con-
centrations, 50 mg/g, 60 mg/g and 70 mg/g, are tried for perfor-
mance comparison. These Re(CO)3(POD)Br doped MCM-41
samples are firstly analyzed by their small angle X-ray diffraction
(SAXRD) patterns shown in Fig. 5. There are three well-resolved
Bragg reflection peaks in each curve, indexed as d100, d110, and
d200, respectively. Despite of their different intensity values, their
2h° are nearly identical to corresponding ones of blank MCM-41
sample [11,12]. We thus tentatively come to a conclusion that
the highly ordered tunnels in MCM-41 supporting matrix have
been perfectly preserved after probe loading procedure.
Re(CO)3(POD)Br excited state is analyzed by its emission decay
dynamics, as shown by the inset of Fig. 4. Clearly, this emission fol-
lows a biexponential decay pattern with
s1 = 0.409 ls (A1 = 0.293)
and 2 = 0.0797 s (A2 = 1.974), respectively. These two compo-
s
l
nents are consistent with the MLCT/LLCT transition nature of Re
(CO)3(POD)Br. Phosphorescent nature of Re(CO)3(POD)Br emission
is thus confirmed by these long-lived decay components. Accord-
ing to the statement by Si and coworkers, the observation of an
To confirm that Re(CO)3(POD)Br has been successfully doped
into MCM-41, absorption and solid state diffuse reflection spectra
of blank MCM-41 and a representative sample (60 mg/g doped)
are shown in Fig. 6. Blank MCM-41 is essentially silica and thus
obvious absorption in
ponent means a potential surface crossing procedure from
p ?
p⁄ region and a short-lived decay com-
p
?
p⁄
state to MLCT state [15]. The long-lived component s1 is thus rec-
ognized as radiative decay of MLCT state, while the short-lived one
is attributed to radiative decay of
p ?
p⁄ state. Such assignment
agrees with electronic nature of each component.
Emission quantum yield of Re(CO)3(POD)Br in CH2Cl2 (10 lM) is
determined by a literature method and obtained as 0.030 [19].
Compared to those of superior emitters, this emission yield is not
satisfactory enough [13–16]. For a better understanding on our
probe, its radiative probability (Kr) and non-radiative probability
(Knr) are calculated as 0.135 ꢀ 106 sꢂ1 and 4.370 ꢀ 106 sꢂ1, respec-
tively, by Formulas (1)–(3). It is observed that Re(CO)3(POD)Br
emissive state is dominated by its non-radiative decay path, lead-
ing to its poor emission yield. Geometric relaxation that happens in
excited state has been proved as a major non-radiative decay path
for MLCT excited state. There is, however, another possible path to
be considered, namely the electronic configuration transition
between 3MLCT state to 3LC (ligand centered) state. As reported
by Zhang and coworkers, a large conjugation plane in ligand can
lead to such electronic configuration transition, even though there
is an electron-pulling group in it [16]. POD 3LC energy is
determined as 394 nm by its low temperature phosphorescence
Fig. 5. SAXRD curves of Re(CO)3(POD)Br doped MCM-41 (50 mg/g, 60 mg/g and
70 mg/g) and blank MCM-41.

L. Daofu / Inorganica Chimica Acta 438 (2015) 212–219
217
Fig. 7A. Emission spectra of Re(CO)3(POD)Br doped MCM-41 sample (50 mg/g)
towards various O2 concentrations from 0% to 100% (interval = 10%).
Fig. 6. Absorption (abs.) and solid state diffuse reflection (Ref.) spectra of blank
MCM-41 and a representative sample (60 mg/g doped).
has slim absorption in the whole UV–Vis region. Correspondingly,
its solid state reflection spectrum shows strong reflection from
200 nm to 600 nm. Absorption spectrum of 60 mg/g doped sample
is consistent with that of Re(CO)3(POP)Br in CH2Cl2, showing
absorption bands peaking at 238 nm, 276 nm and 366 nm, respec-
tively. It is thus confirmed that Re(CO)3(POP)Br has been success-
fully loaded into MCM-41 matrix. Correspondingly, its solid state
reflection spectrum has two minima peaking at 263 nm and
360 nm, respectively, which is consistent with its absorption
bands. There are no new absorption or reflection bands, suggesting
that dopant molecules are simply immobilized in MCM-41 tunnels
with no strong interaction with them. We thus come to a conclu-
sion that Re(CO)3(POD)Br molecules are simply immobilized in
MCM-41 tunnels with no strong interaction with them. The inter-
action between dopant and matrix exerts slim effect on sensing
performance. In other words, sensing performance is dominated
by dopant itself, which is consistent with literature reports
[11,12,17,18].
Fig. 7B. Emission spectra of Re(CO)3(POD)Br doped MCM-41 sample (60 mg/g)
towards various O2 concentrations from 0% to 100% (interval = 10%).
3.5. Sensing performance of Re(CO)3(POD)Br doped MCM-41
3.5.1. Sensitivity
Emission spectra of Re(CO)3(POD)Br doped MCM-41 samples
towards various O2 concentrations are recorded to evaluate their
oxygen sensing performance. As shown in Fig. 7, the presence of
O2 quenches sample emission effectively, giving sensing signals.
For comparison, maximum sensitivity of each sample is defined
as its ratio of I0/I100, where I0 and I100 are emission intensity values
in pure N2 atmosphere and in pure O2 atmosphere, respectively.
Sensitivity values and other key sensing parameters of Re(CO)3
(POD)Br doped MCM-41 samples are listed in Table 4. Our sensitiv-
ity values are found comparable to or even higher than literature
values [11,12,13–18]. For similar sensing systems with Re(CO)3
(IPD-Et)Br
throlin-2-yl)-N,N-diphenylaniline] as dopant, their sensitivity is
3.99 with a long excited state lifetime of 4.68 s [20]. Sensitivity
of our samples is found higher (5.65) even though our excited state
lifetime is only 0.222 s. The positive effect of large conjugation
[IPD-Et = 4-(1-ethyl-1H-imidazo[4,5-f][1,10]phenan-
l
l
chain on sensing performance can be confirmed. As a consequence,
the following factors should be responsible for our improve sens-
ing performance. First, collision probability between probe and
O2 is increased by the large conjugation plane in POD and the
long-lived phosphorescent emissive center. Second, there is roomy
space in Re(CO)3(POD)Br coordination sphere, which favors oxygen
Fig. 7C. Emission spectra of Re(CO)3(POD)Br doped MCM-41 sample (70 mg/g)
towards various O2 concentrations from 0% to 100% (interval = 10%).

218
L. Daofu / Inorganica Chimica Acta 438 (2015) 212–219
Table 4
Key sensing parameters of Re(CO)3(POD)Br doped MCM-41 samples.
Sample
k (nm)
I0/I100
KSV1 (O2%ꢂ1
)
KSV2 (O2%ꢂ1
)
f01
f02
R2
Tres (s)
Trec (s)
50 mg/g
60 mg/g
70 mg/g
545
548
548
5.34
5.65
5.03
0.2004
0.1943
0.2019
0.0004
0.0024
0.0031
0.847
0.831
0.792
0.153
0.169
0.208
0.9990
0.9994
0.9991
9
8
9
21
17
22
attack. Third, MCM-41 matrix guarantees efficient O2 transporta-
tion and diffusion, which is positive for O2 sensing.
Never the less, our sensitivity (5.65) is still underdeveloped
compared to literature values [11,12]. After a careful comparison
between our result and literatures, we come to a conclusion that
probe structure difference between our system and literature ones
should be the major factor leading to our underdeveloped sensitiv-
ity. For a Ru(II)-based probe, all diamine ligands (sensing area) are
open for oxygen attack, showing high sensitivity values [11,12]. For
our Re(I)-based probe, its diamine ligand (sensing area) is covered
by other ligands (CO, Br). O2 molecules thus have to find a way to
attack the excited electrons in diamine ligand, compromising
sensitivity.
Compared to Re(CO)3(POD)Br emission in CH2Cl2 (552 nm), Re
(CO)3(POD)Br emission in MCM-41 matrix shows a slight blue
shift, which is attributed to rigidochromism [11,12]. FWHM values
of these composite samples (ꢁ88 nm) are found larger than that of
Re(CO)3(POD)Br in CH2Cl2 (76 nm). We thus come to a conclusion
that Re(CO)3(POD)Br molecules have been trapped in MCM-41
matrix, depressing geometric relaxation.
It seems that 60 mg/g is the optimal doping concentration,
while both a higher one (70 mg/g) and a lower one (50 mg/g) com-
promise sensitivity. Assuming that there are two opposite factors
affecting sensitivity. One should be probe amount, the other should
be self-quenching between probe molecules. Probe emission is
rather weak when doping concentration is low, showing poor
sensitivity. On the other hand, if probe molecules are more than
enough, probe emission may be compromised by their self-
quenching effect, leading to a decreased sensitivity [14–18]. There
may be both enough probe molecules and limited self-quenching
effect in the 60 mg/g doped sample, showing the highest sensitiv-
ity value.
Fig. 8. Stern–Volmer curves of Re(CO)3(POD)Br doped MCM-41 (50 mg/g, 60 mg/g
and 70 mg/g). Sensing error for each point is presented in cap. Inset: emission decay
dynamics of a representative sample (60 mg/g) under pure N2, pure O2 and air
conditions.
3.5.2. Stern–Volmer working curves
To identify the sensing mechanism in Re(CO)3(POD)Br doped
MCM-41 samples, emission decay dynamics of a representative
sample (60 mg/g) are recorded under pure N2, air and pure O2 con-
ditions, as shown by the inset of Fig. 8. Unlike the case in CH2Cl2
solution, Re(CO)3(POD)Br emission in MCM-41 matrix follows sin-
gle exponential decay mode with lifetimes of 7.79
ls in pure N2,
0.57 s in air and 0.43 s in pure O2 conditions. These long-lived
l
l
lifetimes clearly come from Re(CO)3(POD)Br dopant, indicating its
successful loading in MCM-41 matrix. This single exponential
decay pattern suggests that all probe molecules are uniformly
distributed in matrix [11,12]. These lifetime values are much
longer than that in CH2Cl2 solution, which has been attributed to
rigidochromism as above mentioned. It appears that probe excited
state is vulnerable to O2 molecules, suggesting a dynamic sensing
mechanism described as Formula (4), where ‘‘⁄” means excited
state.
Fig. 9. Emission monitoring of Re(CO)3(POD)Br doped MCM-41 (50 mg/g, 60 mg/g
and 70 mg/g) when atmosphere is periodically switched between 100% N2 and 100%
O2.
ing curve should be a linear one with slope of KSV, where I and I0
are emission intensity and maximum emission intensity under
pure N2 condition, respectively, KSV and [O2] are Stern–Volmer con-
stant and O2 concentration, respectively [11,12]. Our Stern–Volmer
working curves shown in Fig. 8, however, are not linear ones,
which means that the sensing procedure in Re(CO)3(POD)Br doped
MCM-41 samples is a complicated one (Fig. 9).
ꢃ
ꢃ
3
1
Probe þ O2 ! Probe þ O2
ð4Þ
Considering that (1) probe molecules are homogeneously dis-
tributed in matrix; (2) their emission follows single exponential
decay mode; (3) sensing mechanism is a dynamic one, emission
spectral response can be analyzed by Stern–Volmer equation
[11,12,17]. As shown by Formula (5), an ideal Stern–Volmer work-
I0=I ¼ 1 þ KSV½O2ꢄ
ð5Þ

L. Daofu / Inorganica Chimica Acta 438 (2015) 212–219
219
We tentatively assume that there are two kinds of sensing sites
in these samples: only one of them is quenchable by O2 molecules
while the other is not. Given this hypothesis, Stern–Volmer equa-
tion should be modified with contribution of each sensing site
taken into account. This modified equation is described as Formula
was incorporated into POD ligand to increase its conjugation chain.
As a consequence, distribution and lifetime of excited electrons
were increased to meet oxygen sensing collision. Our hypothesis
was confirmed by single crystal analysis, theoretical calculation
and photophysical measurement. By doping Re(CO)3(POD)Br into
MCM-41 matrix, oxygen sensing performance was carefully dis-
cussed. It was found that oxygen sensing was finished through a
dynamic mechanism, showing the highest sensitivity of 5.65 and
the shortest response time of 8 s. This promising performance
was attributed to the following factors. First, the large conjugation
chain in POD and the long-lived MLCT excited state increased col-
lision probability between probe and O2 molecules. Second, steric
hindrance for O2 attack around Re(CO)3(POD)Br was limited owing
to its small ligands. Third, the highly ordered tunnels of MCM-41
matrix guaranteed efficient O2 transportation and diffusion. For
further improvement, composite structure should be modified,
hoping to eliminate or decrease the amount of ‘‘far-corner”.
(6), where f f2, KSV1 and KSV2 stand for fractional contributions
1,
and corresponding Stern–Volmer constants of sensing sites,
respectively [11,12]. From fitting parameters listed in Table 4, it
is observed that this modified formula fits well with our working
curves. For all samples, KSV2 value is smaller than KSV1 value by
two orders of magnitude, which means that sensing site-2 is highly
insensitive to O2 molecules. Non-linearity of our working curves
and compromised sensitivity values are caused by this sensing
site-2. This site is tentatively attributed to the ‘‘far-corner” in
composite samples which is impenetrable by O2 diffusion. For
further improvement, composite structure should be modified to
eliminate or decrease the amount of such ‘‘far-corner”.
I0
I
1
¼
ð6Þ
Acknowledgement
f1
f2
þ
1þKSV1pO2
1þKSV2pO2
The author thanks my university for her support and help.
3.5.3. Response/recovery character and photostability
Appendix A. Supplementary material
For a further confirmation on the correlation between sample
emission and O2 existence, emission of Re(CO)3(POD)Br doped
MCM-41 samples is monitored when atmosphere is periodically
switched between 100% N2 and 100% O2. Clearly, sample emission
is dependent on O2 existence. Under pure N2 condition, sample
emission is maintained well, suffering from no photobleaching or
photodecomposition. When atmosphere is switched to pure O2
condition, sample emission is quenched to minimal level instantly.
Such cycle can be fully repeated with no hysteresis. For comparison
convenience, response (Tres)/recovery (Trec) time is defined as the
time taken by a sample to decrease/increase to 95% of its initial
emission intensity when atmosphere is periodically switched
[11,12]. As shown in Table 4, Tres and Trec values are found compa-
rable to or even smaller than literature values, which can be attrib-
uted to the following factors [11,12]. First, the large POD
conjugation chain and the long-lived MLCT excited state increase
sensing collision probability between probe and O2 molecules.
Second, steric hindrance for O2 attack around Re(CO)3(POD)Br is
limited owing to its small ligands. Third, the highly ordered
MCM-41 tunnels guarantee efficient O2 transportation and diffu-
sion. As shown in Table 4, recovery time values are obviously
longer than response time values, which can be explained by diffu-
sion-controlled dynamic response and recovery behavior [21]. This
observation is consistent with literature reports on similar systems
[11,12,17].
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ica.2015.09.024.
References
[1] Y. Guo, L. Zhang, S. Zhang, Y. Yang, X. Chen, M. Zhang, Biosens. Bioelectron. 63
(2015) 61.
[2] D. Sarkar, A. Pramanik, S. Jana, P. Karmakar, T.K. Mondal, Sens. Actuators, B 209
(2015) 138.
[3] M.R. Awual, M.M. Hasan, M.A. Khaleque, Sens. Actuators, B 209 (2015) 194.
[4] C.S. Chu, C.Y. Chuang, Sens. Actuators, B 209 (2015) 94.
[5] O.S. Wolfbeis, Anal. Chem. 80 (2008) 4269.
[6] C. Baleizao, S. Nagl, M. Schaferling, M.N. Berberan-Santos, O.S. Wolfbeis, Anal.
Chem. 80 (2008) 6449.
[7] L.C. Clark Jr., Trans. Am. Soc. Artif. Intern. Organs 2 (1956) 41.
[8] L.W. Winkler, Ber. Dtsch. Chem. Ges. 21 (1888) 2843.
[9] G. Zhang, J. Chen, S.J. Payne, S.E. Kooi, J.N. Demas, C.L. Fraser, J. Am. Chem. Soc.
129 (2007) 15728.
[10] A. Gulino, S. Giuffrida, P. Mineo, M. Purrazzo, E. Scamporrino, G. Ventimiglia,
M.E. Van der Boom, I. Fragala, J. Phys. Chem. B 110 (2006) 16781.
[11] B. Lei, B. Li, H. Zhang, L. Zhang, W. Li, J. Phys. Chem. C 111 (2007) 11291.
[12] B. Lei, B. Li, H. Zhang, S. Lu, Z. Zheng, W. Li, Y. Wang, Adv. Funct. Mater. 16
(2006) 1883.
[13] L. Yang, J.K. Feng, A.M. Ren, M. Zhang, Y.G. Ma, X.D. Liu, Eur. J. Inorg. Chem. 10
(2005) 1867.
[14] J. Jia, Y. Tian, Z. Li, Synth. Met. 161 (2011) 1377.
[15] Z. Si, X. Li, X. Li, H. Zhang, J. Organomet. Chem. 694 (2009) 3742.
[16] L. Zhang, S. Yue, B. Li, D. Fan, Inorg. Chim. Acta 384 (2012) 225.
[17] L. Shi, B. Li, S. Yue, D. Fan, Sens. Actuators, B 137 (2009) 386.
[18] M. Kranenburg, Y.E.M. van der Brugt, P.C.J. Kamer, P.W.N.M. van Leeuwen,
Organometallics 14 (1995) 3081.
[19] L. Zhang, B. Li, Z. Su, Langmuir 25 (2009) 2068.
[20] X.Y. Xu, A.P. Deng, Y. Xiang, J. Lumin. 160 (2015) 151.
[21] S.M. Kuang, D.G. Cuttell, D.R. McMillin, P.E. Fanwick, R.A. Walton, Inorg. Chem.
41 (2002) 3313.
4. Conclusion
As a conclusion, this paper developed Re(CO)3(POD)Br as a
probe for oxygen sensing. An electron-pulling oxadiazole group